Acetylcholine and Attention: Part Two: The Division of Labour

Abstract

Part One described acetylcholine in the hippocampus as a signal that sets the operating mode of a circuit rather than carrying content. This part follows the same system outward, into cortex.

A 2018 review by Nicola Solari and Balázs Hangya asks what the division of labour is within the cholinergic projection system — and finds that one half of it has been studied extensively while the other has barely been examined.

It also contains a methodological caution about the field's principal tool that changes how a large body of earlier work should be read.

1.Two Projection Systems

The basal forebrain cholinergic system is not one pathway. The medial septum projects to the hippocampus; a posterior division of the basal forebrain projects to neocortical areas.

The septo-hippocampal pathway has been investigated in many studies, which the authors note is unsurprising given the hippocampus receives its main subcortical input from it. Much less is known about the parallel system linking the posterior basal forebrain with neocortex.

That asymmetry matters, because the cortical areas involved — the posterior parietal cortex and the retrosplenial cortex — do work the hippocampus does not.

2.What Acetylcholine Is Implicated In

The list the authors give is broad and worth reproducing, because it shows how little the word cholinergic narrows anything down.

Arousal. Attention. Sensory processing. Reinforcement expectation. Reward and addiction. And learning and plasticity — where septal cholinergic inputs modulate several forms of hippocampal synaptic plasticity with what the review calls remarkable temporal precision, coordinating presynaptic and postsynaptic activity.

Temporal precision is the interesting term there. It sits awkwardly beside the global, diffuse release described in Part One, and the tension between those two characterisations is one of the field's live problems rather than something either review resolves.

Note also what is absent from the list: memory storage. Acetylcholine appears throughout as a modulator of when and how well information is taken in, not as a substrate in which it is held.

3.Navigation as the Test Case

The review uses spatial navigation because it is a behaviour complex enough to require several structures and precise enough to be measured.

Neuroimaging and lesion studies implicate an intricate network: hippocampus, entorhinal cortex, parahippocampal gyrus, medial and right inferior parietal cortex, regions of prefrontal cortex, cerebellum, parts of the basal ganglia, posterior cingulate and retrosplenial cortex.

The authors focus on the differential roles of hippocampus, posterior parietal cortex and retrosplenial cortex — and on how cholinergic input to each may regulate different aspects of learning, memory and navigation.

The framing is regulation of a distributed system, not enhancement of a faculty. Which is the distinction this whole topic turns on.

4.The Tool Problem

Most of what is known about cholinergic function comes from removing cholinergic neurons and observing what breaks. The standard method uses 192 IgG-saporin — a ribosome-inactivating protein linked to an antibody against the p75 nerve growth factor receptor, which cholinergic neurons express.

The review's cautions about it are specific and consequential.

Small injections produce incomplete lesions. Large doses affect other cell types, including parvalbumin-expressing GABAergic neurons. Injections often spread into the striatum, where they could affect cholinergic interneurons. Intraventricular injections destroy tissue more broadly still. And the p75 receptor is expressed by cholinergic neurons but also by Purkinje cells in the cerebellum.

So the method is either too weak to remove what it targets, or strong enough to remove other things as well. Findings attributing a behavioural deficit specifically to loss of acetylcholine inherit that ambiguity, and it compounds the co-release problem from Part One, where the same neurons also release GABA.

This is not a reason to dismiss the literature. It is a reason to hold its conclusions more loosely than review articles written for general audiences usually do.

5.Where Two Reviews Leave It

Acetylcholine sets circuit states — high during theta and exploration, low during ripples and consolidation. It acts through receptors with opposite signs on different cell types. It is co-released with GABA by the same neurons. Its projection system has two arms, one well studied and one not.

And the principal method used to establish all of this removes more than it intends to.

That is a mechanism worth understanding and a poor foundation for a promise. There is no step in it at which more cholinergic activity is straightforwardly better, and no measurement in this literature that anyone can perform on themselves.

Editorial Comment

Clarity contains CDP-choline, and choline is the substrate from which acetylcholine is synthesised. That connection is why this topic sits in our library, and it is exactly the connection we are not permitted — and do not wish — to convert into a claim.

So, precisely. Neither CDP-choline nor any other ingredient we use holds an authorised health claim in the European Union relating to acetylcholine, attention, memory or cognition. Nothing in these two parts shows that supplemental choline changes cortical or hippocampal cholinergic signalling in a healthy adult, and we are not aware of evidence that it does. Describing a neurotransmitter system is not describing our product.

The reason to read this material is the opposite of a reason to buy something. Once you know that acetylcholine selects between encoding and consolidation, the phrase "supports acetylcholine" stops sounding like a benefit and starts sounding like an unanswered question: raised when, and at the cost of which other state?

A note on sources. This topic was filed with three reviews, all closed access with no open copy obtainable — including the 2004 chapter in which Michael Hasselmo set out the high-acetylcholine-for-encoding and low-acetylcholine-for-consolidation hypothesis. We substituted two open reviews that cover the same ground with more recent data, and the encoding-consolidation framework appears here through the theta and sharp-wave-ripple findings described in Part One rather than from Hasselmo directly.