Discovery of MA48, a Small Molecule Inhibitor of CAPON (NOS1AP)-NOS1 Protein-Protein Interaction

This study identifies and validates MA48 as the first small molecule inhibitor of the CAPON-nNOS protein-protein interaction, demonstrating its direct binding affinity and efficacy in disrupting the complex within living cells to support future therapeutic development for neurodegenerative diseases.

The Gist

The Big Picture: Finding a "Key" for a Sticky Lock

Imagine your brain is a bustling city. In this city, there is a very important traffic controller called nNOS (a protein that makes a gas called Nitric Oxide). This gas is essential for learning and memory, kind of like the electricity that keeps the streetlights on.

However, there is a troublemaker named CAPON. Under normal circumstances, CAPON helps regulate the traffic controller. But when things go wrong (like in Alzheimer's or Parkinson's disease), CAPON gets too excited and overpowers the system. It clings too tightly to the traffic controller, causing the "streetlights" to flicker out or burn out, leading to brain cell damage and memory loss.

For a long time, scientists thought CAPON was a "locked door" that couldn't be opened by medicine. Why? Because CAPON isn't a machine with a clear hole to put a key in (like an enzyme); it's more like a sticky piece of tape that holds two things together. It was thought to be impossible to find a small chemical "key" to pry it apart.

This paper is the story of finding the very first "key" that can pry CAPON loose.

---

The Detective Work: How They Found the Key

The researchers (led by Dr. Gabr at Weill Cornell Medicine) didn't just guess; they went on a massive treasure hunt.

1. The Great Fish Net (AS-MS Screening)
They took a library of 10,000 different small chemical compounds—think of this as a giant bucket of 10,000 different puzzle pieces. They threw this bucket into a tank containing the CAPON protein.

• The Analogy: Imagine throwing 10,000 different shapes into a pool of Velcro. Most shapes just float by, but a few might stick to the Velcro.
• The Result: They used a high-tech scanner (Mass Spectrometry) to see which shapes actually stuck. Out of 10,000, they found a few that stuck, and one stood out as the best: a compound they named MA48.

2. The Strength Test (MST)
Finding a stick wasn't enough; they needed to know how strongly it stuck. They used a technique called Microscale Thermophoresis (MST).

• The Analogy: Imagine trying to pull a magnet off a fridge. Some magnets slide off easily; others require a lot of force. They measured exactly how much force was needed to pull MA48 off the CAPON protein.
• The Result: MA48 stuck with a moderate grip (a "dissociation constant" of 11.9 µM). It wasn't a super-strong industrial magnet yet, but it was the first time anyone had ever seen a small chemical stick to CAPON at all. This proved CAPON could be targeted.

3. The Blueprint (Computer Modeling)
Since no one had ever taken a clear photo (X-ray crystal structure) of CAPON, the scientists had to build a 3D model of it using a computer, based on a similar protein they already knew.

• The Analogy: It's like trying to figure out where a key fits in a lock you've never seen, so you build a clay model of the lock based on a sketch.
• The Result: They ran a digital simulation (Molecular Docking) to see how MA48 fit into their clay model. It fit perfectly into a specific "pocket" on the protein, confirming that the chemical was interacting with the right spot.

4. The Real-World Test (NanoBRET in Cells)
Binding to a protein in a test tube is one thing; working inside a living cell is another. They put the MA48 chemical into living cells (CHO-K1 cells) that had both the traffic controller (nNOS) and the troublemaker (CAPON) glowing with different colored lights.

• The Analogy: When the two proteins hug, the lights mix and glow a specific color. If you break the hug, the lights separate.
• The Result: When they added MA48, the "hug" broke. The proteins stopped sticking together as much. This proved that the chemical could actually walk into a living cell and stop the bad interaction.

---

The "Secret Sauce" (Structure-Activity Relationship)

The scientists didn't just stop at MA48. They wanted to know why it worked so they could make it better. They looked at the chemical structure of MA48 and made tiny changes to it, like swapping out a wheel on a car to see if it drives faster.

They found three "ingredients" that were essential for the key to work:

1. The Core: A specific ring-shaped structure (thiazolothiadiazole) that acts as the anchor. If you change this, the key breaks.
2. The Bridge: A specific connection point (an amino group) that acts like a handshake. Without it, the key can't grab on.
3. The Handle: A hydrophobic (water-repelling) tail that fits into a cozy nook in the protein.

By tweaking these parts, they confirmed that MA48 wasn't just a lucky accident; it was a real, designed interaction.

---

Why This Matters

The "First Step" Analogy:
Imagine you are trying to fix a broken car, but you don't have a wrench. For years, everyone said, "You can't fix this car because there's no tool that fits." This paper is the moment someone says, "Actually, I found a tool that fits, even if it's a bit loose right now."

• It proves the impossible is possible: It shows that CAPON, a protein previously thought to be "undruggable," can actually be targeted by small molecules.
• It opens the door: Now that they have the first key (MA48), other scientists can use it as a blueprint to forge better, stronger keys.
• Future Hope: This could eventually lead to new drugs for Alzheimer's, Parkinson's, and other brain diseases by stopping the "traffic jam" caused by CAPON.

In short: The researchers found the first chemical "wrench" that can loosen a sticky protein in the brain that causes disease. It's not a perfect wrench yet, but it's the first one we've ever found, and it proves we can fix the problem.

Technical Summary

1. Problem Statement

Target: CAPON (also known as NOS1AP), an adaptor protein that binds to neuronal nitric oxide synthase (nNOS/NOS1).
Clinical Relevance: Dysregulation of the CAPON-nNOS interaction is implicated in numerous neurodegenerative and psychiatric disorders, including Alzheimer's, Parkinson's, Huntington's disease, ALS, depression, and anxiety. Overexpression of CAPON disrupts normal nNOS-PSD95 signaling, leading to neurotoxicity, Tau phosphorylation, and synaptic dysfunction.
Scientific Challenge: Despite its therapeutic potential, CAPON has remained an unexplored target for small-molecule drug discovery. Unlike enzymes, CAPON functions as an adaptor protein mediating protein-protein interactions (PPIs) via extended, shallow binding interfaces without a well-defined catalytic pocket. The lack of high-resolution experimental structures and the perceived "chemical intractability" of PPIs have hindered rational ligand design. To date, no validated small-molecule inhibitors of CAPON have been reported.

2. Methodology

The study employed a multi-stage workflow combining biophysical screening, structural modeling, and cellular validation:

• Affinity Selection-Mass Spectrometry (AS-MS):
  • A library of 10,000 diverse, drug-like compounds (ThermoScientific HitFinder) was screened against recombinant CAPON protein.
  • The workflow involved incubating protein with compound pools, separating complexes via size-exclusion chromatography (SEC), and identifying bound ligands via HPLC and time-of-flight mass spectrometry.
  • This label-free approach was chosen specifically because it does not require prior knowledge of binding pockets.
• Biophysical Validation (Microscale Thermophoresis - MST):
  • Hit compounds were subjected to concentration-dependent MST assays to quantify binding affinity ($K_d$).
  • Fluorescently labeled CAPON-His was titrated with compounds to measure thermophoretic movement changes.
• Structural Biology & Computational Modeling:
  • Homology Modeling: Since no crystal structure of CAPON exists, a 3D model was generated using SWISS-MODEL with the PTB domain-containing engulfment adapter protein 1 (PDB ID: 6itu.1.A) as a template (29.45% sequence identity).
  • Binding Site Prediction: The PrankWeb server was used to identify potential druggable pockets on the homology model.
  • Molecular Docking: The Schrödinger Suite (Induced Fit protocol) was used to dock the lead compound into the predicted binding site to elucidate interaction mechanisms.
• Cellular Assay (NanoBRET):
  • A NanoBRET assay was developed in CHO-K1 cells co-expressing NanoLuc-tagged NOS1 and Venus-tagged CAPON.
  • This assay measured the proximity of the two proteins in living cells to determine if the small molecule could disrupt the PPI in a physiological context.
• Structure-Activity Relationship (SAR):
  • A focused set of commercially available analogs of the lead compound was tested to identify essential pharmacophoric features.

3. Key Contributions

• First Small Molecule Inhibitor: The study reports the discovery of MA48, the first identified small-molecule binder and inhibitor of CAPON.
• Proof of Tractability: It demonstrates that the CAPON-nNOS interface, traditionally considered "undruggable" due to its adaptor nature, is chemically addressable by small molecules.
• Pharmacophore Definition: The study defines the critical structural features required for CAPON binding, specifically identifying a thiazolothiadiazole core and an amino-based linker as essential.
• Integrated Workflow: It successfully combines AS-MS screening (for unbiased hit identification) with homology modeling and cellular PPI disruption assays to validate a target lacking crystal structures.

4. Key Results

• Hit Identification: From the 10,000-compound screen, 124 potential binders were identified. After filtering and single-dose validation, MA48 was selected as the lead candidate.
• Binding Affinity: MST analysis confirmed direct binding of MA48 to CAPON with a dissociation constant ($K_d$) of 11.9 ± 7 µM. This is consistent with fragment-sized ligands and early-stage PPI inhibitors.
• SAR Findings:
  • Core: The thiazolothiadiazole heterocyclic core is essential; rigidification or extensive substitution (e.g., MA48.13) drastically reduced affinity ($K_d > 200$ µM).
  • Linker: An amino-based linker (NH group) connecting the core to the aromatic ring is critical. Removing or substituting this linker (e.g., MA48.5, MA48.6) resulted in a complete loss of activity.
  • Hydrophobicity: Moderate hydrophobic substituents (methyl, methoxy) on the terminal aromatic ring were tolerated and sometimes enhanced binding, while excessive bulk or rigidity caused steric clashes.
• Structural Insights:
  • The homology model and docking studies suggest MA48 binds to a pocket involving residues Lys70, Lys71, Lys87, Leu85, Val83, Val108, Met109, and Tyr36.
  • The interaction involves a combination of hydrophobic contacts and hydrogen bonding, with the heterocyclic core acting as an anchor.
• Cellular Efficacy: In the NanoBRET assay, MA48 demonstrated a concentration-dependent disruption of the NOS1-CAPON interaction in live cells. At 100 µM, it reduced the interaction signal to ~80% of the vehicle control, confirming that the compound can engage the target and modulate the PPI in a cellular environment.

5. Significance

This work represents a significant conceptual advance in neurodegenerative drug discovery. By identifying MA48, the authors have:

1. Validated CAPON as a Druggable Target: They have overturned the perception that adaptor proteins like CAPON are chemically intractable.
2. Provided a Starting Point for Optimization: The identified chemotype (thiazolothiadiazole with an amino linker) and the defined pharmacophore provide a concrete foundation for structure-guided medicinal chemistry to improve potency and selectivity.
3. Opened Therapeutic Avenues: The ability to modulate the CAPON-nNOS axis offers a potential new mechanism for treating a wide spectrum of neurological disorders linked to nitric oxide signaling dysregulation.
4. Established a Methodological Blueprint: The study demonstrates a viable pipeline for targeting proteins lacking high-resolution structures, utilizing AS-MS for screening and homology modeling for structural insight.

In conclusion, MA48 serves as a critical proof-of-concept that small molecules can directly engage and functionally modulate the CAPON-nNOS interaction, paving the way for the development of novel therapeutics for neurodegenerative diseases.