Optimized molecular design of ADAPT-based HER2-imaging probes

Subscriber access provided by the Henry Madden Library | California State University, Fresno

Article

Optimized molecular design of ADAPT-based HER2-imaging probes labelled with In and Ga 111

68

Sarah Lindbo, Javad Garousi, Bogdan Mitran, Anzhelika Vorobyeva, Maryam Oroujeni, Anna Orlova, Sophia Hober, and Vladimir Tolmachev Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.8b00204 • Publication Date (Web): 04 Jun 2018 Downloaded from http://pubs.acs.org on June 5, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Optimized molecular design of ADAPT-based HER2-imaging probes labelled with 111In and 68Ga

Sarah Lindbo‡☺, Javad Garousi║☺, Bogdan Mitran§, Anzhelika Vorobyeva║, Maryam Oroujeni║, Anna Orlova§, Sophia Hober‡*, Vladimir Tolmachev║ Author Contributions ☺

Equally contributing author.

‡

School of Engineering in Chemistry, Biotechnology and Health (CBH), Division of Protein Science, KTH Royal Institute of Technology, SE-10691, Stockholm, Sweden

║ §

Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden

Department of Medicinal Chemistry, Uppsala University, Uppsala, Sweden

*Corresponding author: Prof. Sophia Hober CBH Division of Protein Science KTH Royal Institute of Technology Stockholm Sweden E-mail [email protected]

1 ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Radionuclide molecular imaging is a promising tool for visualization of cancer associated molecular abnormalities in vivo and stratification of patients for specific therapies. ADAPT is a new type of small engineered proteins based on the scaffold of an albumin binding domain of protein G. ADAPTs have been utilized to select and develop high affinity binders to different proteinaceous targets. ADAPT6 binds to human epidermal growth factor 2 (HER2) with low nanomolar affinity and can be used for its in vivo visualization. Molecular design of 111

In-labeled anti-HER2 ADAPT has been optimized in several earlier studies. In this study,

we made a direct comparison of two of the most promising variants, having either a DEAVDANS or a (HE)3DANS sequence at the N-terminus, conjugated with a maleimido derivative of DOTA to a GSSC amino acids sequence at the C-terminus. The variants (designated DOTA-C59- DEAVDANS-ADAPT6-GSSC and DOTA-C61-(HE)3DANSADAPT6-GSSC) were stably labeled with 111In for SPECT and 68Ga for PET. Biodistribution of labeled ADAPT variants was evaluated in nude mice bearing human tumor xenografts with different levels of HER2 expression. Both variants enabled clear discrimination between tumors with high and low levels of HER2 expression. 111In-labeled ADAPT6 derivatives provided higher tumor-to-organ ratios compared to 68Ga-labeled counterparts. The best performing variant was DOTA-C61-(HE)3DANS-ADAPT6-GSSC, providing tumor-to-blood ratios of 208±36 and 109±17 at 3 h for 111In and 68Ga labels, respectively.

Keywords ADAPT, HER2, radionuclide imaging, imdium-111, gallium-68, DOTA


Page 2 of 32

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents/Abstract Graphic



Introduction The human epidermal growth factor receptor 2 (HER2) is a molecular target for antibodybased treatment of breast and gastroesophageal cancers. A high HER2 expression level in tumors is a precondition for response to treatment with the monoclonal antibody trastuzumab.1 The major issue is that the level of HER2 expression might be different in metastases than in the primary tumor.2, 3, 4, 5 Radionuclide imaging of HER2 expression might be a suitable approach for determination of HER2 expression status in disseminated cancer 6 since expression of the molecular target in all metastases can be evaluated using a single scan. It has to be noted that the so called partial volume effect results in an appreciable underestimation of tracer uptake in metastases with a size that is numerically smaller than resolution of the imaging devices.7, 8 This creates a requirement for very high ratios between radioactivity concentration in tumors and surrounding tissues to avoid false-negative diagnosis.8 One approach for visualization of molecular target expression is the use of therapeutic antibodies conjugated with an appropriate chelator and labeled with a long-lived positronemitting nuclide such as 89Zr.9, 10 This approach utilizes the high sensitivity and resolution of positron emission tomography (PET) and the specificity of monoclonal antibodies. Limitations of this approach include possible false-positive findings due to unspecific accumulation of macromolecules in tumors and low contrast due to slow clearance of large immunoglobulins from blood.10 The use of small proteins as imaging agents seems to be a way to overcome limitations associated with monoclonal antibodies.11 A very promising type of small proteins for this purpose are engineered scaffold proteins.12 Scaffold proteins are characterized by robust frameworks that permit variegation of surface exposed residues, which enables selection of high affinity binders using molecular display approaches. The affibody, knottin, DARPin and 4 ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fibronectin domain scaffolds have shown to be suitable domains for development of radionuclide-based probes.12 Recently, feasibility to use the albumin binding domain (ABD) from streptococcal protein G as scaffold for development of imaging probes has been demonstrated.13 This new class of scaffold proteins is denoted ADAPTs (ABD-Derived Affinity ProTeins). First variants of ADAPT (Supplemental Figures 1A and 1B) has demonstrated capacity to provide high-contrast radionuclide images of HER2-expressing tumor xenografts in mice.13 Experience with affibody molecules demonstrates that optimization of the amino acid composition, charge, and lipophilicity14, 15, together with the use of labels with beneficial physicochemical properties16, 17 increases the imaging contrast provided by the probe. These factors have also shown to influence the biodistribution of ADAPTs. However, the influence is not quite similar to influence on affibody molecules. For example, the use of a HEHEHE tag appreciable increased the blood clearance rate and decreased the hepatic uptake of affibody molecules in comparison with a hexahistidine tag, but this was not the case for ADAPTs (Supplemental Figures 1A, 1C, 1D and 1E).18 This encouraged us to initiate a series of systematic studies concerning influence of various factors on the imaging properties of ADAPTs. Influence of N-terminal composition of ADAPT6 was evaluated with the radionuclide 111In, conjugated site-specifically to the N-terminus using a maleimido derivative of the DOTA chelator (Supplemental Figures 1A, 1F, 1G, 1H and I). The results showed that the N-terminal sequences GC(HE)3DANS (Supplemental Figure 1F), and GCVDANS (Supplemental Figure 1I) provided appreciably higher tumor-to-organ ratios than GCDEAVDANS (Supplemental Figure 1H).19 Further evaluation demonstrated that positioning of the [111In]In-DOTA label at the C-terminus (Supplemental Figure 1J) provided significantly higher tumor-to-lung, tumor-to-liver, tumor-to-spleen and tumor-to-muscle ratios compared with positioning at N-terminus (Supplemental Figure 1F).20 This behavior is 5 ACS Paragon Plus Environment


probably associated with the increase of the local hydrophilicity of the C-terminus. Another study evaluated the feasibility of using a peptide-based cysteine-containing chelator (GSSC) at the C-terminus of ADAPT6 for labeling with 99mTc (Supplemental Figure 1M).21 Since the presence of (HE)3-sequence in the targeting protein results in loss of site-specificity of 99mTc to peptide-based cysteine-containing chelators22, the N-terminal sequence GDEAVDANS was used in that construct. As a comparator, we have used an 111In-labeled counterpart containing DOTA (Supplemental Figure 1N).21 This construct provided a substantial increase in tumorto-organ ratios compared to its N-terminally, [111In]In-DOTA labeled counterpart attached via DOTA to GCDEAVDANS-sequence at N-terminus and not having GSSC at C-terminus (Supplemental Figure 1H).19, 21 Overall, these studies suggested that increase of hydrophilicity of the C-terminus of ADAPT6 improves the tumor-to-organ ratios. Earlier ADAPT6 studies have mainly utilized the relatively long-lived radionuclide 111In (T1/2 = 2.8 d), suitable for use in single photon computed tomography (SPECT). In addition, the use of a long-lived label permits biodistribution experiments several hours after injection, which enables studies in the elimination phase. However, it would be attractive to exploit the advantages of PET for ADAPT-based imaging. Here the radionuclide 68Ga (T1/2 = 67.6 min) would be a suitable label since the half-life of 68Ga is compatible with the rapid in vivo kinetics of ADAPT. Additionally, the short half-life of 68Ga results in a low absorbed dose in patients. Moreover, generator-based production makes 68Ga easily available in clinics23, the labeling chemistry of 68Ga is straightforward and DOTA provides the formation of stable complex with gallium23. It has to be noted that the geometry of gallium in complex with DOTA is different form the geometry of indium complex (Figure 1A).24 This may cause conformational changes in the labeled protein, as well as different local distribution of hydrophilicity, which might influence off-target interactions of imaging probes. Thus, substitution of 111In to 68Ga may trigger considerable modification of the biodistribution and 6 ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


tumor-to-organ ratios of short peptides25,26 and scaffold proteins17,27,28. Hence, it cannot be assumed that the optimal molecular design of an imaging probe labeled with 111In is the optimal for 68Ga labeling. This necessitates a re-evaluation of imaging probes after replacement of 111In to 68Ga. The goals of this study were: - To compare the influence of [111In]In- and [68Ga]Ga-DOTA labels, conjugated sitespecifically to the cysteine in the C-terminus of ADAPT6, on the biodistribution and targeting properties of variants having GDEAVDANS or G(HE)3DANS N-terminal sequences (Figure 1B); - To compare the ability of the tested 68Ga-labelled variants to discriminate between xenografts with high and low levels of HER2 expression using PET imaging.

Figure 1. A. Structure of indium and gallium complexes with monoamide DOTA. Atoms involved in the complex formation are highlighted. B. Amino acid sequences of (HE)3DANSADAPT6-GSSC and DEAVDANS-ADAPT6-GSSC. N-terminal sequences are marked by blue and C-terminal cysteines (sites of the chelator conjugation) are marked by red fonts.

Experimental Section 7 ACS Paragon Plus Environment


Materials and general methods DOTA-C59-DEAVDANS-ADAPT6-GSSC and DOTA-C61-(HE)3DANS-ADAPT6-GSSC were produced, purified and characterized as described earlier20,21. Buffers used for radiolabeling were purified from metal impurities by treatment with ion-exchange resin Chelex 100 (BioRad Laboratories). [68Ga]Ga was obtained by elution from a 68Ge/68Ga generator (Eckert and Ziegler) with 0.1 M hydrochloric acid. Fractions of 200 µl were collected and the fraction with the highest radioactivity concentration was used for radiolabeling. 111In was purchased from Mallinckrodt Sweden AB as a solution in 0.01 M hydrochloric acid. Radioactivity measurement during in vitro cell and biodistribution studies was performed using an automated γ-spectrometer with a NaI(Tl) detector (1480 WIZARD; Wallac Oy). Radioactivity measurements for formulation of injection solution were done using VDC-405 ionization chamber (Veenstra Instruments BV, The Netherlands). Quantitative analysis of radioactivity distribution in SDS-PAGE gels and ITLC strips was performed using Cyclone Storage Phosphor System (PerkinElmer). GraphPad Prism (version 4.00 for Windows; GraphPad Software) was used for statistical analysis of difference in cellular binding and biodistribution data. An unpaired 2-tailed t-test was used when two groups were compared. Analysis of difference between uptake of 68Ga and 111In in co-injection studies was performed using paired t-test. Radiolabeling and in vitro stability studies Labeling of DOTA-C59- DEAVDANS-ADAPT6-GSSC and DOTA-C61-(HE)3DANSADAPT6-GSSC with 68Ga was performed by mixing 50 µg (8 nmol) of DOTA-conjugated protein in 40 µL 1.25 M sodium acetate, pH 3.6, with 100 µL eluate containing ca. 100 MBq 68

Ga. The mixture was incubated at 95°C for 20 min, thereafter tetrasodium salt of ethylene 8 ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


diamine tetraacetic acid (Na4EDTA) (50 µL, 0.1 mg/mL) was added and the mixture was further incubated at 95°C for 5 min. Labeled ADAPTs were purified using size-exclusion NAP-5 columns. Radiochemical yield and radiochemical purity of 68Ga-labeled ADAPTs were determined using radio instant thin layer chromatography (ITLC) eluted with 0.2 M citric acid, as described earlier.13 The ITLC was validated by radio-sodium dodecyl sulfate polyacrylamide gel electrophoresis (radio-SDS-PAGE) at 200 V, using NuPAGE 4–12 % Bis-Tris Gels (Invitrogen AB) in MES buffer (Invitrogen AB). Radiolabeling of DEAVDANS-ADAPT6-GSSC-DOTA and (HE)3DANS-ADAPT6-GSSCDOTA with 111In was performed as describe earlier.20,21 Stability of the radionuclides coupling was tested by incubation with 1000-fold excess of Na4EDTA. Samples diluted with an equal volume of PBS were used as controls.

In vitro binding specificity and cellular processing of bound conjugates by HER2expressing cells For in vitro studies, HER2-expressing ovarian carcinoma SKOV-3 and breast carcinoma BT474 cell lines were used (American Type Culture Collection, Manassas, VA). The studies were performed in the same way as for the 111In-DEAVDANS-ADAPT6-GSSC-DOTA.21 Cells (ca. 1×106 per culture dish) were seeded in cell culture dishes one day before experiment. For each conjugate, a set of six dishes was prepared. The binding specificity was tested by a saturation test. In three dishes, receptors were saturated by adding 2.5 mM of unlabeled conjugate before adding a radiolabeled ADAPT. The labeled ADAPTs were added to all dishes to a concentration of 25 nM, and the cells were incubated at 37°C for 1 h. After



the incubation, the medium was removed, cells were washed twice with PBS and detached by trypsin. The cell-associated radioactivity was measured and percentage of cell-bound radioactivity was calculated. Internalization of conjugates was evaluated using a modified acid wash method described earlier.21

In vivo studies Animal studies were planned in agreement with Swedish national legislation concerning protection of laboratory animals and were approved by the Ethics Committee for Animal Research in Uppsala. Female BALB/C nu/nu mice (Scanbur A/S) bearing SKOV-3 xenografts with high HER2 expression (1.6×106 receptors/cell)29 and DU 145 xenografts with low HER2 expression (5.1×104 receptors/cell)30 were use as models for in vivo studies. Xenografts in hind legs were established by subcutaneous implantation of 107 SKOV-3 or 5×106 DU 145 cells. Experiments were performed three weeks after implantation. The average mouse weight at the time of the study was 16±2 g and the average tumor weights were 0.48±0.13 and 0.28±0.09 g for SKOV-3 and DU 145, respectively. Targeting properties of 111In- and 68Ga-labeled ADAPT6 variants were compared by coinjection of both probes in the same mice. For this, ADAPT6 variants were separately labeled with 111In and 68Ga. The labeled variants were mixed before injection and total injected protein mass was adjusted to 10 µg with non-labeled conjugate. This injected protein dose was selected because higher amount of injected protein provides good discrimination between xenografts with high and low HER2 expression13,31. For animals sacrificed at 1 h after injection, 200 kBq 68Ga-labeled ADAPT and 10 kBq 111In-labeled ADAPT were mixed. For animals sacrificed at 3 h after injection, 700 kBq 68Ga-labeled ADAPT and 10 kBq of 111In10 ACS Paragon Plus Environment

Page 10 of 32

Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


labeled ADAPT were mixed. Biodistribution in mice bearing SKOV-3 xenografts was measured at 1 and 3 h after injection. Biodistribution in mice bearing DU 145 xenografts was measured at 3 h after injection. Mice were sacrificed by i.p. injection of a lethal dose of anesthesia (Ketalar, 10 mg/mL; Rompun, 1 mg/mL, 20 µL/g body weight) followed by heart puncture and exsanguination. Blood and tissue samples were collected and weighed. Radioactivity of each nuclide in each sample was determined as described earlier.28 Radioactivity uptake was calculated as percent of injected dose per gram tissue (%ID/g). Imaging Mice bearing SKOV-3 or DU 145 xenografts were injected with [68Ga]Ga-DOTA-C59DEAVDANS-ADAPT6-GSSC (10 µg, 10-12 MBq) and [68Ga]Ga-DOTA-C61-(HE)3DANSADAPT6-GSSC (10 µg, 15-16 MBq) . Whole body PET/CT scans were performed using the Triumph™ Trimodality System (TriFoil Imaging, Inc., Northridge, CA, USA) at 3 h p.i. The mice were euthanized with CO2 asphyxiation immediately before being scanned. The PET scans were performed for 30 min (field of view (FOV) 8.0 cm) followed by computed tomography (CT) acquisitions at the following parameters: FOV, 8.0 cm; magnification, 1.48; one projection and 512 frames for 2.13 min. The PET data were reconstructed into a static image using an ordered subsets expectation maximization 3D algorithm (20 iterations). The CT raw data were reconstructed using filtered back projection. PET and CT data were fused and analyzed using PMOD v3.510 (PMOD Technologies Ltd., Zurich, Switzerland). Coronal PET/CT images of the scans were presented as SUV in RGB color scale. To assess the capacity of radioconjugates to discriminate between low and high HER2 expression the images were presented at SUV 4 and SUV 0.5.



Page 12 of 32

RESULTS Protein production, purification, conjugation and characterization (HE)3DANS-ADAPT6-GSSC and DEAVDANS-ADAPT6-GSSC were successfully produced in E. coli, harvested and purified to homogeneity. Both protein constructs were efficiently conjugated with maleimide-DOTA and the molecular weights were confirmed by mass spectrometry (Table 1). The final purity for both constructs was determined to above 98% by reverse-phase high-performance liquid chromatography. Circular dichroism measurements showed high α-helical content and the melting temperatures were similar for both conjugates, above 65°C (Table 1). Both constructs also demonstrated excellent refolding ability after thermal denaturation (data not shown). The affinities were determined using surface plasmon resonance to 3.5 and 4.5 nM for DEAVDANS-ADAPT6-GSSC and (HE)3DANS-ADAPT6-GSSC, respectively (Table 1). These values are comparable with the affinity for the parental ADAPT6 molecule (2.5 nM).13 Table 1. Biophysical characteristics of DOTA-labeled ADAPT6 variants

DOTA-C59-

DEAVDANSADAPT6GSSC

DOTA-C61(HE)3DANS-

Number of amino acids 58

61

Theoretical Measured molecular molecular weight (Da) weight (Da)

Tm (C°)

6292.0

6292.6

68

KD (nM) as determined by SPR measurements 3.5

6676.4

6676.6

67

4.5

ADAPT6GSSC

pI

Net charge at pH 7

4.83

-2.9

5.58

-3.6

Tm: melting point; KD: equilibrium dissociation constant; SPR: surface plasmon resonance; pI: isoelectric point.


Page 13 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Radiolabeling The radiochemical yields of [68Ga]Ga-DOTA-C59-DEAVDANS-ADAPT6-GSSC, [68Ga]GaDOTA-C61-(HE)3DANS-ADAPT6-GSSC and [111In]In-DOTA-C61-(HE)3DANS-ADAPT6GSSC labeling were 93.2±0.5%, 96.9±1.1% and 95.5±0.7%, respectively. Size-exclusion chromatography purification provided a radiochemical purity of more than 99% for all conjugates. The isolated decay corrected yields were 88±1%, 90±1% and 90±2% for [68Ga]Ga-DOTA-C59-DEAVDANS-ADAPT6-GSSC, [68Ga]Ga-DOTA-C61-(HE)3DANSADAPT6-GSSC and [111In]In-DOTA-C61-(HE)3DANS-ADAPT6-GSSC, respectively. Apparent molar specific activity of 68Ga-labeled conjugates was up to 11 GBq/µmol, and of [111In]In-DOTA-C61-(HE)3DANS-ADAPT6-GSSC up to 6 GBq/µmol. No release of radioactivity under challenge with 1000-fold excess of EDTA during 2 h was detected for either of the conjugates, as the differences between the control samples were within accuracy of the analytical method (Supplemental Table 1).

In vitro binding specificity and cellular processing of bound conjugates by HER2expressing cells Pre-saturation of HER2 receptors with non-labeled constructs resulted in a significant (p

Optimized molecular design of ADAPT-based HER2-imaging probes

Recommend Documents