PD-1 (CD279) is the defining immune checkpoint of modern oncology — the axis that transformed immunotherapy from niche to standard of care. With 16+ approved mAbs and a patent cliff approaching for first-generation agents, the challenge is no longer validation but differentiation: glycan-specific epitopes, cis-binding mechanisms, and next-generation Fc silencing strategies. MAbSilico engineers differentiated PD-1 candidates from 892 annotated antibodies and 312 affinity measurements.
PD-1 is a 288-amino acid Type I transmembrane protein whose co-option by tumors as an immune evasion mechanism created a multibillion-dollar therapeutic sector. Understanding its structural biology is key to engineering differentiated next-generation candidates.
PD-1 belongs to the CD28 family, sharing ~15% sequence identity with CD28 and ~20% with CTLA-4. Its extracellular domain consists of a single V-set immunoglobulin-like domain. The cytoplasmic tail contains two critical signaling motifs: an immunoreceptor tyrosine-based inhibitory motif (ITIM) and an immunoreceptor tyrosine-based switch motif (ITSM). Upon ligand binding (PD-L1 or PD-L2), the ITSM recruits SHP-1 and SHP-2 phosphatases, suppressing TCR signaling and T-cell activation.
A critical recent discovery is the distinction between trans-interactions (between different cells) and cis-interactions (on the same cell). When tumor cells or APCs co-express PD-1 and PD-L1, they can bind in cis — effectively sequestering PD-L1 and preventing it from engaging PD-1 on T cells in trans. This explains why some patients with high PD-L1 expression fail to respond to PD-L1 inhibitors but respond to PD-1 inhibitors.
In malignant contexts, chronic tumor antigen exposure leads to sustained PD-1 expression and T-cell exhaustion — a dysfunction state characterized by impaired effector function and co-expression of multiple inhibitory receptors (TIM-3, LAG-3, TIGIT). Therapeutic PD-1 blockade restores T-cell function, enabling durable anti-tumor responses in immunogenic tumors with high mutational burden.
| Antibody | Subclass | Epitope Focus | KD |
|---|---|---|---|
| Nivolumab | IgG4 (S228P) | N-terminal loop (N-loop) | 3.0 nM |
| Pembrolizumab | IgG4 (S228P) | C’D loop | 29 pM |
| Cemiplimab | IgG4 (S228P) | BC, C’D and FG loops | 0.6 nM |
| Dostarlimab | IgG4 (S228P) | BC, C’D and FG loops | High affinity |
| Penpulimab | IgG1 (silenced) | N58 glycosylation site | Slow off-rate |
| Prolgolimab | IgG1 (LALA) | BC, C’D loops | High occupancy |
Patent cliff opportunity: First-generation IgG4 (S228P) agents (pembrolizumab, nivolumab) are approaching patent expiry. MAbSilico's platform identifies differentiated epitopes and novel engineering strategies — glycan-specific binding, cis-disruption — to build IP-protected next-generation PD-1 candidates.
Anti-PD-1 antibody engineering has undergone a fundamental shift from "natural" IgG frameworks to highly modified scaffolds designed to optimize safety and biophysical stability.
Wild-type IgG4 molecules exhibit "Fab-arm exchange" — heavy-light chain pairs can swap between molecules in vivo, creating bispecific antibodies with unpredictable pharmacokinetics. The S228P mutation (Serine → Proline at position 228) stabilizes the hinge region, preventing this exchange. All first-generation commercial anti-PD-1 antibodies (pembrolizumab, nivolumab, cemiplimab) incorporate S228P.
Newer entries (penpulimab, prolgolimab) utilize an IgG1 backbone with extensive Fc silencing. IgG1 is generally more stable and less aggregation-prone than IgG4, but its high FcγR affinity necessitates engineering to avoid ADCC-mediated T-cell depletion. Penpulimab silences binding to FcγRI, FcγRIIa and FcγRIIIa entirely. Prolgolimab achieves silencing via the LALA mutation (L234A/L235A).
Penpulimab's binding to the N58 glycosylation site on the BC loop represents a paradigm shift. This glycan-specific engagement provides a significantly slower antigen-binding off-rate than first-generation agents — suggesting that "glyco-specific" targeting may achieve higher receptor occupancy and more durable inhibition, with potential for a differentiated IP position.
Our 6-step workflow is designed to navigate the crowded PD-1 landscape — identifying differentiated epitopes, novel Fc configurations and IP-protected positions.
The PD-1 field is crowded but far from closed. Novel epitopes, glycan-specific binding and next-gen bispecific formats open new IP positions. MAbSilico identifies them from 892 annotated antibody sequences.