AF710B

Anavex's M1-PAM + sigma-1 dual agonist, originally synthesized at Israel's IIBR by Abraham Fisher's group, then licensed worldwide to Anavex in 2014. In aged Alzheimer's-model rats it reverted cognition + amyloid + tau + inflammation, with effects persisting 5 weeks after dosing stopped — the rodent profile is the most "disease-modifying" data in this watch-list cluster. Translation is the problem: the program has now been in human trials since ~2019 and the only efficacy readout (Oct 2025 Phase 2 in schizophrenia) was a safety/biomarker trial that didn't formally hit cognition, while the sister sigma-1 compound blarcamesine got an EMA refusal recommendation Dec 11 2025. For Dylan: SKIP-FOR-NOW — interesting mechanism, no clinical translation, no community dosing, no realistic vendor route.

Two receptors, one molecule — pitched as disease-modifying rather than symptomatic.

1. M1 muscarinic positive allosteric modulator (PAM)

The M1 muscarinic acetylcholine receptor is concentrated on cortical and hippocampal pyramidal neurons and is the primary postsynaptic cholinergic site for memory and attention. Cholinergic deficit — loss of basal-forebrain cholinergic neurons + reduced cortical ACh — is the oldest validated lesion in Alzheimer's disease (the "cholinergic hypothesis," Bartus et al. 1982). Donepezil, rivastigmine, galantamine all work by raising synaptic ACh via cholinesterase inhibition; they help symptomatically but do not slow disease progression.

AF710B takes a different angle: instead of raising ACh, it makes the M1 receptor more responsive to whatever ACh is left. As an allosteric PAM, it binds at a site distinct from the orthosteric ACh binding pocket and increases the affinity + efficacy of native ACh at M1. Two consequences:

• It amplifies endogenous signaling rather than producing tonic activation — so when ACh is low (as in AD), AF710B-bound M1 still fires; when ACh is normal (other brain regions, normal cells), the amplification is proportional.
• It is selective for M1 and avoids M2/M3 — sidestepping the GI / cardiac / sweating side effects that killed earlier orthosteric M1 agonists like xanomeline-monotherapy (xanomeline is now back as KarXT/Cobenfy combined with trospium to block peripheral muscarinic effects, FDA-approved Sept 2024).

Downstream: M1 activation drives Gq/PLC/IP3/DAG → PKC → ERK1/2 → CREB phosphorylation. AF710B has been shown at low concentrations to potentiate carbachol-induced phospho-ERK1/2 and phospho-CREB at M1, and these are exactly the cascades that drive synaptic plasticity, BDNF transcription, and (mechanistically) reduced amyloidogenic APP processing (M1 activation shifts APP cleavage toward the non-amyloidogenic α-secretase pathway).

2. Sigma-1 receptor agonist

Sigma-1 is not a classical neurotransmitter receptor — it's a chaperone protein at the mitochondria-associated ER membrane (MAM). Under cellular stress, sigma-1 dissociates from BiP/GRP78 and stabilizes IP3R3 calcium channels at the ER-mito interface, supporting mitochondrial calcium uptake → ATP production → cell survival. Sigma-1 also chaperones misfolded proteins, promoting their proper folding or degradation, and modulates ER stress / unfolded protein response signaling.

In AD, ALS, Huntington's, and stroke models, sigma-1 agonists (pridopidine, blarcamesine/ANAVEX 2-73, AF710B itself) reduce protein-misfolding stress, restore ER-mito calcium handling, and reduce neuroinflammation. The blarcamesine readouts are the most clinically advanced sigma-1 data — and they're now mixed (positive Phase 2b/3 by company analysis, negative by EMA reading).

Convergence

The dual-receptor pitch: M1 amplification fixes the symptomatic cholinergic deficit and shifts APP toward non-amyloidogenic processing, while sigma-1 agonism addresses the upstream proteostasis / mitochondrial / ER-stress axis. This is why the rat data showed simultaneous improvement in cognition (M1 side) AND amyloid pathology + neuroinflammation + synaptic markers (sigma-1 + M1-APP-processing side) — and why effects persisted after washout: structural disease modification is the claim, not transient receptor activation.

Plain English: It does two things at once. It makes the brain's "memory neurotransmitter" receptor more sensitive (helping cognition while ACh-producing neurons are still alive), and it props up the cellular machinery that handles misfolded proteins and energy production (slowing the underlying disease). The combination in animals looks like the holy grail. In humans, we don't yet know if either half translates.