Rational Design of a Therapeutic Nanobody for the Direct Inhibition of the KRAS Oncoprotein
Description
Executive Summary
Context and Objective
The oncoprotein KRAS is a critical therapeutic target in pancreatic, lung, and colorectal cancers. Existing covalent inhibitors have limitations, such as resistance and applicability restricted to specific mutations (e.g., G12C). This work proposes PIA-KRASv2-Nb, a 100% humanized nanobody designed via the Protein Interaction Architect (PIA) method, which blocks the DEYDPTIEDS epitope in the Switch I region of KRAS with a high predicted affinity (ipTM = 0.78), and whose binding pose has been validated as exceptionally stable by molecular dynamics simulations.
Key Methodology
-
Computational Design: Use of the quantum-harmonic operator to generate CDR sequences with spectral complementarity and intrinsic humanization constraints.
-
Structural Prediction: Directed sampling on AlphaFold-Multimer v3, identifying 12 high-affinity conformations (ipTM ≥ 0.7), with seed 72 as the optimum.
-
Static Validation: Verification of architecture and stability with SCALOP (canonical loops), NanoBodyBuilder2 (RMSD < 0.35 Å), and Hu-mAb (VH3 humanization).
-
Dynamic Validation: Molecular dynamics simulation (500 ps) to confirm the stability of the complex under physiological conditions.
Key Results
-
Interface Affinity (ipTM): 0.78 (seed 72)
-
Interaction Surface Area: 788 Ų
-
Structural Stability (predicted): 0.19 Å (CDR3), 0.35 Å (framework) (Model's RMSD against templates via NanoBodyBuilder2)
-
Dynamic Stability (simulated): Stable and convergent (~2.2 Å @ 500 ps) (Complex's RMSD during MD simulation)
-
Humanization (Hu-mAb): 1.0 (VH3 family)
Implications and Next Steps
-
Advantages: Ab initio humanization without posterior engineering and a pan-mutant mechanism of action.
-
Limitations: Requires experimental validation (SPR, cellular assays).
-
Next Steps: Expression in E. coli SHuffle®, competitive binding assays with RAF, and optimization for intracellular delivery.
Conclusion
PIA-KRASv2-Nb represents an advance in the rational design of therapeutic nanobodies, combining high affinity, intrinsic humanization, and a conformational reproducibility confirmed by molecular dynamics simulations that demonstrate the persistence of the binding pose. Experimental confirmation of these results could lead to the extension of the PIA method to other therapeutic targets.
Note: The full scientific report and 3D models are available at the associated GitHub repository: https://github.com/NachoPeinador/PIA-KRASv2-Nb
Files
PIA-KRASv2-Nb_Report_ESP.pdf
Additional details
Related works
- Is new version of
- Preprint: 10.21203/rs.3.rs-7239936/v1 (DOI)
- Is supplement to
- Software: https://github.com/NachoPeinador/PIA-KRASv2-Nb (URL)
Software
- Repository URL
- https://github.com/NachoPeinador/PIA-KRASv2-Nb
- Programming language
- Python
- Development Status
- Active
References
- Abramson, J., et al. (2024). Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature.
- Eastman, P., et al. (2017). OpenMM 7: Rapid development of high performance algorithms for molecular dynamics. PLoS Computational Biology, 13(7), e1005659.
- Puszkiel, A., et al. (2019). KRAS-Mutant Cancer: A Challenging Target. Cancers (Basel), 11(9), 1277.