Abstract
Adenoviral vectors are the most commonly used delivery vectors for clinical gene therapy due to their favorable characteristics: (i) they do not integrate into the host cell genome and thus harbor a reduced risk of insertional mutagenesis, (ii) they efficiently transduce dividing and non-dividing cells, and (iii) they have a large packaging capacity allowing large and/or multiple cargo delivery. However, site-specific gene delivery in vivo is still compromised because of the endogenous adenoviral tropism and interactions with the host immune system.
To overcome this limitation, our research group developed a generic adenoviral de-/retargeting system consisting of a designed ankyrin repeat protein (DARPin)-based adapter mediating cell-specific transduction and a single-chain variable fragment (scFv)-based vector shield. In an autocrine delivery approach, efficient cancer drug delivery to tumor cells could previously be demonstrated in vivo by applying tumor cell marker-specific DARPin adapters to the adenoviral vector. Building on this approach, we then aimed to expand our system and enable alternative targeting strategies to further improve the efficacy of targeted vector delivery.
The aim of this thesis was to retarget a human adenovirus serotype 5 (HAdV5)-derived vector to stromal cells in the tumor microenvironment in order to deliver therapeutic biomolecules to the tumor microenvironment (TME). The therapeutic biomolecules, initially encoded by the adenoviral vector, would thus be produced by the transduced stromal cells and locally secreted into the TME, where they could exert their anti-tumorigenic effect. This should increase the therapeutic efficacy, reduce off-target effects, and prevent systemic toxicities. Since stromal cells are genetically more stable than tumor cells, in some cancers also more abundant than tumor cells, and theoretically not affected by the vector-encoded cancer drug, we anticipated this paracrine delivery approach to harbor several advantages over the existing, tumor cell- targeted autocrine delivery approach.
In the main part of this thesis (presented in chapter three and Hartmann et al., 2023), the stromal cell-targeted paracrine adenoviral delivery of cancer therapeutics to the TME is presented. To realize the project, a suitable stromal target for adenoviral retargeting, namely fibroblast activation protein (FAP) on cancer-associated fibroblasts (CAFs), was first identified. Aiming to generate FAP-specific adenoviral retargeting adapters, ribosome display selections were performed to obtain FAP-specific DARPins, which were then utilized as targeting domain in the previously developed bispecific modular adapter. Biologically functional adapters capable to successfully mediate retargeting of HAdV5 to CAFs via FAP were subsequently selected in a cell-based screening approach. These adapters were characterized in great detail for several binding characteristics (including binding kinetics, potential binding epitopes, and cross-reactivity to human/mouse FAP), and tested in vitro for specificity and selectivity using various target and non-target cell lines. HAdV5-based vector retargeting to CAFs was furthermore investigated in vivo using a selected, human/mouse FAP cross-reactive adapter and a suitable mouse model of cancer which had been established before. In these in vivo studies, efficient vector retargeting to CAFs by the FAP-specific adapter was demonstrated. Using the same mouse model of cancer, adenoviral delivery of a therapeutic monoclonal antibody with anti-tumor activity to CAFs could subsequently be investigated in vivo, and showed to be successful as evidenced by reduced tumor growth. Importantly, the therapeutic effect of the monoclonal antibody was superior when encoded and delivered by the FAP- retargeted adenoviral vector than when administered directly as recombinant protein. Altogether, it was thus demonstrated that retargeting of HAdV5-based vectors via FAP-specific adapters enables an efficient, stromal cell-targeted paracrine delivery of anti-cancer therapeutics to the TME.
In addition to developing the stromal cell-targeted adenoviral vector for a paracrine cancer drug delivery, it was also possible to contribute to other research projects dealing with the retargeting of HAdV5-based vectors for cancer therapeutic applications.
In one collaboration project (presented in chapter four and Freitag et al., 2023), it was intended to retarget HAdV5 to T cells and achieve an efficient transduction of these immune cells, which are naturally not susceptible to an HAdV5 infection. This goal was achieved using the adapter- based retargeting strategy in combination with the experimental finding that T cells ought to be activated to be adenovirally transduced. As a result, novel scFv-based adapters targeting the T cell co-receptor CD3, the co-stimulatory receptor CD28, and the interleukin (IL)-2 receptor were developed and deployed concurrently to promote cell specificity for adenoviral retargeting while providing T cell activation stimuli. Cell-specific retargeting of HAdV5-based vectors to human T cells and efficient human T cell transduction was shown in vitro as well as in vivo in relevant humanized mouse models. This T cell-retargeted adenoviral vector could now be used to engineer T cells in vivo in order to improve chimeric antigen receptor (CAR)-T cell therapy for cancer treatment.
In two other collaboration projects, the adenoviral delivery of cancer therapeutics to the TME via the previously established, tumor cell-targeted autocrine approach was investigated in vivo. In both these projects the HAdV5-based vector was retargeted to tumor cells via the human epidermal growth factor receptor 2 (HER2) using specific DARPin adapters. However, different therapeutic biomolecules, including the cytokine IL-12 or a bispecific T cell engager (BiTE) aiming to enhance anti-tumor NK cell (presented in chapter six and Kirchhammer et al., 2022) or T cell activity (presented in chapter five and Freitag and Kolibius et al., manuscript in preparation for publication), were encoded by the adenoviral vector. Targeted payload delivery as well as therapeutic effects were investigated in vivo in relevant tumor mouse models. In both projects, target-specific adenovirus-mediated payload delivery with great therapeutic efficacy could be successfully demonstrated.
Hartmann, Karen Patricia