Chemical generation of checkpoint inhibitory T cell engagers for the treatment of cancer

Bispecific antibodies (bsAbs), particularly BiTEs, show promise in cancer treatment by bridging cytotoxic T cells to target cancer cells. CiTEs enhance this approach by incorporating immunomodulatory proteins, reversing cancer's immune-dampening mechanisms. Current methods primarily rely on protein engineering, but chemical methods offer increased modularity, speed, and opportunities for further functionalization. This study presents a chemical method to generate functionalized three-protein CiTE conjugates, utilizing click chemistry to create a stable, homogeneous construct. This method uses a dibromopyridazinedione (Br2PD) scaffold to covalently link the heavy and light chains of Fab fragments. This strategy allows for the incorporation of two distinct Fabs along with an additional immunomodulatory protein and a small molecule, such as biotin for imaging and purification. The chosen immunomodulatory protein in this study is either an anti-PD-1 Fab fragment or *Salmonella typhimurium* sialidase (ST sialidase), an enzyme known for its immunomodulatory activity. The in vitro efficacy of these CiTE constructs is then evaluated through cell cytotoxicity assays.

Several reviews highlight the growing importance and advancements in chemical bispecific antibody synthesis, emphasizing its advantages over traditional expression-based methods. Existing methods, such as using PEG linkers and next-generation maleimides (NGMs), while useful, faced limitations in reaction times, yields, and further functionalization possibilities. Recently developed pyridazinedione-based methods, using click chemistry, offer improvements. These methods have successfully created homogeneous bsAbs and even IgG-like molecules, but many combinations of checkpoint inhibitors and BiTEs remain unexplored. Specifically, the use of ST sialidase as a checkpoint inhibitor in combination with antibody-mediated targeting has shown promising results, motivating the current study to explore the use of this enzyme in a novel CiTE construct.

The researchers explored multiple strategies for creating the three-protein CiTEs. The final, successful method utilized strain-promoted inverse electron-demand Diels-Alder cycloaddition (SPIEDAC) between tetrazine and bicyclo[6.1.0]non-4-yne (BCN) strained alkyne for all protein-protein linkages. The process began by generating a bispecific Fab-Fab construct with an azide handle. This construct was then converted to a BCN-functionalized molecule using a BCN-PEG-BCN linker. The third protein (sialidase or anti-PD-1 Fab) and DBCO-biotin were added simultaneously in a one-pot reaction via tetrazine-BCN and azide-DBCO clicks respectively. The generation of the Fab moieties from full-length antibodies was achieved through enzymatic digestion. Purification steps included monomeric avidin agarose purification (for biotinylated constructs) and size-exclusion chromatography (SEC). Several constructs were synthesized, including FabHER2-Sia-biotin, FabHER2-FabCD3-biotin (BiTE), FabHER2-(biotin)-FabCD20-biotin, FabHER2-FabCD20-Sia-biotin, and FabHER2-FabCD3-Sia-biotin (CiTE). A final CiTE with an anti-PD-1 Fab was also synthesized (FabCD3-FabHER2-FabPD-1-biotin). The purity and molecular weight of all constructs were confirmed using SDS-PAGE and LC-MS. Detailed procedures and spectral data are provided in the Supplementary Information.

The researchers successfully developed a robust and reproducible chemical method for generating functionalized three-protein CiTEs using Cu-free click chemistry. The resulting CiTEs, incorporating either ST sialidase or an anti-PD-1 Fab, demonstrated superior cytotoxicity *in vitro* compared to their BiTE counterparts. Flow cytometry analysis confirmed that the FabHER2-FabCD3-Sia-biotin CiTE retained its ability to bind to both HER2+ cancer cells and T cells. Importantly, the ST sialidase component of the CiTE showed significant desialylation activity on both T cells and HER2+ cancer cells, with activity dependent on HER2 expression. The FabCD3-FabHER2-FabPD-1-biotin CiTE showed somewhat weaker binding to HER2+ cells and T cells than the BiTE control. However, the PD-1 binding ability of this construct was confirmed through assays with anti-CD3 mAb pre-incubated T cells. A cytotoxicity assay on HER2+ MDA-MB-231 cells revealed enhanced cytotoxicity of both CiTEs compared to BiTE at lower concentrations (0.01-1 nM) both with and without IFN-γ treatment. Specifically, the sialidase-containing CiTE (CITE 24) demonstrated significantly higher cytotoxicity than both the PD-1 blocking CiTE (CITE 27) and the BiTE at these concentrations, suggesting a synergistic effect between sialic acid removal and T cell engagement.

This study successfully demonstrated a chemical method for generating functionalized three-protein CiTEs. The superior *in vitro* cytotoxicity of the CiTEs compared to BiTEs highlights the potential advantages of incorporating immunomodulatory proteins to enhance T cell activation and tumor cell killing. The findings suggest that both ST sialidase-mediated desialylation and PD-1 blockade can synergistically enhance BiTE efficacy. The modular nature of the chemical synthesis method enables the exploration of various combinations of BiTEs and checkpoint inhibitors, allowing for optimization of CiTEs for different cancer types and treatment scenarios. The observation of reduced binding affinity for the Fab fragment in the middle of the three-protein construct suggests potential for modulating binding profiles to enhance tumor specificity and reduce side effects.

The researchers successfully developed a novel chemical method for generating functionalized three-protein CiTEs. The in vitro data demonstrates a significant enhancement of T-cell mediated cytotoxicity by incorporating either ST sialidase or an anti-PD-1 Fab into a BiTE scaffold. Future studies should focus on in vivo evaluation of these CiTEs, exploring different combinations of BiTEs and immunomodulatory proteins, and optimizing the synthesis and purification strategies for large-scale production.

The study primarily focuses on in vitro testing, limiting the generalizability of the findings to in vivo settings. The purity of the FabCD3-FabHER2-FabPD-1-biotin CiTE was not completely assured, although the major contaminant was identified and quantified by LC-MS and appears to not significantly affect the biological activity. Further optimization of the synthesis to achieve higher purity and investigate the impact on biological activity would be beneficial. The study used only a limited number of cancer cell lines and T cell donors. More extensive testing with a wider range of cell lines and donors is needed to validate the findings and assess the generalizability across various cancer types and patient populations.