TU Darmstadt / ULB / TUbiblio

Novel technologies for antibody hit discovery and engineering of antibody-like proteins with therapeutic relevance

Pekar, Lukas (2021):
Novel technologies for antibody hit discovery and engineering of antibody-like proteins with therapeutic relevance. (Publisher's Version)
Darmstadt, Technische Universität,
DOI: 10.26083/tuprints-00017513,
[Ph.D. Thesis]

Abstract

In 1975, scientists Köhler and Milstein successfully fused B-lymphocytes from immunized mice with myeloma cells and, through this procedure entitled hybridoma technology, were able to combine the advantages of both precursor cells. These immortalized cells are culturable and able to secrete antigen-specific antibodies. The scientists' invention, revolutionary for the production of monoclonal antibodies, was awarded with a Nobel Prize in 1984. It became apparent, however, that the hybridoma technology is a lengthy process with low efficiency. In addition, the murine origin of the antibodies causes a potential problem with immunogenicity when applied to humans. Several new technologies have been developed in the last decades in order to overcome these limitations. These technologies include the use of human naïve, semi-synthetic or synthetic antibody diversities (to avoid the immunogenicity issue) in combination with cellular or non-cellular in vitro selection systems for the presentation and isolation of antibodies, such as Phage or Yeast Display. By the invention of these methods, large repertoires could be examined for predefined characteristics, thus enabling the isolation of specific molecules, which advanced today's successful use of monoclonal antibodies in biotechnological and medical applications. Despite all improvements in the finding process of antibodies, the generation of initial display libraries is still complex and laborious. It is often a multistage process which, as with Yeast Display, encompasses multiple cloning steps, the generation of separated libraries for heavy and light chains as well as, ultimately, their mating. This process can be simplified and condensed to only one reaction through the application of Golden Gate Cloning (GGC). The cloning reaction within GGC is based on the use of Type IIs restriction enzymes, which digest DNA in a defined distance from their recognition sequences and, thereby, allow for the incorporation of overhangs with requested signatures. Type IIs restriction enzymes enable a targeted, single-stage cloning process in which all recognition sequences are removed during the reaction, thereby causing an equilibrium shift towards the specific cloning product. As part of this work, the feasibility of GGC for the generation of Yeast and Phage Display libraries could be demonstrated for three different selection campaigns. Specific Fab antibodies against CEACAM6, EGFR and hCG were isolated from the immune repertoires of transgenic rats and wild type chickens (Yeast Display), and EGFR-specific scFv and VHH antibodies were isolated from the immune repertoires of chickens and camelids (Phage Display). In comparison to the traditional generation of antibody libraries, GGC could be validated with regards to gene repertoire distribution, the overall size of libraries as well as to the biophysical characteristics of isolated antibodies. The developed single-stage GGC process is able to generate antibody libraries based on combinatorial heavy and light chain diversities with the same quality as the traditional method. In addition, the process is a faster and less laborious method, which parallels the multiple steps of the original approach. Besides classical monoclonal antibodies and antibody fragments, new antibody formats can broaden the therapeutic space of new biological entities. Among these are molecules such as bispecific antibodies (bsAbs) and immunoligands, as wells as other antibody fragments and derivates - such as camelid single domain antibody Fc-fusions. In theory, four different plasmids have to be used during the expression of IgG-based bispecific antibodies in cell culture, as they consist of two different heavy chains and two different light chains. This would lead to a statistical yield of only 12.5 % of the desired assembled protein. The Strand Exchange Engineered Domain (SEED) technology is one possibility to avoid heterogeneity during the expression of several similar polypeptide chains. Through the introduction of alternating IgA and IgG β sheet structures in the CH3 domains, antiparallel immunoglobulin structures occur, which promote a heterodimerization of the heavy chains. In order to avoid heavy chain homodimerization, additional technologies, such as "knob-into-holes", controlled Fab arm exchange or targeted amino acid exchanges used to introduce contrary electrostatic charges into both heavy chains have been developed. Moreover, mispairing of antibodies light chains can be prevented by technologies such as CrossMab or the use of a Common Light Chain. Another possibility is the utilization of camelid VHH domains. VHHs, compared to canonical antibodies, provide specific antigen binding exclusively via the variable domain of the heavy chain. In this context, it was one of the goals of the present work to combine the SEED Technology with a bispecific Fab-like VHH antibody which was conceptionally described by Baty and coworkers in 2013. This combination enabled the generation of a novel bi- and trispecific, IgG-like VHH-based antibody platform with related variable valences. The results of the characterization of this antibody format with regards to its biophysical and biochemical characteristics - such as the specific NK cell-mediated ADCC - verify the versatility of the generic platform in expressing fully functional mono-, bi- and trispecific antibodies with different valences as a "plug and play" application. Furthermore, the use of bispecific molecules for a targeted NK cell recruitment is of substantial interest nowadays. The specific NK cell recruitment towards tumorous or infected cells enables a specific cytolysis of target cells and a simultaneous immune modulation through NK cell-mediated cytokine release. One possibility to address the recruitment of NK cells are the Natural Cytotoxicity Receptors (NCRs), comprehensively expressed by NK cells. For this work, and through the use of Yeast Display, the N-terminal IgV-like domain of B7 H6 (natural ligand of NKp30), which is relevant for receptor binding, could be affinity matured. Via Fluorescence Activated Cell Sorting (FACS)-based selection, B7 H6 variants with significantly increased affinity to NKp30 were achieved. Additionally, these variants showed significantly increased NK cell mediated cytotoxicities and cytokine release when used in the format of bispecific immunoligands, compared to the parental B7-H6 molecule. These results verify the underlying hypothesis that an increased affinity of B7 H6 to NKp30 results in an increased NK cell-based tumor cell cytolysis as well as in increased NK cell-mediated secretion of proinflammatory cytokines. An increased affinity of activating NK cell receptors natural ligands could, therefore, be used in the development of potential immunotherapies for the treatment of patients with various cancerous diseases.

Item Type: Ph.D. Thesis
Erschienen: 2021
Creators: Pekar, Lukas
Status: Publisher's Version
Title: Novel technologies for antibody hit discovery and engineering of antibody-like proteins with therapeutic relevance
Language: English
Abstract:

In 1975, scientists Köhler and Milstein successfully fused B-lymphocytes from immunized mice with myeloma cells and, through this procedure entitled hybridoma technology, were able to combine the advantages of both precursor cells. These immortalized cells are culturable and able to secrete antigen-specific antibodies. The scientists' invention, revolutionary for the production of monoclonal antibodies, was awarded with a Nobel Prize in 1984. It became apparent, however, that the hybridoma technology is a lengthy process with low efficiency. In addition, the murine origin of the antibodies causes a potential problem with immunogenicity when applied to humans. Several new technologies have been developed in the last decades in order to overcome these limitations. These technologies include the use of human naïve, semi-synthetic or synthetic antibody diversities (to avoid the immunogenicity issue) in combination with cellular or non-cellular in vitro selection systems for the presentation and isolation of antibodies, such as Phage or Yeast Display. By the invention of these methods, large repertoires could be examined for predefined characteristics, thus enabling the isolation of specific molecules, which advanced today's successful use of monoclonal antibodies in biotechnological and medical applications. Despite all improvements in the finding process of antibodies, the generation of initial display libraries is still complex and laborious. It is often a multistage process which, as with Yeast Display, encompasses multiple cloning steps, the generation of separated libraries for heavy and light chains as well as, ultimately, their mating. This process can be simplified and condensed to only one reaction through the application of Golden Gate Cloning (GGC). The cloning reaction within GGC is based on the use of Type IIs restriction enzymes, which digest DNA in a defined distance from their recognition sequences and, thereby, allow for the incorporation of overhangs with requested signatures. Type IIs restriction enzymes enable a targeted, single-stage cloning process in which all recognition sequences are removed during the reaction, thereby causing an equilibrium shift towards the specific cloning product. As part of this work, the feasibility of GGC for the generation of Yeast and Phage Display libraries could be demonstrated for three different selection campaigns. Specific Fab antibodies against CEACAM6, EGFR and hCG were isolated from the immune repertoires of transgenic rats and wild type chickens (Yeast Display), and EGFR-specific scFv and VHH antibodies were isolated from the immune repertoires of chickens and camelids (Phage Display). In comparison to the traditional generation of antibody libraries, GGC could be validated with regards to gene repertoire distribution, the overall size of libraries as well as to the biophysical characteristics of isolated antibodies. The developed single-stage GGC process is able to generate antibody libraries based on combinatorial heavy and light chain diversities with the same quality as the traditional method. In addition, the process is a faster and less laborious method, which parallels the multiple steps of the original approach. Besides classical monoclonal antibodies and antibody fragments, new antibody formats can broaden the therapeutic space of new biological entities. Among these are molecules such as bispecific antibodies (bsAbs) and immunoligands, as wells as other antibody fragments and derivates - such as camelid single domain antibody Fc-fusions. In theory, four different plasmids have to be used during the expression of IgG-based bispecific antibodies in cell culture, as they consist of two different heavy chains and two different light chains. This would lead to a statistical yield of only 12.5 % of the desired assembled protein. The Strand Exchange Engineered Domain (SEED) technology is one possibility to avoid heterogeneity during the expression of several similar polypeptide chains. Through the introduction of alternating IgA and IgG β sheet structures in the CH3 domains, antiparallel immunoglobulin structures occur, which promote a heterodimerization of the heavy chains. In order to avoid heavy chain homodimerization, additional technologies, such as "knob-into-holes", controlled Fab arm exchange or targeted amino acid exchanges used to introduce contrary electrostatic charges into both heavy chains have been developed. Moreover, mispairing of antibodies light chains can be prevented by technologies such as CrossMab or the use of a Common Light Chain. Another possibility is the utilization of camelid VHH domains. VHHs, compared to canonical antibodies, provide specific antigen binding exclusively via the variable domain of the heavy chain. In this context, it was one of the goals of the present work to combine the SEED Technology with a bispecific Fab-like VHH antibody which was conceptionally described by Baty and coworkers in 2013. This combination enabled the generation of a novel bi- and trispecific, IgG-like VHH-based antibody platform with related variable valences. The results of the characterization of this antibody format with regards to its biophysical and biochemical characteristics - such as the specific NK cell-mediated ADCC - verify the versatility of the generic platform in expressing fully functional mono-, bi- and trispecific antibodies with different valences as a "plug and play" application. Furthermore, the use of bispecific molecules for a targeted NK cell recruitment is of substantial interest nowadays. The specific NK cell recruitment towards tumorous or infected cells enables a specific cytolysis of target cells and a simultaneous immune modulation through NK cell-mediated cytokine release. One possibility to address the recruitment of NK cells are the Natural Cytotoxicity Receptors (NCRs), comprehensively expressed by NK cells. For this work, and through the use of Yeast Display, the N-terminal IgV-like domain of B7 H6 (natural ligand of NKp30), which is relevant for receptor binding, could be affinity matured. Via Fluorescence Activated Cell Sorting (FACS)-based selection, B7 H6 variants with significantly increased affinity to NKp30 were achieved. Additionally, these variants showed significantly increased NK cell mediated cytotoxicities and cytokine release when used in the format of bispecific immunoligands, compared to the parental B7-H6 molecule. These results verify the underlying hypothesis that an increased affinity of B7 H6 to NKp30 results in an increased NK cell-based tumor cell cytolysis as well as in increased NK cell-mediated secretion of proinflammatory cytokines. An increased affinity of activating NK cell receptors natural ligands could, therefore, be used in the development of potential immunotherapies for the treatment of patients with various cancerous diseases.

Place of Publication: Darmstadt
Collation: II, 284 Seiten
Divisions: 07 Department of Chemistry
07 Department of Chemistry > Fachgebiet Biochemie
Date Deposited: 26 Mar 2021 10:12
DOI: 10.26083/tuprints-00017513
Official URL: https://tuprints.ulb.tu-darmstadt.de/17513
URN: urn:nbn:de:tuda-tuprints-175138
Referees: Kolmar, Prof. Dr. Harald ; Neumann, Prof. Dr. Siegfried
Refereed / Verteidigung / mdl. Prüfung: 7 December 2020
Alternative Abstract:
Alternative abstract Language

Im Jahr 1975 konnten die Wissenschaftler Köhler und Milstein erfolgreich B-Lymphozyten aus immunisierten Mäusen mit Myelomzellen fusionieren und mittels dieses, als Hybridoma Technologie bezeichneten Verfahrens, Zellen erzeugen die Vorteile beider Vorgängerzellen vereinten. Diese immortalisierten Hybridomazellen sind sowohl kultivierbar, als auch in der Lage Antigen spezifische Antikörper zu sezernieren. Ihre bahnbrechende Erfindung für die Herstellung von Antikörpern wurde 1984 mit dem Nobelpreis honoriert. Allerdings stellt die Hybridoma Technologie einen langwierigen Prozess mit geringer Effizienz dar. Zusätzlich ergibt sich durch die murine Herkunft der Antikörper eine potenzielle Immungenitätsproblematik bei Applikation im Menschen. Aus diesen Gründen wurden in den letzten Jahrzehnten verschiedene Technologien zur Überwindung dieser Limitationen entwickelt. Sie umfassen dabei die Nutzung von humanen naiven, semi- oder komplett synthetischen Antikörperdiversitäten (zur Vermeidung der Immungenitätsproblematik) in Verbindung mit zellulären oder nicht-zellulären in vitro Selektionssystemen zur Präsentation und Isolation von Antikörpern, wie z.B. das Phagen- oder Hefe-Display. Durch die Entwicklung dieser Methoden konnten große Repertoires bezüglich vorgegebener Eigenschaften durchmustert und spezifische Moleküle isoliert werden, was den heutigen, erfolgreichen Einsatz von monoklonalen Antikörpern in biotechnologischen und medizinischen Anwendungen unterstützte. Trotz aller Verbesserungen im Antikörperfindungsprozess stellt die Generierung der initialen Display-Bibliotheken immer noch einen aufwendigen und anspruchsvollen Prozess dar. Dieser ist oft mehrstufig und umfasst am Beispiel des Hefe-Displays verschiedene Klonierungsschritte, die Erstellung von getrennten Bibliotheken für schwere und leichte Ketten, sowie letztlich deren Paarung durch Mating. Dieser Ablauf kann durch die Anwendung von Golden Gate Cloning (GGC) vereinfacht werden und in nur einer Reaktion erfolgen. Die Klonierungsreaktion im GGC beruht dabei auf der Nutzung von Typ IIs Restriktionsenzymen. Diese Enzyme schneiden DNA in einer definierten Distanz außerhalb ihrer Erkennungssequenzen und ermöglichen dadurch die Inkorporation von Überhängen mit gewünschter Signatur. Typ IIs Restriktionsenzyme erlauben somit einen gerichteten, einstufigen Klonierungsprozess, bei dem die Erkennungssequenzen während der Reaktion entfernt und das Reaktionsgleichgewicht auf die Seite des spezifischen Klonierungsproduktes verschoben wird. In der vorliegenden Arbeit konnte anhand von drei unterschiedlichen Selektionskampagnen die Anwendbarkeit von GGC für die Bibliotheks-Generierung im Hefen- und Phagen-Display demonstriert werden. Dabei wurden spezifische Fab Antikörper gegen CEACAM6, EGFR und hCG aus Immunrepertoiren von transgenen Ratten und Wildtyp Hühnern (Hefe-Display), sowie EGFR spezifische scFv und VHH Antikörper aus Immunrepertoiren von Hühnern und Kameliden (Phagen-Display) isoliert. GGC konnte dabei im Vergleich zur klassischen Generierung von Antikörper-Bibliotheken bezüglich Verteilungen der Genrepertoires, der allgemeinen Bibliotheks-Größen, sowie biophysikalischer Charakteristika der isolierten Antikörper validiert werden. Der entwickelte, einstufige GGC Prozess ist in der Lage Antikörper Bibliotheken basierend auf kombinatorischen schweren und leichten Ketten Diversitäten mit gleicher Qualität gegenüber der traditionellen Methode zu generieren. Zusätzlich bietet er den Vorteil einer schnelleren und weniger aufwendigen Methodik, da er die multiplen Schritte des traditionellen Ansatzes parallelisiert. Neben klassischen monoklonalen Antikörpern und Antikörperfragmenten bieten neue Antikörperformate Ansätze zur therapeutischen Anwendung. Dazu zählen Moleküle wie bispezifische Antikörper (bsAbs) und Immunoliganden, sowie andere Antikörperfragmente und -derivate, z.B. kamelide Einzeldomänenantikörper Fc Fusionen. Bei der Herstellung von IgG-basierten bispezifischen Antikörpern, bestehend aus zwei unterschiedlichen schweren und zwei unterschiedlichen leichten Ketten, müssen theoretisch vier unterschiedlichen Plasmide für deren Expression in der Zellkultur eingesetzt werden, was zu einer statistischen Ausbeute von lediglich 12.5 % des gewünscht assembliertem Proteins führt. Eine Möglichkeit die Heterogenität bei der Expression von mehreren gleichgearteten Polypeptidketten zu umgehen, ist die Nutzung der Strand-Exchange Engineered Domain (SEED) Technologie. Durch die Einfügung alternierender IgA und IgG β Faltblatt Strukturen in den CH3 Domänen, entstehen anti-parallele Immunglobulinstrukturen, die eine Heterodimerisierung der schweren Ketten begünstigen. Zur Vermeidung von Homodimerisierung der schweren Ketten wurden neben SEED noch andere Technologien wie „knob-into-holes“, controlled Fab arm exchange oder gezielte Aminosäure-Austausche zur Einführung von konträren elektrostatischen Ladungen in den beiden schweren Ketten entwickelt. Um darüber hinaus eine leichte Ketten-Fehlpaarung der Antikörper zu verhindern, wurden ebenfalls diverse Technologien entwickelt, wie z.B. CrossMab oder die Nutzung einer Common Light Chain. Eine weitere Möglichkeit hierfür besteht zusätzlich in der Verwendung von kameliden VHH Domänen. Diese nutzen nicht die für den kanonischen Antikörper notwendige leichte Kette, sondern stellen eine spezifische Antigen-Bindung ausschließlich über die variable Domäne der schweren Kette her. In diesem Kontext war ein Ziel der vorliegenden Arbeit die Kombination der SEED Technologie mit einem 2013 von Baty und Kollegen konzeptionell beschriebenen Fab ähnlichen bispezifischen VHH Antikörper. Dies ermöglichte die Generierung einer neuartigen bi- und trispezifischen, IgG-ähnlichen VHH basierten Antikörperplattform mit damit verbundenen variablen Valenzen im Rahmen dieser Promotion. Die Ergebnisse der Charakterisierung dieses Antikörperformats bezüglich ihrer biophysikalischen und biochemischen Eigenschaften, wie z.B. spezifischem, NK-Zell vermitteltem ADCC, belegten die Vielseitigkeit dieser generischen Plattform zur Expression voll funktioneller mono-, bi und trispezifischer Antikörper mit unterschiedlichen Valenzen als „plug-and-play“ Anwendung. Darüber hinaus ist der Einsatz von bispezifischen Molekülen für die gerichtete NK-Zell Rekrutierung in heutiger Zeit von großem Interesse. Die spezifische NK-Zell Rekrutierung zu tumorösen oder infizierten Zellen ermöglicht eine spezifische Zytolyse der Zielzellen und die gleichzeitige Immunmodulation durch eine NK-Zell vermittelte Zytokin-Freisetzung. Eine Möglichkeit, um die Rekrutierung von NK-Zellen zu adressieren, bieten die von diesen Zellen umfassend exprimierten Natürlichen Cytotoxischen Rezeptoren (NCRs). In dieser Arbeit konnte mittels Hefe-Display die bindungsrelevante N-terminale IgV-ähnliche Domäne des natürliche Liganden B7 H6 für den NCR Rezeptor NKp30 affinitätsmaturiert werden. Mittels Fluorescence Activated Cell Sorting (FACS) basierter Selektion wurden B7 H6 Varianten mit signifikant erhöhter Affinität zu NKp30 erhalten. Auch zeigten diese Varianten bei der Anwendung im Format bispezifischer Immunoliganden signifikant gesteigerte NK-Zell vermittelte Zytotoxizitäten und Zytokin-Freisetzungen gegenüber dem parentalen B7 H6 Molekül. Diese Erkenntnisse belegen die zu Grunde liegende Hypothese, dass eine gesteigerte Affinität von B7 H6 zu NKp30 in einer erhöhten NK-Zell basierten Tumorzell-Zytolyse und einer gesteigerten NK-Zell vermittelten Sekretion von proinflammatorischen Zytokinen resultiert. Eine Affinitätssteigerung von natürlichen Liganden aktivierender NK-Rezeptoren könnte somit zur Entwicklung potenzieller Immuntherapeutika zur Behandlung von Patienten mit unterschiedlichen Krebserkrankungen verwendet werden.

German
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details