Logo Logo
Hilfe
Kontakt
Switch language to English
Development and characterization of universal and modular bispecific immune cell engagers for cancer immunotherapy
Development and characterization of universal and modular bispecific immune cell engagers for cancer immunotherapy
Bispecific immune cell engaging antibodies are highly promising agents for cancer immunotherapy. However, one major obstacle to achieving efficacy across patient populations remains inter- and intra-patient tumor heterogeneity. This heterogeneity has been observed in terms of both tumor antigen profiles and immune cell compartments and immune fitness. As such, one-size-fits-all drug approaches, targeting specific antigens and specific immune cells within an indication, are bound to show limited therapeutic effects across patients. A proposed solution gaining interest in the scientific community is treatment personalization. While displaying high potential, personalization of drugs to each patient has been proven to substantially increase therapy costs, and pose risks of delayed treatment due to production timelines. An attractive approach, albeit not fully explored in previous studies, would be a combination of personalization potential with off-the-shelf drug availability. In this thesis, three universal and modular therapeutic antibody platforms are presented, aiming to address these issues. Firstly, a novel universal P329G-Engager antibody platform is developed. The modular platform is based on two libraries enabling mix-and-match drug assembly. The first is a library of tumor antigen-binding adaptor IgG antibodies, each of which bears P329G mutations in its Fc portion. Upon identifying a tumor target of choice, the specific adaptor is selected. The second library consists of bispecific immune engaging antibodies, recognizing both the P329G mutation and an immune receptor of choice, resulting in P329G-T cell bispecifics (P329G-TCB), ADCC-competent P329G innate cell engagers (P329G-ICE), P329G costimulatory molecules (P329G-CD28/4-1BBL) or P329G immunocytokines (P329G-IL2v). A selection of a desired antigen-targeting adaptor, and a desired immune engager, enables assembly into a functional immunotherapeutic drug with a mode of action of choice. The antigen binders, the P329G mutation, the P329G binder and immune receptor binders are all based on clinically-tested and validated immunotherapeutics. In vitro assays presented here show all P329G-Engager modalities induce anti-tumoral and/or immunomodulatory cell activity. Anti-tumoral efficacy of secondary antibodies is observed only in the presence of tumor-targeted P329G adaptors both in vitro and in vivo, while no effect is observed in their absence. Altogether, experimental evidence is provided for the double universal character of the P329G-Engager antibody platform, whereby tumor binding components and immune engaging components enable assembly of functional drugs from modules of choice. Secondly, in order to tackle on-target off-tumor toxicity linked to classical TCBs, a proteaseactivated pro-P329G-TCB is developed. To ensure drug activation only in the tumor microenvironment, a tumor-activated switch component is engineered and added to the P329G-TCB structure – a tumor- associated protease-cleavable binder mask. Upon encountering the protease-rich tumor microenvironment, the mask can be released, resulting in activation of the drug. To retain P329G-binder universality a P329G-mutated CH2 fragments engineered as a mask. While the inactivating mask has the same structure as the epitope on the P329G-mutated adaptors, protease-dependent release of the mask enables adaptor-engager assembly and functional T cell engagement against tumor cells. Pro-P329G-TCB is shown to release the mask upon cleavage by matriptase only when a specific cleavage site is utilized. Moreover, due to being based on the P329G-Engager, pro-P329G-TCB is able to assemble with different tumor-binding P329G-mutated IgG adaptors and exert proteasedependent antitumoral efficacy in vitro on different cancer cells. Therefore, pro-P329G-TCB enables next-generation tumor activation of a target-agnostic P329G-directed TCB. Additionally, due to the potential of the P329G-masking being transferrable to other P329GEngagers such as P329G-ICE, a triple universality is enabled – in terms of protease masking, antigen choice, and immune engager choice. Thirdly, to capitalize on differences in pharmacological properties of small molecules and antibodies, as well as high affinity hapten binding, an adaptor-based universal DOTAM-TCB platform is developed. The tumor-binding adaptor library consists of a set of small molecules. Their structure is derived from published small molecule ligands to surface tumor targets, and designed to be covalently linked to the Ca2+-loaded chelator DOTAM serving as a hapten. This results in new design of molecules such as Ca-DOTAM-DUPA and Ca-DOTAM-AAZ for targeting PSMA and CAIX expressing tumors, respectively. The immune engager of choice is a bispecific antibody with a hapten binder recognizing Ca-DOTAM and a binder to CD3ε. X-Ray crystallography reveals a deep binding pocket between VH and VL domains of the anti-DOTAM antibody binder. The Ca-DOTAM interaction with the DOTAM antibody binder reaches doubledigit femtomolar affinity, enabling stable assembly of the adaptor with the adaptor-specific DOTAM-TCB. This high affinity characteristic was utilized for the development of a novel protocol of non-covalent small molecule-antibody complex formation. Native mass spectrometry method is utilized and optimized to enable verification of the complexing protocol. In functional in vitro assays, DOTAM-TCB combined with the haptenated small molecule adaptors is shown to exert antitumoral efficacy on two tumorassociated targets, in both complexed and separate formulations. Administration of adaptors only or DOTAM-TCB only does not lead to tumor killing, highlighting the adaptor-TCB assembly as a requirement for functionality. In vitro and ex vivo assessments show a lack of nonspecific binding. While Ca-DOTAM-AAZ is based on published in vivo-tested AAZ ligands, here, Ca-DOTAM-AAZ with DOTAM-TCB leads to rapid-onset toxicity in a HT-29 (CAIX+)bearing humanized CD47-BRGS mouse model. With no toxicity observed for Ca-DOTAM-AAZ only, DOTAM-TCB only, and low toxicity observed in the CAIX-TCB group, the hypothesis is raised that both CAIX expression and Ca-DOTAM-AAZ off-tumor activity may contribute to the observed toxicity. Further directions are proposed for the development of small molecule-based adaptors for use in cancer immunotherapy, and specifically, T cell engagement. Despite the aforementioned in vivo results, all in all, in vitro proof of concept for haptenated small molecule adaptors and universal DOTAM-TCB is achieved, providing in depth biochemical characterization and preliminary data on functionality. Moreover, the Ca-DOTAM high affinity interaction with DOTAM opens avenues to repurposing this interaction for other immunotherapy modalities beyond TCBs. In summary, this thesis describes three universal, target-agnostic, and modular platforms for use with bispecific immune engagers in cancer immunotherapy. All platforms exploit different avenues of antibody engineering, such as precise paratope use against a P329G Fc mutation, conditional activation of prodrugs in the tumor microenvironment, and high affinity hapten binding with stable non-covalent complex formation. Importantly, all three platforms described here enable current or future implementation of double universality in terms of targeting an antigen of choice and engaging an immune cell or pathway of choice. Additionally, the platforms utilize antibody structures and binders based on clinical molecules for the majority of their components. As such, the developments presented in this thesis may ultimately be utilized for mix-and-match cancer immunotherapeutics, targeting heterogeneous tumors with off-the-shelf drugs.
Not available
Surówka, Marlena Katarzyna
2025
Englisch
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Surówka, Marlena Katarzyna (2025): Development and characterization of universal and modular bispecific immune cell engagers for cancer immunotherapy. Dissertation, LMU München: Fakultät für Chemie und Pharmazie
[thumbnail of Surowka_Marlena.pdf]
Vorschau
PDF
Surowka_Marlena.pdf

32MB

Abstract

Bispecific immune cell engaging antibodies are highly promising agents for cancer immunotherapy. However, one major obstacle to achieving efficacy across patient populations remains inter- and intra-patient tumor heterogeneity. This heterogeneity has been observed in terms of both tumor antigen profiles and immune cell compartments and immune fitness. As such, one-size-fits-all drug approaches, targeting specific antigens and specific immune cells within an indication, are bound to show limited therapeutic effects across patients. A proposed solution gaining interest in the scientific community is treatment personalization. While displaying high potential, personalization of drugs to each patient has been proven to substantially increase therapy costs, and pose risks of delayed treatment due to production timelines. An attractive approach, albeit not fully explored in previous studies, would be a combination of personalization potential with off-the-shelf drug availability. In this thesis, three universal and modular therapeutic antibody platforms are presented, aiming to address these issues. Firstly, a novel universal P329G-Engager antibody platform is developed. The modular platform is based on two libraries enabling mix-and-match drug assembly. The first is a library of tumor antigen-binding adaptor IgG antibodies, each of which bears P329G mutations in its Fc portion. Upon identifying a tumor target of choice, the specific adaptor is selected. The second library consists of bispecific immune engaging antibodies, recognizing both the P329G mutation and an immune receptor of choice, resulting in P329G-T cell bispecifics (P329G-TCB), ADCC-competent P329G innate cell engagers (P329G-ICE), P329G costimulatory molecules (P329G-CD28/4-1BBL) or P329G immunocytokines (P329G-IL2v). A selection of a desired antigen-targeting adaptor, and a desired immune engager, enables assembly into a functional immunotherapeutic drug with a mode of action of choice. The antigen binders, the P329G mutation, the P329G binder and immune receptor binders are all based on clinically-tested and validated immunotherapeutics. In vitro assays presented here show all P329G-Engager modalities induce anti-tumoral and/or immunomodulatory cell activity. Anti-tumoral efficacy of secondary antibodies is observed only in the presence of tumor-targeted P329G adaptors both in vitro and in vivo, while no effect is observed in their absence. Altogether, experimental evidence is provided for the double universal character of the P329G-Engager antibody platform, whereby tumor binding components and immune engaging components enable assembly of functional drugs from modules of choice. Secondly, in order to tackle on-target off-tumor toxicity linked to classical TCBs, a proteaseactivated pro-P329G-TCB is developed. To ensure drug activation only in the tumor microenvironment, a tumor-activated switch component is engineered and added to the P329G-TCB structure – a tumor- associated protease-cleavable binder mask. Upon encountering the protease-rich tumor microenvironment, the mask can be released, resulting in activation of the drug. To retain P329G-binder universality a P329G-mutated CH2 fragments engineered as a mask. While the inactivating mask has the same structure as the epitope on the P329G-mutated adaptors, protease-dependent release of the mask enables adaptor-engager assembly and functional T cell engagement against tumor cells. Pro-P329G-TCB is shown to release the mask upon cleavage by matriptase only when a specific cleavage site is utilized. Moreover, due to being based on the P329G-Engager, pro-P329G-TCB is able to assemble with different tumor-binding P329G-mutated IgG adaptors and exert proteasedependent antitumoral efficacy in vitro on different cancer cells. Therefore, pro-P329G-TCB enables next-generation tumor activation of a target-agnostic P329G-directed TCB. Additionally, due to the potential of the P329G-masking being transferrable to other P329GEngagers such as P329G-ICE, a triple universality is enabled – in terms of protease masking, antigen choice, and immune engager choice. Thirdly, to capitalize on differences in pharmacological properties of small molecules and antibodies, as well as high affinity hapten binding, an adaptor-based universal DOTAM-TCB platform is developed. The tumor-binding adaptor library consists of a set of small molecules. Their structure is derived from published small molecule ligands to surface tumor targets, and designed to be covalently linked to the Ca2+-loaded chelator DOTAM serving as a hapten. This results in new design of molecules such as Ca-DOTAM-DUPA and Ca-DOTAM-AAZ for targeting PSMA and CAIX expressing tumors, respectively. The immune engager of choice is a bispecific antibody with a hapten binder recognizing Ca-DOTAM and a binder to CD3ε. X-Ray crystallography reveals a deep binding pocket between VH and VL domains of the anti-DOTAM antibody binder. The Ca-DOTAM interaction with the DOTAM antibody binder reaches doubledigit femtomolar affinity, enabling stable assembly of the adaptor with the adaptor-specific DOTAM-TCB. This high affinity characteristic was utilized for the development of a novel protocol of non-covalent small molecule-antibody complex formation. Native mass spectrometry method is utilized and optimized to enable verification of the complexing protocol. In functional in vitro assays, DOTAM-TCB combined with the haptenated small molecule adaptors is shown to exert antitumoral efficacy on two tumorassociated targets, in both complexed and separate formulations. Administration of adaptors only or DOTAM-TCB only does not lead to tumor killing, highlighting the adaptor-TCB assembly as a requirement for functionality. In vitro and ex vivo assessments show a lack of nonspecific binding. While Ca-DOTAM-AAZ is based on published in vivo-tested AAZ ligands, here, Ca-DOTAM-AAZ with DOTAM-TCB leads to rapid-onset toxicity in a HT-29 (CAIX+)bearing humanized CD47-BRGS mouse model. With no toxicity observed for Ca-DOTAM-AAZ only, DOTAM-TCB only, and low toxicity observed in the CAIX-TCB group, the hypothesis is raised that both CAIX expression and Ca-DOTAM-AAZ off-tumor activity may contribute to the observed toxicity. Further directions are proposed for the development of small molecule-based adaptors for use in cancer immunotherapy, and specifically, T cell engagement. Despite the aforementioned in vivo results, all in all, in vitro proof of concept for haptenated small molecule adaptors and universal DOTAM-TCB is achieved, providing in depth biochemical characterization and preliminary data on functionality. Moreover, the Ca-DOTAM high affinity interaction with DOTAM opens avenues to repurposing this interaction for other immunotherapy modalities beyond TCBs. In summary, this thesis describes three universal, target-agnostic, and modular platforms for use with bispecific immune engagers in cancer immunotherapy. All platforms exploit different avenues of antibody engineering, such as precise paratope use against a P329G Fc mutation, conditional activation of prodrugs in the tumor microenvironment, and high affinity hapten binding with stable non-covalent complex formation. Importantly, all three platforms described here enable current or future implementation of double universality in terms of targeting an antigen of choice and engaging an immune cell or pathway of choice. Additionally, the platforms utilize antibody structures and binders based on clinical molecules for the majority of their components. As such, the developments presented in this thesis may ultimately be utilized for mix-and-match cancer immunotherapeutics, targeting heterogeneous tumors with off-the-shelf drugs.