Understanding glycan-sensitive vs glycan-agnostic PD-1 antibodies using 3decision 2.0

Introduction 

The programmed cell death-1 receptor (PD-1) is a crucial immune checkpoint receptor, primarily expressed on T cells. It binds its native ligand, Programmed-cell death ligand 1 (PD-L1) which is upregulated on cancer cells and thereby weakens immune responses. Blocking this PD-1/PD-L1 axis with antibodies (immune checkpoint inhibitors, ICIs) has revolutionized cancer therapy. However, emerging evidence highlights that post-translational modifications (PTMs) of PD-1, especially glycosylation, significantly influence PD-1’s function and the binding selectivity and efficacy of ICIs. Therefore, understanding the glycosylation landscape of PD-1 is critical for optimizing existing ICIs and guiding the development of next-generation immunotherapeutics. 

In this post, we’ll summarize key findings from three pivotal 2024 papers that investigated PD-1 glycosylation and its impact on ICIs and therapy outcomes. We'll use the new features from the 3decision 2.0 release to highlight the findings. 

Discover 3decision 2.0 release with enhanced structure registration and biologics features

Using 3decision 2.0 workflow, we will describe and visually present: 

  • Compare structural differences between therapeutic antibodies targeting PD-1 that are dependent on its glycosylation at Asn58 for selective binding, versus those that are glycosylation-independent. This comparison is made possible using the new "superimposition by chain" feature ([see release notes]). 

  • Explore how PD-1 glycosylation at Asn58 influences antibody interactions and highlight key structural features, such as an unusually long CDR loop, using 3decision’s updated annotations browser and custom annotations tools. 

Key findings from studies on PD-1 Glycosylation and ICIs 

Glycosylation regulates PD-1 stability and PD-1/PD-L1 interactions 

Glycosylation is an is a critical post-translational modification of PD-1, contributing to its stability and enhancing its ability to bind PD-L1. Stripping PD-1 of its glycans increases ubiquitination, reducing its surface expression and impairing PD-L1 binding. Therefore, glycosylation helps PD-1 remain on the T cell surface, maintaining a proper conformation conducive to its immunoregulatory function, particularly in the tumor microenvironment. The extracellular domain of PD-1 has four N-linked glycosylation sites (Asn49, Asn58, Asn74, and Asn116), that play key roles in regulating the protein stability, conformation and interactions with ligands and therapeutic antibodies. Studies have shown that core fucosylation (the addition of fucose, at certain sites) particularly at Asn49 and Asn74 of PD-1, is crucial for its stabilization. Blocking this modification significantly lowers PD-1 expression on T cells, enhancing their anti-tumor responses. Fucosylated glycans contribute to PD-1 retention at the cell surface, functioning as molecular glue that supports proper folding and stability.  

Although glycosylation does not directly influence PD-L1 binding, since the PD-1/PD-L1 interface is spatially distant from glycosylation sites, these glycans can indirectly modulate binding affinity. For example, PD-L1 mainly contacts the FG loop, while glycans like Asn58 are located on the BC loop, therefore they do not contact PD-L1 but they may influence local conformations that may favor or hinder PD1-PLD1 interaction. 

The role of glycosylation on PD-1 antibody binding 

Two independent studies revealed that therapeutic antibodies camrelizumab and cemiplimab have a unique dependency on the N-linked glycan at Asn58 of PD-1. 

Camrelizumab targets the BC loop of PD-1, with a strong dependency for a high affinity binding on the core-fucosylated glycan at Asn58. Structural analyses show that residues in the antibody’s CDR2 and CDR3 regions, especially a key serine in the CDR2 heavy chain, directly interact with the fucose moiety of the glycan. As a result, camrelizumab's binding affinity is significantly reduced in the absence of this glycan.  

This glyco-selectivity provides an additional layer of specificity, allowing camrelizumab to preferentially engage glycosylated PD-1 and potentially distinguish between different molecular states of the receptor. 

Cemiplimab binds PD-1 similarly to camrelizumab. Discover more about the Pd-1/cemiplimab complex and binding mode in our Protein of the Month article on PD-1.

In contrast, other known therapeutic antibodies like nivolumab and pembrolizumab do not rely on Asn58 glycosylation. Nivolumab binds near PD-1’s N-terminal (N) loop (residues Asn25–Pro34), with additional contacts on the FG and BC loops.  This region lies outside of the primary PD-L1 binding site and does not involve Asn58 or any glycan-mediated interactions. This finding is particularly notable because the ability of nivolumab to engage a distinct glycan-independent epitope on the N-loop of PD-1, allows it to bind with high affinity even in the absence of PD-1 glycosylation. 

These binding differences can be easily visualized with 3decision’s structural superpositions feature. Discover how to align PD-1 chains and use the annotation tool to highlight key regions such as the N-loop and glycosylation sites. This allows quick and intuitive visualizion on how camrelizumab and nivolumab engage PD-1 at distinct and non-overlapping interfaces. 

Insights and Conclusions

The abovementioned advances in the structural design of PD-1 inhibitors have highlighted the critical role of post-translational modifications, particularly glycosylation, in modulating antibody-receptor interactions. In PD-1, the glycan at Asn58 critically influences receptor surface retention and antibody binding. 

Antibodies like camrelizumab and cemiplimab depend on the fucosylated glycan for high-affinity interaction, suggesting that glycan-sensitive designs may offer enhanced targeting in tumors with elevated glycosylated PD-1, such as late-stage lung cancer.  

By contrast, nivolumab, which binds the N-loop independently of glycosylation, represents a distinct binding mode and a robust strategy in contexts where PD-1 glycosylation is variable or reduced. The N-loop remains consistently accessible, making it a promising target for next-generation antibodies with broader therapeutic applicability

These findings reinforce the notion that PTMs are dynamic regulators of therapeutic response. Using 3decision 2.0, researchers can visualize these interactions in detail, enabling drug developers to distinguish between glycan-sensitive and glycan-agnostic strategies and make informed design decisions based on molecular context to accelerate the therapeutic development. 

As immunotherapy continues to evolve through 2025 and beyond, the ability to target and exploit structural diversity will be critical for developing more precise, adaptable, and effective PD-1 inhibitors. The integration of structural insights and interactive visualization tools is foundational to the future of biologics design. 


Reference:

Lu D, Xu Z, Zhang D, Jiang M, Liu K, He J, Ma D, Ma X, Tan S, Gao GF, Chai Y. PD-1 N58-Glycosylation-Dependent Binding of Monoclonal Antibody Cemiplimab for Immune Checkpoint Therapy. Front Immunol. 2022 Mar 2;13:826045. doi: 10.3389/fimmu.2022.826045. PMID: 35309324; PMCID: PMC8924070.

Liu K, Tan S, Jin W, Guan J, Wang Q, Sun H, Qi J, Yan J, Chai Y, Wang Z, Deng C, Gao GF. N-glycosylation of PD-1 promotes binding of camrelizumab. EMBO Rep. 2020 Dec 3;21(12):e51444. doi: 10.15252/embr.202051444. Epub 2020 Oct 15. PMID: 33063473; PMCID: PMC7726772.

Chu CW, Čaval T, Alisson-Silva F, Tankasala A, Guerrier C, Czerwieniec G, Läubli H, Schwarz F. Variable PD-1 glycosylation modulates the activity of immune checkpoint inhibitors. Life Sci Alliance. 2024 Jan 4;7(3):e202302368. doi: 10.26508/lsa.202302368. PMID: 38176728; PMCID: PMC10766783.

Next
Next

Prediction, expertise, and validation: An integrated approach to advancing SBDD