Identifiability Boundaries in Dynamic GPCR Pharmacology: When Desensitization, Internalization and Persistent Signaling Become Mechanistically Indistinguishable
Year: 2026
Type: preprint
Abstract: Dynamic GPCR pharmacology increasingly relies on time-resolved signaling to infer receptor regulation, trafficking, and ligand residence, yet a kinetic response need not uniquely identify the mechanism that produced it. We developed minimal quantitative receptor-pharmacology models to determine when desensitization, internalization, persistent intracellular signaling, and slow receptor occupancy are distinguishable from a common downstream signal. We first derived exact input–output equivalences. Silent internalization is mathematically identical to desensitization for arbitrary time-varying receptor drive, whereas internalization with persistent signaling is exactly equivalent to a desensitization model under any single constant drive after an explicit parameter transformation. Across a 1,000system parameter ensemble, pulse–wash protocols broke the constant-drive equivalence but remained practically ambiguous after adversarial refitting: in 500 held-out systems, the wrong desensitization model had a median normalized error of 1.26% when downstream signal decay (kD) was unconstrained. Independent calibration of kD increased this to 11.90% and increased the fraction exceeding a 5% discrepancy from 11.4% to 80.0%. At 5% Gaussian observation noise, correct-model selection by BIC increased from 48.8% to 95.8%. An analytical kinetic blind manifold, kD = kr + qki, predicted where downstream-only discrimination failed. A single surface-receptor observation supplied substantially greater mechanism-specific information, with 95.5% of systems showing >20% surface loss after 5 min. Slow receptor occupancy also mimicked persistent intracellular signaling, but false trafficking fits made strongly divergent surface-state predictions. Published β2-adrenoceptor, AM2 receptor, and VIPR1 experiments provide independent pharmacological examples of the predicted information constraints and discriminating measurements. The resulting quantitative framework shows that mechanistic discrimination may depend more on independently constraining a downstream nuisance process or hidden receptor state than on collecting additional samples of the same downstream response.
Source: ChemRxiv
Authors: Abdel-Motaal M. Fouda, Dina A. Fouda
Institution: Mansoura University
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7208737529