Basal Forebrain Cholinergic System in the Dementias: Vulnerability, Resilience and Resistance

Abstract

The basal forebrain cholinergic neurons (BFCN) provide the primary source of cholinergic innervation of the human cerebral cortex. They are involved in the cognitive processes of learning, memory and attention. These neurons are differentially vulnerable in various neuropathologic entities that cause dementia. This review summarizes the relevance to BFCN of neuropathologic markers associated with dementias, including the plaques and tangles of Alzheimer’s disease (AD), the Lewy bodies of diffuse Lewy body disease (DLBD), the tauopathy of frontotemporal lobar degeneration (FTLD-TAU) and the TDP-43 proteinopathy of FTLD-TDP. Each of these proteinopathies has a different relationship to BFCN and their corticofugal axons. Available evidence points to early and substantial degeneration of the BFCN in AD and DLBD. In AD, the major neurodegenerative correlate is accumulation of phosphotau in neurofibrillary tangles. However, these neurons are less vulnerable to the tauopathy of FTLD. An intriguing finding is that the intracellular tau of AD causes destruction of the BFCN whereas that of FTLD does not. This observation has profound implications for exploring the impact of different species of tauopathy on neuronal survival. The proteinopathy of FTLD-TDP shows virtually no abnormal inclusions within the BFCN. Thus, the BFCN are highly vulnerable to the neurodegenerative effects of tauopathy in AD, resilient to the neurodegenerative effect of tauopathy in FTLD and apparently resistant to the emergence of proteinopathy in FTLD-TDP and perhaps also in Pick’s disease. Investigations are beginning to shed light on the potential mechanisms of this differential vulnerability and their implications for therapeutic intervention.

Graphical Abstract

This Review is part of the special issue “Cholinergic Mechanisms” and summarizes the relevance to basal forebrain cholinergic neurons (BFCN) of neuropathologic markers associated with dementias. Available evidence points to early and substantial degeneration of the BFCN in Alzheimer’s disease and diffuse Lewy body disease. BFCN are resilient to the neurodegenerative effect of some tauopathies in frontotemporal lobar degeneration (FTLD), such as that is corticobasal degeneration, and they are apparently resistant to the emergence of proteinopathy in FTLD-TDP and perhaps also in Pick’s disease. These findings have important implications for selective neuronal vulnerability and cholinergic based therapies in dementias.

In 1976, two separate laboratories reported major depletions of the cholinergic synthetic enzyme choline acetyltransferase (ChAT), and the cholinergic hydrolytic enzyme acetylcholinesterase (AChE) within many cortical areas in Alzheimer’s disease (AD) (Davies & Maloney 1976; Bowen et al. 1976). Others demonstrated that cortical cholinergic innervation originates in the ChAT-positive neurons of the basal forebrain (Mesulam & Van Hoesen 1976; Mesulam et al. 1983). Soon thereafter, it became clear that the substantial depletion of cortical cholinergic markers in AD is accompanied by an equally substantial loss of basal forebrain cholinergic neurons (BFCN) (Whitehouse et al. 1981). These observations propelled a decades-long exploration of the basal forebrain cholinergic system in health and disease.

Early vulnerability of the basal forebrain cholinergic system to neurofibrillary tangle pathology and its progressive and severe degeneration in AD are now well established (Geula & Mesulam 1999). Significantly, degeneration of this system in AD is the basis of cholinergic therapy. While not as thoroughly investigated, there is evidence that the BFCN are also vulnerable to degeneration in the dementia caused by diffuse Lewy body disease (DLBD). Less is known about this system in other types of dementias such as those caused by frontotemporal lobar degeneration (FTLD).

The purpose of this review is to summarize available information on the degeneration of the BFCN and their corticofugal axons in distinct neuropathologic entities that cause dementia. We first set the stage by summarizing the characteristics of the relevant neuropathologic entities. We then outline the functional anatomy of the normal BFCN, and its response to the neurodegenerative entities associated with dementia. We conclude with a discussion of potential mechanisms underlying the selective vulnerability of this system in some types of dementia and their therapeutic implications.

PATHOLOGIC HALLMARKS OF NEURODEGENERATIVE DEMENTIAS

The vast majority of neurodegenerative dementias seen in clinical practice are caused by four pathologic entities: AD, DLBD, FTLD with tauopathy (FTLD-Tau), and FTLD with 43-kDa transactive response element DNA-binding protein (TDP-43) proteinopathy (FTLD-TDP). In this review, we will use the acronym AD exclusively to refer to AD neuropathologic changes. We will refer to the clinical manifestations of each neurodegenerative condition by its descriptive cognitive and behavioral phenotype, such as amnestic ‘dementia of the Alzheimer type’ (DAT), primary progressive aphasia (PPA), behavioral variant frontotemporal dementia (bvFTD), corticobasal degeneration syndrome (CBDs) and progressive supranuclear palsy syndrome (PSPs).

Alzheimer’s Disease

In the first report on the disease which now bears his name, Alois Alzheimer described two types of lesions in the brain of his patient: “tangled bundle of fibrils” and “miliary foci resulting from the deposit of a unique substance” (Alzheimer 1977). The terms commonly used today to designate these lesions are the neurofibrillary tangle (NFT) and the senile plaque (SP), respectively (Fig. 1). The characteristic distribution and density of these lesions are used by pathologists for diagnosis of AD (Montine et al. 2012; Hyman et al. 2012; Hyman & Trojanowski 1997; Mirra et al. 1993; Khachaturian 1985).

Figure 1. Pathologic entities associated with dementias.

(A) Thioflavin-S stain visualizes mature extracellular senile plaques and intracellular neurofibrillary tangles in Alzheimer’s disease (AD) cortex. (B) Phosphotau immunoreactivity is present in neurofibrillary tangles and neuropil threads in AD. (C) Plaques in AD cortex display immunoreactivity for the amyloid-β peptide. (D) Dystrophic neurites within mature plaques in AD brains display immunoreactivity for phosphotau. (E) A substantia nigra neuron in a case of diffuse Lewy body disease (DLBD) contains a LB in a hemotoxylin & eosin stained section. (F) α-Synuclein immunoreactivity is present in neurons and neurites in DLBD cortex. (G) Phosphotau immunoreactivity in corticobasal degeneration visualizes inclusions in neurons, neurites and astrocytic plaques (arrow). (H) Cortical Pick bodies containing phosphotau immunoreactivity in Pick’s disease. (I) Phosphorylated TDP-43 immunoreactive inclusions in cortical neurons in TDP-43 proteinopathy.

The plaque is a complex structure found in the neuropil and consists of amyloid, abnormal neurites and glial cells (Terry & Wisniewski 1972) (Fig. 1 A&C). The amyloid-β (Aβ) peptide, which is a major component of all plaques, is a protein of 40–43 amino acids (Iizuka et al. 1995). It is clipped out of a larger amyloid precursor protein through a set of complex proteolytic processes (Selkoe 1994). Aß can exist in various physical conformations, which include small soluble aggregates, large non-fibrillar aggregates, and aggregated fibrillar forms (Klein 2006; Podlisny et al. 1995; Lorenzo & Yankner 1994) (Fig. 1 A). In addition to Aß, abnormal (dystrophic) neurites are associated with a subset of SP, and contain fibrillar elements that may represent degenerating axons and dendrites (Terry & Wisniewski 1972; Trojanowski et al. 1995) (Fig. 1D). The plaques with neurites often have activated microglia and astrocytes associated with them (Terry & Wisniewski 1972). Immunohistochemistry using antibodies to Aβ results in staining of two subtypes (Mann et al. 1990; Wisniewski et al. 1989); the diffuse SP are round or amorphous deposits of aggregated, non-fibrillar Aβ with a granular reaction product and without clear borders (Fig 1 C). The compact SP, on the other hand are well-defined round deposits of darkly stained fibrillar Aβ, and stain positively with thioflavin-S (Fig. 1 A). Compact plaques are frequently associated with dystrophic neurites and glial cells. In addition to the presumptive neurodegenerative effect exerted by fibrillar Aβ upon surrounding neurons and neuropil (Shah et al. 2010), soluble Aβ oligomers may have toxic effects on synapses (Klein 2002; Walsh et al. 2002).

A major component of NFT is the abnormally phosphorylated microtubule associated protein tau (phosphotau). The NFT contain paired helical filaments (PHF) (Terry 1998). They are argentophilic, thioflavin S-positive (Fig. 1 A) and stain immunohistochemically with antibodies against PHF and phosphotau (Terry 1998; Terry et al. 1994) (Fig. 1 B). It is important to note that the neurites within SP (Fig. 1 D), as well as fragmented dendrites and axons in the neuropil (Fig. 1 B), contain components that are identical to those in NFT (Terry & Wisniewski 1972; Trojanowski et al. 1995). NFT appear to be preceded by accumulation of soluble phosphotau in neurons, the so-called pre-tangle, which are thioflavin S-negative (Lauckner et al. 2003). Pre-tangles and NFT are thought to damage neurons by disrupting transport of various cellular components and by displacing cytoplasmic elements and thus leading to the degeneration of the neurons within which they are formed.

Careful mapping of the distribution of cortical SPs and NFTs has revealed different patterns of distribution (Arnold et al. 1991; Arriagada et al. 1992). Plaque deposition is likely initiated in association neocortex where SP density is highest, and gradually spreads to paralimbic and limbic regions. NFT appear first, and display highest density in limbic and paralimbic regions of the medial temporal lobe, namely the entorhinal cortex and hippocampus, followed by other paralimbic and then association neocortex (Braak & Braak 1995). Cortical and subcortical areas with high densities of SP, and particularly NFT, display significant loss of neurons, dendrites and synapses (Coleman & Flood 1987; Terry et al. 1991; Scheff et al. 1990). An inevitable consequence of this pathology is the disruption of neural circuits and isolation of affected areas from cortical networks to which they belong. The most common clinical phenotype in AD is the amnestic DAT. However, atypical forms of AD can also cause aphasic, behavioral and visuospatial dementias (Rogalski et al. 2016; Johnson et al. 1999; Hof et al. 1993).

Diffuse Lewy Body Disease (DLBD)

Lewy bodies (LB) are intracytoplasmic inclusions (Hansen 1994) with a dense eosinophilic core surrounded by a less densely stained peripheral halo (Fig. 1 E). They are composed primarily of α-synuclein, a presynaptic protein likely involved in membrane fusion and neurotransmitter release (McKeith et al. 2017; Liu et al. 2021). α-Synuclein also accumulates in neurons with or without mature LB, and in neurites (Fig. 1 F). The LB is a pathological hallmark of Parkinson’s disease (PD), where LB are found most prominently in the substantia nigra and other subcortical nuclei. Sparsely distributed cortical LBs may also be present in PD brains (Braak et al. 2002). A more widespread distribution of LBs, particularly within the cerebral cortex is a hallmark of DLBD (McKeith et al. 2017; McKeith et al. 1996). Brainstem, limbic and neocortical stages of LB pathology have been identified, with some investigators suggesting a sequential spread of pathology along these three groups of regions (McKeith et al. 2017; Braak et al. 2006). Others have pointed out that the three stages may not develop sequentially (Attems et al. 2021). It appears that significant brainstem LB pathology is associated with Parkinsonian signs, and the limbic and neocortical stages with cognitive symptoms (Attems et al. 2021). Some patients will experience dementia without extrapyramidal impairments suggesting that some trajectories of progression may start in limbic and cortical areas without having to go through a brainstem stage (Attems et al. 2021). Similar to AD, DLBD is accompanied by significant neuronal degeneration. The dementia of DLBD can be amnestic, visuospatial, dysexecutive and, rarely, aphasic (Buciuc et al. 2021).

Frontotemporal Lobar Degeneration (FTLD)

Severe atrophy of frontal and anterior temporal cortex characterizes FTLD-Tau and FTLD-TDP. FTLD-Tau is further divided into CBD, PSP and Pick-type neuropathologies (McKeith et al. 2017) whereas FTLD-TDP has been classified into 5 types, A-E (Mackenzie et al. 2011; Lee et al. 2017). Clinical presentations include behavioral variant frontotemporal dementia (bvFTD), primary progressive aphasia (PPA), and three disorders characterized by motor as well as cognitive abnormalities, corticobasal degeneration syndrome (CBDs), progressive supranuclear palsy syndrome (PSPs), and cognitive impairment with motor neuron disease (FTLD-MND). FTLD can also cause the amnestic phenotype of DAT (Nelson et al. 2019). The CBD and PSP neuropathologies have no obligatory relationship to CBDs and PSPs. For example, approximately 40% of CBDs in our brain bank housed within the Northwestern Alzheimer’s Disease Center have AD pathology as the primary neuropathologic diagnosis whereas approximately 30% of PPA cases have CBD or PSP pathology

Frontotemporal Lobar Degeneration with Tauopathy

In FTLD-Tau, the tauopathy is completely different from the tauopathy of AD. In contrast to AD where the NFT usually are flame-shaped and contain a mixture of 3 and 4 microtubule binding repeat tau isoforms (3R and 4R tau), the tauopathy in FTLD has either only 3R tau (Pick’s disease), or only 4R tau (PSP and CBD), with morphologies that are either spheroid as in Pick’s disease or that take the form of tufted astrocytes and astrocytic plaques of PSP and CBD (Fig. 1 G&H). The one common feature of all tauopathies, in AD as well as FTLD, is the presence of phosphotau (Mackenzie et al. 2010; Bigio 2013). Unlike AD where the NFT are beta pleated, tau inclusions in FTLD have a folding pattern that do not bind Thioflavin-S. Phosphotau accumulation is seen in neurons, neurites and glia in AD, in neurons and neurites in Pick’s disease (Fig. 1 H), in neurons, tufted astrocytes and neurites in PSP, and in neurons, astrocytic plaques and neurites in CBD (Fig. 1 G). The distributions of cells with tau accumulation differs in each FTLD-Tau disorder. Pick’s disease is characterized by tau accumulation in neurons and neurites primarily in cortical regions, CBD displays tau accumulation in cortical regions as well as subcortical structures, such as the striatum, and PSP displays additional inclusions in the brainstem. The common denominator of this pathology is neurosynaptic damage and loss.

Frontotemporal Lobar Degeneration with TDP-43 Proteinopathy

TDP-43 is a 414-amino acid protein with two RNA recognition motifs and normally displays a strictly nuclear localization (Chen-Plotkin et al. 2010; Lee et al. 2012), where it binds to a large number of RNA species (>6,000), and thereby controls RNA transcription, splicing and transport, and the levels of microRNAs (Narayanan et al. 2012; Tollervey et al. 2011; Buratti et al. 2010). In FTLD, TDP-43 is mislocalized from the nucleus to the cytoplasm and neurites, is truncated and hyperphosphorylated (Cairns et al. 2007; Chen-Plotkin et al. 2010). Three types of neuronal inclusions are present in FTLD with TDP-43 pathology (Fig. 1 I), neuronal intranuclear inclusions, neuronal cytoplasmic inclusions or extracellular dystrophic neurites (Mackenzie et al. 2011). Based on the predominance of one or more of the three types of inclusions, the length of dystrophic neurites, and the laminar distribution of cortical inclusions, TDP-43 pathology is categorized into subtypes A-E (Mackenzie et al. 2011; Lee et al. 2017). TDP-43 inclusions appear to be preceded by pre-inclusions of smooth or granular staining in neurons (Kim et al. 2019).

ORGANIZATION OF THE BASAL FOREBRAIN CHOLINERGIC SYSTEM

The mammalian nervous system contains several groups of cholinergic (ChAT-positive) neurons. These include projection neurons, such as those within the basal forebrain, brainstem, spinal cord and the periphery, as well as cholinergic neurons involved in intrinsic circuity, such as those in the striatum (Mesulam et al. 1989; Mesulam & Geula 1988; Geula & Mesulam 1999). Due to their vulnerability to some of the pathologies that cause dementia, the focus of this review is on the cholinergic neurons of the basal forebrain. The “Ch” nomenclature (Mesulam & Geula 1988) has been used to refer to the various groups of cholinergic projection neurons. In this review, we will use this designation to refer to specific groups of cholinergic neurons, followed by an abbreviation of the nuclei within which each group is located.

Basal Forebrain Cholinergic Neurons

Based on anatomical location, four groups of cholinergic neurons have been described in the human basal forebrain. The Ch1 neurons are centered in the medial septum (Ch1-ms), the Ch2 around the vertical limb of the diagonal band of Broca (Ch2-dbv), the Ch3 neurons around the horizontal limb of the diagonal band of Broca (Ch3-dbh), and Ch4 neurons around the nucleus basalis of Meynert (Ch4-nbM) (Hedreen et al. 1998; Mesulam & Geula 1988; Mesulam et al. 1983). Among these cell groups, Ch4-nbM is the most extensive and phylogenetically the most progressive (Mesulam & Geula 1988; Geula et al. 1993) (Fig. 2, Fig. 3 A&B). The human Ch4 can be subdivided into anterior (Ch4a, with a medial [Ch4am] and a lateral [Ch4al] sub-sector), anterointermediate (Ch4ai), intermediate (Ch4i, with a dorsal [Ch4id] and a ventral [Ch4iv] sub-sector) and posterior (Ch4p) sectors (Mesulam & Geula 1988) (Fig. 2 A–D).

Figure 2. Anatomical delineation of the nucleus basalis of Meynert (Ch4-nbM) in the human brain.

(Top) Myelin stained whole-brain section, showing the substantia innominata below the crossing anterior commissure within which the anterior aspects of Ch4-nbM neurons are located. (A-D) Acetylcholinesterase stained sections delineating the anterior (A), anterointermediate (B), intermediate (C) and posterior (D) sectors of Ch4-nbM. Ac – anterior commissure; ai – anterointermediate sector of Ch4-nbM; al – anterolateral subsector of Ch4-nbM; am – anteromedial subsector of Ch4-nbM; Am – amygdala; an – ansa lenticularis; ap – ansa peduncularis; bl – basolateral nucleus of amygdala; GP – globus pallidus; GPe – globus pallidus externa; GPi – globus pallidus interna; GPv – ventral globus pallidus; hc – head of caudate; Hp – hippocampus; ic – internal capsule; id – intermediodorsal subsector of Ch4-nbM; iml – internal medullary lamina or globus pallidus; iv – intermedioventral subsector of Ch4-nbM; Nhl – nucleus of the horizontal limb of the diagonal band of Broca; nst – nucleus of stria terminalis; oc – optic chiasm; ot – optic tract; p – posterior sector of Ch4-nbM; Pt – putamen; si – substantia innominata; son – supraoptic nucleus of hypothalamus; v – blood vessel; vt – temporal horn of lateral ventricle; tc – tail of caudate. A-D are reproduced with permission from Wiley and Sons: Mesulam, M. M. and Geula, C. (1988) Nucleus basalis (Ch4) and cortical cholinergic innervation in the human brain: observations based on the distribution of acetylcholinesterase and choline acetyltransferase. J.Comp.Neurol. 275, 216–240.

Figure 3. Cholinergic neurons of the nucleus basalis of Meynert (Ch4-nbM) in dementias.

Low (A) and high (B) power images of choline-acetyltransferase immunoreactive (ChAT) Ch4-nbM neurons in the normal human brain. (C) Ch4-nbM of an AD brain, demonstrating significant loss of ChAT immunoreactive Ch4-nbM neurons. (D) Thioflavin-S staining, demonstrating presence of mature globose neurofibrillary tangles in the remaining Ch4-nbM neurons in AD. (E) Phosphotau immunoreactivity in the remaining Ch4-nbM neurons in AD. (F) α-Synuclein immunoreactivity in Ch4-nbM neurons in diffuse Lewy body dementia. (G) Robust preservation of magnocellular Nissl stained Ch4-nbM neurons in Pick’s disease; Pick bodies are not observed in Ch4-nbM neurons in contrast to abundant Pick bodies in cortex (see Figure 1H). (H) Abundant accumulation of phosphotau accumulation in Ch4-nbM neurons in corticobasal degeneration, despite preservation of Ch4-nbM neurons. (I) Absence of TDP-43 mature inclusions in Ch4-nbM neurons in TDP-43 proteinopathy; only occasional pre-inclusions (arrow) are seen in Ch4-nbM neurons.

In the monkey, the Ch1-ms and Ch2-dbv neurons provide the major cholinergic innervation of the hippocampus; the Ch3-dbh neurons provide the major cholinergic innervation of the olfactory bulb; and the Ch4-nbM neurons provide the major cholinergic innervation of the entire cortical mantle and the amygdala. The Ch4am sector projects mainly to the medial aspects of the hemisphere, including the cingulate cortex, the Ch4al sector to the frontoparietal opercular areas and the amygdala, the Ch4i sector to dorsolateral frontal and lateral parieto-occipital cortex, and the Ch4p sector to the superior temporal gyrus and the temporal pole (Mesulam et al. 1983; Koliatsos et al. 1988). The striatum and thalamus also receive lesser cholinergic inputs from Ch1-Ch4 in the monkey and human brains (Mesulam et al. 1992b; Arikuni & Kubota 1984). It is important to note that cortical ChAT-positive neurons have not been observed in the adult primate brain (Mesulam & Geula 1988). Therefore, the cholinergic innervation of the human cerebral cortex is most likely exclusively extrinsic in origin.

Cortical Cholinergic Axons

Investigations based on AChE histochemistry and ChAT immunohistochemistry have allowed a detailed mapping of cortical cholinergic innervation in the human brain (Mesulam et al. 1992a; Geula & Mesulam 1996). The highest density of cholinergic axons is found within core limbic areas such as the amygdala and hippocampus. Paralimbic cortical areas, such as the entorhinal cortex and cingulate cortex show the next highest density of cholinergic axons. Association cortices contain an intermediate density of cholinergic axons. The primary visual cortex (area 17) contains one of the lowest densities of cholinergic axons. Several studies of biochemically determined ChAT and AChE activities have shown similar gradients of cholinergic markers in the human cerebral cortex (Davies & Maloney 1976; Perry et al. 1977; Mesulam et al. 1986).

Cortical Cholinergic Receptors and Acetylcholinesterase-Rich Cholinoceptive Neurons

Cholinergic axons exert their excitatory neurotransmitter effects through the mediation of nicotinic and muscarinic receptors. Molecular biological studies have identified many distinct nicotinic receptor subunits (at least nine neuronal α subunits and three neuronal β subunits), which enter into different combinations of pentameric oligomers, each with distinct pharmacological, kinetic, and physiological properties (Koukouli & Changeux 2020; Graham et al. 2003). Among the homopentameric nicotinic receptors, the α7-containing subtypes account for the greatest proportion of high affinity binding sites in the mammalian nervous systems. In the cerebral cortex, the highest density of nicotinic receptors is found in the hippocampus and other limbic / paralimbic regions, followed by deeper layers of neocortex (Cimino et al. 1992). Recently, the density of α4β2 subtype of nicotinic receptors has been found to show a strong relationship with cognitive performance (Okada et al. 2013).

Five subtypes of muscarinic cholinergic receptors (M1-M5) have been recognized, each the product of a different gene (Brann et al. 1993). Muscarinic receptors couple to multiple G proteins to modulate various signal transduction pathways. Studies of the human brain have shown the highest density of muscarinic cholinergic receptors within the striatum and hypothalamus, intermediate to high densities in the amygdala, hippocampus and cerebral cortex, and the lowest density in the cerebellum (Cortes et al. 1987; Lin et al. 1986). High concentrations of both M1 and M2 receptors have been reported in the monkey cerebral cortex with different distributions (Mash et al. 1988; Lidow et al. 1989). The M2 receptor is particularly prominent in the thalamus. In the cerebral cortex, the M1 receptors are much more numerous than the M2 subtype and display peak densities in almost all limbic and paralimbic regions.

The cerebral cortex of the human brain contains AChE-rich neurons which represent a subset of cortical cholinoceptive neurons (Mesulam & Geula 1991). Some are polymorphic in shape and distributed preferentially in the deeper cortical laminae and adjacent white matter. A majority are pyramidal in shape and are located in cortical laminae III and V.

Functional Affiliations

In keeping with its widespread innervation of the cerebral cortex, the basal forebrain cholinergic system has been implicated in a number of behaviors including sleep and arousal, mood and affect, and particularly attention and memory (Drachman & Leavitt 1974; Muir et al. 1992; Dingledine & Kelly 1977; Janowsky et al. 1972). Lesions of the cholinergic cells of the basal forebrain deplete the cortex of its cholinergic innervation and impair learning and memory in rodents and primates (Flicker et al. 1983; Aigner et al. 1987; Leanza et al. 1996; Walsh et al. 1996; Fine et al. 1997). Basal forebrain lesions have also been shown to affect expectancy and particularly attention (Muir et al. 1992; Stoehr et al. 1997). Given the very dense cholinergic innervation of the cerebral cortex, it is interesting that our understanding of its behavioral and cognitive correlates is rather primitive.

BASAL FOREBRAIN CHOLINERGIC SYSTEM IN DEMENTIA

Alzheimer’s Disease: Cholinergic Axons, Neurons and Receptors

Cortical Cholinergic Innervation

A dramatic loss of cortical ChAT activity (up to 95%) has been established as one of the most consistent findings in AD (Geula & Mesulam 1999). Biochemical studies have also found an up to 90% loss in the activity of cortical AChE in AD (Davies 1979; Zubenko et al. 1989). In keeping with these biochemical results, studies using AChE histochemistry (Geula & Mesulam 1996) (Geula & Mesulam 1989; McGeer et al. 1986) or ChAT immunohistochemistry (Geula & Mesulam 1996; Ransmayr et al. 1989) have demonstrated a widespread loss of cortical cholinergic innervation in AD. We have demonstrated that cortical areas with the greatest loss (>75% reduction) of cholinergic innervation are all in the temporal lobe and include areas 20, 21, 22 and 28 of Brodmann (Geula & Mesulam 1996). The frontal and parietal association areas as well as the insula and temporal pole showed an intermediate magnitude of loss (45–75%) whereas the anterior cingulate gyrus, primary motor, primary somatosensory and primary visual cortex displayed less than 25% loss of cholinergic fibers. Within the hippocampal formation, fiber density was reduced in all sectors. In the amygdala, ChAT immunoreactive varicosities showed a marked depletion in the cortical, accessory basal and lateral nuclei while the cholinergic innervation in the central nucleus remained relatively preserved (Emre et al. 1993). Biochemical studies have also shown the greatest decrement in cholinergic enzymes in AD to occur in cortical structures within the temporal lobe (Rossor et al. 1981; Wilcock et al. 1982) while cingulate cortex (areas 24, 25 and 32) and cortical areas within the occipital lobe, particularly the primary visual cortex (area 17) have been shown to display the least depletion of cholinergic enzymes (Rossor et al. 1982; Siek et al. 1990).

Biochemical studies of the relationship between residual ChAT activity and SP in AD cortex have produced conflicting results (Zubenko et al. 1989; Perry et al. 1981; Wilcock et al. 1982; DeKosky et al. 1992). We conducted a survey of the relationship between the density of SP in 14 cortical areas of AD brains and the total density of residual AChE-positive cholinergic fibers (Geula et al. 1998). We estimated the percentage of depletion through comparison with age-matched controls. We found a moderate negative correlation (−0.46 – −0.57) between the number of SP and the density of remaining cholinergic fibers. There was a moderate positive correlation between the percentage of fiber loss and the density of SP in AD cortex (0.49–0.53). However, the estimated density of lost cholinergic fibers showed no correlation with the density of plaques. Some investigators have found a small but significant negative correlation (0.38–0.58) between residual cortical ChAT levels and density of NFT (Mountjoy et al. 1984; Wilcock et al. 1982) while others have found no such correlation (Zubenko et al. 1989; Ransmayr et al. 1992). Our quantitative anatomical studies of the density of NFT and AChE-positive cholinergic axons indicate a relatively high correlation between NFT density and the estimated number of lost cholinergic axons (0.52–0.79) (Geula et al. 1998). Thus while SP density shows no relationship with the estimated lost density of cholinergic axons, NFT density shows a strong relationship.

Cholinergic Neurons of the Basal Forebrain

Loss of Ch4-nbM neurons in AD had been reported by Pilleri in 1966 (Pilleri 1966). This finding gained considerable support with the study of Whitehouse and colleagues (Whitehouse et al. 1981) which detected a 75% decrease in the number of hyperchromic magnocellular Ch4-nbM neurons in AD patients. Subsequent reports have produced overwhelming support for a consistent loss of Ch4-nbM neurons ranging in magnitude from 30% to 95% (Geula & Mesulam 1999; Liu et al. 2015) (Fig. 3 C). A large number of biochemical investigations have also reported a significant loss of ChAT activity (30–90%) in the Ch4-nbM of AD patients (Henke & Lang 1983; Bird et al. 1983; Etienne et al. 1986). In situ hybridization studies have shown an overall 60% reduction in ChAT mRNA in the Ch4-nbM of these patients.

Neuronal loss is reported in practically all basal forebrain cholinergic cell groups and in all Ch4-nbM subsectors. In general, neuronal loss is less in ms-Ch1 and dbv-Ch2 than in Ch4-nbM (Arendt et al. 1985; Mufson et al. 1989b). Studies which have compared the subsectors of Ch4-nbM have demonstrated that Ch4-nbMp and Ch4-nbMal neurons show the greatest and most consistent range of loss (39–76%), followed by the Ch4-nbMi (25–62%) and Ch4-nbMam (13–46%) (Geula & Mesulam 1999). Observations from our laboratory have confirmed the much greater loss of neurons in Ch4-nbMp than in Ch4-nbMa in AD (Mesulam & Geula 1988). Like cortical ChAT activity, Ch4-nbM neuronal loss in typical amnestic AD shows no hemispheric asymmetry (Doucette & Ball 1987). However, in AD that causes PPA, the Ch4-nbM loss is more prominent in the language dominant (usually left) hemisphere (Mesulam et al. 2019).

A significant relationship has been reported between cortical SPs and Ch4-nbM neuronal loss (r=0.54) and shrinkage (r=0.81) by some investigators (Mann et al. 1985) and not by others (Rinne et al. 1987). Two studies found a larger correlation between Ch4-nbM neuronal loss and SPs in anatomically related as compared with unrelated cortical sites (Arendt et al. 1985). Some investigators have reported a significant correlation between Ch4-nbM neuronal loss and cortical NFT (Mann et al. 1984; Mann et al. 1985) while others found no such relationship (Rinne et al. 1987).

The Ch4-nbM neurons are highly vulnerable to NFT formation, and many of the remaining neurons contain NFT (Mufson et al. 1989b; Geula et al. 1998; Geula et al. 2008) (Fig. 3 D&E). A moderate density of SP has also been reported in the basal forebrain region within which the Ch4-nbM neurons are located (Arendt et al. 1988). However, we have shown that the nbM region is relatively resistant to formation of neuritic SP (Baker-Nigh et al. 2015). We have also demonstrated significant and early accumulation of soluble Aβ oligomers in human Ch4-nbM neurons in the course of normal aging and AD (Baker-Nigh et al. 2015). The size of soluble Aβ oligomers accumulating in these neurons was increased in the course of aging, and in AD. The relationship between soluble Aβ oligomer accumulation and loss of Ch4-nbM neurons in AD remains to be investigated.

Loss of BFCN is seen in all clinical phenotypes of AD. For example, we found that 80% of PPA patients with AD pathology had moderate / severe Ch4-nbM neuronal loss and gliosis. (Mesulam et al. 2019). Detailed quantitative analysis revealed substantial tangle formation in Ch4-nbM neurons, loss of Ch4-nbM neurons and degeneration of cortical cholinergic axons. Loss of histochemically visualized cholinergic axons was observed in all cortical areas studied, but was more severe in regions affiliated with language function. Consistent with the PPA phenotype, degeneration and pathology were more severe in the language-dominant hemisphere. These observations indicate that the presence of AD pathology is likely to be associated with significant degeneration of the BFCN regardless of clinical presentation and anatomical distribution of SP and NFT.

Cholinergic Receptors

Several studies have reported no change or an increase in the density or affinity of the total population of cortical muscarinic receptors in AD (Zubenko et al. 1989; Araujo et al. 1988; DeKosky et al. 1992; Svensson et al. 1997). Others show a small (20–30%) reduction in the density of these receptors, with the largest and most consistent loss observed within the hippocampal formation (Vanderhayden et al. 1987; Nordberg et al. 1992). This loss was reported to represent a reduction in the density of the M2 subtype (Araujo et al. 1988; Mash et al. 1985). Other investigations, however, have reported loss of M1 receptor binding, immunoreactivity, and mRNA in AD cortex (Flynn et al. 1995; Yi et al. 2020). In contrast, markers for the M3 and particularly M4 receptor subtype remained unchanged (Flynn et al. 1995; Rodriguez-Puertas et al. 1997). A number of investigations have found abnormalities in the coupling of muscarinic receptors to G proteins in AD, implicating alterations in signal transduction following receptor stimulation (Ferrari-DiLeo et al. 1995; Cowburn et al. 1996; Ladner et al. 1995). Thus, the residual muscarinic receptors in AD cortex may be functionally impaired.

With two exceptions (Lang & Henke 1983; Shimohama et al. 1986), all investigations report a significant loss of cortical nicotinic receptors in AD cortex (Svensson et al. 1997; Nordberg et al. 1992; Perry et al. 1995; Whitehouse et al. 1988), especially in the temporal cortex (Flynn & Mash 1986). This decrease appears to result from the loss of the high affinity subtype of this receptor (Nordberg et al. 1988). Significantly, the density of cortical α4β2 nicotinic receptors was found to display a strong correlation with cognitive performance in mild AD, and to be associated with progression of cognitive decline (Colloby et al. 2010; Okada et al. 2013; Sabri et al. 2018).

Our quantitative studies revealed a marked loss of the AChE-rich staining pattern in putatively cholinoceptive cortical pyramidal neurons in AD (Heckers et al. 1992). The absence of correlation with local neuronal loss led to the suggestion that this subset of cortical cholinoceptive neurons may display downregulation of AChE in the course of AD. More recently, we have demonstrated significant reduction in histochemical AChE activity in cortical pyramidal neurons in the course of normal aging, and a further significant reduction in elderly with exceptional memory performance (Janeczek et al. 2018). It is tempting to speculate that these neurons display a reactive downregulation of AChE activity to counteract reduced presynaptic transmitter release in the course of aging, and particularly in AD.

Relation to Cognitive Performance

A significant correlation has been reported between cortical ChAT activity and the degree of dementia as confirmed by objective neuropsychological assessment (Wilcock et al. 1982; DeKosky et al. 1992; Bierer et al. 1995) whereas cortical levels of other neurochemicals, such as norepinephrine do not appear to show a significant relationship with the degree of dementia (Palmer et al. 1987). The extent of neuronal loss and density of tangles in Ch4-nbM (Samuel et al. 1991; Lehericy et al. 1993; Mesulam et al. 2004), but not in the serotonergic dorsal raphe nucleus (Halliday et al. 1992), have also been shown to be correlated with cognitive impairment along the age-Mild Cognitive Impairment-AD continuum. Thus, depletion of cortical cholinergic innervation is a likely contributor to cognitive abnormalities in AD.

Diffuse Lewy Body Disease

Degeneration of BFCN has been reported in the synucleinopathies of PD and DLBD. As indicated earlier, limbic areas are quite vulnerable to LB formation as well as abnormal α-synuclein deposition in neuronal and neuritic components. The vulnerability of BFCN to PD and DLBD may therefore reflect their location along the mediotemporal and basal forebrain band of limbic formations. Significant reductions in ChAT activity in neocortical regions have been reported in PD and DLBD, and particularly in prefrontal cortex, independent of comorbid AD pathology (Perry et al. 1993; Tiraboschi et al. 2000). Compared to DAT-AD, the reduction of ChAT activity is greater in neocortical but not limbic areas (Perry et al. 1994; Sahin et al. 2006). Loss of ChAT activity in neocortical areas in patients with combined AD and DLBD pathology was greater than that in pure AD (Tiraboschi et al. 2002). Positron emission tomography (PET) imaging with the vesicular acetylcholine transporter has confirmed cholinergic loss in the opercular regions, cingulate cortex, and insular cortex in DLBD (Kanel et al. 2020). The Ch4-nbM neurons are vulnerable to accumulation of LB and α-synuclein accumulation in DLBD (Graeber & Muller 2003) (Fig. 3 F), and significant loss of Ch4-nbM neurons has been reported in PD with or without dementia (Whitehouse et al. 1983; Gasper & Gray 1984; Liu et al. 2015).

Some investigations report no change, or an increase in cortical muscarinic receptors in PD and DLBD (Perry et al. 1993; Tagliavini et al. 1984a), while others find reduced cortical muscarinic receptors in DLBD (Shiozaki & Iseki 2004; Tiraboschi et al. 2002). Reductions are also reported in nicotinic receptor subunits in the hippocampus and entorhinal cortex in DLBD (Teaktong et al. 2004). Thus, while not as thoroughly investigated as in AD, the basal forebrain cholinergic system spears to display significant degeneration in DLBD and PD.

Frontotemporal Lobar Degeneration with Tauopathies

Pick’s Disease – A 3R Tauopathy

The relatively few investigations of the basal forebrain cholinergic system in Pick’s disease primarily point to no or mild degeneration (Mizukami & Kosaka 1989; Tagliavini & Pilleri 1983) (Fig. 3 G), with very few studies (Uhl et al. 1983) finding significant loss. All available studies of ChAT activity in cortical regions, including prefrontal cortex, report no loss in Pick’s disease when compared with controls (Wood et al. 1983b; Hansen et al. 1988; Yates et al. 1980; Wood et al. 1983a). Biochemically determined ChAT and AChE activities were reported to be decreased in basal forebrain in Pick’s disease (Sparks & Markesbery 1991). Consistent with degeneration of cerebral cortex, cortical muscarinic receptor binding is decreased in Pick’s disease (Hansen et al. 1988; Yates et al. 1980). However, one study reported discrepant results (Wood et al. 1983b). A survey of three cases from our collection showed the Ch4-nbM to be unaffected at a stage when the cerebral cortex was devastated by neurodegeneration. Tau immunohistochemistry in these cases showed very high numbers of Pick bodies in cortical areas but almost none in the Ch4-nbM. In summary, the basal forebrain cholinergic system remains intact, or displays very mild degeneration in Pick’s disease. It must be noted that many of the available reports are based on very few cases, sometimes a single patient.

Progressive Supranuclear Palsy – A 4R Tauopathy

Little information is available on cortical cholinergic innervation in PSP pathology. The only available study employed PET imaging of vesicular acetylcholine transporter, and found significant loss in the anterior cingulate cortex but not hippocampus (Mazere et al. 2012). A few studies have reported loss of Ch4-nbM neurons in PSP (Kasashima & Oda 2003; Tagliavini et al. 1984b; Tagliavini et al. 1983). This reported loss is variable, small in many cases, and less than that observed in AD. Cortical muscarinic receptors seem to remain unchanged (Asahina et al. 1998; Warren et al. 2008), while one study reported reduced cortical nicotinic receptors in PSP (Whitehouse et al. 1988). Therefore, while PSP appears to be associated with loss of Ch4-nbM neurons, this loss in smaller when compared with AD. Additional experiments are needed to assess the vulnerability of BFCN to phosphotau pathology in PSP.

Corticobasal Degeneration – A 4R Tauopathy

A single study based on one case of CBD found greater loss of BFCN in this tauopathy when compared with PSP (Kasashima & Oda 2003). We conducted an investigation of phosphotau inclusions in the Ch4-nbM and status of cortical cholinergic axons in autopsy confirmed cases of CBD. All CBD cases displayed phosphotau inclusions that encompassed 30–70% of Ch4-nbM neurons (Fig. 3 H). The inclusions were morphologically distinct from NFT of AD. There was no correlation between cellular phosphotau accumulation and neuronal number. Moreover, the density of histochemically visualized cortical cholinergic axons in CBD was similar to those of normal controls and was significantly greater than that of AD cases. Thus, despite CBD-related intracellular 4R tauopathy, the number of BFCN and their cortical cholinergic projections remained intact (Dunlop et al. Basal Forebrain Cholinergic System Resilience to Tauopathy in Corticobasal Degeneration. Soc. Neurosci. Abst., Virtual Global Connectome, Session # P120.02, 2021). The cellular integrity of BFCN is therefore based on resilience to the 4R tauopathy of CBD.

Frontotemporal Lobar Degeneration with TDP-43 Proteinopathies

There is a dearth of information on the status of the basal forebrain cholinergic system in TDP-43 proteinopathies. To fill this gap, we conducted an investigation of susceptibility of Ch4-nbM neurons to accumulation of TDP-43 inclusions and to degeneration. Using cases with clinically diagnosed PPA and pathologically confirmed TDP-43 proteinopathy, we found near absence of TDP-43 neuronal inclusions in Ch4-nbM neurons. Few and very sparsely populated Ch4-nbM neurons (<2%) contained TDP-43 pre-inclusions (Fig 3 I), indicating very early TDP-43 pathology within these neurons even at terminal disease stages (Dunlop et al., Vulnerability of the Cholinergic Basal Forebrain to AD vs FTLD-TDP Pathology in Primary Progressive Aphasia. Alzheimer’s Association International Conference, Chicago, 2018). Additionally, in semiquantitative ratings, 90% of FTLD-TDP cases who had PPA in life displayed no or very mild Ch4-nbM neuronal loss and gliosis (Mesulam et al. 2019). These results suggest that the basal forebrain cholinergic system may be resistant to the accumulation of abnormal TDP-43 inclusions.

SELECTIVE VULNERABILITY OF BASAL FOREBRAIN CHOLINERGIC SYSTEM

The evidence reviewed here clearly demonstrates selective vulnerability of the basal forebrain cholinergic system to AD and DLBD. There is evidence that the BFCN degeneration in AD occurs earlier than in other subcortical nuclei (Geula & Mesulam 1999). We have shown that phosphotau accumulation in Ch4-nbM, and abnormalities in cortical cholinergic axons are observed early in the process of normal aging, and are greatly exacerbated in AD (Geula et al. 2008). The substantial increase in the density of tangles in Ch4-nbM neurons is associated with the degeneration of these neurons. Of interest, as reviewed earlier, basal forebrain cholinergic degeneration is not only a feature of the typical amnestic DAT-AD, but is also present in the aphasic PPA variant (Mesulam et al. 2019). Available evidence suggests that the basal forebrain cholinergic system also undergoes degeneration in the synucleinopathy of DLBD, and most likely in PD, particularly when it is accompanied by cognitive impairment.

In contrast to AD, DLBD and PD, the basal forebrain cholinergic system does not display significant degeneration in 3R or 4R tauopathies or in TDP-43 proteinopathies. In fact, in the tauopathy of CBD, the BFCN displays resilience despite accumulation of 4R phosphotau in Ch4-nbM neurons. This observation has important implications for the mechanisms of neurodegeneration within the spectrum of tauopathies. If the intraneuronal abnormality of tau implies abnormal microtubule stabilization, this loss of function cannot be considered a general mechanism of neurodegeneration since it would have been as relevant to AD as to CBD and lead to degeneration in both conditions. Instead, the neurodegenerative impact of tauopathy may reflect a toxic gain of function that is specifically linked to the mixture of 3R and 4R phosphotau that characterizes AD. This differential vulnerability is most likely due to differences in tau conformations and strains. As reviewed earlier, phosphotau in AD forms PHF that have the β-pleated sheet abnormal protein conformations recognized by Thioflavin-S, whereas tau inclusions in most other tauopathies do not (Mackenzie et al. 2011). Furthermore, pathology in each tauopathy appears to be composed of a distinct tau strain that can be distinguished by specific antibodies, and retain distinct cell-specificity and propagation properties when seeded in mouse brains (Narasimhan et al. 2017; Gibbons et al. 2018; Xu et al. 2021). The results reviewed above also suggest that the BFCN may be resistant to intracellular proteinopathy in the 3R tauopathy of Pick’s disease and in FTLD-TDP.

The propensity of limbic and paralimbic regions to formation of inclusions in AD and DLBD is likely to contribute to the vulnerability of BFCN in these disorders. Molecular characteristics of BFCN may also contribute of their vulnerability in AD and DLBD. In the adult human brain, the Ch1-Ch4 complex is the major cell group which expresses substantial immunoreactivity for the high affinity neurotrophin receptor TrkA and the p75 low affinity neurotrophin receptor (p75^LANTR) (Mufson et al. 1989a; Kordower et al. 1994), which are receptors for nerve growth factor (NGFr). The NGFr are transported anterogradely to terminal fields of Ch1-Ch4 neurons where they bind NGF. The NGFr-NGF complex is then retrogradely transported to the Ch1-Ch4 perikarya that depend on NGF for survival and protection from damage (Ferguson et al. 1991; Ehlers et al. 1995; Tuszynski et al. 1990). It is therefore conceivable that the selective vulnerability of cortical cholinergic innervation may be caused by a deficiency of cortically produced NGF or the failure of Ch1-Ch4 neurons to express NGFr. Studies of NGF levels in the cerebral cortex and hippocampus have shown either no loss or a significant increase in AD, consistent with defective transport (Allen et al. 1991; Scott et al. 1995). Loss of NGFr in Ch4-nbM has been reported at the level of mRNA and protein in mild cognitive impairment and AD (Mufson et al. 2002; Mufson et al. 2000). A significant decrease in NGF immunoreactivity has also been reported in the remaining Ch4-nbM neurons (Mufson et al. 1995). Thus, disruption of retrograde NGF transport is likely to contribute to the loss of Ch1-Ch4 neurons in AD.

Most Ch1-Ch4 neurons in the primate brain show immunoreactivity for the calcium binding protein calbindin-D_28K (Wu et al. 2000; Geula et al. 1993; Celio & Norman 1985). In the human and the non-human primate brain, approximately 75% of Ch4-nbM neurons are calbindin-positive (Wu et al. 2000; Geula et al. 1993). However, none of the ChAT-positive Ch4-nbM in the rodent brain are calbindin-positive (Celio 1990; Geula et al. 1993). In a series of experiments, we have demonstrated a selective, substantial and significant loss of calbindin from the human and monkey BFCN in the course of normal aging (Wu et al. 1997; Geula et al. 2003a; Geula et al. 2003b; Wu et al. 2003). The age-related loss of calbindin from the BFCN was neurochemically and regionally specific, and was not accompanied by loss of ChAT, TrkA, p75^LANTR, or AChE (Wu et al. 1997; Wu et al. 2003).

Calbindin possesses a structural domain which binds Ca²⁺ with high affinity (Miller 1991). Neuronal calbindin regulates intracellular Ca²⁺ and reduces excessive Ca²⁺ levels (Mattson et al. 1991). With few exceptions, the presence of calbindin confers protection to a variety of neurons against deleterious effects of increased intracellular Ca²⁺ (Rintoul et al. 2001). We have hypothesized that the age-related loss of calbindin deprives the BFCN of the capacity to buffer excess Ca²⁺ and leaves them vulnerable to involutional and neurodegenerative processes. We have found that the Ch4-nbM neurons that degenerate in AD belong to the subpopulation that lack calbindin. Nearly all of the surviving BFCN in AD are calbindin-positive (Riascos et al. 2010). We have also found that calbindin-positive BFCN may be relatively more resistant to NFT formation (Riascos et al. 2011; Ahmadian et al. 2015). It appears, therefore, that the age-related loss of calbindin from Ch4-nbM may be involved in the pathologic cascade underlying their selective vulnerability in AD.

THERAPEUTIC IMPLICATIONS

The discovery of the cholinergic denervation in the cerebral cortex and the therapeutic effect of cholinesterase inhibitors led to the formulation of a ‘cholinergic theory’ of Alzheimer’s disease. However, it is important to keep in mind that the cholinergic innervation of the basal ganglia and thalamus remain nearly intact in the initial stages of AD. It would appear that the preferential loss of cortical cholinergic innervation in AD and DLBD reflects the vulnerability of Ch4-nbM, in specific, not of cholinergic neurotransmission in general. As noted above, the vulnerability of the Ch4-nbM may be based on the continuity of the basal forebrain with the hippocampal-amygdaloid-olfactory band of limbic formations. Nonetheless, the significant degeneration of the basal forebrain cholinergic system in AD, DLBD and PD, and the participation of this system in cognitive processes of memory and attention gave rise to the development of cholinergic therapies based on the inhibition of AChE, the major hydrolytic enzyme that terminates the synaptic action of acetylcholine. Treatment with these cholinergic inhibitors has consistently resulted in modest but statistically robust improvement in cognitive abilities in cases of AD, PD and DLBD (Bullock et al. 2005; Farlow 2002; Bhasin et al. 2007; Emre et al. 2004). The information reviewed here helps explain why cholinesterase inhibitors are ineffective in Pick’s disease, CBD, PSP and FTLD-TDP. Direct cholinergic receptor stimulation employing muscarinic and nicotinic agonists are being pursued as potential cholinergic therapies for dementia (Moran et al. 2019; Hoskin et al. 2019). If successful, such agents may further enhance cholinergic transmission in AD and DLBD, in which the pool of acetylcholine is deficient due to degeneration of BFCN axons in cortex. Reducing Aβ production, suppressing inflammation or promoting neuroplasticity may represent additional avenues through which cholinergic receptor agonists or cholinesterase inhibitors may confer disease-modifying benefits (Hoskin et al. 2019; Fisher 2012; Gamage et al. 2020).

Abbreviations:

a

Anterior

Aβ

Amyloid-β Peptide

AChE

Acetylcholinesterase

AD

Alzheimer’s Disease

ai

Anterointermediate

al

Anterolateral

am

Anteromedial

BFCN

Basal Forebrain Cholinergic Neurons

bvFTD

Behavioral Variant Frontotemporal Dementia

CBDs

Corticobasal Degeneration Syndrome

Ch

Cholinergic Cell Group

ChAT

Choline Acetyltransferase

DAT

Dementia of the Alzheimer Type

dbh

Horizontal Limb of the Diagonal Band of Broca

dbv

Vertical Limb of the Diagonal Band of Broca

DLBD

Diffuse Lewy Body Disease

FTLD

Frontotemporal Lobar Degeneration

i

Intermediate

id

Intermediodorsal

iv

Intermedioventral

LANTR

Low Affinity Neurotrophin Receptor

LB

Lewy Body

M

Muscarinic Cholinergic Receptor

ms

Medial Septum

nbM

Nucleus Basalis of Meynert

NFT

Neurofibrillary Tangle

NGF

Nerve Growth Factor

NGFr

Nerve Growth Factor Receptor

p

Posterior

PD

Parkinson’s Disease

PHF

Paired Helical Filaments

PPA

Primary Progressive Aphasia

PSPs

Progressive Supranuclear Palsy Syndrome

3R

3 Microtubule Binding Repeat Isoform of Tau

4R

4 Microtubule Binding Repeat Isoform of Tau

SP

Senile Plaque

TDP-43

43-kDa Transactive Response Element DNA-Binding Protein

TrkA

Tyrosine Kinase A