Decoding the functional impact of the cancer genome through protein-protein interactions

ai
cancer genomics
machine learning
oncogenic mutations
protein-protein interactions
structural biology
targeted therapy
  • Core Concept: This review establishes that oncogenic driver mutations exert their functional impact largely by rewiring molecular signaling networks through the alteration of Protein-Protein Interactions (PPIs).
  • Mechanism: Mutations at the protein surface, particularly at PPI interfaces, can lead to the creation of neomorphic PPIs (neoPPIs) or the loss of existing hypomorphic PPIs (hypoPPIs), necessitating a mutation-focused analysis.
  • Therapeutic Implication: The mutation-directed PPIs are presented as a new class of targets for precision oncology, paving the way for the development of small-molecule modulators that can selectively disrupt oncogenic neoPPIs or restore tumor-suppressive hypoPPIs.
Published

23 January 2026

PubMed: 39810024 DOI: 10.1038/s41568-024-00784-6 Overview generated by: Gemini 2.5 Flash, 28/11/2025

Key Findings: Linking Oncogenic Mutations to PPI Network Rewiring

This review focuses on the paradigm that Protein-Protein Interactions (PPIs) are the critical functional intermediaries between genomic mutations and the resulting oncogenic phenotype. The core assertion is that the functional impact of individual genomic alterations is often transmitted through altered nodes and hubs of PPIs, leading to the rewiring of molecular signaling cascades.

Mutational Effects on PPIs: NeoPPIs and HypoPPIs

Oncogenic mutations frequently map to the protein surface, particularly at or near the hotspot residues of the PPI interface. These mutations can perturb the interactome in two primary ways:

  1. Neomorphic PPIs (neoPPIs): Oncogenic mutations may lead to modified residues that create new contact sites or neo-epitopes, enabling interactions with other proteins that the wild-type protein does not typically bind to.
  2. Hypomorphic PPIs (hypoPPIs): Conversely, mutations can decrease or disrupt the interaction of existing protein complexes, leading to a loss-of-function phenotype.

Understanding the mechanisms of these mutation-driven, differential PPIs is crucial for deciphering tumor heterogeneity and developing precision oncology strategies.

Molecular Basis of Cancer-Associated PPI Perturbation

The functional outcome of a driver mutation is determined by its location and the nature of the amino acid substitution, which is why a mutation-focused approach is necessary, rather than just a driver gene-focused approach.

Site-Specific and Lineage-Dependent Effects

  • Different mutations within the same gene can affect various structural components, including defined protein domains, Short Linear Motifs (SLIMs), and Intrinsically Disordered Regions (IDRs).
  • The same gene can be an oncogenic driver in one cancer type and a tumor suppressor in another. For example, different hotspot mutations in PIK3CA (in the helical vs. kinase domain) are prevalent in different cancer types (cervical vs. breast cancer, respectively), exhibiting differential oncogenic activities.
  • Even different amino acid substitutions at the same hotspot position can lead to distinct functional outcomes, as seen with IDH1 R132H (associated with a less-aggressive phenotype in gliomas) and IDH1 R132C (associated with enhanced proliferation in AML), due to differential neo-enzymatic activity.

Technologies for PPI Identification and Prediction

The review summarizes the experimental and computational methods used to study mutation-affected PPIs:

Experimental Monitoring

Experimental approaches are categorized into:

  • Co-complex affinity-based technologies: These detect co-complex formation and include methods like co-immunoprecipitation, and Affinity Purification coupled with Mass Spectrometry (AP-MS).
  • Proximity-based assays: These detect interactions based on close spatial proximity and include FRET, BRET, Proximity Ligation Assays (PLAs), and yeast two-hybrid (Y2H) systems.

Computational and AI Methods

  • Computational tools (e.g., FoldX, Flex ddG) quantify mutation-induced changes in free binding energy.
  • AI/Machine Learning methods are rapidly advancing the prediction of protein and protein complex structures. Recent models like AlphaFold3 (AF3) and AlphaFold Multimer can predict the structures of protein-protein, protein-DNA, and protein-ligand complexes. However, these models have limitations, such as difficulty in predicting mutational effects in highly flexible regions or capturing dynamic conformational transitions.

Therapeutic Potential

The intersection of cancer variants and altered PPI interfaces opens up a new frontier for developing tumor-selective therapeutic strategies:

  • Targeting Mutant Residues: The development of agents that specifically target the modified residues of the driver protein itself (e.g., KRAS(G12C)-targeted therapies).
  • Modulating NeoPPIs and HypoPPIs: New therapeutic strategies involve designing small-molecule PPI modulators to either disrupt oncogenic neoPPIs or to restore function to lost hypoPPIs (e.g., using small-molecule glues to re-establish a suppressed interaction). This approach represents a path toward personalized medicine by directly addressing the network vulnerabilities created by specific driver mutations.