Aβ-Overlapping Ectodomain Binding of the Clinical-Stage TREM2 Agonist VG-3927

This study reveals that the clinical-stage TREM2 agonist VG-3927 directly binds to the receptor's ectodomain hydrophobic groove, an Aβ-binding site, such that Aβ occupancy interferes with VG-3927 binding and attenuates its signaling activity, suggesting that amyloid levels may modulate agonist efficacy in Alzheimer's disease.

The Gist

The Big Picture: A Broken Lock and a New Key

Imagine your brain is a bustling city, and the immune cells (microglia) are the cleanup crew. Their job is to sweep up trash, specifically a sticky, gooey substance called Amyloid-beta (Aβ) that clumps together to form "plaque" in Alzheimer's disease.

To do their job, these cleanup crews have a special sensor on their surface called TREM2. Think of TREM2 as a doorbell. When the right person rings it, the cleanup crew wakes up, gathers around the trash, and starts cleaning.

For a long time, scientists thought the doorbell only worked when the "trash" (Amyloid) rang it. But recently, a new drug called VG-3927 was developed to act like a super-button that rings the doorbell even if the trash isn't there, hoping to wake up the cleanup crew and clear the Alzheimer's plaque.

This paper asks a simple but crucial question: How exactly does this new drug (VG-3927) ring the doorbell? And does the trash (Amyloid) get in the way?

---

The Discovery: Finding the Hidden Handle

Scientists used a powerful computer program (like a high-tech 3D puzzle solver) to look at the shape of the TREM2 doorbell. They found something surprising.

They discovered that VG-3927 doesn't just push the button from the outside; it actually snugs into a specific groove on the doorbell.

• The Analogy: Imagine the doorbell has a little keyhole (a hydrophobic groove) that was previously known to be the spot where the "trash" (Amyloid) fits.
• The Finding: The computer showed that VG-3927 fits right into this same keyhole. It's like the drug and the trash are fighting for the same parking spot.

The Experiment: The "Greasy" Problem

To prove this, the scientists tried to test if the drug actually sticks to the doorbell in a lab setting.

• The Problem: They first tried using a standard cleaning solution (containing a detergent called Tween-20). It was like trying to find a magnet on a fridge covered in oil. The detergent was so "greasy" that it filled up the keyhole itself, blocking the drug from getting in. The test failed.
• The Fix: They switched to a different solution (PEG-400) that is more like water and doesn't fill the keyhole. Suddenly, the drug stuck perfectly!
• The Lesson: This proved the drug really does go into that specific greasy groove. It also taught scientists that when testing these drugs, you have to be very careful about what "soap" you use, or you might miss the answer.

The Showdown: Drug vs. Trash

Now that they knew the drug and the trash use the same spot, they asked: What happens if both are there?

1. The Competition: When they put the drug and the trash together, the trash pushed the drug out of the way. The drug couldn't ring the doorbell as loudly.
2. The Cell Test: They tested this in a living cell system. When they added the trash (Amyloid), the drug (VG-3927) had to be used in much higher doses to get the same effect. It was like the trash was blocking the door, so you had to ring the bell much harder to get the same attention.
3. The Signal: When the drug was blocked by the trash, the signal to the cleanup crew (a chemical message called p-SYK) became weaker.

What Does This Mean for Patients?

This is a double-edged sword, but mostly a helpful discovery.

1. The "Traffic Jam" Theory: In an Alzheimer's brain, there is a lot of Amyloid trash. This trash might be blocking the drug from working as well as we hoped because they are fighting for the same spot on the doorbell.
2. The "Dual-Mode" Superpower: However, the scientists realized the drug is smart. Even if the trash blocks the "keyhole" (the ectodomain), the drug can still hold onto the doorbell from the other side (the transmembrane part) and keep the doorbell ringing, just not as loudly. It has a backup plan.
3. The Future Strategy: This suggests that the best way to treat Alzheimer's might be a two-pronged attack:
   • Use a drug to lower the trash (Amyloid) so it stops blocking the doorbell.
   • Use the VG-3927 drug to ring the doorbell and wake up the cleanup crew.

The Bottom Line

This paper reveals that the new Alzheimer's drug VG-3927 works by fitting into the exact same spot on the brain's immune sensor that the disease-causing plaque uses. While the plaque can get in the way and make the drug less effective, the drug still works partially.

The takeaway: To get the best results, doctors might need to combine this drug with other treatments that clear away the plaque first, ensuring the drug has a clear path to ring the doorbell and wake up the brain's cleanup crew.

Technical Summary

1. Problem Statement

Triggering Receptor Expressed on Myeloid Cells 2 (TREM2) is a critical microglial receptor linked to Alzheimer's Disease (AD) risk. VG-3927 is the first small-molecule TREM2 agonist to enter clinical development. While it is proposed to function as a "molecular glue" and positive allosteric modulator (PAM) that promotes receptor clustering, the precise atomic-level binding site of VG-3927 on the TREM2 ectodomain has remained unknown.

A key biological question is whether VG-3927 engages the same ligand-recognition surface on the TREM2 ectodomain used by endogenous AD-related ligands, specifically Amyloid-β (Aβ). If VG-3927 and Aβ compete for the same binding site, the high local concentration of Aβ in the AD brain could potentially interfere with the therapeutic efficacy of VG-3927.

2. Methodology

The authors employed a multi-modal approach combining computational modeling, biophysical binding assays, and cell-based functional assays to map the binding site and assess ligand competition.

• Computational Docking: Used DiffDock-L, a deep learning-based blind docking algorithm, to predict VG-3927 binding poses across the entire TREM2 structure without prior knowledge of the binding pocket.
• Microscale Thermophoresis (MST):
  • Buffer Optimization: Tested binding in buffers containing 0.05% Tween-20 (standard non-ionic detergent) versus 0.05% PEG-400 (hydrophilic).
  • Direct Binding: Measured the interaction between VG-3927 and the recombinant TREM2 ectodomain.
  • Competition Assays: Pre-incubated TREM2 with Aβ1–42 to determine if Aβ interferes with VG-3927 binding.
• Cell-Based Functional Assays:
  • NFAT Reporter Assay: Used Jurkat cells expressing human TREM2, DAP12, and an NFAT-luciferase reporter to measure downstream signaling. Dose-response curves for VG-3927 were generated alone and in the presence of 1 µM Aβ.
  • Western Blotting: Measured phosphorylation of Spleen Tyrosine Kinase (p-SYK), an immediate downstream effector of TREM2 signaling, in cells treated with VG-3927 ± Aβ.

3. Key Contributions

• Identification of a Novel Binding Mode: Provided the first experimental evidence that VG-3927 binds directly to the TREM2 ectodomain hydrophobic groove (formed by CDR1, CDR2, and CDR3 loops), a site previously characterized as the binding pocket for Aβ and apolipoproteins (apoE).
• Methodological Insight: Demonstrated that standard detergents like Tween-20 can occupy hydrophobic pockets and mask ligand binding in biophysical assays, advocating for the use of PEG-400 in similar studies involving hydrophobic targets.
• Mechanism of Interference: Established that Aβ and VG-3927 share an overlapping binding surface, leading to competitive interference in the AD microenvironment.

4. Key Results

A. Computational Prediction (DiffDock-L)

• Blind docking predicted a high-confidence binding site for VG-3927 within the TREM2 ectodomain hydrophobic groove.
• Interactions: The bipyridine core of VG-3927 engages in $\pi$–$\pi$ T-shaped stacking with Trp44. The 2-fluoro-4-chlorophenyl moiety forms alkyl/$\pi$-alkyl interactions with Leu71, Phe74, and Leu89.
• Steric Clash: An unfavorable steric contact was noted between the fluorine atom of VG-3927 and Leu89, suggesting a potential site for medicinal chemistry optimization.

B. Biophysical Binding (MST)

• Detergent Sensitivity: No binding was detected in Tween-20 buffer, likely due to detergent competition at the hydrophobic pocket.
• Successful Detection: Robust binding was confirmed in PEG-400 buffer, yielding a measurable interaction.
• Competition: Co-incubation with 10 µM Aβ1–42 significantly reduced the VG-3927 thermophoretic signal, confirming that Aβ competes for the same binding site.
• Aβ Affinity: Direct binding of Aβ1–42 to TREM2 was confirmed with a $K_d$ of 1.85 ± 0.24 µM.

C. Functional Signaling (NFAT & p-SYK)

• Dose-Response Shift: In the NFAT reporter assay, co-treatment with 1 µM Aβ caused a rightward shift in the VG-3927 dose-response curve.
  • $EC_{50}$ increased from 88.2 nM (VG-3927 alone) to 351.2 nM (with Aβ), a ~4-fold increase.
  • Maximal efficacy was preserved at high VG-3927 concentrations, supporting a competitive antagonism mechanism.
• Signaling Attenuation: Western blot analysis showed that Aβ reduced VG-3927-induced phospho-SYK levels, indicating that Aβ occupancy dampens the downstream signaling output of the agonist.

5. Significance and Implications

• Dual Binding Mode: The study proposes that VG-3927 possesses a dual mechanism of action:
  1. The previously proposed transmembrane "molecular glue" activity (promoting clustering).
  2. A newly identified ectodomain binding mode within the hydrophobic groove.
• Context-Dependent Efficacy: In the AD brain, high local concentrations of Aβ near plaques may partially occupy the TREM2 ectodomain, interfering with VG-3927 binding and reducing its potency. However, because VG-3927 retains its transmembrane clustering function, it likely retains partial efficacy even when the ectodomain is occupied.
• Therapeutic Strategy: The findings support combination therapies. Pairing TREM2 agonists with Aβ-lowering agents (e.g., anti-Aβ antibodies like lecanemab) could clear Aβ from the microenvironment, reducing ligand competition at the ectodomain and potentially enhancing the therapeutic efficacy of VG-3927.
• Drug Design: The identification of the specific hydrophobic groove and the steric clash at Leu89 provides a structural basis for designing next-generation agonists that either avoid Aβ competition or optimize binding affinity to this specific pocket.

In conclusion, this work refines the pharmacological model of VG-3927, revealing that its efficacy in AD may be modulated by the amyloid burden via direct competition at the TREM2 ectodomain.