In Silico Bias Profiling of LSD at the 5‑HT₂A Receptor

Background

The 5‑HT₂A receptor (PDB 6A93) is a G protein‑coupled receptor that couples to Gq/11 (activating PLC and Ca²⁺ release) and recruits β‑arrestin (leading to desensitization/internalization). LSD is known to engage both pathways with a behavioral bias toward β‑arrestin in vivo.

Objectives

1. Compare Apo vs. LSD‑Bound Ensembles
   • Generate ≥10 independent 10 ns replicates per condition (minimization, 150 ps NVT, 10 ns NPT).
   • Sample ~5 000 frames per replicate.
2. Extract Structural Markers
   • Ionic Lock (R³·⁵⁰–E⁶·³⁰ Cα distance)
   • TM6 Movement (R³·⁵⁰–L⁶·³⁴ distance)
   • NPxxY Dihedral (N7.49–P7.50–x–Y7.53)
   • H8 Tilt & SASA (helix 8 exposure)
   • C‑tail SASA
3. Principal Component Analysis & Clustering
   • Build feature vectors per frame.
   • Perform PCA on production trajectories.
   • Cluster end‑state conformations, compare populations apo vs. bound.
4. Statistical Analysis
   • Per‑replicate means and 95% CIs for each metric.
   • Endpoint‑state fractions (active‑like vs. inactive) with Wilson intervals.
   • Group comparisons via t‑tests or nonparametric alternatives.
5. Representative Arrestin‑Prone Conformation & Phosphorylation
   • Identify “arrestin‑like” frames (all markers above thresholds).
   • Cluster arrestin‑like pool; select centroid.
   • Model key C‑tail phosphorylations (S380, T382, S384, T386) in CHARMM‑GUI.
   • Run short MD to validate stabilization of arrestin signature.

Expected Outcomes

• Quantitative bias factor (Gq vs. β‑arrestin) for LSD at 5‑HT₂A.
• Structural ensemble shifts in PCA space upon LSD docking.
• In silico demonstration that phosphorylation locks in β‑arrestin–biased conformations.

Future Directions

• Long‑timescale enhanced sampling (umbrella/PLUMED) along TM6 and NPxxY CVs.
• In vitro validation via IP₃ assays and arrestin BRET.
• Extension to other psychedelics and receptor subtypes.

Photos