Introduction: Antibody-drug conjugates (ADC) and immunotherapies are part of the rapidly expanding landscape of targeted cancer treatments, with market sizes reaching over $10 and $150 billion in 2024¹. However, further understanding of tumor biology is crucial to developing effective therapies that minimize resistance. N-linked glycosylation, a post-translational membrane modification, is well known for its critical role in tumor growth and resistance, its impact on antibody tumor accumulation, and its effect on tumor response to targeted therapies^2,3. Despite its significance, methods to image and target N-glycosylation are still not fully developed. Using biorthogonal (click) chemistry, positron emission tomography (PET), six different cancer models, and two different antibody therapies (PD-L1: avelumab and HER2: trastuzumab), our study has two main objectives: image and target N-glycosylation in tumors (Fig. 1).

Methods: In vitro studies: Western blot analyses were performed to detect N-glycosylation for PD-L1 (4T1, B16F20, and CT26 murine cancer cells) and receptor-tyrosine kinases (RTKs; NCIN87, MIAPaCa-2 and A431 human cancer cells).

In vivo studies: 4T1 or MIAPaCa-2 cancer cells were inoculated subcutaneously in Balb/c or nu/nu mice, respectively. Glycans visualization: Mice were administered intravenously with tetraacetyl-N-azidoacetylmannosamine (mannose-N₃; 40 mg/kg) or PBS (control) once daily for three consecutive days. At 24 h post the last injection, AZDye-680 DBCO (5 mg/kg) was intravenously injected. At 24 h post-AZDye-680 DBCO administration, the organs were harvested, scanned, homogenized, and lysed for fluorescence and protein measurements. Glycans targeting to enhance anti-PD-L1 antibody uptake: Mice were administered mannose-N₃ or PBS as described before. Mice were then intravenously injected with [⁶⁴Cu]Cu-NOTA-avelumab-DBCO (6.7 MBq, 40 µg). PET images were acquired at 24 and 48 h. Glycans targeting to reduce RTKs signaling: Mice were intratumorally administered with the oligosaccharyltransferase inhibitor NGI-1 (10 µM, 50 µL) or PBS (control, 50 µL). At 4 h post-administration of NGI-1 or PBS, mice were intravenously administrated with [⁸⁹Zr]Zr-DFO-trastuzumab (7.4 MBq, 50 µg). PET imaging and ex vivo biodistribution were performed at 24 h.

Results:  Glycans visualization: Tumors exhibit high levels of N-glycosylation, as previously described^2-4. To visualize N-glycans within tumors, we employed metabolic glycoengineering using a mannose-N₃ derivative and AZDye-680 DBCO fluorescent dye. First, mannose-N₃ is administered, introducing azide functional groups to cell-surface N-glycoproteins. Next, the AZDye-680 DBCO is administered, clicking with azide groups on the cancer cell’s surface. When compared with PBS, we observed a 3.5-fold increase in tumor uptake in animals administered mannose-N₃ (Sup. Fig. 1A).

Glycans targeting to enhance anti-PD-L1 antibody uptake: Previous studies demonstrated that PD-L1 is highly N-glycosylated in cancer cells^5,6, affecting antibody accumulation in tumors. Using immuno-PET with [⁶⁴Cu]Cu-NOTA-avelumab-DBCO, we showed a 1.9-fold increase in avelumab accumulation in mannose-administered mice compared to PBS controls (Sup. Fig. 1A).

Glycans targeting to reduce RTKs signaling: While testing pharmacological approaches targeting N-glycosylation, we validated that NGI-1 reduces PD-L1 glycosylation in cells (Sup. Fig. 1B), resulting in increased avelumab uptake in tumors (not shown). Additionally, we observed a reduction in multi-RTKs (HER2, EGFR, HER3) in cells administered with NGI-1. These biological findings were further validated in vivo by PET imaging showing a 1.6-fold reduction in HER2 (Sup. Fig. 1C).

Conclusions: In this study, we applied click chemistry to not only visualize N-glycans within tumors but also to exploit their presence to enhance antibody therapy. Furthermore, we tested the use of NGI-1 as a pharmacological approach to disrupt N-glycosylation and RTK signaling in vivo. Considering that clinical trials are shifting from antibody monotherapies to combination therapies, including immune checkpoint inhibitors in combination with ADCs, our ongoing studies are focused on developing a dual ADC to specifically and simultaneously deplete PD-L1 glycosylation and RTK signaling and deliver a potent chemo drug (DXd) to tumors (Fig. 1).

Image/Figure:

Click to view full size

Image/Figure Caption:

Figure 1. Schematic representing the approaches used in this study to image and target N-glycosylation in tumors (panel on the left). First, animals are administered with the unnatural sugar mannose-N₃, which enters the sialic acid metabolic pathway to generate azide (-N₃) groups expressed on cancer cell surface glycans. Subsequently, this approach will enable the enhancement of tumor accumulation of dibenzocyclooctyne (DBCO) compounds (such as small molecules or antibodies) via click chemistry. The panel on the right illustrates our ongoing studies developing a dual ADC combining the HER2-targeting antibody trastuzumab with NGI-1, a partial inhibitor of the oligosaccharyltransferase (OST) complex which catalyzes the transfer of N-glycans in the endoplasmic reticulum, and deruxtecan (Dxd), a potent DNA topoisomerase I inhibitor. The scheme was created using BioRender.

Presentation Poster:
Click to View

Author