Introduction
Protein misfolding is a hallmark of numerous debilitating diseases, including Alzheimer’s and cystic fibrosis. While misfolding is often characterized by non-specific aggregation, a more subtle and complex form involves topological traps known as non-covalent lasso entanglements. In these structures, a loop within the protein chain is closed by a specific chemical contact, and another segment of the chain incorrectly threads through—or fails to thread through—this loop.
These entanglements are particularly problematic because they can result in soluble, yet non-functional, proteins that evade the cell's standard quality control mechanisms. Once a "lasso" is formed incorrectly, the protein becomes kinetically trapped in a misfolded state. This study investigates whether small molecule drugs can be designed to steer the folding process away from these topological dead-ends.
Methodology
The researchers employed a multi-scale computational approach focusing on two bacterial proteins prone to such entanglements: CAT-III and DDLB.
- Coarse-Grained (CG) Modeling: To simulate the extensive timescales of protein folding, the team utilized structure-based "Gō-like" models, representing each amino acid as a single bead to prioritize topological accuracy.
- Drug Design Strategies: Two idealized strategies were evaluated:
- Strategy 1: Stabilizing the native (correctly entangled) state.
- Strategy 2: Stabilizing a non-native intermediate structure to delay the closure of the entanglement loop.
- Virtual Screening: To identify real-world candidates, the team screened 2,056 FDA-approved drugs from the e-Drug3D database using docking software (AutoSite and e-LEA3D) to find molecules that could mimic Strategy 2.
- All-Atom Validation: The most promising candidates were subjected to all-atom Molecular Dynamics (MD) simulations to verify binding stability under realistic conditions.
Key Insights: The "Buying Time" Principle
A central finding of the study is the counter-intuitive success of Strategy 2 over Strategy 1. Stabilizing the native state (Strategy 1) often proved detrimental; by making the native loop more stable, the drug caused it to close prematurely, before the threading segment was correctly positioned. This effectively increased the rate of misfolding.
In contrast, Strategy 2 succeeded by stabilizing a non-native intermediate. This acted as a kinetic "speed bump," delaying loop closure and providing sufficient time for the rest of the protein chain to thread correctly. This shift from "native state stabilizers" to "kinetic pathway chaperones" suggests a new paradigm for treating misfolding diseases by targeting the folding process itself.
Results and Drug Candidates
The "buying time" strategy significantly improved folding outcomes in simulations:
- DDLB: Correct folding increased from 64% to 96%.
- CAT-III: The native population doubled, rising from 23% to 46%.
The virtual screen identified several FDA-approved drugs as potential "topological rescuers":
- For DDLB, candidates include Digoxin (a cardiac glycoside) and Aprepitant.
- For CAT-III, candidates include Latanoprost acid and Ezetimibe (a cholesterol-lowering medication).
All-atom simulations confirmed that these drugs remained stably bound to their target intermediate structures for over 500 nanoseconds in the majority of trials.
Limitations and Future Directions
While these theoretical results are promising, several limitations remain. The coarse-grained models may overlook fine-grained chemical details, and the simulations assume idealized drug behavior without off-target effects. Most importantly, these findings require experimental validation in "wet lab" environments to confirm if these drugs can indeed restore enzymatic activity in CAT-III or DDLB.
Given that approximately half of all globular proteins in humans and E. coli contain similar entanglements, this research opens a vast new landscape for drug discovery targeting topological misfolding.
Summary updated