City University of New York (CUNY) CUNY Academic Works
All Dissertations, Theses, and Capstone Projects Dissertations, Theses, and Capstone Projects
2-2020
Leveraging Antibodies for Positron Emission Tomography and Near-Infrared Fluorescence Imaging of Cancer
Kimberly C. Fung The Graduate Center, City University of New York
How does access to this work benefit ou?y Let us know!
More information about this work at: https://academicworks.cuny.edu/gc_etds/3566 Discover additional works at: https://academicworks.cuny.edu
This work is made publicly available by the City University of New York (CUNY). Contact: [email protected]
LEVERAGING ANTIBODIES FOR POSITRON EMISSION
TOMOGRAPHY AND NEAR-INFRARED FLUORESCENCE
IMAGING OF CANCER
by
KIMBERLY CAROL FUNG
A dissertation submitted to the Graduate Faculty in Chemistry in partial fulfillment of the
requirements for the degree of Doctor of Philosophy, The City University of New York
2020
© 2020
KIMBERLY CAROL FUNG
All Rights Reserved
ii
Leveraging Antibodies for Positron Emission Tomography and Near-Infrared Fluorescence Imaging of Cancer
by
Kimberly Carol Fung
This manuscript has been read and accepted for the Graduate Faculty in Chemistry in satisfaction of the dissertation requirement for the degree of Doctor of Philosophy.
Date Brian M. Zeglis
Chair of Examining Committee
Date Brian R. Gibney
Executive Officer
Supervisory Committee:
Lynn C. Francesconi
Donna McGregor
THE CITY UNIVERSITY OF NEW YORK
iii ABSTRACT
Leveraging Antibodies for Positron Emission Tomography and Near-
Infrared Fluorescence Imaging of Cancer
by
Kimberly Carol Fung
Advisor: Brian M. Zeglis
The high specificity and affinity of antibodies make them attractive for developing drugs to diagnose and treat cancer. The overarching goal of this work is to explore the synthesis and use of antibody-based imaging agents in preclinical models of cancer. This work can be described as two- fold. In the first part, we investigated how the use of a glycans-specific bioconjugation strategy affects FcRI binding and why it results in improved in vivo performance of immunoconjugates.
To do so, we used the clinically relevant positron emission tomography (PET) imaging agent, 89Zr-
DFO-Pertuzumab, in mouse models of human breast cancer. In the second part, we developed two imaging agents for ovarian cancer, 89Zr-DFO-AR20.5 for PET and ssB43.13-IR800 for near- infrared fluorescence imaging and evaluated their in vivo performance in mouse models of human ovarian cancer. With this body of work, we hope to demonstrate the potential of antibodies as effective imaging agents for cancer.
iv ACKNOWLEDGMENTS
First and foremost, I would like to thank my mentor, Prof. Brian Zeglis. I truly appreciate all the
support, encouragement, and help you have given me throughout this roller coaster that is grad
school. Your intelligence, passion for science, and kindness inspires me to achieve more. Thank
you for helping me believe in myself!
I would like to express my sincere gratitude to my committee members:
Prof. Lynn Francesconi: Thank you for giving me my first research experience as an
undergraduate, introducing me to radiochemistry, and continuing to offer your help and advice
throughout grad school.
Prof. Donna McGregor: Thank you for giving me the opportunity to work in your lab as an
undergraduate and continuing to support me and provide valuable feedback in grad school.
Prof. Melissa Deri: Thank you for stepping in last minute during my defense.
I would like to thank my collaborators: Dr. Jason Lewis and Dr. Eric Price for their invaluable
help with my thesis work.
I am also grateful to the IGERT program for providing funding for the first four years of my PhD.
To past and present members of the Zeglis lab: Rosemery, Brendon, Delphine, Pierre, Sai, Cindy,
Outi, Guillaume, Samantha, Doug, Stephen – Thank you for being the best lab mates I could ask
for! I have learned so much from you all and I will always remember how you all looked out for
me during this journey.
To my lab friends with a special shout out to Jeannette, Michelle, Justin, and Sophia – I
appreciate you all for encouraging me, offering words of advice, and support, especially
surrounding my defense. Thank you for keeping me sane during those late nights and weekends
in the lab and the fun times outside the lab. I am so grateful to have met you all and I know you
will all reach your goals and get your dream jobs!
v To the gals from 500 Women Scientists NYC Pod leadership team – I’m grateful to have met you
all during my Ph.D! Thank you for inspiring me to keep going and achieve more.
To my high school advisor, Ms. Sarah Petersen – Thank you for believing in me and encouraging
me to pursue chemistry.
To my longtime friends: Jessica, Tejal, Rubana, Neha, Shima, Jonathan – Thank you for helping
me get through all those of science classes in undergrad, inspiring me with your continued successes and perseverance, and being there for me. I am so proud of you all and grateful for your
friendship!
To my family: My dad and aunt, Manna, thank you for continuously encouraging and supporting
me throughout this process!
Last, but not least, my mom who has been with me every step of way, through all the ups and downs. Thank you for believing in me and inspiring me to be the best that I can be. Thank you for
keeping me company and staying with me during those late nights of work. I would not have
gotten to this point without you. I love you!
This dissertation consists of one published paper and two papers which are under preparation for
publication. Chapter 2 was published as “Vivier, D.; Fung, K.; Rodriguez, C.; Adumeau, P.;
Ulaner, G. A.; Lewis, J. S.; Sharma, S. K.; Zeglis, B. M. The influence of glycans-specific
bioconjugation on the FcɣRI binding and in vivo performance of 89Zr-DFO-pertuzumab.
Theranostics. 2020, 10 (4), 1746-1757.” Chapters 3 and 4 are under preparation and will be
published in the near future.
vi TABLE OF CONTENTS Chapter 1 : Introduction ...... 1 1.1. Molecular Imaging ...... 2 1.1.1. PET Imaging ...... 4 1.1.1.1. Principles of PET Imaging ...... 4 1.1.1.2. Positron Emitters ...... 6 1.1.2. Fluorescence Imaging ...... 7 1.1.2.1. Principles of Fluorescence Imaging ...... 7 1.1.2.2. Fluorophores ...... 9 1.2. Structure of Antibodies ...... 14 1.3. Bioconjugation Strategies ...... 15 1.3.1. Traditional Conjugation Method ...... 15 1.3.2. Site-specific Conjugation Method ...... 16 1.5. References ...... 19 Chapter 2 : The Influence of Glycans-Specific Bioconjugation on the FcɣRI Binding and In Vivo Performance of 89Zr-DFO-Pertuzumab ...... 23 2.1. Introduction ...... 23 2.2. Experimental Section ...... 26 2.2.1. Reagents and General Procedures ...... 26 2.2.2. SDS-PAGE Analysis ...... 26 2.2.3. Size Exclusion Chromatography ...... 27 2.2.4. Flow Cytometry ...... 27 2.2.5. FcRI Binding ELISA ...... 27 2.2.6. Surface Plasmon Resonance ...... 28 2.2.7. Radiolabeling with 89Zr ...... 30 2.2.8. Bead-based Immunoreactivity Assay ...... 31 2.2.9. Radioimmunoconjugate Stability Assays ...... 32 2.2.10. PET Imaging ...... 32 2.2.11. Acute Biodistribution Experiments ...... 32 2.2.12. Statistical Analysis ...... 33 2.3. Results ...... 33 2.3.1. Model System, Synthesis, and Characterization ...... 33 2.3.2. In Vitro Characterization ...... 35 2.3.3. Radiolabeling ...... 39 2.3.4. In Vivo Behavior ...... 39 2.4. Discussion ...... 44 2.5. Conclusion ...... 47 2.6. References ...... 49
Chapter 3 : [89Zr]Zr-DFO-AR20.5: A MUC1-Targeting ImmunoPET Probe ...... 52 3.1. Introduction ...... 52
vii 3.2. Experimental ...... 54 3.2.1. Materials and Methods ...... 54 3.2.2. Instrumentation ...... 55 3.2.3. DFO modification of antibodies ...... 55 3.2.4. Radiolabeling with 89Zr ...... 56 3.2.5. Cell Culture ...... 57 3.2.6. Flow Cytometry ...... 57 3.2.7. Radiolabeled Antibody Stability Assays ...... 57 3.2.8. Xenograft Models ...... 58 3.2.9. PET Imaging ...... 59 3.2.10. Acute Biodistribution ...... 60 3.2.11. Histopathology ...... 60 3.3. Results ...... 61 3.5. Conclusion ...... 71 3.6. References ...... 73 Chapter 4 : A Molecularly-Targeted Intraoperative Imaging Agent for High-Grade Serous Ovarian Cancer ...... 77 4.1. Introduction ...... 77 4.2. Experimental ...... 79 4.2.1. Materials and Methods ...... 79 4.2.2. Synthesis of nssB43.13-IR800 and nssmIgG-IR800 ...... 79 4.2.3. Synthesis of ssB43.13-IR800 ...... 80 4.2.4. Cell Culture ...... 81 4.2.5. Flow Cytometry ...... 82 4.2.6. Xenograft Models ...... 82 4.2.7. NIRF Imaging ...... 83 4.2.8. Histopathology ...... 84 4.3. Results and Discussion ...... 84 4.3.1. PET Imaging of CA125 Expression in Ovarian Cancer ...... 84 4.3.2. Near-Infrared Fluorescence Imaging of CA125 Expression in Ovarian Cancer ...... 88 4.4. Conclusion ...... 102 4.5. References ...... 103 Chapter 5 : Conclusions ...... 106 APPENDICES ...... 108 CHAPTER 2: APPENDIX ...... 108 CHAPTER 3: APPENDIX ...... 123
viii LIST OF FIGURES
MAIN TEXT: Figure 1.1. Principle of positron emission tomography (PET)...... 5 Figure 1.2. Geometric and basic principle of fluorescence imaging ...... 8 Figure 1.3. Absorption of light ...... 9 Figure 1.4. Detailed structural schematic of a full-length IgG ...... 15 Figure 1.5. Two most common types of bioconjugation reactions ...... 16 Figure 1.6. (A) The biantennary structure of the heavy chain glycans; (B) Structures of natural and synthetic monosaccharides ...... 17 Figure 1.7. Schematic of a Gal-T(Y289L)-based site-specific modification procedure ...... 18 Figure 2.1. The construction of the pertuzumab immunoconjugates ...... 35 Figure 2.2. Comparative SPR and ELISA analysis of the interaction between pertuzumab immunoconjugates and recombinant human versus mouse FcRI...... 38 Figure 2.3. (A) Planar and maximum intensity projection PET images of athymic nude mice bearing subcutaneous BT474 xenografts injected with the three 89Zr-DFO-pertuzumab radioimmunoconjugates; (B) Biodistribution data for athymic nude mice bearing HER2-expressing
BT474 xenografts injected with 89Zr-DFO-nsspertuzumab, 89Zr-DFO-sspertuzumab-EndoS, or 89Zr-
DFO-sspertuzumab-Gal ...... 41 Figure 2.4. (A) Planar and maximum intensity projection PET images of huNSG mice bearing subcutaneous BT474 xenografts after the administration of the three radioimmunoconjugates; (B)
Biodistribution data for 89Zr-DFO-nsspertuzumab, 89Zr-DFO-sspertuzumab-βGal, and 89Zr-DFO- sspertuzumab-EndoS following administration in huNSG mice bearing subcutaneous HER2- expressing BT474 xenografts ...... 43
Figure 3.1. Schematic of the bioconjugation and radiosynthesis of [89Zr]Zr-DFO-AR20.5 ...... 61 Figure 3.2. Representative MALDI-ToF spectrum for AR20.5 ...... 62 Figure 3.3. Representative MALDI-ToF spectrum for DFO-AR20.5 ...... 62 Figure 3.4. Flow cytometry data of DFO-AR20.5 with SKOV3 human ovarian cancer cells...... 63
Figure 3.5. Stability of [89Zr]Zr-DFO-AR20.5 in human serum ...... 64 Figure 3.6. Planar and maximum intensity projection PET images of athymic nude mice bearing subcutaneous SKOV3 xenografts following the intravenous tail-vein injection of [89Zr]Zr-DFO-
ix AR20.5 or [89Zr]Zr-DFO-mIgG ...... 65 Figure 3.7. Biodistribution data from athymic nude mice bearing SKOV3 human ovarian cancer xenografts after the intravenous administration of [89Zr]Zr-DFO-AR20.5 ...... 67 Figure 3.8. (A) Bioluminescence, planar, and maximum intensity projection PET images of athymic nude mice bearing orthotopic SKOV3-Red-FLuc xenografts following the intravenous tail-vein injection of [89Zr]Zr-DFO-AR20.5 or [89Zr]Zr-DFO-mIgG; (B) Planar PET image of the representative athymic nude mouse bearing an orthotopic SKOV3-Red-FLuc xenograft at 120 hours post-injection of [89Zr]Zr-DFO-AR20.5; (C) H&E and IHC staining of the peritoneal metastatic lesion from the representative mouse ...... 68 Figure 3.9. (A) Biodistribution data for athymic nude mice bearing orthotopic SKOV3-Red-FLuc xenografts following the intravenous tail-vein injection of [89Zr]Zr-DFO-AR20.5; (B) Planar and maximum intensity projection PET images of a representative athymic nude mouse bearing an orthotopic SKOV3-Red-FLuc xenograft; (C) Biodistribution data for the metastatic lesions collected from the orthotopic tumor-bearing mice ...... 69
Figure 4.1. CA125-targeted PET imaging with 89Zr-DFO-B43.13 ...... 85
Figure 4.2. The visualization of lymph nodes with 89Zr-DFO-B43.13 ...... 86
Figure 4.3. The visualization of lymph nodes with 89Zr-DFO-B43.13. Coronal (A) and maximum intensity projection (B) PET images after the administration of 89Zr-DFO-B43.13 to a mouse bearing a CA125-positive OVCAR3 xenograft; (C) PET/CT image after the administration of 89Zr- DFO-B43.13 to the same mouse ...... 86
Figure 4.4. Comparative analysis of activity concentrations from 89Zr-DFO-B43.13 vs. 89Zr-DFO- Isotype IgG in select tissues of interest ...... 87 Figure 4.5. Histopathological analysis of contralateral and ipsilateral sub-mandibular lymph nodes ...... 88 Figure 4.6. Schematic of the chemoenzymatic method for site-specific labeling of the B43.13 antibody ...... 90 Figure 4.7. In vitro selectivity ...... 91 Figure 4.8. Near infrared fluorescence (NIRF) in vivo imaging of OVCAR3 subcutaneous xenograft with B43.13-AF680 ...... 92
Figure 4.9. Synthesis and in vitro characterization of ssB43.13-IR800 conjugates ...... 93 Figure 4.10. In vivo NIRF imaging with B43.13-IR800 ...... 94
x Figure 4.11. In vivo and ex vivo NIRF imaging of ovarian cancer using IR800-Labeled-B43.13...... 94
Figure 4.12. In vivo and ex vivo NIRF imaging with ssB43.13-IR800, nssB43.13-IR800, and ssmIgG- IR800 in an NSG mouse bearing a subcutaneous OVCAR3 xenograft ...... 97
Figure 4.13. In vivo and ex vivo NIRF imaging with ssB43.13-IR800, nssB43.13-IR800, and ssmIgG- IR800 in an NSG mouse bearing an orthotopic OVCAR3 xenograft...... 98
Figure 4.14. In vivo and ex vivo NIRF imaging with ssB43.13-IR800, nssB43.13-IR800, and ssmIgG- IR800 in an NSG mouse bearing an orthotopic OVCAR3 xenograft...... 99 Figure 4.15. Histopathologic analysis of tumor and lymph node samples from high grade serous ovarian cancer patient ...... 101
APPENDIX: Chapter 2 Figure A.2.1. Representative MALDI-ToF spectrum for native pertuzumab ...... 109
Figure A.2.2. Representative MALDI-ToF spectrum for DFO-nsspertuzumab ...... 109
Figure A.2.3. Representative MALDI-ToF spectrum for N3-sspertuzumab-βGal ...... 110
Figure A.2.4. Representative MALDI-ToF spectrum for DFO-sspertuzumab-βGal ...... 110
Figure A.2.5. Representative MALDI-ToF spectrum for N3-sspertuzumab-EndoS ...... 111
Figure A.2.6. Representative MALDI-ToF spectrum for DFO-sspertuzumab-EndoS ...... 111 Figure A.2.7. Denaturing SDS-PAGE of the immunoconjugates ...... 112 Figure A.2.8. Size exclusion chromatogram of native pertuzumab ...... 113
Figure A.2.9. Size exclusion chromatogram of DFO-nsspertuzumab ...... 113
Figure A.2.10. Size exclusion chromatogram of DFO-sspertuzumab-βGal...... 114
Figure A.2.11. Size exclusion chromatogram of DFO-sspertuzumab-EndoS ...... 114 Figure A.2.12. SPR sensorgrams for the binding of (A) native pertuzumab, (B) DFO- nsspertuzumab, (C) DFO-sspertuzumab-EndoS, and (D) DFO-sspertuzumab-βGal for recombinant HER2 ...... 115 Figure A.2.13. Flow cytometry data showing the binding of the immunoconjugates to HER2- expressing BT474 human breast cancer cells ...... 116 Figure A.2.14. Stability of the three radioimmunoconjugates to demetallation in human serum ...... 117
xi LIST OF TABLES MAIN TEXT: Table 1.1. Features of select molecular imaging modalities ...... 4 Table 1.2. Physical decay characteristics of the most commonly used PET radionuclides...... 6 Table 1.3. Structure, excitation/emission wavelength, and properties of clinically employed fluorophores ...... 14 Table 2.1. SPR-derived binding parameters for the various pertuzumab immunoconjugate and HER2 ...... 35 Table 2.2. SPR-derived binding parameters for the various pertuzumab immunoconjugate and human and murine FcRI ...... 39 Table 3.1. Degree of labeling of DFO-AR20.5 as determined via MALDI-MS ...... 63 Table 3.2. Biodistribution data from athymic nude mice bearing SKOV3 human ovarian cancer xenografts after the intravenous administration of [89Zr]Zr-DFO-AR20.5 ...... 67 Table 3.3. Biodistribution data from athymic nude mice bearing SKOV3-RFluc human ovarian cancer orthotopic xenografts after the intravenous administration of [89Zr]Zr-DFO-AR20.5 or
[89Zr]Zr-DFO-mIgG via the tail vein ...... 69
APPENDIX: Chapter 2 Table A.2.1. Degree of labeling of each immunoconjugate as determined via MALDI-MS ...... 118 Table A.2.2. Biodistribution data for the three constructs in nude mice bearing BT474 subcutaneous xenografts ...... 119 Table A.2.3. Tumor-to-tissue activity concentration ratios for the three constructs in nude mice bearing BT474 subcutaneous xenografts...... 120 Table A.2.4. Biodistribution data for the three radioimmunoconjugates at 144 h post-injection in humanized NSG mice bearing BT474 subcutaneous xenografts ...... 121 Table A.2.5. Tumor-to-tissue activity concentration ratios for the three radioimmunoconjugates in humanized NSG mice bearing BT474 subcutaneous xenografts ...... 122 Chapter 3 Table A.3.1. Tumor-to-background activity concentration ratios calculated using the biodistribution data from athymic nude mice bearing SKOV3 human ovarian cancer xenografts
xii after the intravenous administration of [89Zr]Zr-DFO-AR20.5 ...... 124 Table A.3.2. Tumor-to-organ activity concentration ratios derived from the biodistribution data from athymic nude mice bearing SKOV3-RFluc human ovarian cancer orthotopic xenografts after the intravenous administration of [89Zr]Zr-DFO-AR20.5 or [89Zr]Zr-DFO-mIgG ...... 125 Table A.3.3. Biodistribution data for the metastatic lesions collected from athymic nude mice bearing SKOV3-RFluc human ovarian cancer orthotopic xenografts after the intravenous administration of [89Zr]Zr-DFO-AR20.5 ...... 126
xiii Chapter 1 : Introduction
The unmatched ability of antibodies to specifically target disease-associated antigens makes them desirable for developing drugs to diagnose, treat, and manage cancer.1-4 In recognizing the clinical potential of antibodies for diagnostic imaging, research into antibody-based molecular imaging agents has grown considerably over the past two decades, especially in oncology.2 Two imaging modalities that have been studied extensively are positron emission tomography (PET) and near-infrared fluorescence (NIRF) imaging. PET and NIRF imaging provide noninvasive ways to visualize biochemical processes in vivo. PET is a highly sensitive and quantitative nuclear imaging technique that enables whole-body imaging. NIRF imaging is a clinically relevant technique that holds great promise for image-guided surgery. With these studies, our goal is to leverage antibodies for PET and NIRF imaging in preclinical mouse models of cancer.
This body of work explores the synthesis and application of immunoconjugates for the imaging of cancer. Chapter 2 investigates the effect of bioconjugation strategy on the FcyRI receptor and the in vivo performance of the radioimmunoconjugate, 89Zr-DFO-Pertuzumab, in a human breast cancer mouse model. Chapters 3 and 4 explore the use of antibodies for imaging ovarian cancer, which is the deadliest gynecological malignancy in the United States. Chapter 3 details the development of an immunoPET imaging agent for imaging mucin-1 (MUC1), which is overexpressed in primary epithelial ovarian cancer and >90% of metastatic cancers. Chapter 4 details the development of a NIRF imaging agent that targets cancer antigen 125 (CA125), which is overexpressed in high grade-serous ovarian cancer.
In this work, we hope to demonstrate that combining antibodies with PET and NIRF imaging create powerful tools for diagnostic imaging and image-guided surgery of cancer. Of course, the ultimate goal of this work is to translate these imaging agents into the clinic and support
1 clinicians in the diagnosis, treatment, and management of cancer.
1.1. Molecular Imaging
Molecular imaging is a growing field that is valuable for diagnosing, monitoring, and managing the treatment of disease, especially in the fields of oncology, cardiology, and neurology.
According to the Molecular Imaging Center of Excellence (MICoE) and Society of Nuclear
Medicine (SNM), molecular imaging is defined as “the visualization, characterization, and measurement of biological processes at the molecular and cellular levels in humans and other living systems”.5 Molecular imaging comprise the following imaging modalities: single photon emission computed tomography (SPECT), positron emission tomography (PET), magnetic resonance spectroscopy, optical imaging (bioluminescent and fluorescence imaging), molecular ultrasound imaging, hybrid imaging (e.g. PET-CT, PET-MRI), and others.6, 7 Typically, an imaging agent is used to perform molecular imaging. The imaging agent consists of a targeting moiety (e.g. antibody, peptide, small molecule) that directs the agent to the molecule of interest and a signaling moiety (e.g. radionuclide, fluorophore) that permits visualization of the agent.7 Once the imaging agent has reached its target in the body, it can be imaged with the appropriate device and images can be generated for clinical use.
The benefits of molecular imaging include its noninvasive nature and ability to detect disease at earlier stages. To diagnose diseases, more invasive procedures such as biopsy or surgery are often necessary. Molecular imaging offers a noninvasive way to detect chemical and biological processes that indicate the occurrence and progression of disease. Other clinically employed imaging modalities such as x-rays, computed tomography (CT), and ultrasound primarily provide anatomical information. However, it would take more time before the disease can be detected with these modalities because the disease would have to progress to a stage where it is anatomically visible.7, 8 Lastly, molecular imaging modalities can be coupled with structural imaging modalities,
2 such as CT or MRI to obtain both anatomical and biochemical information.
For the purpose of this work, I will focus my discussion on the imaging modalities, PET and NIRF. The features of select imaging modalities are presented in Table 1.1. PET is a highly sensitive, whole-body imaging technique that is widely used in the clinic. It has an excellent depth of penetration and provides quantitative data. A drawback of PET is that it requires the use of ionizing radiation (i.e. radiation with enough energy to remove electrons from atoms or molecules), which can damage tissues with high levels of exposure. Other limitations are that it is relatively expensive and mostly requires the use of radionuclides that can only be generated via a cyclotron.
NIRF imaging falls under the category of optical fluorescence imaging. Optical fluorescence imaging is a relatively inexpensive imaging technique that does not require the use of ionizing radiation and allows for the imaging of multiple agents in one session. It is generally safe for use in clinical studies, but it is dependent on the fluorophore used and the mass required for imaging
(typically exceeds that required for PET or SPECT). The limitations of fluorescence imaging are the limited depth of penetration, surface weighting issues (i.e., objects closer to the surface appear brighter), and autofluorescence, which compromises the sensitivity of the technique. The use of a near-infrared imaging agent can help remedy some of these issues (will be discussed in the following section) as well as type of imaging instrument used and corrections applied.
Modality Spatial Depth of Sensitivity Cost Safety Profile Resolution Penetration Computed 50-200 µm Limitless Not determined $$ Ionizing radiation tomography (preclinical) (CT) 0.5-1 mm (clinical) Magnetic 25-100 µm Limitless 10-3 to 10-5 M $$$ No ionizing radiation resonance (preclinical) imaging (MRI) ~1 mm (clinical) Positron 1-2 mm Limitless 10-11 to 10-12 M $$$ Ionizing radiation emission (preclinical) tomography 5-7 mm (PET) (clinical)
3 Single photon 1-2 mm Limitless 10-10 to 10-11 M $$ Ionizing radiation emission (preclinical) tomography 8-10 mm (SPECT) (clinical) Ultrasound (US) 0.01-0.1 mm mm-cm Excellent when $ Good safety profile for microbubbles are superficial used (~10-12 M) (few mm depth) applications 1-2 mm for deeper (few cm depth) applications Optical 2-3 mm <1 cm ~10-9 to ~10-12 M $ Good safety profile fluorescence but depends on imaging fluorophore used and mass Optical 3-5 mm 1-2 cm ~10-15 to 10-17 M $ Good safety profile bioluminescence imaging
Table 1.1. Features of select molecular imaging modalities.7 1.1.1. PET Imaging
1.1.1.1. Principles of PET Imaging
Positron emission tomography (PET) is a highly sensitive, quantitative imaging technique with a multitude of clinical applications. PET requires the use of a positron emitter, which is a radionuclide that decays through the emission of a positron (β+). The nucleus of a positron emitter contains an excess of protons, causing instability. To regain stability, a proton in the nucleus is converted to a neutron and a positron along with a neutrino.7
During a PET scan, a positron-emitting imaging agent is administered to the subject, allowed to accumulate at the target site (e.g., tumor), and a positron is emitted from its nucleus
(Figure 1.1). As the positron passes through tissue, it quickly loses kinetic energy, collides with an electron, and they mutually annihilate. The annihilation produces two coincident 511 keV photons
(the rest-masses of the positron and electron are converted to their energy equivalent) that travel in
4 opposite directions. The two coincident photons exit the body and register on a ring of scintillation detectors surrounding the subject.9, 10 When both photons are detected simultaneously by two opposing detectors (typically within 4 to 12 ns), it is recorded as a coincidence event.11 Since the photons are approximately 180° apart, this places the coincidence event along the volume between the opposing detectors, referred to as the line of response. Throughout the course of a PET scan, many coincidence events are recorded, allowing for localization of the radionuclide to a location.
The data are then corrected for photon attenuation, scatter, accidental coincidence events, detector dead time, detector efficiency and ultimately reconstructed into 3D images. For preclinical studies, the same principles apply, but a small animal PET scanner is used (refer to Table 1.1 for properties).7, 9, 10
Figure 1.1. Principle of positron emission tomography (PET). Created with BioRender.com.
It is important to note that PET cannot be used to image multiple radionuclides in one scan.
A PET scanner only detects photons of a single energy because regardless of the positron emitter, only 511 keV photons are emitted during the annihilation event. On the other hand, single photon
5 emission computed tomography (SPECT) can distinguish between the radionuclides based on their emission energies.10 All factors considered, PET offers several advantages over SPECT such as higher sensitivity and spatial resolution (refer to Table 1.1 for values).12
1.1.1.2. Positron Emitters
Radionuclide Half-life (t1/2) Eß+ (keV) Decay mode (% Main production branching ratio) routes 11C 20.4 m 385.6 ß+ (100) 14N(p,ɑ)11C 13N 9.97 m 491.82 ß+ (100) 16O(p,ɑ)13N 15O 2.04 m 735.28 ß+ (100) 15N(p,n)15O 15N(d,n)15O 18F 109.77 m 249.8 ß+ (100) 18O(p,n)18F 20Ne(d,ɑ)18F 124I 4.18 d 687.04, ε + ß+ (100) 124Te(p,n)124I 974.74 ß+ (22.7) 64Cu 12.7 h 278.21 ε + ß+ (61.5) 64Ni(p,n)64Cu ß+ (17.6) ß- (38.5) 68Ga 67.71 m 836.02 ε + ß+ (100) 68Ge/68Ga-generator ß+ (89.14) 86Y 14.74 h 535 ε + ß+ (100) 86Sr(p,n)86Y ß+ (31.9) 89Zr 78.4 h 395.5 ε + ß+ (100) 89Y(p,n)89Zr ß+ (22.74)
Table 1.2. Physical decay characteristics of the most commonly used PET radionuclides.12, 13
When selecting a positron emitter for preclinical and clinical studies, factors such as the half-life, decay characteristics, production routes, and ease of synthesis should be considered. The physical decay characteristics of the most commonly used PET radionuclides are presented in Table
1.2. The non-metallic positron emitters, 11C, 13N, 15O, 18F, and 124I, are typically used with small molecule tracers with rapid pharmacokinetics. With the exception of 124I, they all have short-half lives that are only suitable for investigating biochemical processes over a short period of time, typically minutes or hours at most. Due to their half-lives, use of these radionuclides often requires the presence of a local medical cyclotron. In addition, they typically have to be incorporated into the tracer itself, which can result in complicated and long syntheses. Most of the radiometals, 64Cu,
6 86Y, and 89Zr, have longer half-lives, which are suitable for conjugation to bimolecular targeting moieties such as antibodies and peptides. By using bifunctional chelators that bind the radionuclide, they can easily be appended to these biomolecules.12 For our studies, we chose to use 89Zr because its physical half-life of ~3.3 days is a close match to the biological half-life of antibodies (Table
1.2). It also has a lower energy positron than most of the other commonly used radionuclides, which translates to a higher spatial resolution in the PET image.
1.1.2. Fluorescence Imaging
1.1.2.1. Principles of Fluorescence Imaging
Fluorescence imaging relies on the use of a fluorophore, a molecule capable of absorbing and re-emitting light. The general process of fluorescence imaging is illustrated in Figure 1.1. First, light is filtered to the appropriate wavelength for exciting the fluorophore. The fluorophore can be either a biological molecule located in the tissue (i.e., autofluorescence) or an administered fluorophore. The light must then pass through the tissue to reach the fluorophore. The fluorophore absorbs the photons and reaches an excited state. It remains in the excited state for a very short amount of time, usually a few nanoseconds, depending on the lifetime of the fluorophore. When it returns to its ground state, it emits photons. During the transition between the two states, energy is lost and a shift to longer wavelengths (lower energy) – referred to as the Stokes shift – occurs.
Lastly, only a small portion of the emitted light can be detected with a charged coupled device
(CCD) camera, which converts light into electrons and generates an image.14
Not all of the emitted light will reach the camera because when light travels through the tissue, the following processes occur: reflection at the tissue surface, absorption, scattering, and autofluorescence.15, 16 First, the difference between the refractive indices of air (n ~1.0) and tissue
(n ~ 1.4) causes some of the excitation light to be reflected at the tissue surface.16, 17 Second, components within the tissue – primarily water, lipids, oxygenated hemoglobin, and deoxygenated
7 hemoglobin – can absorb light. At wavelengths below 650 nm, oxyhemoglobin and deoxyhemoglobin absorb a great deal of light, which limits the depth of light penetration into the tissue (Figures 1.2A and B). Above 900 nm, water absorbs light (Figure 1.2A).15 Third, photons can change directions when traveling through tissue, an effect referred to as scattering. As the wavelength increases, scattering is reduced in all types of tissue. Lastly, in the ultraviolet and visible light region (200-700 nm), endogenous fluorophores in the tissue such as reduced nicotinamide adenine dinucleotide (NADH) and flavins can absorb light and result in autofluorescence.14, 16, 18 In fluorescence imaging, it is important to maximize the tumor-to- background ratio (referred to as the signal-to-noise ratio for general imaging purposes) because it allows for proper differentiation between tumor and non-target tissues.14 Autofluorescence results in nonspecific background signal, which decreases the signal-to-noise ratio and hampers one’s ability to clearly visualize the target. Near-infrared (NIR, 700 – 1000 nm) fluorescence imaging takes advantage of the window between 650-900 nm where there is limited interference from autofluorescence and deeper tissue penetration.14, 15
Figure 1.2. Geometric and basic principle of fluorescence imaging. (A) Light of the appropriate excitation wavelength is selected using a filter (Fex), which is located between a light source and the tissue. The excitation light travels through the tissue and is absorbed by a fluorophore, which subsequently emits light of a different wavelength. A small portion of this emitted light will exit
8 the tissue and can be detected with a camera. A filter (Fem) is placed in front of the camera, which allows only the emitted light to pass into the camera. (B) Absorbed light by the fluorophore instigates an electron (e) in the ground state toward an electronically excited state. Upon return to the ground state, the fluorophore emits a photon. The wavelength of this emitted photon is specific for the fluorophore. Fex, excitation filter; Fem, emission filter. Reprinted from Clinical Cancer Research, ©2013, 19, 3745-3754, Keereweer, S. et al., Optical Image-Guided Cancer Surgery: Challenges and Limitations, with permission from AACR.
Figure 1.3. Absorption of light. (A) Absorption of light by various components varies over the wavelength spectrum, resulting in an optimal window for fluorescence imaging in the NIR light region between 650 and 900 nm; (B) Penetration of light range from 0.5 mm to more than 10 mm contingent to the wavelength. OxyHb, oxygenated hemoglobin; DeoxyHb, deoxygenated hemoglobin. Reprinted from Clinical Cancer Research, ©2013, 19, 3745-3754, Keereweer, S. et al., Optical Image-Guided Cancer Surgery: Challenges and Limitations, with permission from AACR. 1.1.2.2. Fluorophores
Fluorophores are compounds, typically polyaromatic hydrocarbons or heterocycles, that are capable of absorbing and re-emitting light upon excitation. As mentioned before, fluorophores in the near-infrared (NIR) region exhibit the best properties for in vivo imaging because the effects of absorption, scattering, and autofluorescence are greatly reduced in this region. In addition, the following properties should be considered when selecting a suitable fluorophore for imaging: solubility, photophysical properties, photostability, non-specific binding, clearance, and toxicity.18-
9 20 Fluorophores should exhibit good solubility in aqueous solutions for administration in vivo. In terms of photophysical properties, fluorophores should possess high brightness, quantum yields, and extinction coefficients. The fluorescence quantum yield is defined as the number of photons emitted per excitation photon absorbed and is measure of how efficient the process of fluorescence is. Ideally, the quantum yield would be 100%, which means that every photon absorbed results in the emission of a photon. The extinction coefficient is a measure of the capacity of the fluorophore to absorb light at a particular wavelength. In terms of photostability, the fluorophore needs to retain its photophysical properties in vivo. A major threat to photostability is photobleaching, which is the photochemical destruction of the fluorophore (likely due to the generation of reactive oxygen species). After repeated exposure to the excitation light, the excited fluorophore can become unstable and permanently lose its ability to fluoresce. The fluorophore should have low binding to non-target tissues, which could result in background signal. Lastly, the fluorophore should clear quickly from the body and be nontoxic in vivo.18-20 All of these properties should be considered to maximize the tumor-to-background ratio during imaging.
A brief list of NIR fluorophores that are FDA-approved or used in clinical trials is presented in Table 1.3. In 1959, indocyanine green (ICG) became the first FDA-approved NIR fluorophore for in-human clinical use. ICG is negatively charged, and thus, binds to proteins in plasma, such as albumin, immediately after administration. The binding of ICG with plasma proteins causes retention within the vascular system, which makes it well-suited for cardiovascular and lymphatic angiography.14, 18, 20 ICG is rapidly cleared through the hepatobiliary system, which permits imaging of the liver and bile duct.14, 16, 20 ICG also has applications in oncology such as detection of the sentinel lymph nodes, liver metastases, hepatocellular carcinoma, and other cancers in image- guided surgery.16, 18, 20-26 Methylene blue (MB) is also FDA-approved, but the FDA issued a warning against its clinical use in 2011, owing to reports of serotonin syndrome.16 MB emits in the
10 far-red region (peak emission = 686 nm). It is positively charged, resulting in non-specific binding to proteins. It is cleared through both the renal and hepatobiliary system with a third in the urine.
In the presence of a reducing agent, the central nitrogen atom of MB loses its aromaticity and as result, its fluorescence; MB is thus converted to leucomethylene blue.16, 20, 21 MB has been used for cardiovascular and lymphatic angiography; bile duct, gastrointestinal tract, and ureter imaging; and image-guided surgery of parathyroid adenoma.16, 20, 21, 27-30 Though they are clinically relevant, ICG and MB can only target tissues passively either through non-specific binding or the enhanced permeability and retention (EPR) effect (for tumor targeting); neither can be conjugated to a targeting moiety.
Two fluorophores, IRDye800CW and ZW800-1, are being tested in clinical trials.
IRDye800CW is a negatively charged dye that exhibits good solubility in water due to the possession four sulfonate groups. Due to its low molecular weight and stability in serum, it is quickly cleared through the renal system. Renal clearance is preferred over hepatobiliary clearance because it does not produce high background signal in the GI tract.31 IRDye800CW is commercially available and can be functionalized with an NHS ester or maleimide for conjugation to a targeting vector. IR800DyeCW has been conjugated to antibodies and peptides for imaging and image-guided surgery for a variety of cancers such as breast, head and neck, and rectal cancer.
A recent article by van Keulen et al. reports the use of panitumumab-IRDye800CW for image- guided surgery in 14 head and neck patients enrolled in clinical trial, NCT02415881.32
Panitumumab targets the epidermal growth receptor factor, a glycoprotein expressed in various types of cancer. NIRF imaging with panitumumab-IRDye800CW identified a close tumor margin in one patient, secondary primary tumor outside of the planned surgical area in a second patient, and residual disease in a third patient. This study thereby demonstrates the construct’s utility for intraoperative decision making.32 In addition to panitumumab, IR800CW has also been conjugated
11 to the vascular endothelial growth factor (VEGF)-targeting antibody, bevacizumab, and EGRF- targeting antibody, cetuximab for use in clinical trials evaluating the dosage, safety, and efficacy of the constructs (NCT01508572, NCT01972373, NCT02129933, NCT02583568, NCT03134846,
NCT03384238, NCT02415881).16 ZW800-1 is a zwitterionic dye that also exhibits good solubility in water and renal clearance. Due to its neutral net charge, it has very low non-specific tissue binding, resulting in higher tumor-to-background ratio when compared to other charged NIR fluorophores.33-35 In various studies, ZW800-1 was conjugated to the cyclic peptide, cRGD, which targets integrin ɑvß3. Integrins are a family of cell surface receptors that promote the attachment of cells to the extracellular matrix (ECM). Studies report that the expression levels of integrin ɑvß3 increase with tumor cell invasion and metastasis as well as the formation of tumor vessels. It is also overexpressed in nearly all solid tumors. cRGD and RGD has been studied extensively in the clinic and used as targeting moieties for PET and SPECT imaging studies, so there has been interest in developing an improved cRGD-based NIR agent for imaging and image-guided surgery of cancer.16, 20 In 2013, Choi et al. demonstrate that cRGD-ZW800-1 can be used to achieve higher tumor-to-background ratios (17.2 ± 1.2) in mice bearing bilateral human melanoma cells (tumors from ɑvß3-positive M21 cell line and ɑvß3-negative M21-L cell line) than conjugates with the NIR fluorophores, IRDye800-CW (5.1 ± 1.2) and Cy5.5 (2.7 ± 1.4).34 Currently, there are two clinical trials in the Netherlands using cRGD-IR800-1, one for squamous cell carcinoma of the oral cavity and the other for colorectal carcinoma (Netherlands Trial Register number NL7724 and NL8049).
Many other NIR fluorophores are being tested in preclinical studies.16 In addition, NIR fluorophores emitting in the 1,000-1,700 nm range, referred to as NIR-II, have gained the interest of researchers and are being developed and investigated for in vivo imaging.16
12 Fluorophore Structure Excitation/ Properties emission wavelength (nm) Indocyanine 807/822 Negatively green (ICG) charged; lipophilic; QY=9.3% (serum)
Methylene 665/686 Positively Blue (MB) charged; QY=9.6% (serum)
IRDye800CW 774/789 Negatively charged; hydrophilic; QY=12% (serum)
Zwitterionic 772/788 Zwitterionic cyanine dyes (ZW800-1) (neutral net charge); hydrophilic; QY=15.1% (serum)
13 Table 1.3. Structure, excitation/emission wavelength, and properties of clinically employed fluorophores.16
1.2. Structure of Antibodies
Antibodies, also known as immunoglobulins, are glycoproteins that are produced by the immune system. Antibodies have a Fab (Fragment antigen binding) region for binding its specific antigen and a Fc (Fragment crystallizable) region for binding receptors on immune effector cells
(e.g. B cells, activated T cells). Each antibody contains a pair of identical light chains (L) and a pair of identical heavy chains (H) that give the antibody its distinctive Y shape. The light chains are ~25 kDa each and the heavy chains are ~55 kDa each.36 These chains are held together by interchain disulfide bonds and non-covalent interactions (i.e., hydrogen bonds, van der Waals interactions) that vary based on the immunoglobulin isotype. The immunoglobulin G (IgG) isotype consists of a total of 16 disulfide bonds, 4 of which are interchain disulfide bonds and 12 of which are intrachain disulfide bonds.36, 37 For the purpose of this work, use of the term, “antibodies” will refer to the IgG isotype (Figure 1.4). The light and heavy chains are composed of regions referred to as the variable (V) and constant (C) domains. The light chain has one variable (VL) and one constant domain (CL). The heavy chain has one variable domain (VH) and three constant domains
(CH1, CH2, CH3). The variable domains have three hypervariable complementarity determining regions (CDR1, CDR2, CDR3) that are responsible for the antibody’s ability to recognize specific antigens. The 2 Fab regions of the antibody are made up of the VL, CL, VH, and CH1 domains, and the Fc region of the antibody is made up of the CH2 and CH3 domains. The Fab and Fc regions are linked by a flexible hinge region that allows the two Fab fragments to move independently.36
Lastly, it is important to note that the Fc region contains two conserved N-linked glycosylation sites at the Asn297 residues, which has important implications for protein folding and interactions with immune effector cells.36, 38
14
Figure 1.4. Detailed structural schematic of a full-length IgG. Reprinted from ref. 36 under the terms of the Creative Commons Attribution International License (http://creativecommons.org/licenses/by/4.0/).
1.3. Bioconjugation Strategies
1.3.1. Traditional Conjugation Method
The traditional conjugation method usually involves conjugating an amine-reactive chelator or fluorophore to the lysines of an antibody (reaction shown in Figure 1.5). However, antibodies possess a vast number of lysine residues, resulting in poorly defined and heterogenous constructs.
The benefits of this approach are that it is easily accessible and facile. On the other hand, it can result in a possible loss of immunoreactivity if the conjugation occurs in the Fab region, and there is heterogeneity at three levels.37, 39 First, the conjugation reaction will produce immunoconjugates with different degrees of labeling (DOL). Second, regioisomers are likely to exist. Lastly, every conjugation is different, so it is unlikely to produce batches with the exact same DOLs and mixture of regioisomers.37
15
Figure 1.5. Two most common types of bioconjugation reactions: Top, reaction of a primary amine with a N-hydroxysuccinimide ester (NHS ester); Bottom, reaction of a primary amine and isothiocyanate (NCS). 1.3.2. Site-specific Conjugation Method
Site-specific conjugation methods have been shown to produce more homogeneous, better- defined, and more reproducible constructs than those synthesized with the traditional conjugation method.40-45 Moreover, several studies have demonstrated that site-specifically modified immunoconjugates exhibit improved in vivo behavior compared with randomly modified constructs.46-51 Four main strategies for site-specifically modifying antibodies have been described in the literature.37, 52 The first strategy incorporates peptide tags into antibodies. Peptide tags either serve as recognition sites for enzyme-mediated conjugation or as the coordination framework for radiometal chelation. The second strategy incorporates non-canonical amino acids into the antibody via genetic engineering. The third approach utilizes cysteine residues to conjugate bifunctional chelators to the antibody. While more accessible than the two aforementioned approaches, there are still several cysteines located on the antibody, though fewer than the number of lysines. As a result, antibodies modified in this manner may still have different degrees of labeling. The last strategy relies on modifying the heavy chain glycans in the Fc region of the antibody (see structure in Figure 1.6). A major advantage of modifying the glycans is that the distance between the Fc and
Fab regions of the antibody reduces the risk that conjugations will inadvertently occur in the
16 antigen-binding regions. Additionally, the heavy chain glycans possess a biantennary oligosaccharide chain, which allows for up to 4 conjugations per antibody. Lastly, this strategy provides us with an accessible and modular way to produce site-specific constructs without utilizing genetic engineering because the chemistry of sugars varies from that of amino acids.37, 40
Figure 1.6. (A) The biantennary structure of the heavy chain glycans; the dotted outlines indicate residues that are not always present in the glycans. (B) Structures of natural and synthetic monosaccharides. Reprinted from ref. 36 under the terms of the Creative Commons Attribution International License (http://creativecommons.org/licenses/by/4.0/). For the purpose of this work, we will focus on the glycan-based conjugation method. There are two main methods for site-specific conjugation on the glycans: oxidation-based and glycoengineering. In the oxidation-based method, the heavy chain glycans are oxidized with periodate (IO4-) to produce aldehydes, allowing for selective reactions with nucleophiles. While facile, this method may require reduction to obtain stable immunoconjugates and possibly result in
17 the oxidation of methionine residues. In the glycoengineering method, natural and engineered enzymes are used to incorporate modified sugars into the antibody in a chemoselective fashion.
The galactosyltransferase enzyme, Gal-T1(Y289L) is frequently used in this method because it enables the transfer of azide-functionalized galactose residues. Bioorthogonal click chemistry, particularly strain-promoted azide-alkyne cycloaddition (SPAAC), can then be used to incorporate a variety of dibenzocyclooctyne (DBCO)-modified chelators and/or drugs into the antibody construct (Figure 1.7).37, 46, 47
Figure 1.7. Schematic of a Gal-T(Y289L)-based site-specific modification procedure. Reprinted from ref. 36 under the terms of the Creative Commons Attribution International License (http://creativecommons.org/licenses/by/4.0/).
18 1.5. References
1. Wu, A. M.; Senter, P. D., Arming antibodies: prospects and challenges for immunoconjugates. Nat Biotechnol 2005, 23 (9), 1137-1146. 2. Wu, A. M.; Olafsen, T., Antibodies for molecular imaging of cancer. Cancer J. 2008, 14 (3), 191-197. 3. Wu, A. M., Antibodies and antimatter: The resurgence of immuno-PET. J. Nucl. Med. 2009, 50 (1), 2-5. 4. Knowles, S. M.; Wu, A. M., Advances in immuno-positron emission tomography: Antibodies for molecular imaging in oncology. J. Clin. Oncol. 2012, 30 (31), 3884-3892. 5. Mankoff, D. A., A definition of molecular imaging. J. Nucl. Med. 2007, 48 (6), 18n, 21n. 6. Willmann, J. K.; van Bruggen, N.; Dinkelborg, L. M.; Gambhir, S. S., Molecular imaging in drug development. Nat. Rev. Drug Discov. 2008, 7 (7), 591-607. 7. James, M. L.; Gambhir, S. S., A molecular imaging primer: Modalities, imaging agents, and applications. Physiol. Rev. 2012, 92 (2), 897-965. 8. Willmann, J. K.; van Bruggen, N.; Dinkelborg, L. M.; Gambhir, S. S., Molecular imaging in drug development. Nature Reviews Drug Discovery 2008, 7 (7), 591-607. 9. Cherry, S. R., Fundamentals of positron emission tomography and applications in preclinical drug development. J. Clin. Pharmacol. 2001, 41 (5), 482-491. 10. Cherry, S. R.; Dahlbom, M., PET: Physics, Instrumentation, and Scanners. In PET: Molecular Imaging and Its Biological Applications, Phelps, M. E., Ed. Springer New York: New York, NY, 2004; pp 1-124. 11. Livieratos, L., Basic principles of SPECT and PET imaging. In Radionuclide and hybrid bone imaging, Fogelman, I.; Gnanasegaran, G.; van der Wall, H., Eds. Springer Berlin Heidelberg: Berlin, Heidelberg, 2012; pp 345-359. 12. Zeglis, B. M.; Lewis, J. S., A practical guide to the construction of radiometallated bioconjugates for positron emission tomography. Dalton Trans. 2011, 40 (23), 6168-6195. 13. Welch, M. J.; Redvanly, C. S., Handbook of radiopharmaceuticals: Radiochemistry and applications. John Wiley & Sons: 2003. 14. Rosenthal, E. L.; Warram, J. M.; Bland, K. I.; Zinn, K. R., The status of contemporary image-guided modalities in oncologic surgery. Ann. Surg. 2015, 261 (1), 46-55. 15. Keereweer, S.; Van Driel, P. B. A. A.; Snoeks, T. J. A.; Kerrebijn, J. D. F.; Baatenburg de Jong, R. J.; Vahrmeijer, A. L.; Sterenborg, H. J. C. M.; Löwik, C. W. G. M., Optical image- guided cancer surgery: Challenges and limitations. Clin. Cancer Res. 2013, 19 (14), 3745. 16. Hong, G.; Antaris, A. L.; Dai, H., Near-infrared fluorophores for biomedical imaging. Nat. Biomed. Eng. 2017, 1 (1), 0010. 17. Bashkatov, A. N.; Genina, E. A.; Tuchin, V. V., Optical properties of skin, subcutaneous, and muscle tissues: A review. J. Innov. Opt. Health Sci. 2011, 04 (01), 9-38. 18. Frangioni, J. V., In vivo near-infrared fluorescence imaging. Curr. Opin. Chem. Biol. 2003, 7 (5), 626-634. 19. te Velde, E. A.; Veerman, T.; Subramaniam, V.; Ruers, T., The use of fluorescent dyes and probes in surgical oncology. Eur. J. Surg. Oncol. 2010, 36 (1), 6-15. 20. 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., Beyond the margins: Real-time detection of cancer using targeted fluorophores. Nat. Rev. Clin. Oncol. 2017, 14 (6), 347-364. 21. Tanaka, E.; Chen, F. Y.; Flaumenhaft, R.; Graham, G. J.; Laurence, R. G.; Frangioni, J. V., Real-time assessment of cardiac perfusion, coronary angiography, and acute intravascular
19 thrombi using dual-channel near-infrared fluorescence imaging. J. Thorac. Cardiovasc. Surg. 2009, 138 (1), 133-140. 22. Troyan, S. L.; Kianzad, V.; Gibbs-Strauss, S. L.; Gioux, S.; Matsui, A.; Oketokoun, R.; Ngo, L.; Khamene, A.; Azar, F.; Frangioni, J. V., The FLARE™ intraoperative near-infrared fluorescence imaging system: A first-in-human clinical trial in breast cancer sentinel lymph node mapping. Ann. Surg. Oncol. 2009, 16 (10), 2943-2952. 23. van der Vorst, J. R.; Schaafsma, B. E.; Hutteman, M.; Verbeek, F. P. R.; Liefers, G.-J.; Hartgrink, H. H.; Smit, V. T. H. B. M.; Löwik, C. W. G. M.; van de Velde, C. J. H.; Frangioni, J. V.; Vahrmeijer, A. L., Near-infrared fluorescence-guided resection of colorectal liver metastases. Cancer 2013, 119 (18), 3411-3418. 24. Tummers, Q. R. J. G.; Verbeek, F. P. R.; Prevoo, H. A. J. M.; Braat, A. E.; Baeten, C. I. M.; Frangioni, J. V.; van de Velde, C. J. H.; Vahrmeijer, A. L., First experience on laparoscopic near-infrared fluorescence imaging of hepatic uveal melanoma metastases using indocyanine green. Surg. Innov. 2014, 22 (1), 20-25. 25. Verbeek, F. P. R.; Tummers, Q. R. J. G.; Rietbergen, D. D. D.; Peters, A. A. W.; Schaafsma, B. E.; van de Velde, C. J. H.; Frangioni, J. V.; van Leeuwen, F. W. B.; Gaarenstroom, K. N.; Vahrmeijer, A. L., Sentinel lymph node biopsy in vulvar cancer using combined radioactive and fluorescence guidance. Int. J. Gynecol. Cancer 2015, 25 (6), 1086. 26. Mieog, J. S. D.; Troyan, S. L.; Hutteman, M.; Donohoe, K. J.; van der Vorst, J. R.; Stockdale, A.; Liefers, G.-J.; Choi, H. S.; Gibbs-Strauss, S. L.; Putter, H.; Gioux, S.; Kuppen, P. J. K.; Ashitate, Y.; Löwik, C. W. G. M.; Smit, V. T. H. B. M.; Oketokoun, R.; Ngo, L. H.; van de Velde, C. J. H.; Frangioni, J. V.; Vahrmeijer, A. L., Toward optimization of imaging system and lymphatic tracer for near-infrared fluorescent sentinel lymph node mapping in breast cancer. Ann. Surg. Oncol. 2011, 18 (9), 2483-2491. 27. Matsui, A.; Tanaka, E.; Choi, H. S.; Winer, J. H.; Kianzad, V.; Gioux, S.; Laurence, R. G.; Frangioni, J. V., Real-time intra-operative near-infrared fluorescence identification of the extrahepatic bile ducts using clinically available contrast agents. Surgery 2010, 148 (1), 87-95. 28. Verbeek, F. P. R.; van der Vorst, J. R.; Schaafsma, B. E.; Swijnenburg, R.-J.; Gaarenstroom, K. N.; Elzevier, H. W.; van de Velde, C. J. H.; Frangioni, J. V.; Vahrmeijer, A. L., Intraoperative near infrared fluorescence guided identification of the ureters using low dose methylene blue: A first in human experience. J. Urol. 2013, 190 (2), 574-579. 29. Ashitate, Y.; Vooght, C. S.; Hutteman, M.; Oketokoun, R.; Choi, H. S.; Frangioni, J. V., Simultaneous assessment of luminal integrity and vascular perfusion of the gastrointestinal tract using dual-channel near-infrared fluorescence. Mol. Imaging 2012, 11 (4), 7290.2011.00048. 30. Tummers, Q. R. J. G.; Schepers, A.; Hamming, J. F.; Kievit, J.; Frangioni, J. V.; van de Velde, C. J. H.; Vahrmeijer, A. L., Intraoperative guidance in parathyroid surgery using near- infrared fluorescence imaging and low-dose methylene blue. Surgery 2015, 158 (5), 1323-1330. 31. Olson, M. T.; Ly, Q. P.; Mohs, A. M., Fluorescence guidance in surgical oncology: Challenges, opportunities, and translation. Mol. Imaging Biol. 2019, 21 (2), 200-218. 32. van Keulen, S.; Nishio, N.; Fakurnejad, S.; Birkeland, A.; Martin, B. A.; Lu, G.; Zhou, Q.; Chirita, S. U.; Forouzanfar, T.; Colevas, A. D.; van den Berg, N. S.; Rosenthal, E. L., The clinical application of fluorescence-guided surgery in head and neck cancer. J. Nucl. Med. 2019, 60 (6), 758-763. 33. Choi, H. S.; Nasr, K.; Alyabyev, S.; Feith, D.; Lee, J. H.; Kim, S. H.; Ashitate, Y.; Hyun, H.; Patonay, G.; Strekowski, L.; Henary, M.; Frangioni, J. V., Synthesis and in vivo fate of zwitterionic near-infrared fluorophores. Angew. Chem. Int. Ed. Engl 2011, 50 (28), 6258-6263. 34. Choi, H. S.; Gibbs, S. L.; Lee, J. H.; Kim, S. H.; Ashitate, Y.; Liu, F.; Hyun, H.; Park,
20 G.; Xie, Y.; Bae, S.; Henary, M.; Frangioni, J. V., Targeted zwitterionic near-infrared fluorophores for improved optical imaging. Nat. Biotechnol. 2013, 31 (2), 148-153. 35. Hyun, H.; Henary, M.; Gao, T.; Narayana, L.; Owens, E. A.; Lee, J. H.; Park, G.; Wada, H.; Ashitate, Y.; Frangioni, J. V.; Choi, H. S., 700-nm zwitterionic near-infrared fluorophores for dual-channel image-guided surgery. Mol. Imaging Biol. 2016, 18 (1), 52-61. 36. Schroeder, H. W., Jr.; Cavacini, L., Structure and function of immunoglobulins. J. Allergy Clin. Immunol. 2010, 125 (2), S41-S52. 37. Adumeau, P.; Sharma, S. K.; Brent, C.; Zeglis, B. M., Site-specifically labeled immunoconjugates for molecular imaging—Part 1: Cysteine residues and glycans. Mol. Imaging Biol. 2016, 18 (1), 1-17. 38. Hayes, J. M.; Cosgrave, E. F. J.; Struwe, W. B.; Wormald, M.; Davey, G. P.; Jefferis, R.; Rudd, P. M., Glycosylation and Fc receptors. In Fc receptors, Daeron, M.; Nimmerjahn, F., Eds. Springer International Publishing: Cham, 2014; pp 165-199. 39. Dobrenkov, K.; Cheung, N.-K. V., GD2-targeted immunotherapy and radioimmunotherapy. Semin. Oncol. 2014, 41 (5), 589-612. 40. Agarwal, P.; Bertozzi, C. R., Site-specific antibody–drug conjugates: The nexus of bioorthogonal chemistry, protein engineering, and drug development. Bioconjug. Chem. 2015, 26 (2), 176-192. 41. Zhou, Q., Site-specific antibody conjugation for ADC and beyond. Biomedicines 2017, 5 (4). 42. Boswell, C. A.; Mundo, E. E.; Zhang, C.; Bumbaca, D.; Valle, N. R.; Kozak, K. R.; Fourie, A.; Chuh, J.; Koppada, N.; Saad, O.; Gill, H.; Shen, B. Q.; Rubinfeld, B.; Tibbitts, J.; Kaur, S.; Theil, F. P.; Fielder, P. J.; Khawli, L. A.; Lin, K., Impact of drug conjugation on pharmacokinetics and tissue distribution of anti-STEAP1 antibody-drug conjugates in rats. Bioconjug. Chem. 2011, 22 (10), 1994-2004. 43. Junutula, J. R.; Raab, H.; Clark, S.; Bhakta, S.; Leipold, D. D.; Weir, S.; Chen, Y.; Simpson, M.; Tsai, S. P.; Dennis, M. S.; Lu, Y. M.; Meng, Y. G.; Ng, C.; Yang, J. H.; Lee, C. C.; Duenas, E.; Gorrell, J.; Katta, V.; Kim, A.; McDorman, K.; Flagella, K.; Venook, R.; Ross, S.; Spencer, S. D.; Wong, W. L.; Lowman, H. B.; Vandlen, R.; Sliwkowski, M. X.; Scheller, R. H.; Polakis, P.; Mallet, W., Site-specific conjugation of a cytotoxic drug to an antibody improves the therapeutic index. Nat. Biotechnol. 2008, 26 (8), 925-932. 44. Pillow, T. H.; Tien, J.; Parsons-Reponte, K. L.; Bhakta, S.; Li, H.; Staben, L. R.; Li, G. M.; Chuh, J.; Fourie-O'Donohue, A.; Darwish, M.; Yip, V.; Liu, L. N.; Leipold, D. D.; Su, D.; Wu, E.; Spencer, S. D.; Shen, B. Q.; Xu, K. Y.; Kozak, K. R.; Raab, H.; Vandlen, R.; Phillips, G. D. L.; Scheller, R. H.; Polakis, P.; Sliwkowski, M. X.; Flygare, J. A.; Junutula, J. R., Site-specific trastuzumab maytansinoid antibody-drug conjugates with improved therapeutic activity through linker and antibody engineering. J. Med. Chem. 2014, 57 (19), 7890-7899. 45. van Geel, R.; Wijdeven, M. A.; Heesbeen, R.; Verkade, J. M.; Wasiel, A. A.; van Berkel, S. S.; van Delft, F. L., Chemoenzymatic conjugation of toxic payloads to the globally conserved N-glycan of native mAbs provides homogeneous and highly efficacious antibody-drug conjugates. Bioconjug. Chem. 2015, 26 (11), 2233-42. 46. Zeglis, B. M.; Davis, C. B.; Abdel-Atti, D.; Carlin, S. D.; Chen, A.; Aggeler, R.; Agnew, B. J.; Lewis, J. S., Chemoenzymatic strategy for the synthesis of site-specifically labeled immunoconjugates for multimodal PET and optical imaging. Bioconjug. Chem. 2014, 25 (12), 2123-2128. 47. Zeglis, B. M.; Davis, C. B.; Aggeler, R.; Kang, H. C.; Chen, A.; Agnew, B. J.; Lewis, J. S., Enzyme-mediated methodology for the site-specific radiolabeling of antibodies based on
21 catalyst-free click chemistry. Bioconjug. Chem. 2013, 24 (6), 1057-1067. 48. Kristensen, L. K.; Christensen, C.; Jensen, M. M.; Agnew, B. J.; Schjoth-Frydendahl, C.; Kjaer, A.; Nielsen, C. H., Site-specifically labeled 89Zr-DFO-trastuzumab improves immunoreactivity and tumor uptake for immunoPET in a subcutaneous HER2-positive xenograft mouse model. Theranostics 2019, 9 (15), 4409-4420. 49. Alvarez, V. L.; Wen, M. L.; Lee, C.; Lopes, A. D.; Rodwell, J. D.; McKearn, T. J., Site-specifically modified 111In labelled antibodies give low liver backgrounds and improved radioimmunoscintigraphy. Nucl. Med. Biol. 1986, 13 (4), 347-352. 50. Tavare, R.; Wu, W. H.; Zettlitz, K. A.; Salazar, F. B.; McCabe, K. E.; Marks, J. D.; Wu, A. M., Enhanced immunoPET of ALCAM-positive colorectal carcinoma using site-specific 64Cu-DOTA conjugation. Protein Eng. Des. Sel. 2014, 27 (10), 317-24. 51. Tinianow, J. N.; Gill, H. S.; Ogasawara, A.; Flores, J. E.; Vanderbilt, A. N.; Luis, E.; Vandlen, R.; Darwish, M.; Junutula, J. R.; Williams, S. P.; Marik, J., Site-specifically 89Zr- labeled monoclonal antibodies for ImmunoPET. Nucl. Med. Biol. 2010, 37 (3), 289-97. 52. Adumeau, P.; Sharma, S. K.; Brent, C.; Zeglis, B. M., Site-specifically labeled immunoconjugates for molecular imaging—Part 2: Peptide tags and unnatural amino acids. Mol. Imaging Biol. 2016, 18 (2), 153-165.
22 Chapter 2 : The Influence of Glycans-Specific Bioconjugation on the FcɣRI Binding and In Vivo Performance of 89Zr-DFO-Pertuzumab
This chapter is an adaptation of accepted work from Vivier, D.; Fung, K.; Rodriguez, C.; Adumeau,
P.; Ulaner, G. A.; Lewis, J. S.; Sharma, S. K.; Zeglis, B. M. The influence of glycans-specific bioconjugation on the FcɣRI binding and in vivo performance of 89Zr-DFO-Pertuzumab.
Theranostics. 2020, 10 (4), 1746-1757.
2.1. Introduction
In recent years, 89Zr-labeled antibodies have emerged as promising agents for the positron emission tomography (PET) imaging of cancer.1 Indeed, 89Zr is nearly ideally suited for immunoPET because its physical half-life (t1/2 = 3.3 days) aligns perfectly with the biological half- life of full-length antibodies.2-4 Recently, two radioimmunoconjugates targeting different epitopes of human epidermal growth factor receptor-2 (HER2) ― 89Zr-DFO-trastuzumab and 89Zr-DFO- pertuzumab ― have shown significant clinical potential in patients with HER2-positive breast cancer.5, 6 Yet despite the promise of 89Zr-labeled antibodies, the current synthetic methodologies used to create them are suboptimal at best. The majority of radioimmunoconjugates are produced via random conjugation methods in which bifunctional chelators ― in the case of 89Zr, desferrioxamine (DFO) ― are attached to the lysines of the antibody.7 Yet because antibodies have several lysines distributed throughout their structure, this approach produces poorly defined and heterogeneous immunoconjugates that can suffer from suboptimal immunoreactivity.
A variety of different site-specific bioconjugation methods have been developed to circumvent these problems, including strategies based on thiol-reactive probes, peptide tags, and non-canonical amino acids.7-9 The benefits of site-specific bioconjugation are clear. To wit, irrespective of the modification method, site-specifically modified immunoconjugates have been shown to be more homogeneous, better-defined, more reproducibly synthesized, and more effective
23 in vivo than their randomly modified cousins.7, 10-14 Moreover, several reports have shown that site-specifically modified radioimmunoconjugates in particular exhibit improved in vivo behavior compared to randomly labeled analogues.15-19
Over the last half-decade, our laboratory and others have worked to develop chemoenzymatic approaches to site-specific bioconjugation capable of selectively appending cargoes — including fluorophores, chelators, and toxins — to the heavy chain glycans on the CH2 domain of an antibody’s Fc region.10, 14, 15, 20-23 These biantennary sugar chains are particularly attractive sites for modification as they offer a pair of biochemically unique handles for manipulation and lie far from the antigen-binding domains of the immunoglobulin. Our strategy approach is predicated on three steps. First, one of two enzymes is used to truncate the glycans: either -galactosidase (Gal, which removes the outermost monosaccharide) or endoglycosidase
(EndoS, which hydrolyzes the chitobiose core of the glycans, leaving only the innermost residue).
Second, a mutant, promiscuous galactosyltransferase [GalT-(Y289L)] is used to install azide- modified galactose residues (GalNAz) into the sugar chains. And finally, dibenzocyclooctyne- bearing cargoes are appended to the azide-bearing sugars via the strain-promoted azide-alkyne click
(SPAAC) reaction. We have demonstrated that this strategy produces well-defined, more homogeneous immunoconjugates with excellent in vivo behavior using several model systems and a range of different payloads, and we are currently bringing this technology to the clinic.8, 15, 20
In the work at hand, we set out to explore the influence of site-specific bioconjugation on the in vivo performance of radioimmunoconjugates. This investigation is fueled in large part by recent results from our laboratory that suggest that the truncation of the heavy chain glycans of radioimmunoconjugates can reduce their retention in healthy non-target organs and boost their accretion in tumor tissue.24, 25 We hypothesize that the root of this phenomenon lies in the conformational change that occurs upon the deglycosylation of the immunoglobulin. This change
24 attenuates the binding of the immunoconjugate to FcRI, an Fc receptor that is expressed on the surface of monocytes, macrophages, and tissue-resident macrophages in the liver.26, 27 It is important to note that we have focused on FcRI in this investigation because it is the only member of the FcR family that is able to bind monomeric IgGs; the others ⎯ FcRII and FcRIII ⎯ prefer to bind to immune-complexes.28 Due to its clinical relevance, we selected pertuzumab — a monoclonal antibody that targets the HER2 antigen over-expressed in 20-30% of breast cancers — for this proof-of-concept study.29, 30 In fact, the first-in-human clinical trial of 89Zr-DFO- pertuzumab in patients with HER2-positive breast cancer was conducted in 2017.5 Here, we synthesized three desferrioxamine-labeled pertuzumab immunoconjugates: one using traditional, random bioconjugation methods (DFO-nsspertuzumab) and two using our chemoenzymatic protocol (DFO-sspertuzumab-EndoS and DFO-sspertuzumab-Gal). Subsequently, we structurally characterized the trio of constructs, interrogated their ability to bind recombinant HER2 as well as human and mouse FcRI, explored their in vitro behavior with HER2-positive BT474 human breast cancer cells, and evaluated their in vivo performance in two different mouse models of HER2- expressing human breast cancer, including one using humanized NSG (huNSG) mice. Taken together, the data clearly illustrate that site-specific bioconjugation not only produces more homogeneous and well-defined radioimmunoconjugates than traditional methods but may also improve their in vivo performance in certain mouse models and — potentially — patient populations. While our laboratory and others have indeed previously noted the improved in vivo performance of radioimmunoconjugates that have been modified on the heavy chain glycans, this investigation represents the first exploration of why this may be the case.15, 16, 20
25 2.2. Experimental Section
2.2.1. Reagents and General Procedures
All chemicals, unless otherwise noted, were acquired from Sigma-Aldrich or Fisher
Scientific and used as received without further purification. All water used was ultra-pure (>18.2
MΩcm-1), and dimethylsulfoxide was of molecular biology grade (>99.9%). DFO-nsspertuzumab,
DFO-sspertuzumab-βGal, and DFO-sspertuzumab-EndoS were prepared according to published protocols and characterized via SDS-PAGE, MALDI-ToF mass spectrometry, and ELISA according to previously reported methods.15, 20 [89Zr]Zr(oxalate) was produced and purified via the
89Y(p,n)89Zr reaction at Memorial Sloan Kettering Cancer Center as previously described. Activity measurements were made using a CRC-15R Dose Calibrator (Capintec, Inc.), and experimental samples were counted on an Automatic Wizard2 γ–counter (PerkinElmer, Inc.). The radiolabeling of the pertuzumab immunoconjugates with [89Zr]Zr4+ was performed according to published procedures and monitored using silica-impregnated instant thin-layer chromatography (iTLC) paper (Pall Corp.) on an AR-2000 radio-TLC plate reader (Bioscan, Inc.). Mice were implanted with HER2-positive BT474 human breast cancer xenografts as previously reported, and PET imaging and acute biodistribution experiments were performed according to published protocols approved by the Institutional Animal Care and Use Committees of Hunter College, Weill Cornell
Medical College, and Memorial Sloan Kettering Cancer Center.15, 20
2.2.2. SDS-PAGE Analysis
The purified constructs ⎯ pertuzumab, DFO-nsspertuzumab, DFO-sspertuzumab-βGal, and
DFO-sspertuzumab-EndoS ⎯ were characterized via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Briefly, 2 μg antibody (0.85 μL of a 5.89 mg/mL stock) were combined with 31.65 μL H2O, 5 μL 500 mM dithiothreitol (NuPAGE® 10X Sample Reducing
26 Agent, Life Technologies), and 12.5 μL 4X electrophoresis buffer (NuPAGE® LDS Sample
Buffer, Thermo Fisher, Eugene, OR). This mixture was then denatured by heating to 95°C for 10 min using a heat block. Subsequently, 20 μL of each sample was then loaded alongside an appropriate molecular weight marker (Mark12TM stained standard, Life Technologies) onto a 1 mm, 10 well 4-12% Bis-Tris protein gel (Life Technologies) and run for ~4 h at 10 V/cm in MOPS buffer. The completed gel was washed 3 times with H2O, stained using SimplyBlueTM SafeStain
(Life Technologies) for 1 h, and destained overnight in H2O. The gel was then analyzed using an
Odyssey Infrared Gel Scanner (Li-Cor Biosciences, Lincoln, NE).
2.2.3. Size Exclusion Chromatography
Size exclusion chromatography was performed on a Shimadzu UFLC HPLC system. 100
µg (0.67 nmol) of each construct was run on an SEC column (SuperdexTM 200 Increase 10/300 GL,
GE Healthcare) in phosphate buffered saline for 40 min (flow rate: 0.75 mL/min) and the absorbance was measured at 280 nm.
2.2.4. Flow Cytometry
Flow cytometry experiments were performed with HER2-positive BT474 cells. Native pertuzumab and the immunoconjugates were incubated at 6 μg/mL in suspension with 106 cells/mL for 30 min on ice. Cells were washed by pelleting and resuspension and then incubated with a goat anti-human IgG-AlexaFluor568 secondary antibody (Thermo Fisher Scientific) at 4
μg/mL. Subsequently, the cells were again washed by pelleting and resuspension three times and then analyzed on a BD LSR II (BD Biosciences, San Jose, CA). Binding data was collected in triplicate, averaged, and plotted.
2.2.5. FcRI Binding ELISA
Recombinant human Fc gamma RI/CD64, CF (R&D Systems # 1257-FC-050) was diluted to
27 10 μg/mL in sterile PBS and 100 μL/well was coated overnight at 4°C onto an ELISA plate (Nunc
MaxiSorp® flat-bottom 96 well plate, Fisher Scientific). After a brief blocking period (40 min with
PBS containing 10% FCS), the immunoconjugates were diluted in a blocking buffer (0.5 or 50
μg/mL) and 100 μL/well were applied for 2 h at room temperature. As not to disrupt the Fc-FcγRI interactions, the bound immunoconjugates were detected using 1: 5000 HRP-labeled anti human
IgG (JacksonImmunoResearch Laboratories, West Grove, PA). After a final wash step, TMB substrate was used to develop the bound HRP secondary antibody, and the color reaction was stopped with 2N H2SO4. Optical Densities at 450 nm were determined using a SpectraMax i3 plate reader (Molecular Devices, San Jose, CA). Binding data was collected in triplicate, averaged, and plotted.
2.2.6. Surface Plasmon Resonance
Binding affinities (KD) and kinetic constants (ka and kd) for pertuzumab as well as the three
DFO-modified immunoconjugates – DFO-nsspertuzumab, DFO-sspertuzumab-Gal and DFO- sspertuzumab-EndoS — were determined via surface plasmon resonance (SPR) on a Biacore T200 instrument (GE Healthcare).
First, in order to characterize the impact of bioconjugation on the immunoreactivity of pertuzumab for its cognate antigen, SPR experiments (n = 3 per construct) with HER2 were performed. To this end, pertuzumab or its DFO-modified immunoconjugates were captured as the ligand on a Protein A sensor chip (29-1275-56, GE Healthcare) by diluting the IgG to a concentration of 1 μg/mL in HBS-EP+ buffer (BR100188, GE Healthcare) and injecting it over a
Series S protein A sensor chip for 30 seconds at a flow rate of 5 μL/min. Next, purified recombinant human HER2 protein (HE2-H822R Acro Biosystems) was used as the analyte and flowed over the functionalized sensor chip. The binding kinetics were evaluated over a range of HER2
28 concentrations in HBS-EP+ buffer (25 nM, 12.5 nM, 6.25 nM, 3.13 nM, 1.56 nM, 0.78 nM, 0.39 nM and 0.195 nM), with each concentration injected for 5 minutes at a flow rate of 5 µL/min to facilitate binding with the immunoconjugate captured on the sensor chip. The dissociation phase was evaluated by allowing the binding buffer (HBS-EP+) to flow (5 µL/min) over the sensor chip for 900 s for the highest analyte concentration (25 nM) and 300 s for all other concentrations. The regeneration buffer (10 mM Glycine-HCl pH 1.5) was then passed over the chip surface for 1 min at a flow rate of 5 µL/min to achieve complete dissociation of the captured antibody and any remaining immunocomplexes. And finally, HBS-EP+ buffer was flowed over the chip for 2 min at
5 µL/min to stabilize the protein A chip surface prior to the injection of the next sample.
Next, to investigate the impact of bioconjugation and deglycosylation on the interaction between the immunoconjugates and Fc receptors, we turned to SPR experiments with the human and murine variants of the high affinity Fc-receptor: FcRI. With respect to the former, a histidine- tagged variant of recombinant human FcRI (500238; NovoPro Labs) was used as the ligand. A
Series S sensor chip CM5 (29401988; GE Healthcare) was functionalized with an anti-histidine antibody using components from the His-capture kit (28995056; GE Healthcare) and the amine coupling kit (BR-1000-50; GE Healthcare) following the standard procedure prescribed by the application wizard on the Biacore T200. After the chip was functionalized, a 0.8 nM solution of huFcR1 in running buffer (HBS-P+ buffer containing 50 M EDTA) was injected over flow cell 2 for 60 s at a flow rate of 10 L/min. Subsequently, high performance injections of various concentrations of pertuzumab or its DFO-modified conjugates (100 nM, 50 nM, 25 nM, 12.5 nM,
6.25 nM, 3.13 nM, 1.56 nM, and 0.78 nM) were performed over flow cells 1 and 2 for 5 min at a flow-rate of 30 L/min. The dissociation of the analyte was evaluated by allowing the running buffer to flow over the chip surface for 5 min. Finally, the chip surface was regenerated using a 60
29 s injection of 10 mM Glycine-HCl (His-capture kit) at a flow-rate of 30 L/min, followed by an extra wash with the running buffer. A similar — though slightly modified — experimental set up was used to evaluate the interaction between the immunoconjugates and murine FcRI (muFcR1).
Here, a 2.4 nM solution of histidine-tagged mouse FcRI (2074-FC; R&D Systems) was captured on the same series S CM5 sensor chip as used above. Based on preliminary indications suggesting rapid on- and off-rates, the association and dissociation phases for all of the pertuzumab-muFcRI interactions were evaluated over a shorter time window: 120 s. The BIAcore T200 evaluation software was used to analyze the kinetic data. A 1:1 fit (RI set to 0) was used to derive kinetic constants for the interactions between the immunoconjugates and both HER2 and huFcRI. In contrast, the rapid on- and off-rates observed in case of muFcRI warranted analysis using steady- state kinetics.
2.2.7. Radiolabeling with 89Zr
For each antibody construct, 400 µg (2.67 nmol) of immunoconjugate solution was diluted in
400 µL PBS, pH 7.4. [89Zr]Zr-oxalate (1300 µCi, 48.1 MBq) in 100 µL of 1.0 M oxalic acid was adjusted to pH 7.0-7.5 with 1.0 M Na2CO3. After the bubbling of CO2 stopped, the 89Zr solution was added to the antibody solution, and the resulting mixture was incubated at room temperature for 1 h. The reaction progress was then assayed using iTLC using an eluent of 50 mM EDTA (pH
5). Subsequently, the reaction was quenched with 10 µL of 50mM of EDTA (pH = 5), and the antibody construct was purified using size exclusion chromatography (Sephadex G-25 M, PD-10 column, GE Healthcare; dead volume = 2.5 mL, eluted with 500 µL fractions of PBS, pH 7.4). If necessary, the antibody was concentrated via centrifugal filtration units with a 50,000 Da molecular weight cut off (AmiconTM Ultra 4 Centrifugal Filtration Units, Millipore Corp. Billerica, MA). The radiochemical purity of the final radiolabeled bioconjugate was assayed by iTLC using 50 mM
30 EDTA (pH 5) as an eluent. In the iTLC experiments, free [89Zr]Zr4+ cations and [89Zr]-EDTA elute with the solvent front, while the radioimmunoconjugate remains at the baseline.
2.2.8. Bead-based Immunoreactivity Assay
A bead-based assay was used to determine the immunoreactive fraction of each of the immunoconjugates. To this end, 20 µL of Ni-NTA beads (Thermo Fisher Scientific # 88831) were transferred to EppendorfTM LoBind microcentrifuge tubes, and 380 µL of PBS-T (PBS with 0.05%
Tween-20) was added. The tubes were vortexed for 5 sec, spun down, and placed on a DynaMag-
2 magnetic rack (Invitrogen) for 30-45 sec. Once the beads settled, the supernatant was aspirated and discarded. The tubes were then placed on the magnetic rack, and 400 µL of PBS-T was added.
3 of the tubes were capped, labeled C1-C3, and set aside as controls.
Next, to prepare the antigen-coupled Ni-NTA beads, 1 µg (10 µL of 0.1 mg/mL dilution) of the
His-tagged HER2 antigen (Acro Biosystems, Newark, DE) was added to the rest of the tubes
(labeled T1-T3), and the tubes were placed on a rotating platform for 15 min. The tubes were then spun down and placed onto the magnetic rack, and the supernatant was aspirated. The same sequence of steps was repeated to wash the beads with 400 µL of PBS-T. Prior to addition of the radioimmunoconjugate in the next step, the beads were resuspended in 399 µL of 1% BSA-PBS.
Subsequently, each radioimmunoconjugates was diluted to 1 ng/µL with PBS, and 1 ng of the diluted radioimmunoconjugate was added to all the tubes. The tubes were vortexed for 5 seconds and placed on the rotating platform for 30 min. The tubes were once again spun down and placed on the magnetic. The supernatant and two washes (400 µL each) were collected. Finally, the beads, supernatant, and washes were measured for radioactivity on a gamma counter (Automatic Wizard2
ɣ-counter, Perkin Elmer, Inc., Waltham, MA). The immunoreactive fraction was then calculated as the fraction of activity associated with the beads corrected for the fraction of non-specific binding observed in the control samples.
31 2.2.9. Radioimmunoconjugate Stability Assays
The stability of the radioimmunoconjugates with respect to radiochemical purity and loss of radioactivity from the antibody was investigated via incubation of the antibodies in human serum for 7 days at 37°C (n = 3). At predetermined time intervals, the radiochemical purity of the radioimmunoconjugates was determined via iTLC with an eluent of 50 mM EDTA pH 5.0.
2.2.10. PET Imaging
PET imaging experiments were conducted on a microPET Focus rodent scanner (Concorde
Microsystems). Mice (athymic nude, NSG, or humanized NSG mice) bearing subcutaneous BT474 xenografts (left shoulder, 60-120 mm3, 25-30 days after inoculation) were administered the radioimmunoconjugates [175-210 μCi, 6.5-7.8 MBq (40-80 μg) in 200 μL of saline] via tail vein injection. Approximately 5 min before PET imaging, the mice were anesthetized by inhalation of a 2% isoflurane (Baxter Healthcare):oxygen gas mixture and placed on the scanner bed. Anesthesia was maintained using a 1% isoflurane mixture. PET data for each mouse were recorded via static scans at 24, 48, 96 and 144 h post-injection (n = 4-5 per group).
2.2.11. Acute Biodistribution Experiments
Mice (athymic nude or humanized NSG mice) bearing subcutaneous BT474 xenografts (left shoulder; 60-120 mm3) were randomized before the study and were administered the radioimmunoconjugates [20 µCi, 0.74 MBq (10 µg) in 200 µL of saline] via tail vein injection.
Subsequently, the animals (n = 5 per group) were euthanized by CO2(g) asphyxiation at 24, 48, and
120 h p.i., and 13 tissues (including the tumor) were removed, washed, dried, weighed, and counted in a gamma counter (PerkinElmer, Inc.). The number of counts in each tissue were background and decay corrected to the time of injection and converted to activity units (µCi) using a calibration curve generated from known standards. The %ID/g for each tissue sample was then calculated by
32 normalization to the total activity injected and the mass of each tissue.
2.2.12. Statistical Analysis
Data were analyzed by the unpaired, two-tailed Student’s t test. Differences at the 95% confidence level (p < 0.05) were considered to be statistically significant.
2.3. Results
2.3.1. Model System, Synthesis, and Characterization
Assembling the three components of the model system for this proof-of-concept study was fairly straightforward. Zirconium-89 (t1/2 ~ 3.3 d) is the radionuclidic gold standard for immunoPET, and desferrioxamine (DFO) remains the only clinically-employed chelator for the radiometal (though several other options are on the cusp of being tested in the clinic).1 The identity of the antibody presented a wider array of options, but the HER2-targeting pertuzumab emerged as the ideal candidate. Indeed, the recent clinical translation of 89Zr-DFO-pertuzumab at Memorial
Sloan Kettering Cancer Center not only illustrated the clinical promise of the radiotracer but also
⎯ and just as importantly ⎯ paved the way for future clinical comparisons with a site-specifically labeled variant of the radioimmunoconjugate.
The non-site-specifically modified immunoconjugate ⎯ DFO-nsspertuzumab ⎯ was prepared according to published procedures via the random conjugation of an isothiocyanate- bearing variant of DFO (p-SCN-Bn-DFO) to the lysines of the antibody (Figure 2.1A).31
MALDI-ToF analysis of the resulting immunoconjugate – DFO-nsspertuzumab – revealed an average of 1.4 ± 0.4 DFO attached per antibody (Figures A.2.1-2.6 and Table A.2.1). The site- specifically modified immunoconjugates were synthesized using a chemoenzymatic method developed in our laboratory.15, 20 Briefly, pertuzumab was first treated with one of two enzymes in order to expose terminal N-acetylglucosamine residues: β-1,4-galactosidase, which removes the
33 terminal galactose residues of the glycans, or EndoS, which hydrolyzes the chitobiose core of the asparagine-linked glycans (Figure 2.1B-C). The resulting constructs were then incubated with the promiscuous galactosyltransferase Gal-T1(Y289L) and the monosaccharide UDP-GalNaz to incorporate azides into the remaining glycans. Finally, the chelator desferrioxamine (DFO) was introduced via the strain promoted alkyne-azide cycloaddition between dibenzocyclooctyne
(DBCO)-DFO and the azide-bearing glycans, ultimately providing DFO-sspertuzumab-βGal or
DFO-sspertuzumab-EndoS. Sodium dodecyl sulfate-polyacrylamine gel electrophoresis (SDS-
PAGE) confirmed the site-specificity of the modification (Figure A.2.7), size exclusion chromatography illustrated that none of the immunoconjugates formed aggregates (Figures A.2.8-
2.11), and MALDI-ToF mass spectrometry revealed degrees of labeling of 2.6 ± 0.1 DFO/mAb for
DFO-sspertuzumab-βGal and 1.3 ± 0.2 DFO/mAb for DFO-sspertuzumab-EndoS (Table A.2.1).
34 Figure 2.1. The construction of the pertuzumab immunoconjugates: DFO-nsspertuzumab (top), DFO-sspertuzumab-βGal (middle), and DFO-sspertuzumab-EndoS (bottom).
2.3.2. In Vitro Characterization
The ability of the immunoconjugates to bind HER2 was first interrogated via SPR assays.
These experiments revealed that native pertuzumab, DFO-nsspertuzumab, DFO-sspertuzumab-βGal,
and DFO-sspertuzumab-EndoS all exhibit nearly identical values for KD (~0.14-0.16 nM), ka (1.9-
2.1 105 M-1s-1), and kd (2.7-3.1 10-5 M-1) (Table 2.1 and Figure A.2.12). These results were
confirmed with flow cytometry experiments using HER2-expressing BT474 human breast cancer
cells (Figure A.2.13). Taken together, these data suggest that the switch to site-specific
bioconjugation does not perturb the ability of the immunoconjugates to bind their target antigen.
pertuzumab DFO-nsspertuzumab DFO-sspertuzumab-βGal DFO-sspertuzumab-EndoS
Dissociation constant (KD) 0.16 0.01 nM 0.15 0.01 nM 0.14 0.01 nM 0.14 0.04 nM
On-rate (ka) 2.02 105 M-1s-1 1.97 105 M-1s-1 2.09 105 M-1s-1 1.92 105 M-1s-1
Off-rate (kd) 3.14 x 10-5 s-1 3.01 10-5 s-1 2.90 10-5 s-1 2.72 10-5 s-1
Table 2.1. SPR-derived binding parameters for the various pertuzumab immunoconjugate and HER2.
35 In light of our recent work on the interplay between the glycosylation, Fc receptor binding, and in vivo performance of radioimmunoconjugates (see Discussion), we also probed the binding of the three immunoconjugates to murine and human FcRI via ELISA and SPR. The ELISA data illustrate that DFO-sspertuzumab-EndoS exhibits attenuated binding to huFcRI compared to DFO- nsspertuzumab (Figure 2.2A). Somewhat surprisingly, DFO-sspertuzumab-βGal also exhibits decreased binding to huFcRI, though not to the same degree as its more truncated counterpart.
These data are generally reinforced by the ELISA experiments with muFcRI, though a higher concentration of immunoconjugate was needed to generate meaningful data due to the lower affinity of the murine receptor for the human IgGs. Here, DFO-sspertuzumab-EndoS again displays reduced binding to muFcRI compared to DFO-nsspertuzumab, while DFO-sspertuzumab-βGal occupies a middle ground.
Broadly speaking, the SPR data confirm the ELISA findings (Figure 2.2B-D and Table
2.2) and are in strong agreement with our recent findings on deglycosylated radioimmunoconjugates.24 With respect to huFcRI, it is clear that the truncation of the heavy glycans exerts a strong influence on binding. More specifically, native pertuzumab (4.7 0.2 10-
9 M), DFO-nsspertuzumab (4.1 0.1 10-9 M), and DFO-sspertuzumab-βGal (4.7 0.2 10-9 M) exhibit similar binding affinities for the receptor, while that of DFO-sspertuzumab-EndoS is weaker: 17.4 0.3 10-9 M (Table 2.2). A similar trend is seen in the kinetic parameters. Here, the EndoS-modified immunoconjugate has a slower on-rate and an accelerated off-rate than the other immunoconjugates. These differences translate to a shorter interaction half-time for DFO- sspertuzumab-EndoS (3.0 0.1 min) compared to native pertuzumab (7.3 0.1 min), DFO- nsspertuzumab (7.4 0.4 min), and DFO-sspertuzumab-βGal (8.0 0.3 min).
The parallel examination of the interactions between the pertuzumab immunoconjugates
36 and murine FcRI revealed a similar overall trend but also a few key differences. While all four immunoconjugates displayed 50-fold lower binding constants for muFcRI compared to huFcRI, unmodified pertuzumab (2.5 0.1 10-7 M), DFO-nsspertuzumab (2.5 0.2 10-7 M), and DFO- sspertuzumab-βGal (2.1 0.1 10-7 M) still retained higher affinities than DFO-sspertuzumab-
EndoS (1.1 0.1 10-6 M). Furthermore, the quartet of immunoconjugates displayed rapid on- and off-rates for binding to muFcRI, thereby necessitating the analysis of the interaction using steady- state kinetics and thus abrogating the determination of kinetic parameters.
37
Figure 2.2. Comparative SPR and ELISA analysis of the interaction between pertuzumab immunoconjugates and recombinant human versus mouse FcRI. (A) ELISA data demonstrating the binding of the immunoconjugates to huFcγRI (10 μg/mL; left) and muFcγRI (10 μg/mL; right); (B) Sensorgrams showing robust dose-response curves and kinetic profiles for the binding of various concentrations of native pertuzumab and the DFO-bearing immunoconjugates to huFcγRI; (C) Bar graphs demonstrating the correlation between deglycosylation and binding affinity (KD), on-rate (ka), off-rate (kd) and half-life (t1/2). The binding affinity, kinetic constants and half-lives for each of the DFO-conjugates were compared with those obtained for unmodified pertuzumab; (D) Sensorgrams showing robust dose-response curves and kinetic profiles for the binding of various concentrations of native pertuzumab and the DFO-bearing immunoconjugates to muFcγRI. Statistically significant relationships are indicated with asterisks. * = p < 0.05, ** = p < 0.005, *** = p < 0.0005, **** = p < 0.00001.
38
DFO-sspertuzumab- DFO-sspertuzumab- pertuzumab DFO-nsspertuzumab βGal EndoS huFcRI Dissociation 4.70 0.18 4.08 0.09 4.68 0.18 17.43 0.25 constant (KD), nM On-rate (ka), M-1s-1 3.36 0.08 105 3.83 0.22 105 3.08 0.21 105 2.22 0.09 105
Off-rate (kd), s-1 1.58 0.03 10-3 1.56 0.08 10-3 1.44 0.05 10-3 3.87 0.19 10-3 Half-time (t1/2), min 7.3 0.1 7.4 0.4 8.0 0.3 3.0 0.1 muFcRI Dissociation 0.25 0.01 0.25 0.02 0.21 0.02 1.08 0.11 constant (KD), M
Table 2.2. SPR-derived binding parameters for the various pertuzumab immunoconjugate and human and murine FcRI 2.3.3. Radiolabeling
The immunoconjugates were subsequently radiolabeled with 89Zr using standard protocols, producing a trio of radioimmunoconjugates in >95% radiochemical yield and >99% radiochemical purity with similar specific activities of 96.2 ± 1.1 MBq/mg (14.4 ± 0.2 GBq/μmol) for 89Zr-DFO- nsspertuzumab, 96.2 ± 1.5 MBq/mg (14.4 ± 0.2 GBq/μmol) for 89Zr-DFO-sspertuzumab-βGal, and
97.7 ± 0.7 MBq/mg (14.7 ± 0.1 GBq/μmol) for 89Zr-DFO-sspertuzumab-EndoS. A subsequent stability study in human serum at 37 °C revealed that over the course of a week, the three radioimmunoconjugates demonstrated >85% stability (Figure A.2.14). Similarly, a bead-based
HER2 binding assay yielded nearly identical immunoreactive fractions for the radioimmunoconjugates: 79 ± 4, 92 ± 2, 88 ± 3 for 89Zr-DFO-nsspertuzumab, 89Zr-DFO- sspertuzumab-βGal, and 89Zr-DFO-sspertuzumab-EndoS, respectively.
2.3.4. In Vivo Behavior
In order to compare the in vivo performance of the radioimmunoconjugates, PET imaging and biodistributions experiments were first conducted in athymic nude mice bearing subcutaneous
HER2-expressing BT474 xenografts. PET images were collected at 24, 48, 96 and 144 h after the
39 intravenous administration of each radioimmunoconjugate (200 μCi, 7.4 MBq, 70-80 μg). As early as 24 h post injection (p.i.) the tumors could easily be delineated in the images acquired with all three radioimmunoconjugates (Figure 2.3A). Over the course of 6 days, the activity concentrations in the tumor increased dramatically, while the uptake in the blood decreased apace. The biodistribution data confirm the observations made via PET: as the experiment progressed, activity concentrations in the tumor increased, ultimately producing excellent tumor-to-healthy organ activity concentration ratios (Figure 2.3B; Tables A.2.2 and A.2.3). Critically, very few qualitative or quantitative differences were found between the images and uptake values obtained with each of the radioimmunoconjugates.
40
Figure 2.3. (A) Planar (left) and maximum intensity projection (MIP, right) PET images of athymic nude mice bearing subcutaneous BT474 xenografts injected with the three 89Zr-DFO-pertuzumab radioimmunoconjugates (179 – 192 μCi, 6.6 – 7.1 MBq, 85 – 92 μg, in 200 μl 0.9% sterile saline). (B) Biodistribution data for athymic nude mice bearing HER2-expressing BT474 xenografts injected with 89Zr-DFO-nsspertuzumab, 89Zr-DFO-sspertuzumab-EndoS, or 89Zr-DFO- sspertuzumab-Gal (lateral tail vein injection, 15 – 20 µCi, 0.56 – 0.74 MBq, 6 – 9 µg). As we have noted, the initial in vivo experiments did not reveal any differences between the
41 performance of the three radioimmunoconjugates. However, pertuzumab is a humanized antibody, athymic nude mice express murine (rather than human) FcRI, and — as we have seen — the affinity of the pertuzumab-based immunoconjugates for muFcRI is rather low. In order to investigate this phenomenon in a more appropriate mouse model, we turned to humanized NOD scid gamma (huNSG) mice. HuNSG mice are NSG mice that were sub-lethally irradiated 3 weeks after birth and then reconstituted with human hematopoietic stem cells in order to facilitate the expression of a functional human immune system, including human NK cells, dendritic cells, T cells, B cells, and monocytes.32-35 PET imaging was performed using huNSG mice bearing subcutaneous HER2-expressing BT474 xenografts (Figure 2.4A). The images reveal stark differences between the randomly-labeled radioimmunconjugate and the two site-specifically labeled variants. To wit, 89Zr-DFO-nsspertuzumab produces lower tumor uptake and higher activity concentrations in healthy organs — specifically, the liver, spleen, and bones — than 89Zr-DFO- sspertuzumab-EndoS and 89Zr-DFO-nsspertuzumab-Gal. The biodistribution data at 120 h p.i. tell a slightly more nuanced story (Figure 2.4B; Tables A2.4 and A2.5). 89Zr-DFO-sspertuzumab-
EndoS produces higher activity concentrations in the tumor compared to 89Zr-DFO-nsspertuzumab
(111.8 ± 39.9 %ID/g vs. 46.1 ± 16.5 %ID/g; p < 0.05) as well as lower uptake in the liver (4.7 ±
0.8 %ID/g vs 10.3 ± 2.7 %ID/g; p < 0.05) and spleen (13.1 ± 4.0 %ID/g vs 44.8 ± 7.9 %ID/g; p <
0.05). Interestingly, the biodistribution of 89Zr-DFO-nsspertuzumab-Gal presents an intermediate case. Its activity concentration in the tumor (77.5 ± 12.6 %ID/g) is greater than that of 89Zr-DFO- nsspertuzumab, though statistically significant differences in the uptake of the two radioimmunoconjugates in the liver and spleen were not observed.
42
Figure 2.4. (A) Planar (left) and maximum intensity projection (MIP, right) PET images of huNSG mice bearing subcutaneous BT474 xenografts collected between 24 and 144 h after the administration of the three radioimmunoconjugates (209 – 218 μCi, 7.7 – 8.1 Mbq, 80 – 83 μg, in 200 μl 0.9% sterile saline); (B) Biodistribution data for 89Zr-DFO-nsspertuzumab, 89Zr-DFO- sspertuzumab-βGal, and 89Zr-DFO-sspertuzumab-EndoS 144 hours following administration in huNSG mice bearing subcutaneous HER2-expressing BT474 xenografts. T = tumor; L = liver; S = spleen; B = bone. * = p < 0.05.
43 2.4. Discussion
While the advent of radioimmunoconjugates for nuclear imaging and therapy has coincided with the rise of ‘precision medicine’, the strategies used to create these tools have long remained surprisingly imprecise. The random conjugation of bifunctional chelators to the lysine residues of immunoglobulins is an undeniably simple approach that nonetheless produces poorly defined and heterogeneous immunoconjugates and risks compromising their bioactivity. The chemoenzymatic bioconjugation methodology that we have developed over the last five years offers a facile — though admittedly slightly more complex — route to better defined, more homogeneous, and highly immunoreactive agents. The key, of course, lies in leveraging the heavy chain glycans, as these biantennary sugar chains provide a handle for selective manipulation, a limited number (i.e. 2 or 4) of conjugation sites, and distance from the immunoglobulin’s antigen-binding domains.
In the case at hand, the very nature of the modification strategy dictates that both DFO- sspertuzumab-EndoS and DFO-nsspertuzumab-Gal are better defined and more homogeneous than
DFO-nsspertuzumab. Yet beyond this fundamental — and important — difference, the three immunoconjugates proved very similar according to the metrics upon which immunoconjugates are traditionally evaluated. SPR assays revealed that each boasts sub-nanomolar KD values for
HER2 (0.14 nM – 0.16 nM) that are essentially identical to that of the parent antibody, findings that were reinforced by FACS with HER2-expressing BT474 cells. The similarities continued upon radiolabeling: 89Zr-DFO-nsspertuzumab 89Zr-DFO-sspertuzumb-EndoS, and DFO-nsspertuzumab-
Gal were produced in similar radiochemical yields and radiochemical purities and exhibited nearly identical specific activities, stabilities, and immunoreactivities. Clearly, this approach to bioconjugation provides structural benefits without any attendant costs when it comes to antigen binding or radiosynthesis.
44 Things become more interesting — and more complicated — when we consider that the manipulation of the heavy chain glycans is not necessarily benign. As we have noted, the heavy chain glycans play an important role in the structure of immunoglobulins and, as a result, their ability to bind Fc receptors, most notably FcRI. We recently explored the interplay between the truncation of the heavy chain glycans, FcRI binding, and the in vivo behavior of randomly- modified radioimmunoconjugates using a series of 89Zr-labeled variants of the HER2-targeting antibody trastuzumab.24 More specifically, we observed that the truncation (via EndoS) or removal
(via PNGaseF) of the heavy chain glycans of 89Zr-DFO-trastuzumab attenuated the binding of the radioimmunoconjugates to FcRI and improved in vivo performance in tumor-bearing NSG and huNSG mice. Critically, our results are corroborated by the work of others on antibody-drug conjugates and fluorophore-modified antibodies.36, 37
This investigation extends our previous work to include a systematic evaluation of immunoconjugates created using both chemoenzymatic methods of bioconjugation. Both of the site-specifically modified immunoconjugates in this study have altered glycans: those of DFO- sspertuzumab-Gal have simply been modified by the addition (via SPAAC) of DFO, while those of DFO-sspertuzumab-EndoS have been truncated and modified with the chelator. These two bioconjugation pathways had little influence on the ability of the immunoconjugates to bind HER2
(vide supra), but they did impact their ability to bind human and murine FcγRI. Both the ELISA and SPR data illustrate that DFO-sspertuzumab-EndoS exhibits attenuated binding to huFcγRI compared to native pertuzumab and DFO-nsspertuzumab, with the SPR data revealing both thermodynamic and kinetic effects. While the immunoconjugates displayed lower binding affinities for muFcγRI across the board, DFO-sspertuzumab-EndoS again exhibited attenuated binding compared to the trio of fully glycosylated analogues according to both SPR and ELISA.
45 Things become more complicated with DFO-sspertuzumab-Gal. In this case, both the ELISA and the SPR data agree that the immunoconjugate exhibits similar affinity to native pertuzumab and
DFO-nsspertuzumab for muFcγRI. In the case of huFcγRI, however, the ELISA and SPR data diverge slightly. The ELISA data shows that DFO-sspertuzumab-Gal displays a decreased affinity for huFcγRI compared to DFO-nsspertuzumab and DFO-sspertuzumab-EndoS, while the SPR data suggest that its binding to huFcγRI is nearly identical to that of the two fully glycosylated analogues. These slight deviations may simply be the result of inherent differences between the assays, especially since some small differences in the kinetic parameters of the binding of DFO- sspertuzumab-Gal to huFcγRI can be seen in the SPR data.
With the relationship between glycans-specific bioconjugation and FcγRI binding largely established, the next step was to investigate whether these differences translate into changes in in vivo performance. As we have discussed (see Results), the three radioimmunoconjugates ⎯ 89Zr-
DFO-nsspertuzumab, 89Zr-DFO-sspertuzumab-EndoS, 89Zr-DFO-sspertuzumab-Gal ⎯ exhibited all but identical behavior in athymic nude mice, suggesting that neither the increased homogeneity nor the attenuated FcγRI binding conferred by site-specific bioconjugation translates to significant in vivo improvements in this particular mouse model. These results are similar to those observed in our previous work and likely stem from the presence of endogeneous murine IgG, the model’s expression of muFcγRI (rather than huFcγRI), and the immunoconjugates’ lower binding affinities for muFcγRI compared to huFcγRI.24, 38, 39
The results in the huNSG mice ⎯ which, of course, express huFcγRI rather than muFcγRI
⎯ tell a more interesting story. In this model, 89Zr-DFO-sspertuzumab-EndoS exhibited lower activity concentrations in the liver and spleen and higher uptake in the tumor than 89Zr-DFO- nsspertuzumab, results consistent with reductions in the interaction between 89Zr-DFO-
46 sspertuzumab-EndoS and FcγRI receptors expressed by monocytes, macrophages, and tissue- resident macrophages in the liver and spleen. The more subtle differences between the tumoral, hepatic, and splenic activity concentrations of 89Zr-DFO-sspertuzumab-Gal and 89Zr-DFO- nsspertuzumab suggest an intermediate case that may relate to the less pronounced differences between the FcγRI binding of the two radioimmunoconjugates or to the increased homogeneity of the former relative to the latter. In light of these data, however, it is important to reinforce that huNSG mice are far from a flawless animal model. To wit, while they do express endogenous IgG
(202 ± 67 µg/mL), the lack of class switching from the isotype IgM to the isotype IgG results in significantly lower titers than in immunocompetent strains and most human patients. This phenomenon can most likely be linked to the lack of B-cell maturation in huNSG mice due to the absence of follicular dendritic cells, germinal centers, and functional lymph nodes.32, 33, 40 As a result, huNSG mice may not provide the most accurate reflections of the behavior of these
‘immune-silent’ radioimmunoconjugates in immunocompetent patients. To remedy this limitation, we are currently working with collaborators to build and utilize murine models of disease that offer even more clinically relevant FcR expression profiles and recapitulations of the human immune environment.
2.5. Conclusion
In the preceding pages, we have demonstrated that the site-specific bioconjugation of antibodies via the manipulation of the heavy chain glycans not only facilitates the reproducible synthesis of better-defined and more homogenous immunoconjugates but can also provide radiotracers with improved in vivo performance. More specifically, compared to a traditional approach to antibody modification, the site-specific bioconjugation of pertuzumab using EndoS produces well-defined and highly immunoreactive immunoconjugates with attenuated binding to
47 murine and human FcRI. This approach improves in vivo behavior in certain immunocompromised mouse models (i.e. huNSG), a result that can have important implications for preclinical research. In retrospect, the lack of any fundamental differences in the antigen- binding behavior of the radioimmunoconjugates upon site-specific bioconjugation was actually advantageous, as it allowed us to parse between changes in biodistribution due to differences in immunoreactivity and changes in biodistribution due to differences in FcRI binding (vide infra).
It remains unclear whether this phenomenon could prove beneficial in the context of clinical immunoPET, though we hypothesize that it is most likely to improve imaging results in immunosuppressed patients with low titers of endogeneous IgG. Moving forward, we plan to interrogate the utility of this phenomenon in the context of radioimmunotherapeutics and antibody- drug conjugates. Furthermore ⎯ and undeniably more importantly ⎯ an upcoming clinical trial comparing the in vivo performance of 89Zr-DFO-nsspertuzumab and 89Zr-DFO-sspertuzumab-EndoS will certainly help resolve the question of the applicability of this work to human immunoPET.
48 2.6. References
1. Deri, M. A.; Zeglis, B. M.; Francesconi, L. C.; Lewis, J. S., PET imaging with ⁸⁹Zr: From radiochemistry to the clinic. Nucl. Med. Biol. 2013, 40 (1), 3-14. 2. Dijkers, E. C.; Oude Munnink, T. H.; Kosterink, J. G.; Brouwers, A. H.; Jager, P. L.; Jong, J. R.; Dongen, G. A.; Schröder, C. P.; Lub-de Hooge, M. N.; Vries, E. G., Biodistribution of 89Zr-trastuzumab and PET imaging of HER2-positive lesions in patients with metastatic breast cancer. Clin. Pharmacol. Ther. 2010, 87 (5), 586-592. 3. Gaykema, S. B. M.; Brouwers, A. H.; Lub-de Hooge, M. N.; Pleijhuis, R. G.; Timmer- Bosscha, H.; Pot, L.; van Dam, G. M.; van der Meulen, S. B.; de Jong, J. R.; Bart, J.; de Vries, J.; Jansen, L.; de Vries, E. G. E.; Schröder, C. P., 89Zr-bevacizumab PET imaging in primary breast cancer. J. Nucl. Med. 2013, 54 (7), 1014-1018. 4. Marquez, B. V.; Ikotun, O. F.; Zheleznyak, A.; Wright, B.; Hari-Raj, A.; Pierce, R. A.; Lapi, S. E., Evaluation of 89Zr-pertuzumab in breast cancer xenografts. Mol. Pharm. 2014, 11 (11), 3988-3995. 5. Ulaner, G. A.; Lyashchenko, S. K.; Riedl, C.; Ruan, S.; Zanzonico, P. B.; Lake, D.; Jhaveri, K.; Zeglis, B.; Lewis, J. S.; O’Donoghue, J. A., First-in-human human epidermal growth factor receptor 2–targeted imaging using 89Zr-pertuzumab PET/CT: Dosimetry and clinical application in patients with breast cancer. J. Nucl. Med. 2018, 59 (6), 900-906. 6. Ulaner, G. A.; Hyman, D. M.; Ross, D. S.; Corben, A.; Chandarlapaty, S.; Goldfarb, S.; McArthur, H.; Erinjeri, J. P.; Solomon, S. B.; Kolb, H.; Lyashchenko, S. K.; Lewis, J. S.; Carrasquillo, J. A., Detection of HER2-positive metastases in patients with HER2-negative primary breast cancer using 89Zr-trastuzumab PET/CT. J. Nucl. Med. 2016, 57 (10), 1523-1528. 7. Agarwal, P.; Bertozzi, C. R., Site-specific antibody–drug conjugates: The nexus of bioorthogonal chemistry, protein engineering, and drug development. Bioconjug. Chem. 2015, 26 (2), 176-192. 8. Adumeau, P.; Sharma, S. K.; Brent, C.; Zeglis, B. M., Site-specifically labeled immunoconjugates for molecular imaging—Part 1: Cysteine residues and glycans. Mol. Imaging Biol. 2016, 18 (1), 1-17. 9. Adumeau, P.; Sharma, S. K.; Brent, C.; Zeglis, B. M., Site-specifically labeled immunoconjugates for molecular imaging—Part 2: Peptide tags and unnatural amino acids. Mol. Imaging Biol. 2016, 18 (2), 153-165. 10. Zhou, Q., Site-specific antibody conjugation for ADC and beyond. Biomedicines 2017, 5 (4). 11. Boswell, C. A.; Mundo, E. E.; Zhang, C.; Bumbaca, D.; Valle, N. R.; Kozak, K. R.; Fourie, A.; Chuh, J.; Koppada, N.; Saad, O.; Gill, H.; Shen, B. Q.; Rubinfeld, B.; Tibbitts, J.; Kaur, S.; Theil, F. P.; Fielder, P. J.; Khawli, L. A.; Lin, K., Impact of drug conjugation on pharmacokinetics and tissue distribution of anti-STEAP1 antibody-drug conjugates in rats. Bioconjug. Chem. 2011, 22 (10), 1994-2004. 12. Junutula, J. R.; Raab, H.; Clark, S.; Bhakta, S.; Leipold, D. D.; Weir, S.; Chen, Y.; Simpson, M.; Tsai, S. P.; Dennis, M. S.; Lu, Y. M.; Meng, Y. G.; Ng, C.; Yang, J. H.; Lee, C. C.; Duenas, E.; Gorrell, J.; Katta, V.; Kim, A.; McDorman, K.; Flagella, K.; Venook, R.; Ross, S.; Spencer, S. D.; Wong, W. L.; Lowman, H. B.; Vandlen, R.; Sliwkowski, M. X.; Scheller, R. H.; Polakis, P.; Mallet, W., Site-specific conjugation of a cytotoxic drug to an antibody improves the therapeutic index. Nat. Biotechnol. 2008, 26 (8), 925-932. 13. Pillow, T. H.; Tien, J.; Parsons-Reponte, K. L.; Bhakta, S.; Li, H.; Staben, L. R.; Li, G. M.; Chuh, J.; Fourie-O'Donohue, A.; Darwish, M.; Yip, V.; Liu, L. N.; Leipold, D. D.; Su,
49 D.; Wu, E.; Spencer, S. D.; Shen, B. Q.; Xu, K. Y.; Kozak, K. R.; Raab, H.; Vandlen, R.; Phillips, G. D. L.; Scheller, R. H.; Polakis, P.; Sliwkowski, M. X.; Flygare, J. A.; Junutula, J. R., Site-specific trastuzumab maytansinoid antibody-drug conjugates with improved therapeutic activity through linker and antibody engineering. J. Med. Chem. 2014, 57 (19), 7890-7899. 14. van Geel, R.; Wijdeven, M. A.; Heesbeen, R.; Verkade, J. M.; Wasiel, A. A.; van Berkel, S. S.; van Delft, F. L., Chemoenzymatic conjugation of toxic payloads to the globally conserved N-glycan of native mAbs provides homogeneous and highly efficacious antibody-drug conjugates. Bioconjug. Chem. 2015, 26 (11), 2233-42. 15. Zeglis, B. M.; Davis, C. B.; Aggeler, R.; Kang, H. C.; Chen, A.; Agnew, B. J.; Lewis, J. S., Enzyme-mediated methodology for the site-specific radiolabeling of antibodies based on catalyst-free click chemistry. Bioconjug. Chem. 2013, 24 (6), 1057-1067. 16. Kristensen, L. K.; Christensen, C.; Jensen, M. M.; Agnew, B. J.; Schjoth-Frydendahl, C.; Kjaer, A.; Nielsen, C. H., Site-specifically labeled 89Zr-DFO-trastuzumab improves immunoreactivity and tumor uptake for immunoPET in a subcutaneous HER2-positive xenograft mouse model. Theranostics 2019, 9 (15), 4409-4420. 17. Alvarez, V. L.; Wen, M. L.; Lee, C.; Lopes, A. D.; Rodwell, J. D.; McKearn, T. J., Site-specifically modified 111In labelled antibodies give low liver backgrounds and improved radioimmunoscintigraphy. Nucl. Med. Biol. 1986, 13 (4), 347-352. 18. Tavare, R.; Wu, W. H.; Zettlitz, K. A.; Salazar, F. B.; McCabe, K. E.; Marks, J. D.; Wu, A. M., Enhanced immunoPET of ALCAM-positive colorectal carcinoma using site-specific 64Cu-DOTA conjugation. Protein Eng. Des. Sel. 2014, 27 (10), 317-24. 19. Tinianow, J. N.; Gill, H. S.; Ogasawara, A.; Flores, J. E.; Vanderbilt, A. N.; Luis, E.; Vandlen, R.; Darwish, M.; Junutula, J. R.; Williams, S. P.; Marik, J., Site-specifically 89Zr- labeled monoclonal antibodies for ImmunoPET. Nucl. Med. Biol. 2010, 37 (3), 289-97. 20. Zeglis, B. M.; Davis, C. B.; Abdel-Atti, D.; Carlin, S. D.; Chen, A.; Aggeler, R.; Agnew, B. J.; Lewis, J. S., Chemoenzymatic strategy for the synthesis of site-specifically labeled immunoconjugates for multimodal PET and optical imaging. Bioconjug. Chem. 2014, 25 (12), 2123-2128. 21. van Berkel, S. S.; van Delft, F. L., Enzymatic strategies for (near) clinical development of antibody-drug conjugates. Drug Discov. Today Technol. 2018, 30, 3-10. 22. Manabe, S.; Yamaguchi, Y.; Matsumoto, K.; Fuchigami, H.; Kawase, T.; Hirose, K.; Mitani, A.; Sumiyoshi, W.; Kinoshita, T.; Abe, J.; Yasunaga, M.; Matsumura, Y.; Ito, Y., Characterization of antibody products obtained through enzymatic and nonenzymatic glycosylation reactions with a glycan oxazoline and preparation of a homogeneous antibody–drug conjugate via Fc N-glycan. Bioconjug. Chem. 2019, 30 (5), 1343-1355. 23. Qasba, P. K., Glycans of antibodies as a specific site for drug conjugation using glycosyltransferases. Bioconjug. Chem. 2015, 26 (11), 2170-2175. 24. Vivier, D.; Sharma, S. K.; Adumeau, P.; Rodriguez, C.; Fung, K.; Zeglis, B., The impact of FcγRI binding on immunoPET. J. Nucl. Med. 2019, 60, 1174-1182. 25. Sharma, S. K.; Chow, A.; Monette, S.; Vivier, D.; Pourat, J.; Edwards, K. J.; Dilling, T. R.; Abdel-Atti, D.; Zeglis, B. M.; Poirier, J. T.; Lewis, J. S., Fc-mediated anomalous biodistribution of therapeutic antibodies in immunodeficient mouse models. Cancer Res. 2018, 78 (7), 1820-1832. 26. Baruah, K.; Bowden, T. A.; Krishna, B. A.; Dwek, R. A.; Crispin, M.; Scanlan, C. N., Selective deactivation of serum IgG: A general strategy for the enhancement of monoclonal antibody receptor interactions. J. Mol. Biol. 2012, 420 (1), 1-7. 27. Borrok, M. J.; Jung, S. T.; Kang, T. H.; Monzingo, A. F.; Georgiou, G., Revisiting the
50 role of glycosylation in the structure of human IgG Fc. ACS Chem. Biol. 2012, 7 (9), 1596-1602. 28. Bruhns, P.; Iannascoli, B.; England, P.; Mancardi, D. A.; Fernandez, N.; Jorieux, S.; Daeron, M., Specificity and affinity of human Fcgamma receptors and their polymorphic variants for human IgG subclasses. Blood 2009, 113 (16), 3716-25. 29. Slamon, D. J.; Clark, G. M.; Wong, S. G.; Levin, W. J.; Ullrich, A.; McGuire, W. L., Human breast cancer: Correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987, 235 (4785), 177-82. 30. Slamon, D. J.; Leyland-Jones, B.; Shak, S.; Fuchs, H.; Paton, V.; Bajamonde, A.; Fleming, T.; Eiermann, W.; Wolter, J.; Pegram, M.; Baselga, J.; Norton, L., Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N. Engl. J. Med. 2001, 344 (11), 783-792. 31. Zeglis, B. M.; Lewis, J. S., The bioconjugation and radiosynthesis of 89Zr-DFO-labeled antibodies. J. Vis. Exp. 2015, (96), e52521. 32. Shultz, L. D.; Ishikawa, F.; Greiner, D. L., Humanized mice in translational biomedical research. Nat. Rev. Immunol. 2007, 7, 118. 33. Brehm, M. A.; Shultz, L. D.; Luban, J.; Greiner, D. L., Overcoming current limitations in humanized mouse research. J. Infect. Dis. 2013, 208 (suppl_2), S125-S130. 34. Ishikawa, F.; Yasukawa, M.; Lyons, B.; Yoshida, S.; Miyamoto, T.; Yoshimoto, G.; Watanabe, T.; Akashi, K.; Shultz, L. D.; Harada, M., Development of functional human blood and immune systems in NOD/SCID/IL2 receptor γ mice. Blood 2005, 106 (5), 1565-73. 35. Baerenwaldt, A.; Lux, A.; Danzer, H.; Spriewald, B. M.; Ullrich, E.; Heidkamp, G.; Dudziak, D.; Nimmerjahn, F., Fcγ receptor IIB (FcγRIIB) maintains humoral tolerance in the human immune system in vivo. Proc. Natl. Acad. Sci. U.S.A. 2011, 108 (46), 18772-7. 36. Gao, P.; Pinkston, K. L.; Wilganowski, N.; Robinson, H.; Azhdarinia, A.; Zhu, B.; Sevick-Muraca, E. M.; Harvey, B. R., Deglycosylation of mAb by EndoS for improved molecular imaging. Mol. Imaging Biol. 2015, 17 (2), 195-203. 37. Li, F.; Ulrich, M. L.; Shih, V. F.-S.; Cochran, J. H.; Hunter, J. H.; Westendorf, L.; Neale, J.; Benjamin, D. R., Mouse strains influence clearance and efficacy of antibody and antibody-drug conjugate via Fc-FcγR interaction. Mol. Cancer Ther. 2019, 18 (4), 780-787. 38. Dekkers, G.; Bentlage, A. E. H.; Stegmann, T. C.; Howie, H. L.; Lissenberg- Thunnissen, S.; Zimring, J.; Rispens, T.; Vidarsson, G., Affinity of human IgG subclasses to mouse Fc gamma receptors. MAbs 2017, 9 (5), 767-773. 39. Overdijk, M. B.; Verploegen, S.; Ortiz Buijsse, A.; Vink, T.; Leusen, J. H. W.; Bleeker, W. K.; Parren, P. W. H. I., Crosstalk between human IgG isotypes and murine effector cells. J. Immunol. 2012, 189 (7), 3430-8. 40. Chen, Q.; He, F.; Kwang, J.; Chan, J. K.; Chen, J., GM-CSF and IL-4 stimulate antibody responses in humanized mice by promoting T, B, and dendritic cell maturation. J. Immunol. 2012, 189 (11), 5223-9.
51 Chapter 3 : [89Zr]Zr-DFO-AR20.5: A MUC1-Targeting ImmunoPET Probe
This chapter is an adaptation of unpublished work from Fung, K.; Vivier, D.; Sharma, S. K.;
Sarbisheh, E. K.; Keinänen, O.; Price, E. W.; Zeglis, B. M. [89Zr]Zr-DFO-AR20.5: A MUC1-
Targeting ImmunoPET Probe.
3.1. Introduction
Mucins are high molecular weight transmembrane glycoproteins that are expressed on the surface of normal and malignant epithelial cells and play a diverse collection of roles, ranging from the protection of the cell surface to the regulation of cellular adhesion.1-3 MUC1 (also known as
CA15.3 or polymorphic epithelial mucin) is a tumor-associated mucin that is over-expressed in most adenocarcinomas, including breast, pancreatic, and epithelial ovarian cancer.4-9 The overexpression and aberrant expression of MUC1 has been linked to tumor aggressiveness and metastasis, poor response to therapy, and poor survival in several tumor types.10 The role of MUC1 in ovarian cancer provides a representative case.11 To wit, MUC1 is highly expressed in both primary epithelial ovarian cancer as well as the vast majority of cases of metastatic disease.7, 10
Moreover, patients with metastatic, treatment-resistant ovarian cancer present elevated levels of
MUC1, with >90% producing antibodies against the antigen.10, 12, 13 Given its role in both transformation and metastatic progression, MUC1 has remained an enticing therapeutic target for almost three decades.10 A wide variety of MUC1-targeted therapies have been developed over the years, including peptide-, protein-, and antibody-based vaccines, therapeutic antibodies, CAR-T cells, and alpha and beta particle-emitting radioimmunoconjugates.3, 10, 14-23 Yet the field’s interest in MUC1-targeted therapeutics has ebbed and flowed during this period, as preclinical successes have proven difficult to recapitulate in the clinic. Recent years, however, have played witness to a
52 reinvigoration of the area in response to the advent of checkpoint inhibitor therapy.24-26
The last ten years have played witness to a steady increase in the use of 89Zr-labeled antibodies as
PET imaging agents for the staging, treatment planning, and treatment monitoring of a range of cancers.27-33 In light of this movement, the therapeutic interest in MUC1, and its role as a prognostic biomarker, we set out to create an immunoPET probe that could be used for clinical diagnostic and theranostic imaging of MUC1-expressing malignancies. The monoclonal antibody that forms the foundation of our imaging agent is AR20.5, a murine IgG capable of binding MUC1 with high affinity and specificity.9 While MUC1 is expressed by normal epithelial cells and cancer cells, the protein is aberrantly underglycosylated in cancer cells.34 This difference in glycosylation state between the epitopes of MUC1 expressed by healthy and malignant tissues has been exploited to create several antibodies capable of specifically binding the tumor-associated antigen, including
AR20.5.9 To wit, AR20.5 binds in a glycosylation-dependent manner to tandem repeat peptide sequence ⎯ i.e. DTRPAP ⎯ of MUC1, a site that is only accessible in tumor-associated epitopes of the mucin.35 The therapeutic mechanism of AR20.5 is predicated on the antibody forming immune complexes with circulating MUC1 or MUC1-expressing tumor cells and subsequently inducing an immune response to the tumor itself.3 A Phase I clinical trial employing AR20.5 as a monotherapy in 17 patients with MUC1-expressing cancers found that the antibody was generally well tolerated and induced MUC1-specific immune responses but did not produce any anti-tumor activity.3 More recently, a therapeutic regimen of AR20.5 in combination with anti-programmed death-ligand (PD-L1) and polyinosinic-polycytidylic acid (PolyICLC) was explored in mouse models of pancreatic cancer and found to produce MUC1-specific immune responses that suppress tumor growth.36
In the manuscript at hand, we report the synthesis, characterization, and in vivo evaluation of [89Zr]Zr-DFO-AR20.5 in two murine models of MUC1-expressing ovarian cancer. We have
53 chosen ovarian cancer as the model system for this investigation because it is the fifth leading cause of cancer deaths amongst women ⎯ with ~14,000 deaths in the United States alone in 2018 ⎯ and, as we have mentioned above, because the overexpression of MUC1 has been linked to aggressiveness, invasiveness, metastatic potential, and resistance to therapy in the disease.11, 13, 37
Ultimately, we envision deploying [89Zr]Zr-DFO-AR20.5 for the immunoPET of all MUC1- expresing malignancies, both as a standalone diagnostic and prognostic tool and as a theranostic imaging companion to AR20.5 itself and other MUC1-targeted therapeutics. Finally, it is important to note that this investigation is admittedly not the first attempt to create a MUC1-targeted radiopharmaceutical for nuclear imaging. Indeed, several reports of MUC1-targeted PET and
SPECT agents appear in the literature, including radioimmunoconjugates labeled with 111In, 99mTc, and 64Cu, a pretargeted approach based on 68Ga, and ⎯ most recently ⎯ a 89Zr-labeled antibody.38-
43 Nonetheless, we believe that the limitations of these earlier imaging strategies, the recent resurgence in MUC1-targeted therapeutics, and the therapeutic potential of AR20.5 combine to form the basis of a strong rationale for the development of [89Zr]Zr-DFO-AR20.5.
3.2. Experimental
3.2.1. Materials and Methods
Unless otherwise noted, all chemicals were purchased from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific (Pittsburgh, PA) and were used without further purification. All water used was ultrapure (>18.2 MΩcm-1 at 25 °C), and dimethylsulfoxide was of molecular biology grade
(>99.9%). p-SCN-Bn-DFO was purchased from Macrocyclics, Inc. (Plano, TX). MALDI-ToF mass spectrometry was performed according to previously reported methods, and service was provided by the Alberta Proteomics and Mass Spectrometry Facility (University of Alberta,
Alberta, Canada). 89Zr was produced and purified via the 89Y(p,n)89Zr reaction at Memorial Sloan
54 Kettering Cancer Center as [89Zr]Zr-oxalate in 1.0 M oxalic acid. The AR20.5 antibody was obtained from Oncoquest, Inc. (Edmonton, Alberta), and mouse immunoglobulin G (mIgG) was purchased from Thermo Fisher Scientific (Waltham, MA). All in vivo experiments were performed in accordance with published protocols approved by the Institutional Animal Care and Use
Committees of Hunter College, Weill Cornell Medical College, and Memorial Sloan Kettering
Cancer Center.
3.2.2. Instrumentation
All instruments were calibrated and maintained according to standard quality control practices and procedures. UV-Vis measurements were taken on a Shimadzu BioSpec-nano Micro- volume UV-Vis Spectrophotometer (Shimadzu Scientific Instruments, Kyoto, Japan). Activity was measured using a CRC-15R Dose Calibrator (Capintec, Inc., Ramsey, NJ), and biodistribution samples were counted on a calibrated Automatic Wizard2 ɣ-counter (PerkinElmer, Inc., Waltham,
MA). The radiolabeling of the immunoconjugate was monitored using glass-fiber, silica- impregnated instant thin-layer chromatography (iTLC) paper (Pall Corp., East Hills, NY) and analyzed on an AR-2000 radio-TLC plate reader using Winscan Radio-TLC software (Bioscan,
Inc., Washington, D.C.).
3.2.3. DFO modification of antibodies
AR20.5 (800 µg, 5.4 nmol) was dissolved in 500 µL of Chelex-treated (Chelex® 100 Resin,
Bio-Rad Laboratories, Inc.) phosphate-buffered saline (Chelex PBS, pH 7.4), and the pH of the solution was adjusted to 8.8-9.0 with Na2CO3 (0.1 M). 5 equivalents of p-SCN-Bn-DFO (13.5 µL,
1.5 mg/mL in DMSO) was added to the solution in small aliquots. The resulting solution was incubated at 37°C for 1 hour with shaking at 500 rpm. The DFO-modified antibody was then purified using size exclusion chromatography (Sephadex G-25 M, PD-10 column, GE Healthcare;
55 dead volume: 2.5 mL, eluted with 2 mL of Chelex PBS, pH 7.4) and concentrated using centrifugal filtration units with a 50,000 Da molecular weight cut-off (AmiconTM Ultra 2 mL Centrifugal
Filtration Units, MilliporeSigma Corp., Burlington, MA). DFO-mIgG was prepared as previously described with the following modifications: 500 µg (3.3 nmol) of mIgG and 8.4 µL of p-SCN-Bn-
DFO was used.
3.2.4. Radiolabeling with 89Zr
590 µg of DFO-AR20.5 was diluted in 400 µL of Chelex PBS, pH 7.4. [89Zr]Zr-oxalate (1640
µCi) in 1.0 M oxalic acid was adjusted to pH 7.0-7.5 with 1.0 M Na2CO3, resulting in a total volume of 95 µL. After the bubbling of CO2 ceased, the 89Zr solution was added to the antibody solution, and the resulting mixture was placed on an agitating thermomixer at 500 rpm for 1 h at 25 °C. The progress of the reaction was then assayed using radio-TLC with an eluent of 50 mM EDTA (pH
5). Subsequently, the reaction was quenched with 10 µL of 50 mM EDTA (pH 5), and the immunoconjugate was purified using size exclusion chromatography (Sephadex G-25 M, PD-10 column, GE Healthcare; dead volume: 2.5 mL, eluted with 500 µL fractions of Chelex PBS, pH
7.4) and concentrated using centrifugal filtration units with a 50,000 Da molecular weight cut-off
(AmiconTM Ultra 2 mL Centrifugal Filtration Units, MilliporeSigma Corp., Burlington, MA). The radiochemical purity of the final radiolabeled construct was assayed via radio-TLC using 50 mM
EDTA (pH 5) as an eluent. In the radio-TLC experiments, the radioimmunoconjugate remains at the baseline, while free 89Zr4+ cations and [89Zr]-EDTA travel with the solvent front. For the radiolabeling of DFO-mIgG, 340 µg of the construct and 950 µCi of [89Zr]Zr-oxalate (total volume of 50 µL) was used.
56 3.2.5. Cell Culture
Human ovarian cancer cell line SKOV3 was purchased from the American Type Culture
Collection (ATCC, Manassas, VA) and maintained in McCoy’s 5A Medium, supplemented with
10% heat-inactivated fetal calf serum, 100 units/mL penicillin, and 100 units/mL streptomycin in an incubator (HeracellTM 150i, ThermoFisher Scientific) set to 37 °C and 5% CO2. Human ovarian cancer cell line expressing the red-shifted firefly luciferase gene SKOV3-Red-FLuc was purchased from PerkinElmer, Inc. (Waltham, MA) and maintained in McCoy’s 5A Medium, supplemented with 10% heat-inactivated fetal calf serum. The cell lines were harvested and passaged upon reaching 80% confluency using 0.25% trypsin/0.53 mM EDTA in Hank’s Buffered Salt Solution without calcium and magnesium. All media was purchased from the Media Preparation Facility at
Memorial Sloan Kettering Cancer Center.
3.2.6. Flow Cytometry
The MUC1-expressing human ovarian cancer cell lines SKOV3 was used for flow cytometry experiments. 50 µL of DFO-AR20.5 at 24, 12, 6, 3, 1.5, 0.75, 0.375, and 0.1875 µg/µL was incubated with 1 x 106 cells/mL for 30 min on ice. The cells were washed with 1 mL of ice-cold
PBS three times by pelleting and resuspension. The cells were then incubated with 50 µL of a goat anti-human IgG-AlexaFluor568 secondary antibody (ThermoFisher Scientific) at 8 µg/µL for 30 min on ice. Subsequently, the cells were again washed with ice-cold PBS and then analyzed on a
BD LSR II flow cytometer (BD Biosciences, San Jose, CA). Binding data was collected and then plotted using FlowJoTM software (BD, Franklin Lakes, NJ).
3.2.7. Radiolabeled Antibody Stability Assays
The stability of the radioimmunoconjugate with respect to radiochemical purity and loss of radioactivity from the antibody was investigated by incubating the antibody in human serum for 7
57 days at 37 °C (n = 3) with shaking at 500 rpm. At predetermined times, the radiochemical purity of the radiolabeled antibody was determined via radio-TLC with an eluent of 50 mM EDTA pH
5.0.
3.2.8. Xenograft Models
Six to eight-week-old female athymic nude mice were obtained from either Charles River
Laboratories (Wilmington, MA) or The Jackson Laboratory (Bar Harbor, ME) and allowed to acclimatize for approximately 1 week prior to inoculation. Animals were housed in ventilated cages and given water and food ad libitum.
Subcutaneous Xenograft Model: Mice were anaesthetized by inhalation of 2% isoflurane (Baxter
Healthcare, Deerfield, IL)/oxygen gas mixture and xenografted subcutaneously on the left shoulder with 5 x 106 SKOV3 cells in a 150 µL cell suspension of a 1:1 mixture of fresh media: Matrigel
(Corning Life Sciences, Corning, NY). The SKOV3 tumors reached the ideal size for imaging and biodistribution studies (~100-150 cm3) after approximately 6 weeks.
Orthotopic Xenograft Model: Mice were anaesthetized by inhalation of 1.5% isoflurane/oxygen gas mixture, and surgery was performed on a heated surface to maintain body temperature. One dose of meloxicam (2 mg/kg) and buprenorphine (0.5 mg/kg) was given preemptively via subcutaneous injection. Bupivacaine, a local anesthetic agent, was injected into the tissue adjacent to the incision line. The skin was then prepped for surgery by alternating scrubs of povidone-iodine and 70% ethanol. A dorsolateral incision (1-2 cm in length) was made on the skin on the top right of the spleen, and the retroperitoneum was dissected to expose the fat pad surrounding the ovary.
8 x 105 cells SKOV3-RedFluc in 15 µL of PBS were injected into the left ovary. The retroperitoneum was closed using Vicryl sutures. The skin was closed with sterile wound clips and the wound edges were sealed with a few drops of tissue adhesive. The tumors reached the ideal size for experiments after approximately 10 weeks.
58 To monitor tumor growth, mice were monitored using bioluminescence imaging at weeks 4 and
8. Approximately 5 min prior to bioluminescence imaging, mice were anesthetized by inhalation of 2% isoflurane/oxygen gas mixture, administered 150 µL luciferin (PerkinElmer, Inc.) via intraperitoneal injection, and placed on the scanner bed. Anesthesia was maintained by inhalation of 1% isoflurane/oxygen gas mixture. Bioluminescent images were acquired with an IVIS
Spectrum Preclinical Imaging System (PerkinElmer). The resulting images were processing using the Living Image (v4.4) software.
3.2.9. PET Imaging
PET imaging was conducted on a microPET Focus 120 small-animal scanner (Siemens Medical
Solutions, Malvern PA). Approximately 5 min prior to PET image acquisition, mice were anaesthetized by inhalation of 2% isoflurane/oxygen gas mixture and kept under anesthesia for the duration of the scan. Static scans were recorded at 24, 72, and 120 h after intravenous administration of either [89Zr]Zr-AR20.5 or [89Zr]Zr-mIgG for a total scan time of 10 min. An energy window of
350-700 keV and a coincidence timing window of 6 ns were used. Data were sorted into 2- dimensional histograms by Fourier re-binning, and transverse images were reconstructed by filtered back-projection (FBP). The imaging data were then normalized to correct for non- uniformity of response of the detector, physical decay of the radioisotope to the time of injection, dead-time count losses, and positron-branching ratio, but no attenuation, scatter, or partial-volume averaging correction was applied. The counting rates in the reconstructed images were converted to activity concentrations (percentage injected dose per gram of tissue [%ID/g]) using a system calibration factor derived from the imaging of a mouse-sized water-equivalent phantom containing
89Zr. Maximum intensity projection (MIP) images were generated from 3-dimensional ordered subset expectation maximization reconstruction (3D-OSEM). The resulting images were analyzed using ASIPro VMTM software (Concorde Microsystems).
59 3.2.10. Acute Biodistribution
Athymic nude mice bearing subcutaneous SKOV3 (left shoulder, 100-150 mm3) were randomized prior to the study and were warmed gently with a heat lamp for 5 min prior to the administration of [89Zr]Zr-AR20.5 or [89Zr]Zr-mIgG via tail vein injection (t = 0). Subsequently, the mice were euthanized via CO2(g) asphyxiation at 24, 72, and 120 h post-injection, and 13 tissues
(including tumor) were collected, rinsed in water, dried, weighed, and counted using a gamma counter calibrated for 89Zr. Counts were converted into activity units (µCi) using a calibration curve generated from known standards. The number of counts per minute in each tissue was background and decay corrected to the injection time. The %ID/g for each sample was calculated by normalization to the total injected activity.
3.2.11. Histopathology
The tumors and metastases harvested for the biodistribution were fixed in 10% neutral buffered formalin and stored until the radioactivity decayed for 10 half-lives. The formalin-fixed tissue samples were processed in ethanol and xylene, embedded in paraffin, sectioned into 5 µm thick sections, and stained with hematoxylin and eosin (H&E). Sections from the tumors and metastases were also stained by immunohistochemistry for MUC1 on a Leica Bond RX automated staining platform (Leica Biosystems). Subsequently, heat-induced epitope retrieval was conducted in a pH
9.0 buffer, and the primary antibody (MUC1 antibody, ThermoFisher Scientific, MA532265) was applied using a 1:250 dilution. Lastly, a polymer detection system (Novocastra Bond Polymer
Refine Detection, Leica Biosystems, DS9800) was applied. The resulting slides were interpreted by a board-certified veterinary pathologist from the Laboratory of Comparative Pathology at
Memorial Sloan Kettering Cancer Center. The slides were digitized using Pannoramic Flash
60 scanners (3DHistech). The scanned images were processed using Pannoramic Viewer software
(3DHistech).
3.3. Results
For the investigation at hand, AR20.5 was first modified with the desferrioxamine (DFO), the ‘gold-standard’ chelator for the positron-emitting radiometal zirconium-89 (t1/2 ~ 3.3 d). To this end, a commercially available isothiocyanate-bearing derivative of DFO ⎯ p-SCN-Bn-DFO ⎯ was appended to the antibody via the e-amines of solvent-exposed lysine residues, according to published procedures (Figure 3.1).44 This approach to bioconjugation yielded an immunoconjugate with an average of 1.2 ± 0.1 DFO/mAb as determined via MALDI-ToF analysis (Figures 3.2-3.3 and Table 3.1). Subsequently, the in vitro behavior of DFO-AR20.5 was evaluated via flow cytometry. More specifically, these experiments confirmed the sensitive and specific binding of
DFO-AR20.5 to the MUC1 antigen on SKOV3 human ovarian cancer cells (Figure 3.4).
Figure 3.1. Schematic of the bioconjugation and radiosynthesis of [89Zr]Zr-DFO-AR20.5
61 x104
74345.0 Intens. [a.u.] Intens.
5
148680.2
4
3
2
1
99536.6
0 60000 80000 100000 120000 140000 160000 180000 m/z
Figure 3.2. Representative MALDI-ToF spectrum for AR20.5.
x104
74898.0 Intens. [a.u.] Intens.
3.0
2.5 149671.8
2.0
1.5
1.0
0.5
100307.6
0.0 60000 80000 100000 120000 140000 160000 180000 m/z Figure 3.3. Representative MALDI-ToF spectrum for DFO-AR20.5.
62 Degree of Labeling Constructs Average mass (Da) (DFO/mAb) AR20.5 148716 ± 49 1.2 ± 0.1 DFO-AR20.5 149652 ± 53
Table 3.1. Degree of labeling of DFO-AR20.5 as determined via MALDI-MS.
Figure 3.4. Flow cytometry analysis of DFO-AR20.5 with SKOV3 human ovarian cancer cells.
63 Stability in Serum (89Zr-AR20.5) 100
80
60 % 40
20
0 0 24 48 72 96 120 144 168 Time (h)
Figure 3.5. Stability of [89Zr]Zr-DFO-AR20.5 in human serum at 37°C. Measurements were collected using a radio-iTLC and performed in triplicate.
After confirming the antigen binding of DFO-AR20.5, the next step was to radiolabel the immunoconjugate with zirconium-89, a radionuclide selected for the task due to the advantageous match between its 3.3 d physical half-life and the multi-day circulation time of IgGs.45 Standard published protocols were employed for the radiosynthesis, ultimately producing [89Zr]Zr-DFO-
AR20.5 in >99% radiochemical purity and a specific activity of 92.9 ± 10.0 MBq/mg (Figure
3.1).46 Finally, a stability study in human serum at 37 °C revealed that 98.3 ± 0.2% of [89Zr]Zr-
DFO-AR20.5 remains intact of an incubation period of 168 hours (Figure 3.5).
To evaluate the in vivo performance of our new probe, small animal PET imaging and biodistribution studies were performed in athymic nude mice bearing subcutaneous SKOV3 human ovarian cancer xenografts. PET images were obtained at 24, 72, and 120 h after the intravenous administration of [89Zr]Zr-DFO-AR20.5 (6.9 – 7.5 MBq, 93 – 102 μg) or an isotype control radioimmunoconjugate [89Zr]Zr-DFO-mIgG (6.9 – 7.1 MBq, 77 – 79 μg) via the tail vein. The PET images reveal that [89Zr]Zr-DFO-AR20.5 clearly delineates the MUC1-expressing tumor, the
64 tumoral activity concentrations of [89Zr]Zr-DFO-AR20.5 growing over the course of the 120 h experiment (Figure 3.6, top). Furthermore, the minimal tumoral accumulation of the control radioimmunoconjugate ⎯ [89Zr]Zr-DFO-mIgG ⎯ underscores that the uptake of [89Zr]Zr-DFO-
AR20.5 in the tumor is not predicated on the enhanced permeability and retention effect alone
(Figure 3.6, bottom).
Figure 3.6. Planar (left) and maximum intensity projection (right; scaled to a minimum of 0% and a maximum of 100%) PET images of athymic nude mice bearing subcutaneous SKOV3 xenografts at 24, 72, and 120 h following the intravenous tail-vein injection of [89Zr]Zr-DFO-AR20.5 (6.9 – 7.5 MBq; 186 – 204 μCi; in 200 μL 0.9% sterile saline) or [89Zr]Zr-DFO-mIgG (6.9 – 7.1 MBq, 187 – 192 μCi). The white arrows mark the tumors.
All of these observations are supported by the biodistribution results. More specifically, the tumoral activity concentrations of [89Zr]Zr-DFO-AR20.5 grew throughout the study, from 11.8 ±
4.1 %ID/g at 24 h post-injection to 22.3 ± 4.6 %ID/g at 72 h p.i. to 33.4 ± 11.2 %ID/g at 120 h p.i
(Figure 3.7 and Table 3.2). A blocking experiment in which the tumor-bearing mice are administered a mixture of [89Zr]Zr-DFO-AR20.5 and a vast excess (500 g) of unlabeled AR20.5 effectively demonstrated the specificity of the former for the MUC1-expressing ovarian cancer cells: the tumoral activity concentration was 6.9 ± 2.5 %ID/g with blocking versus 22.3 ± 4.6 %ID/g
65 without (P = 0.0006; Figure 3.7 and Table 3.2). The biodistribution data also reveals that the background activity concentration in the blood decreases from 15.5 ± 2.9 %ID/g at 24 h p.i. to 9.2
± 0.6 %ID/g at 120 h p.i, as is typical for radioimmunoconjugates. In contrast, the activity concentrations in the bone increase slightly over the course of the experiment ⎯ from 4.8 ± 1.8
%ID/g at 24 h p.i. to 6.6 ± 3.0 %ID/g at 120 h p.i ⎯ in another phenomenon frequently observed with 89Zr-labeled antibodies. The activity concentrations in other healthy organs – including the liver, spleen, and kidneys – remain fairly constant in the range of 2-7% ID/g throughout. While
MUC1 is expressed on the cell surface, the antigen can be shed into circulation as well, whereupon the formation of immune complexes with [89Zr]Zr-DFO-AR20.5 may contribute to high retention in circulation and high liver and spleen uptake. Taken together, these biodistribution data yields generally favorable ⎯ though admittedly not extraordinary ⎯ tumor-to-healthy organ activity concentration ratios.
50 24 hr 72 hr 40 72 hr block 120 hr
30
g
/
D I
% 20
10
0 r t r e e d o r g e n h y le e o a n v e c in in e c n lo m e u i le a t t n s o u H L L p m s s id u B B T S o te te t n n K M S I I ll e a rg m a S L
66 Figure 3.7. Biodistribution data from athymic nude mice (n = 5 per time point) bearing SKOV3 human ovarian cancer xenografts collected 24, 72, and 120 h after the intravenous administration of [89Zr]Zr-DFO-AR20.5 (0.65 – 0.69 MBq; 17.6 – 18.6 μCi; 6.6 – 7.0 μg). For the 72 h blocking experiment, the mice were administered the same amount of [89Zr]Zr-DFO-AR20.5 mixed with a excess of unmodified AR20.5 (~500 μg per mouse).
24 h 72 h 72 h block 120 h Blood 15.5 ± 2.9 11.9 ± 2.9 9.8 ± 2.9 9.2 ± 0.6 Tumor 11.8 ± 4.1 22.3 ± 4.6 6.9 ± 2.5 33.4 ± 11.2 Heart 4.2 ± 1.1 3.2 ± 0.7 3.0 ± 0.8 2.7 ± 0.2 Lungs 8.3 ± 1.9 5.8 ± 2.1 5.3 ± 1.9 4.6 ± 0.3 Liver 6.5 ± 1.6 6.0 ± 1.1 5.6 ± 2.3 6.2 ± 1.1 Spleen 4.5 ± 1.0 4.7 ± 1.6 3.5 ± 1.1 4.8 ± 0.7 Stomach 1.0 ± 0.3 1.0 ± 0.3 0.8 ± 0.3 0.9 ± 0.2 Small Intestine 1.8 ± 0.2 1.4 ± 0.4 1.1 ± 0.3 1.2 ± 0.2 Large Intestine 1.2 ± 0.1 0.9 ± 0.3 0.7 ± 0.2 0.8 ± 0.2 Kidneys 6.7 ± 2.9 4.6 ± 1.1 4.9 ± 1.1 4.4 ± 0.3 Muscle 1.3 ± 0.2 0.9 ± 0.2 0.9 ± 0.3 0.8 ± 0.1 Bone 2.4 ± 0.4 3.7 ± 0.9 3.3 ± 0.7 5.0 ± 1.5
Table 3.2. Biodistribution data from athymic nude mice (n = 5 per time point) bearing SKOV3 human ovarian cancer xenografts collected 24, 72, and 120 h after the intravenous administration of [89Zr]Zr-DFO-AR20.5 (0.65 – 0.69 MBq; 17.6 – 18.6 μCi; 6.6 – 7.0 μg). For the 72 h blocking experiment, the mice were administered the same amount of [89Zr]Zr-DFO-AR20.5 mixed with a excess of unmodified AR20.5 (~500 μg per mouse). The values are %ID/g ± SD. Stomach, small intestine, and large intestine values include contents.
With the subcutaneous xenograft data in hand, the next step was to evaluate [89Zr]Zr-DFO-
AR20.5 in a more realistic orthotopic xenograft model. To this end, orthotopic human ovarian cancer xenografts were established in athymic nude mice via the injection of MUC1- and luciferase-expressing SKOV3-RedFluc cells into the fat pad surrounding the ovary. As early as 24 h, the xenograft in the left ovary can be clearly seen in the PET images, and the uptake in the tumor increases throughout the experiment (Figure 3.8). As in the experiments with the subcutaneous xenograft model, PET imaging using an isotype control radioimmunoconjugate ⎯ [89Zr]Zr-DFO-
67 mIgG ⎯ revealed little tumoral accumulation, reinforcing the specificity of the MUC1-targeting imaging agent. In this case, the biodistribution data is again consistent with the imaging results, revealing a tumoral activity concentration of 11.3 ± 7.1 %ID/g at 120 p.i. but also significant accumulation of [89Zr]Zr-DFO-AR20.5 in the liver (10.5 ± 2.4 %ID/g) and spleen (4.5 ± 1.6 %ID/g) as well (Figure 3.8 and Table 3.3). The latter is most likely the result of the formation of immune complexes between shed MUC1 and circulating radioimmunoconjugate that were subsequently deposited in the liver and spleen. Interestingly, focal uptake of [89Zr]Zr-DFO-AR20.5 was also observed in several lesions that appeared to be metastases in abdomen, peritoneum, liver, and right ovary of the orthotopic tumor-bearing mice (Figure 3.9). Histopathological analyses revealed that these lesions were composed of neoplastic cells arranged in solid nests and scattered tubules and confirmed them to be ovarian cell carcinoma (Figure 3.8). Furthermore, immunohistochemical staining revealed that that the metastatic cells do, indeed, express MUC1, explaining their uptake of the radioimmunoconjugate (Figure 3.8).
Figure 3.8. (A) Bioluminescence (left), planar (center) and maximum intensity projection (right; scaled to a minimum of 30% and maximum of 100%) PET images of athymic nude mice bearing
68 orthotopic SKOV3-Red-FLuc xenografts at 24, 72, and 120 h following the intravenous tail-vein injection of [89Zr]Zr-DFO-AR20.5 (5.6 – 6.0 MBq, 150.6 – 161.4 μCi, 59.1 – 63.3 μg; in 200 μL 0.9% sterile saline) or [89Zr]Zr-DFO-mIgG (6.1 – 6.4 MBq, 163.7 – 171.7 μCi, 61.5 – 70.1 μg). The white arrows mark the tumors; (B) Planar PET image of the representative athymic nude mouse bearing an orthotopic SKOV3-Red-FLuc xenograft at 120 hours post-injection of [89Zr]Zr-DFO- AR20.5 (5.6 – 6.0 MBq, 150.6 – 161.4 μCi, 59.1 – 63.3 μg; in 200 μL 0.9% sterile saline). The white arrows mark the tumor (T) and peritoneal metastatic lesion (Met); (C) H&E (10x magnified; left) and IHC (10x magnified; right) staining of the peritoneal metastatic lesion from the representative mouse, with brown staining indicating the expression of MUC1.
[89Zr]Zr-DFO-AR20.5 [89Zr]Zr-DFO-mIgG Blood 2.4 ± 2.1 4.2 ± 1.1 Tumor 11.3 ± 7.1 3.1 ± 0.8 Heart 0.9 ± 0.7 1.2 ± 0.2 Lungs 1.5 ± 1.4 1.9 ±0.6 Liver 10.5 ± 2.4 5.1 ± 0.1 Spleen 6.1 ± 0.3 4.5 ± 1.6 Stomach 0.3 ± 0.1 0.3 ± 0.1 Small Intestine 0.4 ± 0.1 0.5 ± 0.2 Large Intestine 0.4 ± 0.3 0.5 ± 0.1 Kidneys 2.1 ± 0.4 5.2 ± 0.5 Muscle 0.5 ± 0.2 0.6 ± 0.2 Bone 1.7 ± 0.2 1.8 ± 0.4
Table 3.3. Biodistribution data from athymic nude mice bearing SKOV3-RFluc human ovarian cancer orthotopic xenografts collected 120 h after the intravenous administration of [89Zr]Zr-DFO- AR20.5 (5.6 – 6.0 MBq; 150.6 – 161.4 μCi; 59.1 – 63.3 μg; n = 3) or [89Zr]Zr-DFO-mIgG (6.1 – 6.4 MBq; 163.7 – 171.7 μCi; 61.5 – 70.1 μg; n = 3) via the tail vein. The values are %ID/g ± SD, and the stomach, small intestine, and large intestine values include contents.
Figure 3.9. (A) Biodistribution data for athymic nude mice bearing orthotopic SKOV3-Red-FLuc xenografts collected 120 h following the intravenous tail-vein injection of [89Zr]Zr-DFO-AR20.5 (5.6 – 6.0 MBq, 150.6 – 161.4 μCi, 59.1 – 63.3 μg; in 200 μL 0.9% sterile saline); (B) Planar (left) and maximum intensity projection (MIP, right) PET images of a representative athymic nude
69 mouse bearing an orthotopic SKOV3-Red-FLuc xenograft at 120 hours post-injection of [89Zr]Zr- DFO-AR20.5 (5.6 – 6.0 MBq, 150.6 – 161.4 μCi, 59.1 – 63.3 μg; in 200 μL 0.9% sterile saline); (C) Biodistribution data for the metastatic lesions collected from the orthotopic tumor-bearing mice.
3.4. Discussion In the preceding pages, we have described the synthesis, characterization, and in vivo evaluation of [89Zr]Zr-DFO-AR20.5, a radioimmunoconjugate for the PET imaging of MUC1-expressing tumors. The chemical and biological characterization experiments clearly demonstrate that the bioconjugation and radiolabeling of AR20.5 did not adversely affect the ability of the antibody to bind its target antigen. Moving on to the in vivo experiments, the data show that [89Zr]Zr-DFO-
AR20.5 is capable of clearly visualizing MUC1-expressing tumor tissue in mice bearing both subcutaneous and orthotopic ovarian cancer xenografts. With respect to the former, the in vivo behavior of [89Zr]Zr-DFO-AR20.5 in mice bearing MUC1-positive orthotopic ovarian cancer xenografts mirrors that of other 89Zr-labeled radioimmunoconjugates. To wit, while the activity concentration in tumor tissue increased over the course of the experiment, that in the blood and several other well-perfused organs (e.g. the lungs) decreased in kind. A small amount of radioactivity was observed in the bones of the mice, likely the result of the in vivo demetallation of the radioimmunoconjugate and the subsequent deposition of the osteophilic [89Zr]Zr4+ in the bone.
While the tumor-to-muscle activity concentration ratios in these mice were quite high (42.7 ±
14.6 at 120 h p.i.), several key tumor-to-healthy organ activity concentration ratios ⎯ most notably tumor-to-blood (3.6 ± 1.2 at 120 h p.i.), tumor-to-liver (5.4 ± 2.0 at 120 h p.i.), and tumor-to-spleen
(6.9 ± 2.5 at 120 h p.i.) ⎯ lie below the impressive, ‘double digit’ values observed for several other
89Zr-labeled radioimmunoconjugates such a [89Zr]Zr-DFO-J591, [89Zr]Zr-DFO-trastuzumab, and
[89Zr]Zr-DFO-pertuzumab. This phenomenon is most likely related to the shedding of MUC1 and the formation of MUC1-[89Zr]Zr-DFO-AR20.5 immune complexes that can persist in the blood
70 and subsequently accumulate in the liver and spleen. However, our previous preclinical data with the CA125-targteted radioimmunoconjugate [89Zr]Zr-DFO-B43.13 as well as the preclinical and clinical performance of the CA19.9-targeted radioimmunoconjugate [89Zr]Zr-DFO-5B1 reinforce the feasibility of the immunoPET imaging of shed antigens.33, 47, 48 Furthermore, the in vivo performance of [89Zr]Zr-DFO-AR20.5 compares favorably to that of two recently reported MUC1- targeted radioimmunoconjugates: [64Cu]Cu-DOTA-PR81 and [89Zr]Zr-DFO-GGSK-1/30.38, 43
While direct comparisons are ⎯ of course ⎯ dubious due to the use of different tumor models, the tumoral activity concentrations and tumor-to-healthy organ activity concentration ratios of
[89Zr]Zr-DFO-AR20.5 exceeded those created by the former and were roughly equivalent to those produced by the latter. Finally, the in vivo data in mice bearing orthotopic xenografts largely reflects the data collected in the subcutaneous tumor-bearing animals. However, the ability of
[89Zr]Zr-DFO-AR20.5 to visualize several abdominal, peritoneal, hepatic, and ovarian metastatic lesions underscores its potential as a tool for the non-invasive staging of the disease.
3.5. Conclusion
Moving forward, we plan to continue to explore the use of [89Zr]Zr-DFO-AR20.5 for both diagnostic and theranostic imaging in ovarian cancer. Along these lines, we are currently formulating plans to explore the in vivo performance of [89Zr]Zr-DFO-AR20.5 in mice bearing patient-derived ovarian cancer xenografts. Furthermore, we plan to interrogate the utility of
[89Zr]Zr-DFO-AR20.5 as a predictive theranostic imaging agent in human MUC1 transgenic mice
(MUC.Tg) undergoing AR20.5/anti-PD-L1/PolyICLC combination therapy. Finally, we have also begun exploring whether the in vivo performance of [89Zr]Zr-DFO-AR20.5 could benefit from the removal of the heavy chain glycans ⎯ and thus the abrogation of the radioimmunoconjugate’s interactions with murine FcRI receptors ⎯ a phenomenon that we have previously observed and
71 studied. In this case, we believe that the modulation of the interactions between [89Zr]Zr-DFO-
AR20.5 and the murine immune system may be particularly important given the therapeutic mechanism of the antibody. In the end, this pilot investigation has clearly demonstrated that
[89Zr]Zr-DFO-AR20.5 is an effective tool for the non-invasive delineation of MUC1-expressing tumor tissue, and we are hopeful that in the future, this radioimmunoconjugate can play an important role in the clinical diagnostic and theranostic imaging of ovarian cancer.
72 3.6. References
1. Gendler, S. J.; Lancaster, C. F.; Taylor-Papadimitriou, J.; Duhig, T.; Peat, N.; Burchell, J.; Pemberton, L.; Lalani, E. N.; Wilson, D., Molecular cloning and expression of human tumor- associated polymorphic epithelial mucin. J. Biol. Chem. 1990, 265, 15286-15293. 2. Finn, O. J.; Jerome, K. R.; Henderson, R. A.; Pecher, G.; Domenech, N.; Magarian-Blander, J.; Barratt-Boyes, S. M., MUC-1 epithelial tumor mucin-based immunity and cancer vaccines. Immunol. Rev. 1995, 145, 61-89. 3. de Bono, J. S.; Rha, S. Y.; Stephenson, J.; Schultes, B. C.; Monroe, P.; Eckhardt, G. S.; Hammond, L. A.; Whiteside, T. L.; Nicodemus, C. F.; Cermak, J. M.; Rowinsky, E. K.; Tolcher, A. W., Phase I trial of a murine antibody to MUC1 in patients with metastatic cancer: Evidence for the activation of humoral and cellular antitumor immunity. Ann. Oncol. 2004, 15 (12), 1825-1833. 4. Taylor-Papadimitriou, J.; Burchell, J.; Miles, D. W.; Dalziel, M., MUC1 and cancer. Biochim. Biophys. Acta. 1999, 1455, 301-313. 5. Gendler, S. J., MUC1, the renaissance molecule. J. Mammary Gland Biol. Neoplasia 2001, 6, 339-353. 6. Dong, Y.; Walsh, M. D.; Cummings, M. C.; Wright, R. G.; Khoo, S. K.; Parsons, P. G.; McGuckin, M. A., Expression of MUC1 and MUC2 mucins in epithelial ovarian tumors. J. Pathol. 1997, 183 (3), 311-317. 7. Wang, L.; Ma, J.; Liu, F.; Yu, Q.; Chu, G.; Perkins, A. C.; Li, Y., Expression of MUC1 in primary and metastatic human epithelial ovarian cancer and its therapeutic significance. Gynecol. Oncol. 2007, 105 (3), 695-702. 8. Van Elssen, C. H.; Frings, P. W.; Bot, F. J.; Van de Vijver, K. K.; Huls, M. B.; Meek, B.; Hupperets, P.; Germeraad, W. T.; Bos, G. M., Expression of aberrantly glycosylated Mucin-1 in ovarian cancer. Histopathology 2010, 57 (4), 597-606. 9. Qi, W.; Schultes, B. C.; Liu, D.; Kuzma, M.; Decker, W.; Madiyalakan, R., Characterization of an anti-MUC1 monoclonal antibody with potential as a cancer vaccine. Hybrid. Hybridomics 2001, 20 (5-6), 313-324. 10. Horm, T. M.; Schroeder, J. A., MUC1 and metastatic cancer: Expression, function and therapeutic targeting. Cell Adh. Migr. 2013, 7 (2), 187-198. 11. Siegel, R. L.; Miller, K. D.; Jemal, A., Cancer statistics, 2019. CA Cancer J. Clin. 2019, 69 (1), 7-34. 12. Budiu, R. A.; Mantia-Smaldone, G.; Elishaev, E.; Chu, T.; Thaller, J.; McCabe, K.; Lenzner, D.; Edwards, R. P.; Vlad, A. M., Soluble MUC1 and serum MUC1-specific antibodies are potential prognostic biomarkers for platinum-resistance ovarian cancer. Cancer Immunol. Immunother. 2011, 60 (7), 975-984. 13. Deng, J.; Wang, L.; Chen, H.; Li, L.; Ma, Y.; Ni, J.; Li, Y., The role of tumour-associated MUC1 in epithelial ovarian cancer metastasis and progression. Cancer Metastasis Rev. 2013, 32, 535-551. 14. Bitler, B. G.; Menzl, I.; Huerta, C. L.; Sands, B.; Knowlton, W.; Chang, A.; Schroeder, J. A., Intracellular MUC1 peptides inhibit cancer progression. Clin. Cancer Res. 2009, 15 (1), 100-109. 15. Bitler, B. G.; Schroeder, J. A., Anti-cancer therapies that utilize cell penetrating peptides. Recent Pat. Anticancer Drug Discov. 2010, 5, 99-108. 16. Hisatsune, A.; Nakayama, H.; Kawasaki, M.; Horie, I.; Miyata, T.; Isohama, Y.; Kim, K. C.; Katsuki, H., Anti-MUC1 antibody inhibits EGF receptor signaling in cancer cells. Biochem.
73 Biophys. Res. Commun. 2011, 405 (3), 377-381. 17. Oei, A. L.; Moreno, M.; Verheijen, R. H.; Sweep, F. C.; Thomas, C. M.; Massuger, L. F.; von Mendsorff-Pouilly, S., Induction of IgG antibodies to MUC1 and survival in patients with epithelial ovarian cancer. Int. J. Cancer 2008, 123 (8), 1848-1853. 18. Apostolopoulos, V.; Pietersz, G. A.; Tsibanis, A.; Tsikkinis, A.; Drakaki, H.; Loveland, B. E.; Piddlesden, S. J.; Piebanski, M.; Pouniotis, D. S.; Alexis, M. N.; McKenzie, I. F.; Vassilaros, S., Pilot phase III immunotherapy study in early-stage breast cancer patients using oxidized mannan-MUC1. Breast Cancer Res. 2006, 8 (3), R27. 19. Maher, J.; Wilkie, S.; Davies, D. M.; Arif, S.; Picco, G.; Julien, S.; Foster, J.; Burchell, J.; Taylor-Papadimitriou, J., Targeting of tumor-associated glycoforms of MUC1 with CAR T cells. Immunity 2016, 45 (5), 945-946. 20. Song, E. Y.; Qu, C. F.; Rizvi, S. M. A.; Raja, C.; Beretov, J.; Morgenstern, A.; Apostolidis, C.; Bruchertseifer, F.; Perkins, A.; Allen, B. J., Bismuth-213 radioimmunotherapy with C595 anti-MUC1 monoclonal antibody in an ovarian cancer ascites model. Cancer Biol. Ther. 2008, 7 (1), 76-80. 21. Salouti, M.; Babaei, M. H.; Rajabi, H.; Rasaee, M. J., Preparation and biological evaluation of 177Lu conjugation PR81 for radioimmunotherapy of breast cancer. Nucl. Med. Biol. 2011, 38, 849-855. 22. Hughes, O. D. M.; Bishop, M. C.; Perkins, A. C.; Frier, M.; Price, M. R.; Denton, G.; Smith, A.; Rutherford, R.; Schubiger, P. A., Preclinical evaluation of copper-67 labelled anti-MUC1 mucin antibody C595 for therapeutic use in bladder cancer. Eur. J. Nucl. Med. 1997, 24 (4), 439-443. 23. Gold, D. V.; Modrak, D. E.; Schutsky, K.; Cardillo, T. M., Combined 90yttrium-DOTA-labeled PAM4 antibody radioimmunotherapy and gemcitabine radiosensitization for the treatment of a human pancreatic cancer xenograft. Int. J. Cancer 2004, 109, 618-626. 24. Rivalland, G.; Loveland, B.; Mitchell, P., Update on Mucin-1 immunotherapy in cancer: A clinical perspective. Expert Opin. Biol. Ther. 2015, 25 (12), 1-15. 25. Syrkina, M. S.; Rubtsov, M. A., MUC1 in cancer immunotherapy: New hope or phantom menace? Biochemistry (Mosc.) 2019, 84 (7), 773-781. 26. Taylor-Papadimitriou, J.; Burchell, J. M.; Graham, R.; Beatson, R., Latest developments in MUC1 immunotherapy. Biochem. Soc. Trans. 2018, 46, 659-668. 27. Richard, S. D.; Bencherif, B.; Edwards, R. P.; Elishaev, E.; Krivak, T. C.; Mountz, J. M.; DeLoia, J. A., Noninvasive assessment of cell proliferation in ovarian cancer using [18F] 3'deoxy-3-fluorothymidine positron emission tomography/computed tomography imaging. Nucl. Med. Biol. 2011, 38 (4), 485-491. 28. Sharma, S. K.; Nemieboka, B.; Sala, E.; Lewis, J. S.; Zeglis, B. M., Molecular imaging of ovarian cancer. J. Nucl. Med. 2016, 57 (6), 827-833. 29. Parent, E.; Sanders, V.; Dehdashti, F.; Fowler, K., PET/MRI for gynecological malignancies. In PET/MRI in oncology: Current clinical applications, Iagaru, A.; Hope, T.; Veit-Haibach, P., Eds. Springer International Publishing: Cham, 2018; pp 321-339. 30. Lee, H. J.; Ehlerding, E. B.; Jiang, D.; Barnhart, T. E.; Cao, T.; Wei, W.; Ferreira, C. A.; Huang, P.; Engle, J. W.; Cai, W., Dual-labeled pertuzumab for multimodality image-guided ovarian tumor resection. Am. J. Cancer Res. 2019, 9 (7), 1454-1468. 31. Belhocine, T.; Spaepen, K.; Dusart, M.; Castaigne, C.; Muylle, K.; Bourgeois, P.; Bourgeois, D.; Dierickx, L.; Flamen, P., 18FDG PET in oncology: The best and the worst (Review). Int. J. Oncol. 2006, 28 (5), 1249-1261. 32. Sharma, S. K.; Wuest, M.; Wang, M.; Glubrecht, D.; Andrais, B.; Lapi, S. E.; Wuest, F.,
74 Immuno-PET of epithelial ovarian cancer: Harnessing the potential of CA125 for non-invasive imaging. EJNMMI Res. 2014, 4 (1), 60. 33. Sharma, S. K.; Sevak, K. K.; Monette, S.; Carlin, S. D.; Knight, J. C.; Wuest, F. R.; Sala, E.; Zeglis, B. M.; Lewis, J. S., Preclinical 89Zr immuno-PET of high-grade serous ovarian cancer and lymph node metastasis. J. Nucl. Med. 2016, 57 (5), 771-776. 34. Nath, S.; Mukherjee, P., MUC1: A multifaceted oncoprotein with a key role in cancer progression. Trends Mol. Med. 2014, 20 (6), 332-342. 35. Movahedin, M.; Brooks, T. M.; Supekar, N. T.; Gokanapudi, N.; Boons, G.-J.; Brooks, C. L., Glycosylation of MUC1 influences the binding of a therapeutic antibody by altering the conformational equilibrium of the antigen. Glycobiology 2017, 27 (7), 677-687. 36. Mehla, K.; Tremayne, J.; Grunkemeyer, J. A.; O’Connell, K. A.; Steele, M. M.; Caffrey, T. C.; Zhu, X.; Yu, F.; Singh, P. K.; Schultes, B. C.; Madiyalakan, R.; Nicodemus, C. F.; Hollingsworth, M. A., Combination of mAb-AR20.5, anti-PD-L1 and PolyICLC inhibits tumor progression and prolongs survival of MUC1.Tg mice challenged with pancreatic tumors. Cancer Immunol. Immunother. 2018, 67 (3), 445-457. 37. Torre, L. A.; Trabert, B.; DeSantis, C. E.; Miller, K. D.; Samimi, G.; Runowicz, C. D.; Gaudet, M. M.; Jemal, A.; Siegel, R. L., Ovarian cancer statistics, 2018. CA Cancer J. Clin. 2018, 68 (4), 284-296. 38. Alirezapour, B.; Rasaee, J. M.; Jalilian, A. R.; Rajabifar, S.; Mohammadnejad, J.; Paknejad, M.; Maadi, E.; Moradkhani, S., Development of [64Cu]-DOTA-PR81 radioimmunoconjugate for MUC-1 positive PET imaging. Nucl. Med. Biol. 2016, 43 (1), 73-80. 39. Garkavij, M.; Samarzija, M.; Ewers, S. B.; Jakopovic, M.; Tezak, S.; Tennvall, J., Concurrent radiotherapy and tumor targeting with 111In-HMFG1-F(ab’)2 in patients with MUC1-positive non-small cell lung cancer. Anticancer Res. 2005, 25, 4663-4672. 40. Goldenberg, D. M.; Nabi, H. A., Breast cancer imaging with radiolabeled antibodies. Semin. Nucl. Med. 1999, 39 (1), 41-48. 41. Goldenberg, D. M.; Rossi, E. A.; Sharkey, R. M.; McBride, W. J.; Chang, C.-H., Multifunctional antibodies by the dock-and-lock method for improved cancer imaging and therapy by pretargeting J. Nucl. Med. 2008, 49 (1), 158-163. 42. Schuhmacher, J.; Kaul, S.; Klivenyi, G.; Junkermann, H.; Magener, A.; Henze, M.; Doll, J.; Haberkorn, U.; Amelung, F.; Baster, G., Immunoscintigraphy with positron emission tomography: Gallium-68 chelate imaging of breast cancer pretargeted with bispecific anti- MUC1/anti-Ga chelate antibodies. Cancer Res. 2001, 61, 3712-3717. 43. Stergiou, N.; Nagel, J.; Pektor, S.; Haimes, A.-S.; J., J.; Brenner, W.; M., S.; Miederer, M.; Kunz, H.; Roesch, F.; Schmitt, E., Evaluation of a novel monoclonal antibody against tumor- associated MUC1 for diagnosis and prognosis of breast cancer. Int. J. Med. Sci. 2019, 16 (9), 1188-1198. 44. Zeglis, B. M.; Lewis, J. S., The bioconjugation and radiosynthesis of 89Zr-DFO-labeled antibodies. J. Vis. Exp. 2015, (96), 52521. 45. Deri, M. A.; Zeglis, B. M.; Francesconi, L. C.; Lewis, J. S., PET imaging with 89Zr: From radiochemistry to the clinic. Nucl. Med. Biol. 2013, 40 (1), 3-14. 46. Holland, J. P.; Sheh, Y.; Lewis, J. S., Standardized methods for the production of high specific- activity zirconium-89. Nucl. Med. Biol. 2009, 36 (7), 729-739. 47. Lohrmann, C.; O’Reilly, E. M.; O’Donoghue, J. A.; Pandit-Taskar, N.; Carrasquillo, J. A.; Lyashchenko, S. K.; Ruan, S.; Teng, R.; Scholz, W. W.; Maffuid, P. W.; Lewis, J. S.; Weber, W. A., Retooling a blood-based biomarker: Phase I assessment of the high-affinity CA19-9 antibody HuMab-5B1 for immunoPET Clin. Cancer Res. 2019, 25 (23), 7014-7023.
75 48. Houghton, J. L.; Zeglis, B. M.; Abdel-Atti, D.; Aggeler, R.; Sawada, R.; Agnew, B. J.; Scholz, W. W.; Lewis, J. S., Site-specifically labeled CA19.9-labeled immunoconjugate for the PET, NIRF, and multimodal PET/NIRF imaging of pancreatic cancer. Proc. Natl. Acad. Sci. of the U.S.A. 2015, 112 (52).
76 Chapter 4 : A Molecularly-Targeted Intraoperative Imaging Agent for High-Grade Serous Ovarian Cancer
This chapter is an adaptation of unpublished work from Fung, K.; Sharma, S. K.; Keinänen, O.;
Lewis, J. S.; Zeglis, B. M. A Molecularly-Targeted Intraoperative Imaging Agent for High-Grade
Serous Ovarian Cancer.
4.1. Introduction
Ovarian cancer is the most lethal gynecologic malignancy and the fifth leading cause of cancer-related deaths in women.1 Cytoreductive surgery (CRS) remains the most effective first- line treatment for this disease. However, the complete removal of malignant tissue during surgery can be extremely challenging. In patients with unresectable disease, neoadjuvant chemotherapy followed by interval CRS is a valid alternative to primary CRS and is supported by multicenter randomized controlled trials.2, 3 Indeed, the volume of residual disease (RD) after CRS is one of the strongest prognostic indicators of progression-free and overall survival in ovarian cancer patients.4 While optimal CRS is defined as RD <1cm, each incremental decrease in RD below 1 cm improves outcomes.2, 3 In spite of this, the determination of RD is based solely on the surgeon’s visual and manual assessment. The limitations of this approach are underscored by a study that discovered high inter-observer variability, with surgeons more often underestimating rather than overestimating the amount of RD.5 Furthermore, several studies have examined the value of postoperative CT and found that anatomical imaging revealed RD >1 cm in up to 40% of patients that had been deemed ‘optimally resected’.6, 7 Collectively, this makes a clear case for the need to improve the intraoperative assessment of RD in the treatment of ovarian cancer patients. Therefore, the development of tools that can improve the assessment of RD during CRS is a vitally important unmet clinical need. The development of this technology could aid clinicians in achieving greater
77 reduction of disease burden via NIRF-guided resection of residual tumors to improve patient outcomes.
High-grade serous ovarian cancer (HGSOC) comprises the vast majority (~70%) of ovarian adenocarcinomas.8 A vital molecular signature of this malignant histotype is the over-expression of MUC16, a mucinous glycoprotein that carries the Cancer Antigen 125 (CA125) epitope.9, 10 A common telltale sign of ovarian cancer is elevated serum CA125 levels, and malignant cells themselves are known to profusely express this tumor-associated antigen to their advantage.11
Despite being a shed antigen, high concentrations of MUC16/CA125 remain on the surface of cancer cells, enabling the imaging of ovarian cancer.12 A CA125-targeted intraoperative imaging agent could thus be a valuable surgical tool for the identification of microscopic or metastatic deposits within the peritoneum and in clinically normal-appearing lymph nodes.
Over the past three years, our team has been at the forefront of the development and preclinical validation of various MUC16-targeted molecular imaging agents.13, 14 We recently reported the creation of a PET imaging agent for the in vivo delineation of CA125 expression in ovarian cancer based on the CA125- targeting B43.13 antibody.15 The CA125-targeting murine antibody at the center of this proposal — B43.13 (KD for CA125 = 1.2 nM) — has already been the subject of clinical trials as a cancer vaccine as well as a vector for immunoscintigraphy in ovarian cancer patients.16-22 Furthermore, B43.13-based radioimmunoconjugates labeled with 64Cu and 89Zr have shown significant potential for the PET imaging of CA125 expression in murine models of ovarian cancer.12, 23 In vivo imaging experiments in murine xenograft models of HGSOC clearly demonstrate that this imaging agent is not only capable of visualizing primary tumor tissue but also provides sensitive maps of the metastatic spread of ovarian cancer through the ipsilateral chain of lymph nodes.23
78 Herein, we synthesized a B43.13-based immunoconjugate labeled with the near-infrared dye – IRDye800CW (IR800) for NIRF imaging of HGSOC during CRS. Subsequently, we characterized the ability of B43.13-IR800 to bind to CA125-expressing OVCAR3 human ovarian cancer cells, evaluated their in vivo performance in subcutaneous, orthotopic, and patient-derived xenograft mouse models of HGSOC, and performed ex vivo histopathology analysis. In the short term, this work could have a tremendous impact on patient outcomes by helping clinicians to better assess the peritoneum for residual disease and thereby facilitate a more complete resection of malignant tissue during CRS. In the long term, we believe that this work could help usher in a new era for imaging within the operating room to ultimately improve the surgical treatment of a wide variety of tumors.
4.2. Experimental
4.2.1. Materials and Methods
Unless otherwise noted, all chemicals were purchased from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific (Pittsburgh, PA) and were used without further purification. All water used was ultrapure (>18.2 MΩcm-1 at 25 °C), and dimethylsulfoxide was of molecular biology grade
(>99.9%). IRDye® 800CW NHS ester and IRDye® 800CW DBCO was purchased from LI-COR
Biosciences (Lincoln, Nebraska). The B43.13 antibody was obtained from Oncoquest, Inc.
(Edmonton, Alberta), and mouse immunoglobulin G (mIgG) was purchased from Thermo Fisher
Scientific (Waltham, MA). All in vivo experiments were performed in accordance with published protocols approved by the Institutional Animal Care and Use Committees of Hunter College, Weill
Cornell Medical College, and Memorial Sloan Kettering Cancer Center.
4.2.2. Synthesis of nssB43.13-IR800 and nssmIgG-IR800
B43.13 (1.5 mg, 10.03 nmol) was dissolved in 100 µL of phosphate-buffered saline (PBS,
79 pH 7.4), and the pH of the solution was adjusted to 8.8-9.0 with Na2CO3 (0.1 M). 2 equivalents of
IRDye® 800CW NHS Ester (4 µL, 5.9 mg/mL in DMSO) was added to the solution in small aliquots. The resulting solution was incubated at 37°C for 1 hour with shaking at 500 rpm. The
IR800CW-modified antibody was then purified using size exclusion chromatography (Sephadex
G-25 M, PD-10 column, GE Healthcare; dead volume: 2.5 mL, eluted with 2 mL of Chelex PBS, pH 7.4) and concentrated using centrifugal filtration units with a 50,000 Da molecular weight cut- off (AmiconTM Ultra 2 mL Centrifugal Filtration Units, MilliporeSigma Corp., Burlington, MA). nssmIgG-IR800CW was prepared as previously described, except 6 equivalents of IR800CW (11.9
µL) was used instead. To determine the degree of labeling (DOL), UV-Vis measurements were obtained using a Shimadzu BioSpec-Nano Micro-Volume UV-Vis Spectrophotometer (Shimadzu,
Kyoto, Japan). Absorbance measurements were taken at 280 nm and 774 nm, and the DOL was calculated using the following formulas:
AmAb = A280 – Amax(CF)
DOL = [Amax*MWmAb]/[[mAb]*εIR800CW]] where the correction factor (CF) for IR800CW is 0.03 as provided by the supplier, MWmAb =
150,000 Da, εIR800CW = 240,000 cm-1M-1, and ε280, mAb = 210,000 cm-1M-1.
4.2.3. Synthesis of ssB43.13-IR800
Glycans Modification: For the deglycosylation of B43.13, B43.13 (4.5 mg) was combined with 1.5 units of recombinant EndoS enzyme (New England BioLabs, Ipswich, MA) per 1 µg of antibody, 50 µL 500 mM sodium phosphate, pH 7.5 (10X Glycobuffer 1 from New England
Biolabs), and H2O to make up a total reaction volume of 500 µL. The reaction was incubated at
37°C for 24 h and then purified using magnetic chitin beads (New England Biolabs). The deglycosylated B43.13 antibody was buffer exchanged into pre-treatment (50 mM Bis-Tris, 100 mM NaCl, pH 6.0) using a centrifugal filter (AmiconTM Ultra 2 mL Centrifugal Filtration Units,
80 MilliporeSigma Corp., Burlington, MA). To perform the buffer exchange, the filter was first equilibrated with 50 mM Bis-Tris, pH 6.0 and then centrifuged for 6 minutes at 5,000 g. After the final spin, the deglycosylated antibody was isolated in 100 µL (22.58 mg/mL).
GalNAz Labeling: After the EndoS treatment and buffer exchange, 75 μL H2O, 15 μL 1M Tris buffer pH 7.6, 2.5 μL 1M MnCl2, 5.3 μL UDP-GalNAz (from a stock solution at 40 mM in H2O) and 13.2 μL Gal-T1(Y289L) (stock solution at 2 mg/mL) were added to the reaction solution for a final volume of 265 μL. This resultant solution contained final concentrations of 13.12 mg/mL antibody, 9.5 mM MnCl2, 0.8 mM UDP-GalNAz, and 0.1 mg/mL Gal-T1(Y289L) and was incubated overnight at 30°C.
DBCO-IR800 Ligation: After the GalNAz labeling step, the resulting solution was purified via centrifugal filtration using 2 mL Amicon filtration units and tris-buffered saline (TBS, pH 7.4).
After centrifugation, the modified B43.13 (2 sets of 1,613 μg in 88 μL of TBS) was combined with
86.0 μL DBCO-IR800 each (2 mM stock solution, LI-COR Biosciences) and TBS (pH 7.4) to yield solutions with a final volume of 1.5 mL containing 1.08 mg/mL B43.13 and 0.1 mM DBCO. The solutions were incubated overnight at 25°C.
Purification: After labeling the antibody with DBCO-IR800, the site-specifically modified antibody was purified via size exclusion chromatography (PD-10 column, GE Healthcare) and concentrated using centrifugal filtration units with a 50,000 Da molecular weight cut off
(AmiconTM Ultra 2 mL Centrifugal Filtration Units, MilliporeSigma Corp., Burlington, MA) and
PBS (pH 7.4).
4.2.4. Cell Culture
Human ovarian cancer cell line SKOV3 was purchased from the American Type Culture
Collection (ATCC, Manassas, VA) and maintained in McCoy’s 5A Medium, supplemented with
10% heat-inactivated fetal calf serum, 100 units/mL penicillin, and 100 units/mL streptomycin in
81 an incubator (HeracellTM 150i, ThermoFisher Scientific) set to 37 °C and 5% CO2. Human ovarian cancer cell OVCAR3 was purchased from the ATCC and maintained in RPMI 1640 Medium, supplemented with heat inactivated fetal bovine serum (20% v/v), 2 mM L-glutamine, 10 mM
HEPES, 1 mM sodium pyruvate, 4.5 g/L glucose, 1.5 g/L sodium bicarbonate, 0.01 mg/mL bovine insulin (Gemini Bio-Products, 700-112P), 100 units/mL penicillin, and 100 units/mL streptomycin.
The cell lines were harvested and passaged upon reaching 80% confluency using 0.25% trypsin/0.53 mM EDTA in Hank’s Buffered Salt Solution without calcium and magnesium. All media was purchased from the Media Preparation Facility at Memorial Sloan Kettering Cancer
Center.
4.2.5. Flow Cytometry
Flow cytometry experiments were performed with two cell lines, CA125-positive OVCAR3 and CA125-negative SKOV3 human ovarian cancer cells. Either 106 OVCAR3 or SKOV3 cells were incubated with 1 µg/mL, 4 µg/mL, 8 µg/mL, and 16 µg/mL ssB43.13-AF488 immunoconjugate for 30 min on ice. A blocking condition was also created by incubating the
OVCAR3 or SKOV3 cells with 2 µg/mL ssB43.13-AF488 and a 75-fold excess of unmodified
B43.13 (150 µg/mL). Cells were washed by pelleting, resuspension, and analyzed on a BD LSR II
(BD Biosciences, San Jose, CA). Binding data was collected in triplicate, averaged, and plotted.
4.2.6. Xenograft Models
Ten to twelve-week-old female NSG mice were obtained from The Jackson Laboratory
(Bar Harbor, ME) and allowed to acclimatize for approximately 1 week prior to inoculation.
Animals were housed in ventilated cages and given water and food ad libitum.
Subcutaneous Xenograft: Mice were anaesthetized by inhalation of 2% isoflurane (Baxter
Healthcare, Deerfield, IL)/oxygen gas mixture and xenografted subcutaneously on the right
82 shoulder with 5 x 106 OVCAR3 cells in a 150 µL cell suspension of a 1:1 mixture of fresh media:
Matrigel (Corning Life Sciences, Corning, NY). The OVCAR3 tumors reached the ideal size for imaging and biodistribution studies (~200-300 cm3) after approximately 4 weeks.
Orthotopic Xenograft: Mice were anaesthetized by inhalation of 1.5% isoflurane/oxygen gas mixture, and surgery was performed on a heated surface to maintain body temperature. One dose of meloxicam (2 mg/kg) and buprenorphine (0.5 mg/kg) was given preemptively via subcutaneous injection. Bupivacaine, a local anesthetic agent, was injected into the tissue adjacent to the incision line. The skin was then prepped for surgery by alternating scrubs of povidone-iodine and 70% ethanol. A dorsolateral incision (1-2 cm in length) was made on the skin on the top right of the spleen, and the retroperitoneum was dissected to expose the fat pad surrounding the ovary. 3 x 106 cells OVCAR3-Luc in 15 µL of PBS were injected into the left ovary. The retroperitoneum was closed using Vicryl sutures. The skin was closed with sterile wound clips and the wound edges were sealed with a few drops of tissue adhesive. The tumors reached the ideal size for experiments after approximately 8 weeks.
Patient-derived Xenograft (PDX): The PDX seed was revived from frozen stocks and xenografted subcutaneously on the right shoulder. The tumors reached the ideal size for experiments after approximately 4 weeks.
4.2.7. NIRF Imaging
The IR800-labeled immunoconjugate was administered 72 h prior the NIRF imaging.
Approximately 5 min prior to imaging, mice were anesthetized by inhalation of 2% isoflurane/oxygen gas mixture and placed on the scanner bed. Anesthesia was maintained by inhalation of 1% isoflurane/oxygen gas mixture. Fluorescent images were acquired with an IVIS
Spectrum Preclinical Imaging System (PerkinElmer) using an excitation wavelength of 740 nm and emission wavelength of 820 nm. For ex vivo imaging, the mice were sacrificed via CO2
83 asphyxiation. Select tissues (including tumor) were collected, rinsed in water, dried, and placed on either a black sheet of paper or petri dish. Fluorescent images were acquired with the same wavelength settings on the IVIS Spectrum. The resulting images were processing using the Living
Image (v4.4) software.
4.2.8. Histopathology
The tumors and metastases harvested for the biodistribution were fixed in 10% neutral buffered formalin. The formalin-fixed tissue samples were processed in ethanol and xylene, embedded in paraffin, sectioned into 5 µm thick sections, and stained with hematoxylin and eosin
(H&E). Sections from the tumors and metastases were also stained by immunohistochemistry for
MUC16/CA125 on a Leica Bond RX automated staining platform (Leica Biosystems).
Subsequently, heat-induced epitope retrieval was conducted in a pH 9.0 buffer, and the primary antibody (Novus Biologicals, NBP1-96619) was applied using a 1:250 dilution. This was followed by applying a polymer detection system (Novocastra Bond Polymer Refine Detection, Leica
Biosystems, DS9800). The resulting slides were interpreted by a board-certified veterinary pathologist from the Laboratory of Comparative Pathology at Memorial Sloan Kettering Cancer
Center. The slides were digitized using Pannoramic Flash scanners (3DHistech). The scanned images were processed using Pannoramic Viewer software (3DHistech).
4.3. Results and Discussion
4.3.1. PET Imaging of CA125 Expression in Ovarian Cancer
We have recently reported the synthesis, characterization, and in vivo preclinical validation of
[89Zr]Zr-DFO-B43.13, a CA125-targeted radioimmunoconjugate for the PET imaging of CA125- positive ovarian cancer. B43.13 was conjugated to desferrioxamine (DFO) using p-SCN-Bn-DFO, and the resulting immunoconjugate was radiolabeled with [89Zr]Zr4+ to create a [89Zr]Zr-DFO-
84 B43.13 radioimmunoconjugate that was shown to have high radionuclidic purity (>99%), specific activity (5.5 0.8 mCi/mg), and immunoreactivity (>92%).
Preclinical experiments in murine models of ovarian cancer clearly demonstrate that [89Zr]Zr-
DFO-B43.13 is a highly effective imaging agent for the non-invasive visualization of CA125- expressing tissues. PET imaging experiments in athymic nude mice bearing bilateral OVCAR3
(CA125+) and SKOV3 (CA125-) human ovarian cancer xenografts illustrated the selectivity of the radiotracer. While high activity concentrations were observed in the CA125-positive OVCAR3 xenografts, only low levels of uptake were seen in the CA125-negative SKOV3 tumors. Follow- up imaging and biodistribution experiments in mice bearing only OVCAR3 xenografts revealed activity concentrations in the tumor and blood of 24.7 7.5 %ID/g and 4.0 3.0 %ID/g, respectively, at 120 h post-injection (Figure 4.1). The non-target organ with the highest activity concentration was the liver, with approximately 15 %ID/g at all time points. All other organs ⎯ including the ovaries ⎯ displayed activity concentrations <5 %ID/g.
Figure 4.1. CA125-targeted PET imaging with 89Zr-DFO-B43.13. Serial PET images of an athymic nude mouse bearing a CA125-positive OVCAR3 xenograft after the administration of 89Zr-DFO-B43.13 via tail vein injection. The coronal planar images intersect the middle of the tumor; T = tumor; L = liver; LN = lymph node.
Easily the most intriguing facet of the in vivo experiments with 89Zr-DFO-B43.13 was the revelation of the radiotracer’s ability to track the metastatic spread of the disease through the ipsilateral lymph node chain. In mice bearing OVCAR3 xenografts, high activity concentrations
85 were observed in the axillary (104.8 45.1 %ID/g at 72 h p.i.) and brachial (56.6 39.4 %ID/g at
72 h p.i.) lymph nodes proximal to the tumor. In these same mice, the corresponding contralateral lymph nodes displayed far lower uptake of the radioimmunoconjugate: 3.7 3.2 %ID/g (axillary) and 6.4 3.7 %ID/g (brachial). Furthermore, in a subset of mice, high activity concentrations were observed in the ipsilateral sub-mandibular lymph node — but not the corresponding contralateral node (Figures 4.2-4.4). Control experiments using a 89Zr-labeled isotope IgG revealed minimal uptake in all ipsilateral and contralateral nodes (Figure 4.4).
Figure 4.2. The visualization of lymph nodes with 89Zr-DFO-B43.13. Coronal PET images (120 h p.i) of three OVCAR3 tumor-bearing mice demonstrating the uptake of the radioimmunoconjugate in the ipsilateral chain of lymph nodes. T = tumor; ALN = axillary lymph nodes; BLN = brachial lymph nodes; SMLN = sub-mandibular lymph nodes; L = liver.
Figure 4.3. The visualization of lymph nodes with 89Zr-DFO-B43.13. Coronal (A) and maximum intensity projection (B) PET images obtained 72 h after the administration of 89Zr-DFO- B43.13 to a mouse bearing a CA125-positive OVCAR3 xenograft; (C) PET/CT image obtained 120 h after the administration of 89Zr-DFO-B43.13 to the same mouse; T = tumor; ALN = axillary lymph nodes; BLN = brachial lymph node; SMLN = sub-mandibular lymph node; L = liver.
86
Figure 4.4. Comparative analysis of activity concentrations from 89Zr-DFO-B43.13 vs. 89Zr- DFO-Isotype IgG in select tissues of interest. Biodistribution of 89Zr-DFO-B43.13 and 89Zr- DFO-Isotype IgG in OVCAR3 xenograft-bearing mice (n = 4 per group) comparing the activity concentrations of the two radioimmunoconjugates in the tumor, relevant lymph nodes, and liver. *Indicates p value < 0.05. Ex vivo histopathological analyses of the resected lymph nodes clearly confirmed the presence of metastatic disease in the ipsilateral nodes. The contralateral nodes, on the other hand, appeared normal. More specifically, in mice displaying high concentrations of activity in the ipsilateral axillary and brachial nodes but not the ipsilateral submandibular node, it was found that the axillary and brachial nodes showed evidence of metastatic spread while the submandibular node did not.
However, in mice displaying high activity concentrations in all three ipsilateral lymph nodes, evidence of metastases was found in the ipsilateral submandibular node as well (Figure 4.5). The correlation between the histopathology and PET imaging was remarkable: lymph nodes delineated in the PET scans were pathologically confirmed to contain metastatic disease, while lymph nodes not visualized via PET were proven negative for the presence of ovarian cancer. Simply put, neither false negatives nor false positives were observed.
87
Figure 4.5. Histopathological analysis of contralateral and ipsilateral sub-mandibular lymph nodes. (O) Magnified coronal PET image of the primary OVCAR3 tumor, ipsilateral lymph nodes, and contralateral lymph nodes of a mouse injected with 89Zr-DFO-B43.13; (A) 20X- and (B) 10X- magnified images of H&E-stained sections of the PET-negative sub-mandibular lymph node; (C) 20X- and (D) 10X-magnified images of the same lymph node, analyzed by IHC staining for CA125 expression. In these images, the lymphocytes (stained blue) are seen in typical clusters along the sides of the node; (E) 20X- and (F) 10X-magnified images of H&E-stained sections of the PET- positive sub-mandibular lymph node; (G) 20X- and (H) 10X-magnified images of the same lymph node, analyzed by IHC staining for CA125 expression. This node demonstrated gross invasion by metastatic ovarian cancer cells, which stained positive (dark brown) for the cell surface expression of CA125. 4.3.2. Near-Infrared Fluorescence Imaging of CA125 Expression in Ovarian Cancer
Cytoreductive surgery with complete gross resection of disease remains the most effective first-line treatment strategy for patients with advanced ovarian cancer. However, the complete removal of all malignant tissue can be extremely challenging. The volume of residual disease after
CRS is one of the strongest prognostic indicators for progression free and overall survival. An intraoperative imaging agent capable of the real-time identification of tumor tissue during surgery could aid clinicians in the identification and removal of residual disease, thus increasing the volume of disease resected and improving patient outcomes. This technology would be especially useful
88 with regard to clinically normal-appearing lymph nodes for which there is no agreed upon standard management. We have gathered some preliminary data in the use of and IR800CW-labeled variant of B43.13 (B43.13-IR800) for the near-infrared fluorescence (NIRF) imaging of CA125 expression in ovarian cancer.
To evaluate B43.13 antibody for in vitro and in vivo optical imaging, we synthesized more than one variant of fluorophore-labeled that was utilized for proof-of-concept studies. For example, we chose to append Alexa Fluor-488 (AF488) for in vitro flow cytometry experiments, whereas
Alexa Fluor-680 (AF680) was chosen for conducting preliminary in vivo fluorescence imaging studies in relevant xenograft models of high grade serous ovarian cancer. However, since the ultimate goal of the project was to obtain well-defined fluorophore-labeled-conjugates of the antibody for potential clinical use, we used a site-specific bioconjugation approach instead of the traditional approach (Figure 4.6).24-26 While inexpensive and facile, the traditional bioconjugation method relies on the reaction between the amine-reactive variant of the dye and lysine residues that are randomly distributed on the surface of the antibody. Due to the lack of site-specificity, this bioconjugation approach often yields heterogenous immunoconjugates, which could impair the immunoreactivity of an antibody if the bioconjugation inadvertently occurs in or near the antigen- binding domains.24, 27 To circumvent this potential pitfall, we used a click chemistry-based chemoenzymatic method developed in our laboratory to site-specifically append the dyes to the heavy chain glycans of the antibody.24-26, 28-30 Briefly, the methodology is predicated on three steps:
(1) the use of the EndoS enzyme to cleave the heavy chain glycans; (2) the harnessing of a mutant, promiscuous β-galactosyltransferase [GalT-(Y289L)] to transfer an azide-modified galactose residue (GalNaz) onto the truncated glycans; and (3) the click ligation between a dibenzocyclooctyne (DBCO)-modified IR800CW dye and the azide-modified antibody. It has been demonstrated that this strategy produces more homogenous and well-defined constructs with
89 improved in vivo performance compared to randomly labeled analogues.25, 26, 31-34 Using the site- specific labeling strategy that would theoretically yield 2 fluorophores per antibody, we were able to obtain a degree of labeling (DOL) of 1 AF488 dye molecule per B43.13 antibody, whereas a
DOL of 1.5 per B43.13 antibody was obtained with the AF680 dye. In vitro experiments with
CA125-positive OVCAR3 and CA125-negative SKOV3 human ovarian cancer cells clearly demonstrate that ssB43.13-AF488-EndoS retained the selectivity and affinity of the parent B43.13 antibody for its target antigen (Figure 4.7).
Figure 4.6. Schematic of the chemoenzymatic method for site-specific labeling of the B43.13 antibody. This approach for site-specific labeling of antibodies relies on the modification of the biantennary glycan chain appended to the asparagine residues at position 297 (N297) on the heavy chain glycans of immunoglobulin G (IgG) molecules. Briefly, the B43.13 antibody was treated with a recombinant endoglycosidase (EndoS), which hydrolyses the chitobiose core of the asparagine-linked glycans to produce a partially deglycosylated antibody. This is followed by appending an azide-modified sugar (UDP-GalNAz) residue to the partially deglycosylated antibody using a promiscuous -galactosyltransferase – GalT(Y289L). In the last step, click chemistry is used to append a dibenzocyclooctene-modified variant of the fluorophore of choice to the partially deglycosylated and azide modified B43.13 antibody.
90
Figure 4.7. In vitro selectivity. FACS data illustrating the selectivity of an AlexaFluor-488-labeled B43.13 immunoconjugate (B43.13-AF488) using CA125-positive OVCAR3 (left) and CA125- negative SKOV3 (right) ovarian cancer cells. Red = unstained cells; Green = blocking condition created via co-incubation using 2 g/mL B43.13-AF488 and 150 g/mL unlabeled B43.13; Dark Blue = cells incubated with 1 g/mL B43.13-AF488; Tangerine = cells incubated with 4 g/mL B43.13-AF488; Turquoise = cells incubated with 8 g/mL B43.13-AF488; Magenta = cells incubated with 16 g/mL B43.13-AF488. B43.13-AF488-staining of CA125-positive OVCAR3 cells showed a rightward shift of fluorescence from the baseline (unstained OVCAR3 cells) on the histogram. Blockade of B43.13-AF488-staining by a vast excess of unlabeled B43.13 antibody also exhibited a lack of a rightward shift in fluorescence – similar to unstained cells. On the other hand, the histograms for staining CA125-negative SKOV3 cells with similar concentrations of the B43.13-AF488 showed no shift in fluorescence signal in the histograms suggesting a lack of antibody binding to this CA125-negative cell line. Preliminary in vivo NIRF imaging experiments were performed in athymic nude mice bearing subcutaneous CA125-expressing OVCAR3 human ovarian cancer xenografts (Figure 4.8).
The results from the preliminary studies encouraged us to append a clinically relevant fluorophore that can be used for NIRF imaging. Thus, IR800CW was adopted as the fluorophore of choice for site-specific conjugation of the CA125-targeting antibody B43.13 (Figure 4.9). These experiments revealed that high concentrations of B43.13-IR800 can be seen in the tumor as early as ~24 h after
91 the injection of the immunoconjugate. The uptake in the tumor increases with time, ultimately plateauing around 96-120 hours post-injection (p.i.). Background uptake in the liver was observed as well, likely due to the formation of immune complexes with CA125 in the blood. Similar to what was observed previously in previous immunoPET studies (Figures 4.1-4.5) using the CA125- targeted B43.13 antibody in subcutaneous OVCAR3 xenografts, and orthotopic ovarian xenografts uptake of B43.13-IR800 was observed in lymph nodes proximal to the tumor in the NIRF imaging scans (Figures 4.10-4.11).
Figure 4.8. Near infrared fluorescence (NIRF) in vivo imaging of OVCAR3 subcutaneous xenograft with B43.13-AF680. Schema outlining the subcutaneous implantation of 106 OVCAR3 cells in immunodeficient mice, which were allowed to grow tumors (200 – 300 mm3) for 5 weeks before the mice were switched to alfalfa-free low fluorescence diet one week prior to NIRF imaging. The tumors measured ~300 – 350 mm3 at the time of imaging. Tumors were allowed to grow slightly larger to allow for potential spread along the ipsilateral lymph node chain as observed in previous experiments with this model. Longitudinal NIRF imaging after injecting OVCAR3 xenografts with B43.13-AF680 revealed relatively higher intensity fluorescence signal as early as
92 D1 after injection of the CA125-targeted NIRF imaging agent. High contrast images were observed between D3-D7 after injection.
Figure 4.9. Synthesis and in vitro characterization of ssB43.13-IR800 conjugates. (A) Schematic showing the stepwise synthesis of B43.13 antibody site-specifically labeled with IR800CW; (B) (L-R) SDS-PAGE carried out under reducing conditions to analyze the shift in the mobility of immunoglobulin heavy chains of the unmodified B43.13 antibody versus the partially deglycosylated synthetic intermediate (EndoS-B43.13) resulting from step 1 (shown in A) and the IR800 click-conjugated EndoS-treated B43.13 antibody (ssB43.13-IR800) resulting from steps 2 and 3 (shown in A); (C) Further resolved SDS-PAGE of the immunoglobulin heavy chains of the starting material and the intermediates from the immunoconjugate synthesis process as well as the end-product. A clear downward shift is observed in the EndoS-treated B43.13 due to truncation of the glycans resulting in a loss of molecular weight (MW). Appending the IR800 dye to the partially deglycosylated antibody results in a gain in MW leading to an upward shift in mobility of the IgG heavy chain for ssB43.13-IR800 on SDS-PAGE.
93
Figure 4.10. In vivo NIRF imaging with B43.13-IR800. NIRF images of a nude mouse in (A) prone position and (B) supine position bearing a CA125-positive OVCAR3 xenograft (right flank) 72 h after the administration of B43.13-IR800 (100 g; 0.66 nmol). Note the clear visualization of the primary tumor as well as the ipsilateral axillary lymph node. (C) NIRF image of the same mouse with the internal organs exposed, further illustrating the NIRF signal in the tumor and the IALN as well as the liver. (D) Ex vivo NIRF image illustrating the degree of uptake of B43.13-IR800 in the tumor (T), ipsilateral axillary lymph nodes (IALN), contralateral axillary lymph nodes (CALN), ipsilateral inguinal lymph nodes (IILN), contralateral inguinal lymph nodes (CILN), and liver (L).
Figure 4.11. In vivo and ex vivo NIRF imaging of ovarian cancer using IR800-labeled-B43.13. Bioluminescence image (A) and NIRF image (B) of a mouse, whose left ovary was orthotopically
94 xenografted with a CA125-expressing cell line (luciferase-transfected OVCAR3) representative of human high grade serous ovarian cancer. The mouse was imaged 120 h after the injection of IR800- labeled-B43.13 (100 g; 0.66 nmol). Upon necropsy (C), high NIRF signal was observed in the liver (L) and the left ovary (LO). While the entire liver was fluorescent (D), only the left ovary (primary tumor) showed high fluorescence in the resected female reproductive tract (E). Resection of the liver and the female reproductive tract revealed (F) high fluorescence signal in a peritoneal implant (PI) and renal and lumbar lymph nodes (RLN and LLN). Further resection of the peritoneal implant (G) allowed better delineation (H) and successful resection (I) of the renal (RLN) and lumbar lymph nodes (LLN). Finally, all resected tissues including the ipsilateral (IILN) and contralateral inguinal lymph nodes (CILN) were imaged ex vivo with NIRF as shown in (J). Histopathological analysis of the resected tissues identified the primary tumor and confirmed the presence of neoplastic cells bearing characteristics of ovarian adenocarcinoma, thus leading to the accumulation of IR800-labeled B43.13 in the peritoneal implant (PI), the renal (RLN) and lumbar lymph nodes (LLN). The inguinal lymph nodes that were negative in the ex vivo NIRF imaging proved negative upon histopathological analysis for presence of neoplastic cells. The promising results of the preliminary in vivo experiments prompted us to explore the use of the highly immunodeficient Nod-Scid-Gamma (NSG) mice to develop subcutaneous, orthotopic, and patient-derived xenograft (PDX) models of high grade serous ovarian cancer
(HGSOC). We changed the strain of mice from nude to NSG for development of the mouse xenograft for the following reasons: low take-rate (<25%) of OVCAR3 tumors in nude mice, highly variable tumor growth – ranging from 4-8 weeks to obtain appropriately sized tumors, and unevenly sized tumors in <25% of tumor-bearing mice. The use of NSG mice was preferred because the absence of functional NK cells in this strain allowed for efficient take of the OVCAR3 tumors as witnessed by ~100% take-rate. Additionally, NSG mice are a preferred strain for propagating frozen parent PDX tumor seeds to generate larger cohorts of PDX mice for use in preclinical studies. In these experiments, we added a non-site-specifically modified construct – nssB43.13-
IR800 – to compare with the site-specifically modified construct – ssB43.13-IR800 and a negative control – nssmIgG-IR800 – to validate the specific uptake of ssB43.13-IR800 in CA125-positive tumors. In NSG mice with subcutaneous OVCAR3 tumors, higher levels of tumoral uptake were observed with ssB43.13-IR800 (Figure 4.12A) compared to the non-specific uptake of the negative control, nssmIgGIR800 (Figure 4.12C). This observation was supported by ex vivo imaging, which
95 likewise showed higher uptake in the tumor for ssB43.13-IR800 compared to the negative control.
The same trend was observed with ssB43.13-IR800 compared to nssB43.13-IR800 (Figure 4.12B).
In NSG mice with orthotopic OVCAR3 tumors and with PDXs, higher levels of tumoral uptake were also observed with ssB43.13-IR800 compared to nssB43.13-IR800 and nssmIgG-IR800 (Figure
4.13 and Figure 4.14). Background uptake of ssB43.13-IR800 in the liver was observed as well, which is consistent with our observations in the nude mice with subcutaneous and orthotopic
OVCAR3 tumors.
96
Figure 4.12. In vivo and ex vivo NIRF imaging with ssB43.13-IR800, nssB43.13-IR800, and ssmIgG-IR800 in an NSG mouse bearing a subcutaneous OVCAR3 xenograft. Representative NIRF images were obtained in an NSG mouse bearing a CA125-positive OVCAR3 human ovarian subcutaneous xenograft 72 hours after the intravenous administration of (A) ssB43.13-IR800 (50 µg; 0.33 nmol), (B) nssB43.13-IR800 (50 µg; 0.33 nmol), and (C) nssmIgG-IR800 (50 µg; 0.33 nmol). (Left to right) Representative NIRF images of the live mouse were obtained. Upon necropsy of the mouse, NIRF images of the harvested tissues and select tissues of interest – tumor (T), liver (Liv), spleen (Sp), and ovaries (Ov) – were obtained. For the harvested tissues shown above, T = tumor; H = heart; Lu = lungs; Liv = liver; Sp = spleen; St = stomach; Pa = pancreas; L.I. = large intestine; S.I. = small intestine; Ov = ovaries; Ki = kidneys; Mu = muscle; Bo = bone.
97
Figure 4.13. In vivo and ex vivo NIRF imaging with ssB43.13-IR800, nssB43.13-IR800, and ssmIgG-IR800 in an NSG mouse bearing an orthotopic OVCAR3 xenograft. Representative NIRF images were obtained in an NSG mouse bearing a CA125-positive OVCAR3 human ovarian orthotopic xenograft 72 hours after the intravenous administration of (A) ssB43.13-IR800 (50 µg; 0.33 nmol), (B) nssB43.13-IR800 (50 µg; 0.33 nmol), and (C) nssmIgG-IR800 (50 µg; 0.33 nmol) (Left to right). Representative NIRF images of the live mouse were obtained. Upon necropsy of the mouse, NIRF images of the harvested tissues and select tissues of interest – tumor (T), liver (Liv), spleen (Sp), and ovaries (Ov) – were obtained. For the harvested tissues shown above, T = tumor;
98 H = heart; Lu = lungs; Liv = liver; Sp = spleen; St = stomach; Pa = pancreas; L.I. = large intestine; S.I. = small intestine; Ov = ovaries; Ki = kidneys; Mu = muscle; Bo = bone.
Figure 4.14. In vivo and ex vivo NIRF imaging with ssB43.13-IR800, nssB43.13-IR800, and nssmIgG-IR800 in an NSG mouse bearing an orthotopic OVCAR3 xenograft. Representative NIRF images were obtained in an NSG mouse bearing a CA125-positive OVCAR3 human ovarian orthotopic xenograft 72 hours after the intravenous administration of (A) ssB43.13-IR800 (50 µg; 0.33 nmol) and (B) nssB43.13-IR800 (50 µg; 0.33 nmol). Representative NIRF images of the live mouse were obtained. Upon necropsy of the mouse, NIRF images of the harvested tissues were obtained. For the harvested tissues shown above, T = tumor; H = heart; Lu = lungs; Liv = liver; Sp = spleen; St = stomach; Pa = pancreas; L.I. = large intestine; S.I. = small intestine; Ov = ovaries; Ki = kidneys; Mu = muscle; Bo = bone.
99 Finally, to test the translatability of the IR800-labeled B43.13 antibody, the ssB43.13-IR800 was tested on sections obtained from tumor as well as metastatic lymph node samples harvested during surgical debulking of high grade serous ovarian patients at Memorial Sloan Kettering
Cancer Center (Figure 4.15). Manual immunohistochemical analysis of adjacent sections of these tissues with unmodified B43.13 antibody and the ssB43.13-IR800 construct versus a commercially available anti-CA125 antibody optimized for automated staining on a Ventana instrument revealed identical staining patterns for CA125 expression using all 3 antibody variants. Importantly, tissues from patients with CA125-negative tumors (clear cell carcinoma of the ovary) stained negative with all three antibodies used in this comparative assay. Together, these results suggest minimal if any loss in the ability of the B43.13 antibody to bind and identify CA125-expressing cells in human patients – thus making a strong case for clinical translation.
100
Figure 4.15. Histopathologic analysis of tumor and lymph node Samples from high grade serous ovarian cancer patient. Immunohistochemical staining of CA125 – a tumor biomarker for high-grade serous ovarian cancer – in tumor and lymph node samples obtained from a patient who underwent surgical debulking to remove peritoneal tumor masses and metastatic lymph nodes. A high concordance was observed in the staining patterns when equivalent dilutions of unmodified B43.13, B43.13-IR800, and a commercially available anti-CA125 antibody (Novus Biologicals, NBP1-96619) were employed for immunohistochemistry. These data clearly demonstrate the ability of B43.13 to bind human CA125+ HGSOC lesions and underscore that the conjugation of IR800 to the antibody to create B43.13-IR800 did not impair the immunoconjugate’s ability to bind CA125-expressing tissues.
101 4.4. Conclusion
In this investigation, we have demonstrated that ssB43.13-IR800 can be used to image
CA125-expressing high-grade serous ovarian cancer (HGSOC) tumors in subcutaneous, orthotopic, and patient-derived xenograft mouse models. In the short term, we believe that the clinical translation of ssB43.13-IR800 could have a great impact on the health of patients with
HGSOC. A molecularly-targeted imaging agent — such as ssB43.13-IR800 — capable of identifying small volume macroscopic or microscopic disease during primary and interval cytoreductive surgery will facilitate: (a) the accurate assessment of residual disease, (b) the discrimination between active disease versus radiological (CT/MRI) artifacts due to neoadjuvant chemo/radiation treatment, and (c) the targeted resection of metastatic abdomino-pelvic lymph nodes to limit perioperative morbidity such as lower extremity lymphedema by sparing healthy nodal tissue. Indeed, this technology could push the limits of the surgical treatment of ovarian cancer beyond the nearly complete ‘gross’ resection of disease.
102 4.5. References
1. Siegel, R. L.; Miller, K. D.; Jemal, A., Cancer statistics, 2019. CA Cancer J. Clin. 2019, 69 (1), 7-34. 2. van der Burg, M. E. L.; van Lent, M.; Buyse, M.; Kobierska, A.; Colombo, N.; Favalli, G.; Lacave, A. J.; Nardi, M.; Renard, J.; Pecorelli, S., The effect of debulking surgery after induction chemotherapy on the prognosis in advanced epithelial ovarian cancer. N. Engl. J. Med. 1995, 332 (10), 629-634. 3. Vergote, I.; Tropé, C. G.; Amant, F.; Kristensen, G. B.; Ehlen, T.; Johnson, N.; Verheijen, R. H. M.; van der Burg, M. E. L.; Lacave, A. J.; Panici, P. B.; Kenter, G. G.; Casado, A.; Mendiola, C.; Coens, C.; Verleye, L.; Stuart, G. C. E.; Pecorelli, S.; Reed, N. S., Neoadjuvant chemotherapy or primary surgery in stage IIIC or IV ovarian cancer. N. Engl. J. Med. 2010, 363 (10), 943-953. 4. Chi, D. S.; Liao, J. B.; Leon, L. F.; Venkatraman, E. S.; Hensley, M. L.; Bhaskaran, D.; Hoskins, W. J., Identification of prognostic factors in advanced epithelial ovarian carcinoma. Gynecol. Oncol. 2001, 82 (3), 532-537. 5. Préfontaine, M.; Gelfand, A. T.; Donovan, J. T.; Powell, J. L., Reproducibility of tumor measurements in ovarian cancer: A study of interobserver variability. Gynecol. Oncol. 1994, 55 (1), 87-90. 6. Sala, E.; Mannelli, L.; Yamamoto, K.; Griffin, M.; Griffin, N.; Grant, L.; Parker, R.; Crawford, R., The value of postoperative/preadjuvant chemotherapy computed tomography in the management of patients with ovarian cancer. Int. J. Gynecol. Cancer 2011, 21 (2), 296. 7. Lakhman, Y.; Akin, O.; Sohn, M. J.; Zheng, J.; Moskowitz, C. S.; Iyer, R. B.; Barakat, R. R.; Sabbatini, P. J.; Chi, D. S.; Hricak, H., Early postoperative CT as a prognostic biomarker in patients with advanced ovarian, tubal, and primary peritoneal cancer deemed optimally debulked at primary cytoreductive surgery. AJR Am. J. Roentgenol. 2012, 198 (6), 1453-1459. 8. Mutch, D. G.; Prat, J., 2014 FIGO staging for ovarian, fallopian tube and peritoneal cancer. Gynecol. Oncol. 2014, 133 (3), 401-404. 9. Davis, H. M.; Zurawski, V. R.; Bast, R. C.; Klug, T. L., Characterization of the CA 125 Antigen associated with human epithelial ovarian carcinomas. Cancer Res. 1986, 46 (12 Part 1), 6143. 10. Yin, B. W. T.; Dnistrian, A.; Lloyd, K. O., Ovarian cancer antigen CA125 is encoded by the MUC16 mucin gene. Int. J. Cancer 2002, 98 (5), 737-740. 11. Felder, M.; Kapur, A.; Gonzalez-Bosquet, J.; Horibata, S.; Heintz, J.; Albrecht, R.; Fass, L.; Kaur, J.; Hu, K.; Shojaei, H.; Whelan, R. J.; Patankar, M. S., MUC16 (CA125): Tumor biomarker to cancer therapy, a work in progress. Mol. Cancer 2014, 13 (1), 129. 12. Sharma, S. K.; Wuest, M.; Wang, M.; Glubrecht, D.; Andrais, B.; Lapi, S. E.; Wuest, F., Immuno-PET of epithelial ovarian cancer: Harnessing the potential of CA125 for non-invasive imaging. EJNMMI Res. 2014, 4 (1), 60. 13. Sharma, S. K.; Pourat, J.; Carlin, S.; Abdel-Atti, D.; Bankovich, A.; Sisodiya, V.; Liu, D.; Rudin, C.; Dylla, S.; Zeglis, B.; Poirier, J.; Lewis, J., A DLL3-targeted theranostic for small cell lung cancer: Imaging a low density target with a site-specifically modified radioimmunoconjugate. J. Nucl. Med. 2016, 57 (supplement 2), 50. 14. Rao, T. D.; Fernández-Tejada, A.; Axelrod, A.; Rosales, N.; Yan, X.; Thapi, S.; Wang, A.; Park, K. J.; Nemieboka, B.; Xiang, J.; Lewis, J. S.; Olvera, N.; Levine, D. A.; Danishefsky, S. J.; Spriggs, D. R., Antibodies against specific MUC16 glycosylation sites inhibit ovarian cancer growth. ACS Chem. Biol. 2017, 12 (8), 2085-2096.
103 15. Sharma, S. K.; Nemieboka, B.; Sala, E.; Lewis, J. S.; Zeglis, B. M., Molecular imaging of ovarian cancer. J. Nucl. Med. 2016, 57 (6), 827-833. 16. Phase Ib/IIa trial to evaluate oregovomab and nivolumab in epithelial cancer of ovarian, tubal or peritoneal origin (ORION-01). https://clinicaltrials.gov/ct2/show/NCT03100006 (accessed November 18, 2019). 17. McQuarrie, S. A.; Baum, R. P.; Niesen, A.; Madiyalakan, R.; Korz, W.; Sykes, T. R.; Sykes, C. J.; Hör, G.; McEwan, A. J.; Noujaim, A. A., Pharmacokinetics and radiation dosimetry of 99Tcm-labelled monoclonal antibody B43.13 in ovarian cancer patients. Nucl. Med. Commun. 1997, 18 (9), 878-886. 18. Schultes, B. C.; Baum, R. P.; Niesen, A.; Noujaim, A. A.; Madiyalakan, R., Anti-idiotype induction therapy: Anti-CA125 antibodies (Ab3) mediated tumor killing in patients treated with Ovarex mAb B43.13 (Ab1). Cancer Immunol. Immunother. 1998, 46 (4), 201-212. 19. Noujaim, A. A.; Schultes, B. C.; Baum, R. P.; Madiyalakan, R., Induction of CA125-specific B and T cell responses in patients injected with MAb-B43.13—Evidence for antibody-mediated antigen-processing and presentation of CA125 in vivo. Cancer Biother. Radiopharm. 2001, 16 (3), 187-203. 20. Gordon, A. N.; Schultes, B. C.; Gallion, H.; Edwards, R.; Whiteside, T. L.; Cermak, J. M.; Nicodemus, C. F., CA125- and tumor-specific T-cell responses correlate with prolonged survival in oregovomab-treated recurrent ovarian cancer patients. Gynecol. Oncol. 2004, 94 (2), 340-351. 21. The safety and antitumor activity of the combination of oregovomab and hiltonol in recurrent advanced ovarian cancer. https://clinicaltrials.gov/ct2/show/NCT03162562 (accessed November 18, 2019). 22. A controlled study of the effectiveness of oregovomab (antibody) plus chemotherapy in advanced ovarian cancer. https://clinicaltrials.gov/ct2/show/NCT01616303 (accessed November 18, 2019). 23. Sharma, S. K.; Sevak, K. K.; Monette, S.; Carlin, S. D.; Knight, J. C.; Wuest, F. R.; Sala, E.; Zeglis, B. M.; Lewis, J. S., Preclinical 89Zr immuno-PET of high-grade serous ovarian cancer and lymph node metastasis. J. Nucl. Med. 2016, 57 (5), 771-776. 24. Adumeau, P.; Sharma, S. K.; Brent, C.; Zeglis, B. M., Site-specifically labeled immunoconjugates for molecular imaging—Part 1: Cysteine residues and glycans. Mol. Imaging Biol. 2016, 18 (1), 1-17. 25. Zeglis, B. M.; Davis, C. B.; Aggeler, R.; Kang, H. C.; Chen, A.; Agnew, B. J.; Lewis, J. S., Enzyme-mediated methodology for the site-specific radiolabeling of antibodies based on catalyst-free click chemistry. Bioconjug. Chem. 2013, 24 (6), 1057-1067. 26. Zeglis, B. M.; Davis, C. B.; Abdel-Atti, D.; Carlin, S. D.; Chen, A.; Aggeler, R.; Agnew, B. J.; Lewis, J. S., Chemoenzymatic strategy for the synthesis of site-specifically labeled immunoconjugates for multimodal PET and optical imaging. Bioconjug. Chem. 2014, 25 (12), 2123-2128. 27. Dobrenkov, K.; Cheung, N.-K. V., GD2-targeted immunotherapy and radioimmunotherapy. Semin. Oncol. 2014, 41 (5), 589-612. 28. Houghton, J. L.; Zeglis, B. M.; Abdel-Atti, D.; Aggeler, R.; Sawada, R.; Agnew, B. J.; Scholz, W. W.; Lewis, J. S., Site-specifically labeled CA19.9-targeted immunoconjugates for the PET, NIRF, and multimodal PET/NIRF imaging of pancreatic cancer. Proc. Natl. Acad. Sci. U.S.A. 2015, 112 (52), 15850. 29. Adumeau, P.; Carnazza, K. E.; Brand, C.; Carlin, S. D.; Reiner, T.; Agnew, B. J.; Lewis, J. S.; Zeglis, B. M., A pretargeted approach for the multimodal PET/NIRF imaging of colorectal
104 cancer. Theranostics 2016, 6 (12), 2267-2277. 30. Cook, B. E.; Adumeau, P.; Membreno, R.; Carnazza, K. E.; Brand, C.; Reiner, T.; Agnew, B. J.; Lewis, J. S.; Zeglis, B. M., Pretargeted PET imaging using a site-specifically labeled immunoconjugate. Bioconjug. Chem. 2016, 27 (8), 1789-1795. 31. Kristensen, L. K.; Christensen, C.; Jensen, M. M.; Agnew, B. J.; Schjoth-Frydendahl, C.; Kjaer, A.; Nielsen, C. H., Site-specifically labeled 89Zr-DFO-trastuzumab improves immunoreactivity and tumor uptake for immunoPET in a subcutaneous HER2-positive xenograft mouse model. Theranostics 2019, 9 (15), 4409-4420. 32. Alvarez, V. L.; Wen, M. L.; Lee, C.; Lopes, A. D.; Rodwell, J. D.; McKearn, T. J., Site- specifically modified 111In labelled antibodies give low liver backgrounds and improved radioimmunoscintigraphy. Nucl. Med. Biol. 1986, 13 (4), 347-352. 33. Tavare, R.; Wu, W. H.; Zettlitz, K. A.; Salazar, F. B.; McCabe, K. E.; Marks, J. D.; Wu, A. M., Enhanced immunoPET of ALCAM-positive colorectal carcinoma using site-specific 64Cu- DOTA conjugation. Protein Eng. Des. Sel. 2014, 27 (10), 317-24. 34. Tinianow, J. N.; Gill, H. S.; Ogasawara, A.; Flores, J. E.; Vanderbilt, A. N.; Luis, E.; Vandlen, R.; Darwish, M.; Junutula, J. R.; Williams, S. P.; Marik, J., Site-specifically 89Zr- labeled monoclonal antibodies for ImmunoPET. Nucl. Med. Biol. 2010, 37 (3), 289-97.
105 Chapter 5 : Conclusions
This work demonstrates how antibodies can be leveraged for positron emission tomography
(PET) and near-infrared fluorescence imaging of breast and ovarian cancers. Taking advantage of the specificity and affinity of antibodies for its antigen, we were able to explore the effects of a glycans-based site-specific bioconjugation approach on in vivo performance and develop two imaging agents for ovarian cancer.
In chapter 2, we investigated how the glycans-based site-specific bioconjugation of antibodies can impact FcRI binding and in vivo performance. We demonstrated that this approach results in the attenuation of binding to murine and human FcRI and improves in vivo behavior in certain immunocompromised mouse models of human breast cancer, particularly the humanized
NSG mouse model. An upcoming clinical trial will help evaluate whether this work is applicable to human immunoPET.
In chapter 3, we developed the PET imaging agent, [89Zr]Zr-DFO-AR20.5 and evaluated its in vivo behavior in mice bearing subcutaneous and orthotopic human ovarian cancer xenografts.
Our work clearly demonstrated that [89Zr]Zr-DFO-AR20.5 could be used to clearly delineate the
MUC1-expressing tumor. We plan to explore the use of [89Zr]Zr-DFO-AR20.5 for both diagnostic and theranostic imaging of ovarian cancer in the future.
In chapter 4, we synthesized the NIRF imaging agent, ssB43.13-IR800 and evaluated its ability to image CA125-expressing high-grade serous ovarian cancer (HGSOC) tumors in subcutaneous, orthotopic, and patient-derived xenograft mouse models. Due to its ability to clearly assess residual disease and delineate lymph nodes and tumor tissue, we believe it has potential for image-guided surgery in the near future. Taken together, these data indicate the utility of antibodies for medical imaging applications and clinical potential of our work.
106
107 APPENDICES
CHAPTER 2: APPENDIX
MALDI-ToF spectrum for native pertuzumab (Figure A.2.1)
MALDI-ToF spectrum for DFO-nsspertuzumab (Figure A.2.2)
MALDI-ToF spectrum for N3-sspertuzumab-βGal (Figure A.2.3)
MALDI-ToF spectrum for DFO-sspertuzumab-βGal (Figure A.2.4)
MALDI-ToF spectrum for N3-sspertuzumab-EndoS (Figure A.2.5)
MALDI-ToF spectrum for DFO-sspertuzumab-EndoS (Figure A.2.6)
SDS-PAGE of the immunoconjugates before and after deglycosylation (Figure A.2.7)
Size exclusion chromatogram of native pertuzumab (Figure A.2.8)
Size exclusion chromatogram of DFO-nsspertuzumab (Figure A.2.9)
Size exclusion chromatogram of DFO-sspertuzumab-βGal (Figure A.2.10)
Size exclusion chromatogram of DFO-sspertuzumab-EndoS (Figure A.2.11)
SPR sensorgrams of the binding for recombinant HER2 (Figure A.2.12)
Flow cytometry data with HER2-expressing BT474 cells (Figure A.2.13)
Stability of the three radioimmunoconjugates in human serum (Figure A.2.14)
Degree of labeling of each immunoconjugate as determined via MALDI-MS (Table A.1)
Biodistribution data in nude mice (Table A.2)
Tumor-to-tissue activity concentration ratios in nude mice (Table A.3)
Biodistribution data in humanized NSG mice (Table A.4)
Tumor-to-tissue activity concentration ratios in humanized NSG mice (Table A.5)
108 x104
74100.7 Intens. [a.u.] Intens.
148109.5
4
3
2
1
98906.9 124655.0
0 60000 80000 100000 120000 140000 160000 180000 m/z Figure A.2.1. Representative MALDI-ToF spectrum for native pertuzumab.
x104 74755.4
Intens. [a.u.] Intens. 1.2
149385.0 1.0
0.8
0.6
0.4
0.2
99901.5
0.0
60000 80000 100000 120000 140000 160000 180000 m/z
Figure A.2.2. Representative MALDI-ToF spectrum for DFO-nsspertuzumab.
109
148661.5
Intens. [a.u.] 74518.4
4000
3000
2000
1000
0 60000 80000 100000 120000 140000 160000 m/z
Figure A.2.3. Representative MALDI-ToF spectrum for N3-sspertuzumab-βGal.
150684.3 Intens. [a.u.]
2500 75723.6
2000
1500
1000
500
0 60000 80000 100000 120000 140000 160000 m/z Figure A.2.4. Representative MALDI-ToF spectrum for DFO-sspertuzumab-βGal.
110
73502.1 Intens. [a.u.] Intens.
3000
147071.1
2000
1000
49039.9
98359.8
0 40000 60000 80000 100000 120000 140000 160000 180000 200000 m/z
Figure A.2.5. Representative MALDI-ToF spectrum for N3-sspertuzumab-EndoS.
73923.4 Intens. [a.u.] Intens.
5000
147737.7
4000
3000
2000
1000
98699.7
0 60000 80000 100000 120000 140000 160000 180000 m/z
Figure A.2.6. Representative MALDI-ToF spectrum for DFO-sspertuzumab-EndoS.
111
Figure A.2.7. Denaturing SDS-PAGE of (1/5) unmodified pertuzumab, (2/6) DFO-nsspertuzumab,
(3/7) DFO-sspertuzumab-βGal, and (4/8) DFO-sspertuzumab-EndoS before and after (before/after) deglycosylation. After deglycosylation, all the heavy glycan chains have the same molecular weight, except DFO-sspertuzumab-EndoS, because PNGaseF is unable to digest the immunoconjugate’s truncated glycans. The red dotted line denotes the mass of the heavy chain glycans of native pertuzumab.
112
Figure A.2.8. Size exclusion chromatogram of native pertuzumab
Figure A.2.9. Size exclusion chromatogram of DFO-nsspertuzumab.
113
Figure A.2.10. Size exclusion chromatogram of DFO-sspertuzumab-βGal.
Figure A.2.11. Size exclusion chromatogram of DFO-sspertuzumab-EndoS.
114
Figure A.2.12. SPR sensorgrams for the binding of (A) native pertuzumab, (B) DFO- nsspertuzumab, (C) DFO-sspertuzumab-EndoS, and (D) DFO-sspertuzumab-βGal for recombinant HER2.
115
Figure A.2.13. Flow cytometry data showing the binding of the immunoconjugates to HER2- expressing BT474 human breast cancer cells.
116
Figure A.2.14. Stability of the three radioimmunoconjugates to demetallation in human serum at 37°C. Measurements were analyzed using radio-iTLC and performed in triplicate.
117
Degree of Labeling Constructs Average mass (Da) (DFO/mAb) pertuzumab 148264 ± 110 1.4 ± 0.4 DFO-nsspertuzumab 149300 ± 185 N3-sspertuzumab-βGal 148645 ± 12 2.6 ± 0.1 DFO-sspertuzumab-βGal 150574 ± 86 N3-sspertuzumab-EndoS 147145 ± 59 1.3 ± 0.2 DFO-sspertuzumab-EndoS 146998 ± 60
Table A.2.1. Degree of labeling of each immunoconjugate as determined via MALDI-MS.
118
0.53 3.00 2.12 0.79 1.21 1.17 0.27 0.32 0.15 1.57 0.16 0.96
144 144 h 28.67
1.78± 4.86± 4.70± 2.84± 2.47± 1.77± 0.54± 0.63± 0.48± 0.49± 6.50± 4.09±
76.68±
EndoS
-
96 96 h
23.38
64.43 ±64.43
1.75 ±1.75 0.34 ±3.59 2.50 2.12 ±2.12 0.77 ±3.55 1.10 ±2.47 1.02 ±0.67 0.38 ±0.78 0.24 ±0.92 0.22 ±4.25 1.24 ±0.51 0.14 ±4.06 1.32
±6.12 2.11
pertuzumab
ss
48 48 h
18.38 -
±54.64
1.91 ±1.91 0.62 ±5.06 2.37 3.02 ±3.02 1.30 ±4.14 1.61 ±2.43 0.65 ±0.90 0.31 ±1.03 0.35 ±0.73 0.22 ±4.49 1.16 ±0.91 0.23 ±3.00 0.69
±8.33 3.31
DFO
-
Zr
9.12
24 24 h
89
35.60 ±35.60
1.87 ±1.87 0.14 ±3.77 0.66 3.39 ±3.39 0.47 ±4.13 0.65 ±1.96 0.08 ±0.70 0.27 ±1.13 0.11 ±0.99 0.71 ±3.63 0.41 ±1.08 0.31 ±1.81 0.60
±9.84 0.84
32.99
144 h
77.15 ±77.15
1.99 ±1.99 0.43 ±5.38 2.56 1.86 ±1.86 0.30 ±3.72 0.97 ±5.22 2.04 ±0.53 0.15 ±0.63 0.05 ±0.66 0.16 ±5.72 1.21 ±0.71 0.34 ±2.68 0.66
±5.77 1.53
βGal
SI = small intestine; LI = large intestine. The
-
.
96 96 h 10.67
±61.05
1.94 ±1.94 0.59 ±4.78 1.58 2.75 ±2.75 0.94 ±4.36 0.88 ±4.92 2.21 ±0.62 0.16 ±1.03 0.19 ±0.82 0.07 ±5.40 0.54 ±0.84 0.29 ±2.24 0.44
±7.86 1.36
pertuzumab
ss
-
48 h
2.11 ±2.11 0.65 ±4.51 1.97 8.18 ±8.18 3.29 ±2.77 1.18 ±3.79 1.53 ±3.56 1.26 ±0.78 0.44 ±0.98 0.36 ±0.90 0.20 ±4.10 0.90 ±0.76 0.21 ±1.78 0.50
47.88 ±47.88 14.39 DFO
-
Zr
89
radioimmunoconjugate
0.95 2.10
24 24 h 31.60 ±31.60
±10.41
1.71 ±1.71 0.21 ±4.64 0.59 4.09 ±4.09 0.55 ±2.51 0.33 ±0.79 0.35 ±1.11 0.20 ±1.02 0.47 ±3.81 0.35 ±1.08 0.11 ±1.28 0.41
±3.88 0.75
144 144 h 52.16
93.15 ±93.15
1.68 ±1.68 0.53 ±2.55 0.94 1.90 ±1.90 0.80 ±2.90 0.98 ±4.50 2.38 ±0.45 0.18 ±0.63 0.27 ±0.69 0.15 ±3.47 0.85 ±0.60 0.21 ±8.42 2.36 ±4.88 2.09
h
96 96
10.54
41.30 ±41.30
1.96 ±1.96 0.97 ±3.15 1.07 1.89 ±1.89 0.71 ±3.06 1.01 ±4.10 1.19 ±0.63 0.22 ±0.88 0.42 ±0.96 0.43 ±3.56 0.59 ±0.60 0.14 ±5.25 3.39
±5.42 2.16 pertuzumab
nss
include contents. include
-
0.22 3.32
48 h
DFO
10.54
45.36 ±45.36
-
2.41± 1.05 4.44± 0.55 2.84± 1.11 3.65± 1.04 3.86± 1.76 0.77± 1.22± 0.51 1.40± 0.39 3.52± 0.64 0.90± 0.32 3.94±
7.17± 2.68
Zr
89
± ±
Biodistribution Biodistribution data for the three constructs in nude mice bearing BT474 subcutaneous xenografts. The values
± 0.34 ± 0.37 ± 0.72 ± 0.85 ± 0.65 ± 0.50 ± 0.71 ± 0.19 ± 0.31 ± 0.12 ± 0.41 ± 0.48
1.84 24 24 h
.
30.99
2
2.04 3.73 3.11 4.40 3.37 1.35 1.21 1.47 3.74 0.81 1.87
9.22
SI
LI
Skin
Bone
Lung
Liver
Blood Heart
Spleen
Tumor
Muscle Kidney
Stomach
are are %ID/g ± SD; n = 5 mice for each time point and stomach, SI, and LI values LI SI, and stomach, Table A.2.
119
24.95 16.66 19.78 24.22 84.49 86.03 94.37 71.93 14.03
144h
27.04± 43.06± 31.02± 43.33± 18.75±
120.89± 161.07± 157.71±
143.29±
15.78 ±15.78 7.06 ±11.81 4.18
±16.31 9.55
EndoS
-
96 h 96
15.62 14.27 15.14 64.52 39.71 30.47 57.76 14.14
30.43 ±30.43 ±26.05 ±36.71 ±95.78 ±82.89 ±70.36 ±17.97
±126.32
18.17 ±18.17 8.69 ±15.15 7.06 ±15.86 7.71
±10.52 5.27
pertuzumab
9.88
ss -
SI = small intestine; LI =
48 h 48
13.29 29.09 25.39 34.16 24.99
.
28.55 ±28.55 ±60.40 ±52.87 ±74.96 ±59.73
±6.56 3.42
18.10 ±18.10 ±13.20 6.80 ±22.51 9.66 ±12.16 5.15 ±18.24 7.42 ±10.79 6.23 DFO
-
Zr
89
4.73 5.08 8.66 8.29
3.06
24 h 24
23.73 27.37 12.79
10.51 ±10.51 ±18.20 ±19.08 ±51.06 ±31.63 ±35.97 ±32.99 ±19.71
8.62 ±8.62 2.59 ±9.79 2.74 ±9.44 2.93
±3.62 0.98
19.02 10.39 18.51 75.09 53.13 58.16 70.84 14.17
h 144
20.73 ±20.73 41.58 ±41.58 ±38.71 ±28.75
146.09 ±146.09 ±121.94 ±117.47 ±109.22
13.37 ±13.37 6.73 ±14.79 8.58 ±13.49 6.44 ±14.34 9.18
βGal
- radioimmunoconjugate
2.27 8.50 3.75 5.98 7.23 4.78
96 h 96
11.12 30.03 14.88 14.36 28.14
14.01± 12.40± 31.53± 98.09± 59.25± 74.19± 11.30± 72.85± 27.31± 12.78± 22.19±
±7.76 1.91
pertuzumab
ss
-
4.34 6.38 6.26 9.77 5.62
9.02
48 h 48
39.57 22.90 19.72 25.70 11.11
12.65 ±12.65 17.30 ±17.30 ±13.46 ±22.66 ±61.42 ±48.68 ±52.97 ±11.67 ±63.04 ±26.93 ±10.61
±5.86 2.94
ents.
DFO
-
Zr
z
89
ratios ratios for the three constructs in nude mice bearing BT474 subcutaneous
1.85 2.59 3.64 8.06
24 h 24
5.54
17.87 14.45
18.47 ±18.47 ±40.16 ±28.43 ±31.09 ±29.13 ±24.72
±12.60
8.14 ±8.14 1.67 ±7.73 1.16 ±8.30 0.95 ±6.81 0.97
±3.04 0.34
16.41 13.47 34.41 21.04 15.94 35.63 81.33 24.46
144 h 144 142.31 103.77 104.07
32.14 ±32.14 19.10 ±19.10 ±49.05 ±20.68 ±55.54 ±26.84 ±36.52
147.32 ±147.32 ±135.04 ±156.35 ±206.49
±11.06 6.93
3.61
3.53 5.65 3.89 5.58
9.95
96 h 96
11.71 28.60 25.34 22.25 24.18
13.50 ±13.50 21.84 ±21.84 ±10.07 ±21.05 ±66.01 ±46.97 ±42.93 ±11.59 ±69.04 ±13.11
7.62 ±7.62 ±7.86 5.46 activity activity concentration
pertuzumab
nss
-
tissue
3.80 7.24 4.56 6.02 9.28 2.70
48 h 48
21.81 17.98 11.70 21.37 10.07
-
DFO
12.44 ±12.44 11.75 ±11.75 ±18.79 ±58.67 ±37.31 ±32.33 ±12.90 ±50.43 ±11.51 ±10.22
±15.95
-
6.32 ±6.32 2.78 to
-
Zr
89
2.67 4.26 4.66 4.32
24 h 24
12.14 19.36
Tumor
15.17 ±15.17 ±22.98 ±25.69 ±21.15 ±38.25 ±16.55
. .
9.97 ±9.97 2.79 ±7.05 1.12 ±9.19 1.48 ±8.28 0.56 ±8.31 0.95 ±3.36 0.33
3
SI
LI
Skin
Bone
Lung
Liver
Blood Heart
Spleen
Muscle
Kidney Tumor/
Stomach
xenografts. xenografts. The values are %ID/g ± SD; n = 5 mice for each time point and Table Table A.2. values include LI SI, stomach, and cont intestine. The large
120
89Zr-DFO- 89Zr-DFO- 89Zr-DFO-
nsspertuzumab sspertuzumab-βGal sspertuzumab-EndoS 144 h 144 h 144 h Blood 1.03 ± 0.63 4.51 ± 3.00 8.76 ± 2.31 Tumor 46.13 ± 16.46 77.53 ± 12.56 111.84 ± 39.88 Heart 0.90 ± 0.04 1.76 ± 0.78 2.56 ± 0.23 Lung 2.40 ± 0.50 4.41 ± 1.49 6.55 ± 1.07 Liver 10.25 ± 2.74 8.87 ± 3.53 4.71 ± 0.77 Spleen 44.77 ± 7.92 22.39 ± 17.37 13.09 ± 3.99 Stomach 0.74 ± 0.22 1.08 ± 0.78 0.88 ± 0.25 SI 2.69 ± 0.53 2.44 ± 0.62 1.99 ± 0.49 LI 1.00 ± 0.07 0.92 ± 0.44 1.20 ± 0.73 Kidney 3.94 ± 1.97 5.44 ± 0.65 4.43 ± 0.72 Muscle 0.63 ± 0.17 0.69 ± 0.11 0.90 ± 0.18 Bone 9.14 ± 3.05 5.84 ± 2.11 7.24 ± 2.04 Skin 2.44 ± 0.29 5.34 ± 3.57 4.64 ± 2.20
Table A.2.4. Biodistribution data for the three radioimmunoconjugates at 144 h post-injection in humanized NSG mice bearing BT474 subcutaneous xenografts. The values are %ID/g ± SD; n =
4 mice for 89Zr-DFO-nsspertuzumab, n = 5 mice for 89Zr-DFO-sspertuzumab-βGal, and n = 5 mice for 89Zr-DFO-sspertuzumab-EndoS. SI = small intestine; LI = large intestine. The stomach, SI, and LI values include contents.
121
89Zr-DFO- 89Zr-DFO- 89Zr-DFO-
nsspertuzumab sspertuzumab-βGal sspertuzumab-EndoS Tumor/ 144 h 144 h 144 h Blood 44.91± 19.91 17.21 ± 11.78 12.77 ± 5.66 Heart 51.30 ± 18.42 43.95 ± 20.71 43.73 ± 16.08 Lung 19.21 ± 7.93 17.57 ± 6.58 17.09 ± 6.70 Liver 4.50 ± 2.01 8.74 ± 3.75 23.72 ± 9.30 Spleen 1.03 ± 0.41 3.46 ± 2.75 8.54 ± 4.01 Stomach 62.14 ± 28.82 71.89 ± 53.04 127.34 ± 57.69 SI 17.13 ± 6.98 31.84 ± 9.60 56.26 ± 24.41 LI 46.08 ± 16.79 84.15 ± 42.52 93.44 ± 65.84 Kidney 11.72 ± 7.20 14.26 ± 2.87 25.23 ± 9.90 Muscle 73.33 ± 32.53 112.22 ± 25.95 124.41 ± 50.67 Bone 5.05 ± 2.47 13.28 ± 5.27 15.45 ± 7.02 Skin 18.90 ± 7.11 14.53 ± 10.01 24.12 ± 14.33
Table A.2.5. Tumor-to-tissue activity concentration ratios for the three radioimmunoconjugates at 144 h post-injection in humanized NSG mice bearing BT474 subcutaneous xenografts. The values are %ID/g ± SD; n = 4 mice for 89Zr-DFO-nsspertuzumab, n = 5 mice for 89Zr-DFO- sspertuzumab-βGal, and n = 5 mice for 89Zr-DFO-sspertuzumab-EndoS. SI = small intestine; LI = large intestine. The stomach, SI, and LI values include contents.
122
CHAPTER 3: APPENDIX
Tumor-to-background activity concentration ratios from nude mice bearing SKOV3 subcutaneous xenografts (Table A.3.1)
Tumor-to-organ activity concentration ratios from nude mice bearing SKOV3-RFluc orthotopic xenografts (Table A.3.2)
Biodistribution data for the metastatic lesions collected from nude mice bearing SKOV3-RFluc orthotopic xenografts (Table A.3.3)
123
24 h 72 h 72 h block 120 h Blood 0.8 ± 0.3 1.9 ± 0.6 0.7 ± 0.3 3.6 ± 1.2 Heart 2.9 ± 1.2 7.1 ± 2.2 2.3 ± 1.0 12.5 ± 4.3 Lungs 1.4 ± 0.6 3.9 ± 1.6 1.3 ± 0.7 7.3 ± 2.5 Liver 1.8 ± 0.8 3.7 ± 1.0 1.2 ± 0.7 5.4 ± 2.0 Spleen 2.7 ± 1.1 4.8 ± 1.9 2.0 ± 1.0 6.9 ± 2.5 Stomach 11.5 ± 5.4 21.9 ± 8.2 8.7 ± 4.4 35.7 ± 14.1 Small Intestine 6.6 ± 2.5 15.7 ± 5.2 6.4 ± 3.0 27.7 ± 9.9 Large Intestine 10.0 ± 3.6 24.8 ± 9.1 9.8 ± 4.4 43.6 ± 18.3 Kidneys 1.8 ± 1.0 4.9 ± 1.6 1.7 ± 0.8 7.5 ± 2.6 Muscle 9.3 ± 3.5 24.6 ± 8.0 7.4 ± 3.5 42.7 ± 14.6 Bone 4.8 ± 1.8 6.1 ± 2.0 2.1 ± 0.9 6.6 ± 3.0
Table A.3.1. Tumor-to-background activity concentration ratios calculated using the biodistribution data from athymic nude mice (n = 5 per time point) bearing SKOV3 human ovarian cancer xenografts collected 24, 72, and 120 h after the intravenous administration of [89Zr]Zr- DFO-AR20.5 (0.65 – 0.69 MBq; 17.6 – 18.6 μCi; 6.6 – 7.0 μg). For the 72 h blocking experiment, the mice were administered the same amount of [89Zr]Zr-DFO-AR20.5 mixed with a excess of unmodified AR20.5 (~500 μg per mouse). Stomach, small intestine, and large intestine values include contents.
124
[89Zr]Zr-DFO-AR20.5 [89Zr]Zr-DFO-mIgG Blood 4.6 ± 4.9 0.7 ± 0.3 Heart 12.9 ± 12.6 2.5 ± 0.8 Lungs 7.4 ± 8.0 1.6 ± 0.6 Liver 1.1 ± 0.7 0.6 ± 0.2 Spleen 1.9 ± 1.2 0.7 ± 0.3 Stomach 33.2 ± 23.5 11.9 ± 5.1 Small Intestine 30.4 ± 19.7 6.5 ± 2.8 Large Intestine 27.5 ± 25.9 6.8 ± 2.4 Kidneys 5.3 ± 3.5 0.6 ± 0.2 Muscle 24.6 ± 19.3 5.1 ± 2.2 Bone 6.6 ± 4.2 1.8 ± 0.6
Table A.3.2. Tumor-to-organ activity concentration ratios derived from the biodistribution data from athymic nude mice bearing SKOV3-RFluc human ovarian cancer orthotopic xenografts collected 120 h after the intravenous administration of [89Zr]Zr-DFO-AR20.5 (5.6 – 6.0 MBq; 150.6 – 161.4 μCi; 59.1 – 63.3 μg; n = 3) or [89Zr]Zr-DFO-mIgG (6.1 – 6.4 MBq; 163.7 – 171.7 μCi; 61.5 – 70.1 μg; n = 3) via the tail vein. The stomach, small intestine, and large intestine values include contents.
125
[89Zr]Zr-DFO-AR20.5 Abdominal Met 9.32 Mouse 1 Peritoneal Met 1 6.53 Peritoneal Met 2 7.44 Mouse 2 Peritoneal Met 28.99 Peritoneal Met 5.22 Hepatic Met 23.02 Mouse 3 Right Ovarian Met 1 13.50 Right Ovarian Met 2 5.48
Table A.3.3. Biodistribution data for the metastatic lesions collected from athymic nude mice bearing SKOV3-RFluc human ovarian cancer orthotopic xenografts 120 h after the intravenous administration of [89Zr]Zr-DFO-AR20.5 (5.6 – 6.0 MBq; 150.6 – 161.4 μCi; 59.1 – 63.3 μg; n = 3) via the tail vein. The values are %ID/g ± SD; Met = metastatic lesion.
126