Cholinergic modulation of the hippocampal region and memory function

Abstract

Acetylcholine (ACh) plays an important role in memory function and has been implicated in aging-related dementia, in which the impairment of hippocampus-dependent learning strongly manifests. Cholinergic neurons densely innervate the hippocampus, mediating the formation of episodic as well as semantic memory. Here, we will review recent findings on acetylcholine’s modulation of memory function, with a particular focus on hippocampus-dependent learning, and the circuits involved. In addition, we will discuss the complexity of ACh actions in memory function to better understand the physiological role of ACh in memory.

The neuromodulator acetylcholine (ACh) plays a critical role in memory function, especially in the hippocampus-dependent learning. The cholinergic system is severely affected in Alzheimer’s disease, implicating its role in memory. In this review, the current knowledge on the cholinergic system and its modulation of hippocampal circuits have been reviewed. Furthermore, we describe factors that contribute to the complexity of ACh actions in memory function.

Acetylcholine and Alzheimer’s disease

Since it was first discovered as a neurotransmitter at the neuromuscular junction, ACh has received significant attention as a critical modulator of cognitive functions. One particular reason is that impairment of the cholinergic system often manifests in patients with dementia, including Alzheimer’s disease (AD) (Davies and Maloney, 1976; Whitehouse et al., 1982). AD is a neurodegenerative disease which is characterized by a progressive decline in cognitive functions. AD mainly affects mid- to late-age adults, the impairment of episodic memory is a sign that is prominent from the early stages of AD (Gold and Budson, 2008). A number of studies have shown the atrophy of the cholinergic system in the basal forebrain in early AD patients or subjects with a high risk of developing AD (Grothe et al., 2012; Grothe et al., 2010; Grothe et al., 2014; Teipel et al., 2014). It has been reported that AD not only causes a decrease in the number of cholinergic neurons but also in the levels of choline acetyltransferase (ChAT), an enzyme necessary for synthesizing ACh in the basal forebrain (Davies and Maloney, 1976; Francis et al., 1999; Perry et al., 1977; Whitehouse et al., 1982). In addition, alterations in the function of muscarinic as well as nicotinic ACh receptors have been implicated in the pathophysiology of AD (Ikonomovic et al., 2009; Jiang et al., 2014; Wang et al., 2009; Zuchner et al., 2005). As such, administration of nicotine improves cognitive functions in elderly subjects who are prone to memory problems (Howe and Price, 2001; Min et al., 2001; White and Levin, 2004). Furthermore, elevation of ACh levels via blockade of acetylcholinesterase (the enzyme that breaks down ACh) is a method often used to treat AD patients (Ehret and Chamberlin, 2015). On the other hand, anticholinergic medications, which are often prescribed for gastrointestinal disorders and dizziness, can cause dementia, suggesting a role of the cholinergic system in memory (Gray et al., 2015; Kalisch Ellett et al., 2014).

Although the exact pathophysiology of AD is still not clear, extensive human and animal studies have suggested that accumulation of amyloid β peptide (Aβ) in the extracellular space, and neurofibrillary tangles in the intracellular space, are strongly related to the development of AD (for reviews, see Huang and Mucke, 2012; Kumar et al., 2015). In humans, it has been shown that Aβ₄₂, which is the predominant form of the amyloid β peptide in humans (Gouras et al., 2000), accumulates in cholinergic neurons of the basal forebrain even in young brains, and that the intermediate and heavy forms of Aβ₄₂ is increased with aging and in AD (Baker-Nigh et al., 2015). Interestingly, previous studies have indicated that there is a close relationship between Aβ accumulation and cholinergic dysfunction; Aβ suppresses the synthesis and release of ACh (Pedersen et al., 1996), interferes with cholinergic receptor signaling (Janickova et al., 2013; Mura et al., 2012), and causes a decrease in the number of cholinergic neurons (Zheng et al., 2002). In addition, Aβ has been shown to act as an allosteric modulator that facilitates the acetylcholine hydrolyzing enzyme butylcholinesterase (Darreh-Shori et al., 2011; Kumar et al., 2016). Dysfunction of cholinergic signaling has also been linked to impaired RNA processing, leading to the loss of dendrites in cortical neurons and the upregulation of BACE1 protein, which has been shown to be elevated in late stages of AD (Berson et al., 2012; Kolisnyk et al., 2016).

The hippocampus and its function are significantly affected by cholinergic dysfunction (Berger-Sweeney et al., 2001; Blokland et al., 1992; Opello et al., 1993; von Linstow Roloff et al., 2007). It has been shown that cholinergic projections from the medial septum area to the hippocampus are significantly reduced in AD patients and in a mouse model of AD (Belarbi et al., 2011; Davies and Maloney, 1976). In addition, a reduction in the levels of cholinergic receptors in the hippocampus and decreased ACh binding in this region have been observed in patients with AD (Parent et al., 2013; Shiozaki et al., 2001).

Acetylcholine and its receptors

Acetylcholine is a versatile molecule that acts not only as a neurotransmitter, but also as a neuromodulator in the nervous system (for review, see Picciotto et al., 2012). Its role as a neuromodulator has received particular attention because of the significant implication for cognitive functions. Neurons that produce ACh (i.e. cholinergic neurons) are found throughout the brain, in particular in the brain stem, basal forebrain, stratum and medial habenular nucleus (Frahmm et al., 2015). In the brain stem, cholinergic neurons are located in the pedunculopontine nucleus and laterodorsal tegmental nucleus, which project to the thalamus, basal ganglia, tectum and basal forebrain (Armstrong et al., 1983; Mesulam et al., 1983; Woolf and Butcher, 1986). In the basal forebrain, cholinergic cell bodies are found not only in the medial septum and diagonal band of Broca, which provide major inputs to the hippocampus, but also in the nucleus basalis in the substantia innominata, which project to the neocortex (Armstrong et al., 1983; Kimura et al., 1980; Mesulam et al., 1983; Rye et al., 1984). The selective lesion of cholinergic neurons in the basal forebrain significantly impairs hippocampus-dependent memory function, suggesting a role for septohippocampal cholinergic projections in memory formation (Berger-Sweeney et al., 2001; Nilsson et al., 1992). Consistently, selective elimination of the vesicular acetylcholine transporter (VAChT) in the forebrain, which is required for the release of ACh at synaptic terminals, causes an impairment in synaptic plasticity in the hippocampus and RNA processing in the prefrontal cortex (Kolisnyk et al., 2013a; Martyn et al., 2012). Furthermore, these VAChT-deficient animals show an impairment in the acquisition and extinction of spatial memory as well as in high attention tasks (Kolisnyk et al., 2013a; Martyn et al., 2012; Prado et al., 2017).

There are two main classes of cholinergic receptors: nicotinic receptors, which are ligand-gated ion channels in the cys-loop receptor family, and muscarinic receptors, which are G protein-coupled receptors (for review, see Albuquerque et al., 2009; Fryer et al., 2012). Both nicotinic and muscarinic receptors are expressed in the central as well as in the peripheral nervous system (Rubboli et al., 1994; Tice et al., 1996). In the brain, the ACh receptors are expressed both pre- and postsynaptically to mediate the neuromodulatory functions of ACh (Disney et al., 2006; Xu et al., 2006).

Nicotinic receptors are excitatory cationic ion channels; the activation of these receptors induces membrane depolarization (for review, see Albuquerque et al., 2009). Nicotinic receptors consist of five subunits, the combination of which determines the pharmacological and physiological characteristics of the receptor subtype (for review, see Dineley et al., 2015). Currently twelve different subunits (α2-α10 and β2-β4) are known to exist (Elgoyhen et al., 1994; Elgoyhen et al., 2001; Sargent, 1993), among which nine subunits (α2-α7 and β2-β4) are expressed in the hippocampus (Sudweeks and Yakel, 2000). In the hippocampus, the most abundant nicotinic receptor subtypes are α7 homomeric (although these α7 receptors are known to be heteromeric under certain conditions) and α4β2 heteromeric receptors, which are expressed in pyramidal neurons as well as in GABAergic interneurons (Alkondon and Albuquerque, 1993; Dineley et al., 2015; Jones and Yakel, 1997; McQuiston and Madison, 1999b; Sudweeks and Yakel, 2000). The α4β2 nicotinic receptors have the highest affinity to nicotine and have been shown to mediate nicotine-induced synaptic plasticity in the hippocampus (Alkondon et al., 1997; Bell et al., 2011; Flores et al., 1992; Sudweeks and Yakel, 2000). The α7 nicotinic receptor is different from other nicotinic receptors in that it is highly permeable to Ca²⁺, which plays a unique role in synaptic modulation (Albuquerque et al., 1997; Cheng and Yakel, 2014; Frazier et al., 1998; Gu and Yakel, 2011; Ji et al., 2001; Patrick et al., 1993).

There are five subtypes of muscarinic receptors, M1 through M5, among which M1, M3, and M5 muscarinic receptors are coupled to G_q and are excitatory (for review, see Fryer et al., 2012). On the contrary, M2 and M4 muscarinic receptors are coupled to G_i/o and are inhibitory (Akam et al., 2001). In the hippocampus, it has been shown that the M1, M2 and M4 muscarinic receptors are expressed pre- or post-synaptically (Levey, 1996). M1 muscarinic receptors are the most abundant subtype in the hippocampus and are mainly expressed in the dendrites or somas (Flynn et al., 1995; Levey, 1996; Wall et al., 1991); the M1 muscarinic receptors play a critical role in regulating the excitability of hippocampal neurons (Dasari and Gulledge, 2011; Lawrence et al., 2006; Yi et al., 2014). M2 and M4 muscarinic receptors are mainly expressed at the synaptic terminals and modulate neurotransmitter release (Dasari and Gulledge, 2011; Levey, 1996; Zheng et al., 2012).

Acetylcholine modulation of circuits in the medial temporal lobe

Acetylcholine has been shown to have extensive effects on neuronal circuits by affecting neurogenesis, spine formation, and synapse formation; these long-term changes require protein synthesis (Collo et al., 2013; Lozada et al., 2012; Paez-Gonzalez et al., 2014). Acetylcholine also exerts acute effects on synaptic plasticity by modulating spiking activity of neurons and neurotransmitter release (Aracri et al., 2010; Dannenberg et al., 2015; Maggi et al., 2001; McQuiston and Madison, 1999a). In this section, we will focus on the rapid actions of ACh, which are more closely correlated with changes in extracellular ACh levels in the medial temporal lobe during learning.

Within the hippocampus

It has been shown that cholinergic neurons in the medial septum regulate hippocampal circuits. Optogenetic stimulation of cholinergic neurons in the medial septum area not only causes changes in the firing activity of hippocampal neurons but also modulates theta-band oscillations in the hippocampus in vivo (Dannenberg et al., 2015). Experimental and computer modeling studies have shown that ACh specifically inhibits intrinsic pathways, which are part of the memory consolidation circuits (Hasselmo, 1999), while facilitating afferent projections, which are part of the encoding pathway (Newman et al., 2012). Acetylcholine inhibits the recurrent pathway in the CA3 region via the activation of muscarinic ACh receptors in interneurons (Hasselmo and Bower, 1992; Hasselmo et al., 1995). This ensures that the circuits that carry extrinsic information are preferentially activated, while the intrinsic projections are toned down (Hasselmo, 1999).

In the hippocampal CA1 region, ACh is known to potentiate the Schaffer collateral pathway, via the activation of α7 or non- α7 nicotinic ACh receptors located in pyramidal neurons and GABAergic interneurons (Fujii et al., 1999; Fujii et al., 2000a; Fujii et al., 2000b; Leao et al., 2012; Mann and Greenfield, 2003; Nakauchi et al., 2007). However, these results are controversial. For example, other studies have shown that the Schaffer collateral pathway is instead inhibited by ACh (Dasari and Gulledge, 2011; Hasselmo and Schnell, 1994; Herreras et al., 1988; Mans et al., 2014; Sheridan and Sutor, 1990). One explanation for this discrepancy is that the effect of ACh on synaptic plasticity is timing-dependent. It has been shown that cholinergic input can cause either long-term potentiation or short-term depression, depending on the timing of cholinergic input relative to glutamatergic input to the CA1 (Gu and Yakel, 2011; Ji et al., 2001). In addition, the effect of ACh may vary depending on which cholinergic receptor subtype is activated in different conditions. Furthermore, an electrophysiological study has shown that high frequency stimulation of cholinergic neurons can induce a depolarizing response in hippocampal interneurons, whereas low frequency stimulation of cholinergic activations leads to a hyperpolarizing response via activation of different subtypes of ACh receptors (Bell et al., 2013).

In the dentate gyrus, ACh has been shown to increase long-term potentiation via activation of nicotinic and muscarinic receptors (Luo et al., 2008; Matsuyama et al., 2000; Sawada et al., 1994; Wang et al., 2006; Welsby et al., 2009). In addition, septal cholinergic projections have been shown to activate astrocytes to modulate dentate granule cells (Pabst et al., 2016).

Parahipocampal region

Acetylcholine also has been shown to modulate the neuronal activity in the entorhinal cortex (EC), which acts as the interface between the neocortex and hippocampus. ACh exerts differential effects depending on the circuits in the EC. In the superficial layers of the EC (layer I, II and III), to where cortical neurons that receive extrinsic signals project, ACh has been shown to cause an increase in neuronal spiking and facilitate synchronized oscillatory activity (Dickson and Alonso, 1997; Klink and Alonso, 1997). While the superficial layer projects to the hippocampal area as a part of the memory encoding pathway and is activated during waking, the deep EC layers (layer V and VI in rodents) receive hippocampal inputs and project back to the neocortex for the consolidation of memory. The activity of deep layer EC neurons is increased during sharp wave sleep when memory consolidation is the most active (Chrobak and Buzsaki, 1994). Extracellular recording studies also suggest that the rhythmic activity in the deep EC layers is correlated with the hippocampal activity during slow wave sleep, but not during waking. Our recent unpublished study suggests that ACh activates hippocampal interneurons to suppress hippocampal outputs to the deep layers of EC (Haam and Yakel, 2015, 2016). How cholinergic inputs to the hippocampal area modulate the activity will be an interesting question to explore.

Complexity of acetylcholine modulation of memory function

Previous studies have shown that drugs that modulate extracellular ACh levels or its receptor activity have significant effects on memory functions, but these effects are rather complex. Examples are the drugs for treating Alzheimer’s Disease such as donepezil, rivastigmine and galantamine, which block the breakdown of ACh by inhibiting acetylcholinesterase. These drugs improve memory function in some AD patients but have minimal effects in others (Bartus, 2000; Ehret and Chamberlin, 2015; Giacobini, 2000). ACh’s effect on memory is complex for several reasons (Fig. 1). First, ACh modulation of memory function is selective to only certain types of learning (i.e. hippocampus-dependent learning). Generally declarative memory (such as episodic and semantic memory) has been shown to be dependent on the hippocampus whereas procedural memory (i.e. implicit memory) is independent of the hippocampus (for review, see (Bird and Burgess, 2008)). Previous studies using cholinergic receptor antagonists or choline transport inhibitors suggest that cholinergic receptor activation is critical for episodic and spatial memory, but not for procedural memory (Blokland et al., 1992; Caine et al., 1981; Opello et al., 1993; von Linstow Roloff et al., 2007). In addition, when cholinergic neurons in the medial septum are specifically depleted, it causes impairment of spatial memory while sparing non-spatial memory (Berger-Sweeney et al., 2001; Nilsson et al., 1992). These findings are consistent with the fact that basal forebrain cholinergic populations, which provides major inputs to the hippocampus (Armstrong et al., 1983; Kimura et al., 1980; Mesulam et al., 1983; Rye et al., 1984), are selectively impacted in AD (Whitehouse et al., 1982), which manifests the impairment of hippocampus-dependent memory (Hirono et al., 1997; Libon et al., 1998), supporting the idea that ACh in the basal forebrain modulates hippocampus-dependent learning.

Fig.1. Complexity of acetylcholine (ACh) modulation on cognitive function.

(a) ACh’s modulation of cognitive function is selective to hippocampus-dependent memory, specifically affecting spatial but not procedural memory. (b) The method of cholinergic modulation can result in different outcomes. For example, optogenetic manipulation of cholinergic neurons is more physiological but is not as selective as antagonism of ACh receptors (AChRs). (c) ACh has differential effects on memory encoding and consolidation, favoring the pathways involved in encoding. (d) ACh can exert different responses depending on which receptor subtypes ACh activates, each of which has distinct desensitization characteristics. (e) ACh enhancement of memory function depends on the baseline cholinergic activity, having a larger improvement in subjects with low ACh tone.

Second, many studies have manipulated ACh release using various techniques which can lead to different findings. Cholinergic neurons often co-release other neurotransmitters such as GABA and glutamate (Ren et al., 2011; Saunders et al., 2015), complicating the interpretation of the results with manipulation of cholinergic neurons. To selectively eliminate ACh release without disrupting other co-released molecules from cholinergic neurons, VAChT was selectively targeted (Kolisnyk et al., 2013a; Martyn et al., 2012; Prado et al., 2017) or selective ACh receptor antagonists were used (Blokland et al., 1992; Caine et al., 1981; Felix and Levin, 1997; Roloff et al., 2007), which provided distinct results from studies that ablated cholinergic neurons using a cell-type specific immunotoxin (Cai et al., 2012; Moreau et al., 2008; Opello et al., 1993). In addition, in more recent studies, to mimic more physiological manipulation of cholinergic neurons, optogenetic tools were used (Gu and Yakel, 2011; Mamad et al., 2015; Vandecasteele et al., 2014); this allowed the temporary excitation or silencing of neurons in a specific target area without knocking out any cellular components permanently. The optogenetic approach is distinct from other pharmacological, chemogenetic or gene knockout approaches in that it can selectively manipulate cholinergic cell bodies or axonal terminals, the effect of which may be different from the activation of cholinergic tone in a broader area.

Third, the effect of ACh on memory is not always facilitatory but is dependent on the memory stage as mentioned above. Memory formation consists of two stages, memory encoding and consolidation (Buzsaki, 1989; Hasselmo, 1999). During memory encoding, the cortex sends sensory inputs to the hippocampus, whereas the temporary memory is transferred back to the cortex for long-term storage during memory consolidation (Buzsaki, 1989; Chrobak et al., 2000; Hasselmo, 1999; Witter et al., 2000). Memory encoding and consolidation processes can interfere with each other and are therefore temporally separated (Hasselmo, 1999; Rasch and Born, 2013; Stickgold and Walker, 2013). Previous physiological studies suggest that ACh is an important neuromodulator that switches between the two memory formation modes. It has been suggested that ACh favors memory encoding while suppressing the consolidation process (Hasselmo, 1999). The levels of ACh in the hippocampus are elevated during memory encoding, whereas the levels are low during memory consolidation (Hasselmo and McGaughy, 2004; Kametani and Kawamura, 1990; Marrosu et al., 1995). Furthermore, when muscarinic receptors are pharmacologically blocked with scopolamine (either injected peripherally or locally into the medial septum area), it causes impairment in the encoding of memory, but not the maintenance of long-term memory (Bartolini et al., 1992; Givens and Olton, 1994; Green et al., 2005; Newman and Gold, 2016), suggesting the differential effects of ACh on memory encoding and consolidation. (Rasch et al., 2006; Rogers and Kesner, 2003). Consistent with this, when administered after memory encoding, scopolamine did not impair memory retention in human subjects (Ghoneim and Mewaldt, 1977; Petersen, 1977). Likewise, an artificial increase in ACh levels by physostigmine, an acetylcholinesterase inhibitor that increases the extracellular ACh levels, impairs memory consolidation and retrieval in rodent and human subjects (Gais and Born, 2004; Kukolja et al., 2009; Rogers and Kesner, 2003). The confusion often arises from the difficulty in measuring performance in memory consolidation; the total amount of memory encoded would eventually affect the amount of memory consolidated subsequently. ACh’s effect on long-term memory shown in previous studies (Broks et al., 1988; Davis et al., 1978; Drachman, 1977) might be due to the fact that the total amount of memory consolidated is affected by the prerequisite step memory encoding.

Fourth, the complexity may arise from the various receptor subtypes which have different desensitization characteristics. Previous studies have shown that nicotinic ACh receptors exist in at least four different conformations, and prolonged ligand binding eventually renders the receptors to the inactive, desensitized form (Katz and Thesleff, 1957; Ochoa et al., 1989). Among the nicotinic receptors, α7-containing nicotinic ACh receptors, compared to non-α7 nicotinic ACh receptors, have been shown to be particularly susceptible to rapid desensitization. The onset of desensitization of α7 nicotinic receptors is ~ 100 msec whereas that of non- α7 nicotinic receptors is typically 2–20 s (Giniatullin et al., 2005; McGehee and Role, 1995). In addition, desensitization is dependent on ligand concentration as well as the duration of ligand application; the higher the concentration of agonist, the more rapid the onset of desensitization of the receptors (Adams, 1975; Katz and Thesleff, 1957; Sakmann et al., 1980). As such, the effects of high concentrations of nicotinic receptor agonists can sometimes appear to be similar to those of antagonists (Buccafusco et al., 2009), contributing to the complexity of ACh actions in the brain. For example, it has been shown that some of the nicotine-induced health effects, especially at high concentrations, are due to receptor desensitization (Picciotto et al., 2008). Therefore, exogenous nicotine administration may activate nicotinic receptor signaling at low concentrations while inhibiting the signaling via receptor desensitization at high concentrations.

It has been shown that muscarinic ACh receptors can be desensitized as well; however, the desensitization kinetics are significantly slower than that of nicotinic ACh receptors, typically requiring 20 minutes or longer. Applications of high concentrations of muscarinic ACh receptor agonists causes a decrease in sensitivity of the receptor without affecting the maximum response (Griffin et al., 2004). Desensitization of muscarinic ACh receptors has been characterized to be due to the internalization and downregulation of these receptors, and a decrease in the downstream signaling (el-Fakahany and Lee, 1986; Pals-Rylaarsdam et al., 1995; Tobin et al., 1992). Given that the desensitization of muscarinic ACh receptors has slower kinetics without changes in the maximum response, it is less likely that muscarinic ACh receptor agonists and antagonists will produce similar effects as shown in nicotinic receptor signaling.

Last, the cholinergic enhancement of memory function depends on basal cholinergic activity. It has been shown that individuals with a lower baseline performance show more significant cognitive improvement in response to a pharmacological increase in ACh levels, whereas individuals who have a higher baseline performance show less significant effect (Kukolja et al., 2009). In addition, the AD treatment drug donepezil, which increases ACh levels by blocking acetylcholinesterase, has a minimal effect on healthy subjects (Morasch et al., 2015), supporting the idea that the effect of ACh is dependent on basal cholinergic activity (Kukolja et al., 2009). These findings are in line with the notion that many neuromodulators show the inverted U-type concentration-effect curve; the effect is inversely correlated to the concentration when the concentration is higher than the optimal concentration (Diamond et al., 2007; Floresco and Magyar, 2006; Lupien and McEwen, 1997; Williams and Castner, 2006). Furthermore, the inverted U curve has been shown in muscarinic responses (Flood and Cherkin, 1986; Wanibuchi et al., 1994) as well as in nicotinic responses (Flood and Cherkin, 1986; Kolisnyk et al., 2015; Kolisnyk et al., 2013d; Picciotto, 2003; Wallace et al., 2011; Wanibuchi et al., 1994).

However, the cellular mechanism for ACh’s inverted U-type curve remains unclear. The desensitization of receptors has been suggested for the U-type response, especially regarding nicotinic responses (Picciotto, 2003). Furthermore, it has been suggested that an agonist may trigger at least two sets of signaling cascades, with a different sensitivity to the agonist, that oppose each other (De Biasi and Dani, 2011; Woody et al., 1988). Interestingly, differential effects of ACh on memory encoding and consolidation have been shown (Bartolini et al., 1992; Gais and Born, 2004; Givens and Olton, 1994; Green et al., 2005; Kukolja et al., 2009; Newman and Gold, 2016; Rasch et al., 2006; Rogers and Kesner, 2003), indicating that ACh induces multiple signaling pathways that have opposing effects on overall memory function. When ACh levels are significantly low, it will hinder memory encoding whereas memory consolidation will be compromised when ACh levels are too high.

Concluding Remarks

ACh is a neuromodulator that plays a critical role in cognitive function. Previous immunohistochemical, electrophysiological, pharmacological and behavioral studies have shown that ACh modulates the circuits involved in hippocampus-dependent memory. Due to the complexity of ACh actions in the brain, studies on ACh’s role in memory function have often produced controversial results. Comprehending the complexity of ACh actions will allow us to better understand the physiological roles of ACh in cognitive functions and help us to design better treatment options for age-related dementia. During the past couple of decades, there has been great progress in understanding how ACh modulates memory circuits. With the novel tools now available, examining ACh actions at several different levels (cellular, network, and behavioral) will be critical for understanding the complexity of ACh actions.

Abbreviations

Aβ

amyloid β peptide

ACh

acetylcholine

AD

alzheimer’s disease

ChAT

choline acetyltransferase

EC

entorhinal cortex

VAChT

vesicular acetylcholine transporter