The amyloid precursor protein controls PIKfyve function

While the Amyloid Precursor Protein (APP) plays a central role in Alzheimer's disease, its cellular function still remains largely unclear. It was our goal to establish APP function which will provide insights into APP's implication in Alzheimer's disease. Using our recently developed proteo-liposome assay we established the interactome of APP's intracellular domain (known as AICD), thereby identifying novel APP interactors that provide mechanistic insights into APP function. By combining biochemical, cell biological and genetic approaches we validated the functional significance of one of these novel interactors. Here we show that APP binds the PIKfyve complex, an essential kinase for the synthesis of the endosomal phosphoinositide phosphatidylinositol-3,5-bisphosphate. This signalling lipid plays a crucial role in endosomal homeostasis and receptor sorting. Loss of PIKfyve function by mutation causes profound neurodegeneration in mammals. Using C. elegans genetics we demonstrate that APP functionally cooperates with PIKfyve in vivo. This regulation is required for maintaining endosomal and neuronal function. Our findings establish an unexpected role for APP in the regulation of endosomal phosphoinositide metabolism with dramatic consequences for endosomal biology and important implications for our understanding of Alzheimer's disease.

APP controls the formation of PI(3,5)P2 vesicles through its binding of the PIKfyve complex

Phosphoinositides are signalling lipids that are crucial for major signalling events as well as established regulators of membrane trafficking. Control of endosomal sorting and endosomal homeostasis requires phosphatidylinositol-3-phosphate (PI(3)P) and phosphatidylinositol-3,5-bisphosphate (PI(3,5)P2), the latter a lipid of low abundance but significant physiological relevance. PI(3,5)P2 is formed by phosphorylation of PI(3)P by the PIKfyve complex which is crucial for maintaining endosomal homeostasis. Interestingly, loss of PIKfyve function results in dramatic neurodegeneration. Despite the significance of PIKfyve, its regulation is still poorly understood. Here we show that the Amyloid Precursor Protein (APP), a central molecule in Alzheimer’s disease, associates with the PIKfyve complex (consisting of Vac14, PIKfyve and Fig4) and that the APP intracellular domain directly binds purified Vac14. We also show that the closely related APP paralogues, APLP1 and 2 associate with the PIKfyve complex. Whether APP family proteins can additionally form direct protein–protein interaction with PIKfyve or Fig4 remains to be explored. We show that APP binding to the PIKfyve complex drives formation of PI(3,5)P2 positive vesicles and that APP gene family members are required for supporting PIKfyve function. Interestingly, the PIKfyve complex is required for APP trafficking, suggesting a feedback loop in which APP, by binding to and stimulating PI(3,5)P2 vesicle formation may control its own trafficking. These data suggest that altered APP processing, as observed in Alzheimer’s disease, may disrupt PI(3,5)P2 metabolism, endosomal sorting and homeostasis with important implications for our understanding of the mechanism of neurodegeneration in Alzheimer’s disease.

The role of DNA methylation in aging, rejuvenation, and age-related disease

DNA methylation is a major control program that modulates gene expression in a plethora of organisms. Gene silencing through methylation occurs through the activity of DNA methyltransferases, enzymes that transfer a methyl group from S-adenosyl-l-methionine to the carbon 5 position of cytosine. DNA methylation patterns are established by the de novo DNA methyltransferases (DNMTs) DNMT3A and DNMT3B and are subsequently maintained by DNMT1. Aging and age-related diseases include defined changes in 5-methylcytosine content and are generally characterized by genome-wide hypomethylation and promoter-specific hypermethylation. These changes in the epigenetic landscape represent potential disease biomarkers and are thought to contribute to age-related pathologies, such as cancer, osteoarthritis, and neurodegeneration. Some diseases, such as a hereditary form of sensory neuropathy accompanied by dementia, are directly caused by methylomic changes. Epigenetic modifications, however, are reversible and are therefore a prime target for therapeutic intervention. Numerous drugs that specifically target DNMTs are being tested in ongoing clinical trials for a variety of cancers, and data from finished trials demonstrate that some, such as 5-azacytidine, may even be superior to standard care. DNMTs, demethylases, and associated partners are dynamically shaping the methylome and demonstrate great promise with regard to rejuvenation. © Copyright 2012, Mary Ann Liebert, Inc. 2012.

A comparison of spatial patterns of TDP-43 cellular inclusions in familial and sporadic frontotemporal lobar degeneration with TDP-43 proteinopathy

Abnormal protein aggregates of transactive response (TAR) DNA-binding protein (TDP-43) in the form of neuronal cytoplasmic inclusions (NCI), oligodendroglial inclusions (GI), neuronal internuclear inclusions (NII), and dystrophic neurites (DN) are the pathological hallmark of frontotemporal lobar degeneration with TDP-43 proteinopathy (FTLD-TDP). To investigate the role of phosphorylated TDP-43 (pTDP-43) in neurodegeneration in FTLD-TDP, the spatial patterns of the pTDP-43-immunoreactive NCI, GI, NII, and DN were studied in frontal and temporal cortex in three groups of cases: (1) familial FTLD-TDP caused by progranulin (GRN) mutation, (2) a miscellaneous group of familial cases containing cases caused by valosin-containing protein (VCP) mutation, ubiquitin associated protein 1 (UBAP1) mutation, and cases not associated with currently known genes, and (3) sporadic FTLD-TDP. In a significant number of brain regions, the pTDP-43-immunoreactive inclusions developed in clusters and the clusters were distributed regularly parallel to the tissue boundary. The spatial patterns of the inclusions were similar to those revealed by a phosphorylation-independent anti-TDP-43 antibody. The spatial patterns and cluster sizes of the pTDP-43-immunoreactive inclusions were similar in GRN mutation cases, remaining familial cases, and in sporadic FTLD-TDP. Hence, pathological changes initiated by different genetic factors in familial cases and by unknown causes in sporadic FTLD-TDP appear to follow a parallel course resulting in very similar patterns of degeneration of frontal and temporal lobes.

Extra-nuclear telomerase reverse transcriptase (TERT) regulates glucose transport in skeletal muscle cells

Telomerase reverse transcriptase (TERT) is a key component of the telomerase complex. By lengthening telomeres in DNA strands, TERT increases senescent cell lifespan. Mice that lack TERT age much faster and exhibit age-related conditions such as osteoporosis, diabetes and neurodegeneration. Accelerated telomere shortening in both human and animal models has been documented in conditions associated with insulin resistance, including T2DM. We investigated the role of TERT, in regulating cellular glucose utilisation by using the myoblastoma cell line C2C12, as well as primary mouse and human skeletal muscle cells. Inhibition of TERT expression or activity by using siRNA (100. nM) or specific inhibitors (100. nM) reduced basal 2-deoxyglucose uptake by ~. 50%, in all cell types, without altering insulin responsiveness. In contrast, TERT over-expression increased glucose uptake by 3.25-fold. In C2C12 cells TERT protein was mostly localised intracellularly and stimulation of cells with insulin induced translocation to the plasma membrane. Furthermore, co-immunoprecipitation experiments in C2C12 cells showed that TERT was constitutively associated with glucose transporters (GLUTs) 1, 4 and 12 via an insulin insensitive interaction that also did not require intact PI3-K and mTOR pathways. Collectively, these findings identified a novel extra-nuclear function of TERT that regulates an insulin-insensitive pathway involved in glucose uptake in human and mouse skeletal muscle cells. © 2014 Elsevier B.V.

Oxidised LDL lipids, statins and a blood-brain barrier

Elevated cholesterol in mid-life has been associated with increased risk of dementia in later life. We have previously shown that low density lipoprotein (LDL) is more oxidised in the plasma of dementia patients although total cholesterol levels remained unchanged. Increased systemic oxidative modification (oxLDL) and nitration is also observed during hypercholesterolemia. We have investigated the hypothesis that disruption of blood brain barrier (BBB) function by oxLDL and their lipids may increase risk of neurodegeneration in later life and that statin intervention can mitigate the effects of hyperlipidaemia in mid-life. LDL isolated from statin-naïve hypercholesterolaemic subjects had higher mobility by agarose gel electrophoresis (Rf;0.53±0.06) and 8-isoprostane F2α concentration (43.5±8.42pg/ml) compared to control subjects (Rf; 0.46±0.05 and 24.2±5.37pg/ml respectively; p<0.05). Compared to HMVEC treatment with the LDL-lipids (5μM) from normolipidaemic subjects, LDL-lipids from hypercholesterolaemic subjects increased barrier permeability (103.4±12.5 Ωcm2 v 66.7±7.3 Ωcm2,P<0.01) and decreased cellular glutathione levels (18.5nmol/mg v 12.3nmol/mg) compared to untreated cells (26.2±3.6nmol/mg). LDL-lipids isolated from normolipidaemic subjects shows reduced risk to damage a BBB model compared with LDL-lipids from hypercholesterolaemic subjects. Moreover, a three month statin-intervention reduced the propensity for LDL-lipids from subjects with hyperlipidaemia to damage HMVEC. Post-statin treatment the cytotoxic and pro-inflammatory effects of LDL lipids disappeared. These data support the hypothesis that in vivo intervention with statins modifies LDL lipid oxidation, exerting a protective effect against in microvascular damage independent of cholesterol concentration.

The role of amyloid precursor protein in neuronal and non-neuronal cell lines

Models of Alzheimer’s disease (AD) have provided useful insights into the pathogenesis and mechanistic pathways that lead to its development. One emerging idea about AD is that it may be described as a hypometabolic disorder due to the reduction of glucose uptake in AD brains. Inappropriate processing of Amyloid Precursor Protein (APP) is considered central to the initiation and progression of the disease. Although the exact role of APP misprocessing is unclear, it may play a role in neuronal metabolism before the onset of neurodegeneration. To investigate the potential role of APP in neuronal metabolism, the SHSY5Y neuroblastoma cell line was used to generate cell lines that stably overexpress wild type APP695 or express Swedish mutated-APP observed in familial AD (FAD), both under the control of the neuronal promoter, Synapsin I. The effects of APP on glucose uptake, cellular stress and energy homeostasis were studied extensively. It was found that APP-overexpressing cells exhibited decreased glucose uptake with changes in basal oxygen consumption in comparison to control cell lines. Similar studies were also performed in fibroblasts taken from FAD patients compared with control fibroblasts. Previous studies found FAD-derived fibroblasts displayed altered metabolic profiles, calcium homeostasis and oxidative stress when compared to controls. As such, in this study fibroblasts were studied in terms of their ability to metabolise glucose and their mitochondrial function. Results show that FAD-derived fibroblasts demonstrate no differences in mitochondrial function, or response to oxidative stress compared to control fibroblasts. However, control fibroblasts treated with Aβ1-42 demonstrated changes in glucose uptake. This study highlights the importance of APP expression within non-neuronal cell lines, suggesting that whilst AD is considered a brain-associated disorder, peripheral effects in non-neuronal cell types should also be considered when studying the effects of Aβ on metabolism.

Oxidised LDL-lipids alter redox ratio, lipid raft formation and increase amyloid beta production by SHSY-5Y cells

Elevated cholesterol in mid-life has been associated with increased risk of dementia in later life. We have previously shown that low density lipoprotein (LDL) is more oxidised in the plasma of dementia patients although total cholesterol levels remained unchanged [1]. We have investigated the hypothesis that amyloid beta production and neurodegeneration can be driven by oxidised lipids derived from LDL following the loss of blood brain barrier integrity with ageing. Therefore, we have investigated amyloid beta formation in SHSY5Y cells treated with LDL, minimally modified (ox) LDL, and lipids extracted from both forms of LDL. LDL-treated SHSY-5Y cell viability was not significantly decreased with up to 8 μg LDL/2 × 104 cells compared to untreated cells. However, 8 μg oxLDL protein/2 × 104 cells decreased the cell viability significantly by 33.7% (P < 0.05). A more significant decrease in cell viability was observed when treating cells with extracted lipids from 8 μg of LDL (by 32.7%; P < 0.01) and oxLDL (by 41%; P < 0.01). In parallel, the ratio of reduced to oxidised GSH was decreased; GSH concentrations were significantly decreased following treatment with 0.8 μg/ml oxLD-L (7.35 ± 0.58;P < 0.01), 1.6 μg/ml (5.27 ± 0.23; P < 0.001) and 4 μg/ml (5.31 ± 0.31; P < 0.001). This decrease in redox potential was associated with an increase acid sphingomyelinase activity and lipid raft formation which could be inhibited by desipramine; SHSY5Y cells treated with oxLDL, and lipids from LDL and oxLDL for 16 h showed significantly increased acid sphingomyelinase activity (5.32 ± 0.35; P < 0.05, 5.21 ± 0.6; P < 0.05, and 5.58 ± 0.44; P < 0.01, respectively) compared to control cells (2.96 ± 0.34). As amyloid beta production is driven by the activity of beta secretase and its association with lipid rafts, we investigated whether lipids from ox-LDL can influence amyloid beta by SHSY-5Y cells in the presence of oxLDL. Using ELISA and Western blot, we confirmed that secretion of amyloid beta oligomers is increased by SHSY-5Y cells in the presence of oxLDL lipids. These data suggest a mechanism whereby LDL, and more significantly oxLDL lipids, can drive amyloid beta production and cytotoxicity in neuronal cells. [1] Li L, Willets RS, Polidori MC, Stahl W, Nelles G, Sies H, Griffiths HR. Oxidative LDL modification is increased in vascular dementia and is inversely associated with cognitive performance. Free Radic Res. 2010 Mar; 44(3): 241–8.

Modified lipids from LDL in the blood during mid-life increase blood brain barrier permeability

Low density lipoprotein levels (LDL) are consistently elevated in cardiovascular disease. It has been suggested that those with high circulating LDL levels in mid-life may be susceptible to develop neurodegenerative diseases in later life. In the circulation, high levels of LDL are associated with increased oxidative modification (oxLDL) and nitration. We have investigated the hypothesis that disruption of blood brain barrier function by oxLDL and their lipids may increase risk of neurodegeneration in later life and that statin intervention in mid-life can mitigate the neurodegenerative effects of hyperlipidaemia. Blood from statin-naïve, normo- and hyperlipidaemic subjects (n=10/group) was collected at baseline. Hyperlipidaemic subjects received statin-intervention whereas normolipidaemic subjects did not prior to a second blood sampling, taken after 3 months. The intervention will be completed in June 2013. Plasma was separated by centrifugation (200g, 30min) and LDL was isolated by potassium bromide density gradient ultracentrifugation. Total homocysteine, LDL cholesterol, 8-isoprostane F2α levels were measured in plasma using commercial kits. LDL were analysed by agarose gel electrophoresis. LDL-lipids were extracted by partitioning in 1:1 chloroform:methanol (v/v) and conjugated to fatty acid free-BSA in serum-free EGM-2 medium (4hrs, 370C) for co-culture with human microvascular endothelial cells (HMVEC). HMVEC were maintained on polycarbonate inserts for two weeks to create a microvascular barrier. Change in barrier permeability was measured by trans-endothelial electrical resistance (TER), FITC-dextran permeability and immunohistochemistry. HMVEC glutathione (GSH) levels were measured after 2 hours by GSH-glo assay. LDL isolated from statin-naïve hyperlipidaemic subjects had higher mobility by agarose gel electrophoresis (Rf;0.53±0.06) and plasma 8-isoprostane F2α (43.5±8.42 pg/ml) compared to control subjects (0.46±0.05 and 24.2±5.37 pg/ml; p<0.05). Compared to HMVEC treatment with the LDL-lipids (5μM) from normolipidaemic subjects, LDL-lipids from hyperlipidaemic subjects increased barrier permeability (103.4±12.5 Ωcm2 v 66.7±7.3 Ωcm2,P<0.01) and decreased GSH (18.5 nmol/mg v 12.3 nmol/mg; untreated cells 26.2±3.6 nmol/mg).

Oligodendrocyte pathology in neurodegenerative disease

Oligodendrocytes have multiple functions in the central nervous system including mechanical support of neurons, production of myelin sheaths, and uptake and inactivation of chemical neurotransmitters released by neurons. Consequently, oligodendrocytes could be involved in the pathology of a number of neurodegenerative diseases. Although, the molecular mechanisms involved require further elucidation, it is likely that oligodendrocyte dysfunction is important in Alzheimer’s disease (AD), amyotrophic lateral sclerosis (ALS), and multiple sclerosis (MS). In addition, abnormal protein aggregates in the form of oligodendrocyte inclusions (OI) have been observed in several other disorders, most notable in multiple system atrophy (MSA), in which the glial cytoplasmic inclusion (GCI) is the ‘signature’ pathology of the disease. OI have also been identified in argyrophilic grain disease (AGD), progressive supranuclear palsy (PSP) (Armstrong et al 2007), and various forms of frontotemporal lobar degeneration (FTLD) (Armstrong et al 2010), although their role in the pathology of these disorders is less clear. It is likely that future research will expand the range of disorders in which oligodendrocytes play a significant role in neurodegeneration.

The development of functional astrocyte-neuron networks derived from stem cells

There is currently great scientific and medical interest in the potential of tissue grown from stem cells. These cells present opportunities for generating model systems for drug screening and toxicological testing which would be expected to be more relevant to human outcomes than animal based tissue preparations. Newly realised astrocytic roles in the brain have fundamental implications within the context of stem cell derived neuronal networks. If the aim of stem cell neuroscience is to generate functional neuronal networks that behave as networks do in the brain, then it becomes clear that we must include and understand all the cellular components that comprise that network, and which are important to support synaptic integrity and cell to cell signalling. We have shown that stem cell derived neurons exhibit spontaneous and coordinated calcium elevations in clusters and in extended processes, indicating local and long distance signalling (1). Tetrodotoxin sensitive network activity could also be evoked by electrical stimulation. Similarly, astrocytes exhibit morphology and functional properties consistent with this glial cell type. Astrocytes also respond to neuronal activity and to exogenously applied neurotransmitters with calcium elevations, and in contrast to neurons, also exhibited spontaneous rhythmic calcium oscillations. Astroctyes also generate propagating calcium waves that are gap junction and purinergic signalling dependent. Our results show that stem cell derived astrocytes exhibit appropriate functionality and that stem cell neuronal networks interact with astrocytic networks in co-culture. Using mixed cultures of stem cell derived neurons and astrocytes, we have also shown both cell types also modulate their glucose uptake, glycogen turnover and lactate production in response to glutamate as well as increased neuronal activity (2). This finding is consistent with their neuron-astrocyte metabolic coupling thus demonstrating a tractable human model, which will facilitate the study of the metabolic coupling between neurons and astrocytes and its relationship with CNS functional issues ranging from plasticity to neurodegeneration. Indeed, cultures treated with oligomers of amyloid beta 1-42 (Aβ1-42) also display a clear hypometabolism, particularly with regard to utilization of substrates such as glucose (3). Both co-cultures of neurons and astrocytes and purified cultures of astrocytes showed a significant decrease in glucose uptake after treatment with 2 and 0.2 μmol/L Aβ at all time points investigated (p <0.01). In addition, a significant increase in the glycogen content of cells was also measured. Mixed neuron and astrocyte co-cultures as well as pure astrocyte cultures showed an initial decrease in glycogen levels at 6 hours compared with control at 0.2 μmol/L and 2 μmol/L P <0.01. These changes were accompanied by changes in NAD+/NADH (P<0.05), ATP (P<0.05), and glutathione levels (P<0.05), suggesting a disruption in the energy-redox axis within these cultures. The high energy demands associated with neuronal functions such as memory formation and protection from oxidative stress put these cells at particular risk from Aβ-induced hypometabolism. As numerous cell types interact in the brain it is important that any in vitro model developed reflects this arrangement. Our findings indicate that stem cell derived neuron and astrocyte networks can communicate, and so have the potential to interact in a tripartite manner as is seen in vivo. This study therefore lays the foundation for further development of stem cell derived neurons and astrocytes into therapeutic cell replacement and human toxicology/disease models. More recently our data provides evidence for a detrimental effect of Aβ on carbohydrate metabolism in both neurons and astrocytes. As a purely in vitro system, human stem cell models can be readily manipulated and maintained in culture for a period of months without the use of animals. In our laboratory cultures can be maintained in culture for up to 12 months months thus providing the opportunity to study the consequences of these changes over extended periods of time relevant to aspects of the disease progression time frame in vivo. In addition, their human origin provides a more realistic in vitro model as well as informing other human in vitro models such as patient-derived iPSC.

The amyloid precursor protein (APP) binds the PIKfyve complex and modulates its function

Phosphoinositides are important components of eukaryotic membranes that are required for multiple forms of membrane dynamics. Phosphoinositides are involved in defining membrane identity, mediate cell signalling and control membrane trafficking events. Due to their pivotal role in membrane dynamics, phosphoinositide de-regulation contributes to various human diseases. In this review, we will focus on the newly emerging regulation of the PIKfyve complex, a phosphoinositide kinase that converts the endosomal phosphatidylinositol-3-phosphate [PI(3)P] to phosphatidylinositol-3,5-bisphosphate [PI(3,5)P2)], a low abundance phosphoinositide of outstanding importance for neuronal integrity and function. Loss of PIKfyve function is well known to result in neurodegeneration in both mousemodels and human patients. Our recent work has surprisingly identified the amyloid precursor protein (APP), the central molecule in Alzheimer s disease aetiology, as a novel interaction partner of a subunit of the PIKfyve complex, Vac14. Furthermore, it has been shown that APP modulates PIKfyve function and PI(3,5)P2 dynamics, suggesting that the APP gene family functions as regulator of PI(3,5)P2 metabolism. The recent advances discussed in this review suggest a novel, unexpected, â-amyloid-independent mechanism for neurodegeneration in Alzheimer s disease.

A cell-permeable tool for analysing APP intracellular domain function and manipulation of PIKfyve activity

The mechanisms for regulating PIKfyve complex activity are currently emerging. The PIKfyve complex, consisting of the phosphoinositide kinase PIKfyve (also known as FAB1), VAC14 and FIG4, is required for the production of phosphatidylinositol-3,5-bisphosphate (PI(3,5)P2). PIKfyve function is required for homeostasis of the endo/lysosomal system and is crucially implicated in neuronal function and integrity, as loss of function mutations in the PIKfyve complex lead to neurodegeneration in mouse models and human patients. Our recent work has shown that the intracellular domain of the Amyloid Precursor Protein (APP), a molecule central to the aetiology of Alzheimer's disease binds to VAC14 and enhances PIKfyve function. Here we utilise this recent advance to create an easy-to-use tool for increasing PIKfyve activity in cells. We fused APP's intracellular domain (AICD) to the HIV TAT domain, a cell permeable peptide allowing proteins to penetrate cells. The resultant TAT-AICD fusion protein is cell permeable and triggers an increase of PI(3,5)P2. Using the PI(3,5)P2 specific GFP-ML1Nx2 probe we show that cell-permeable AICD alters PI(3,5)P2 dynamics. TAT-AICD also provides partial protection from pharmacological inhibition of PIKfyve. All three lines of evidence show that the APP intracellular domain activates the PIKfyve complex in cells, a finding that is important for our understanding of the mechanism of neurodegeneration in Alzheimer's disease.

Spatial patterns of neuronal cytoplasmic inclusions in the tauopathies: A comparative study of five disorders

The tauopathies are a major molecular group of neurodegenerative disorders characterised by the deposition of abnormal cellular aggregates of the microtubule associated protein (MAP) tau in the form of neuronal cytoplasmic inclusions (NCI). Recent research suggests that cell to cell propagation of pathogenic tau may be involved in the neurodegeneration of these disorders. If pathogenic tau spreads along anatomical pathways it may give rise to specific spatial patterns of the NCI in brain tissue. To test this hypothesis, the spatial patterns of NCI in cerebral cortical regions were compared in tissue sections taken from five major tauopathies: (1) argyrophilic grain disease (AGD), (2) Alzheimer's disease (AD), (3) corticobasal degeneration (CBD), (4) Pick's disease (PiD), and (5) progressive supranuclear palsy (PSP). In the cerebral cortex of these disorders, NCI were frequently aggregated into clusters and the clusters were regularly distributed parallel to the pia mater. In a significant proportion of regions, the mean size of the regularly distributed clusters of NCI was in the range 400 – 800 m, measured parallel to the pia mater, approximating to the dimension of cell columns associated with the cortico-cortical anatomical pathways. Hence, the data suggest that cortical NCI in the tauopathies exhibit a spatial pattern in the cortex which could result from the spread of pathogenic tau along anatomical pathways. Treatments designed to protect the cortex from tau propagation may therefore be applicable across several different disorders within this molecular group.

Can neurodegenerative disease be defined by four 'primary determinants':anatomy, cells, molecules, and morphology?

Traditional methods of describing and classifying neurodegenerative disease are based on the clinico-pathological concept supported by molecular pathological studies and defined by 'consensus criteria'. Disease heterogeneity, overlap between disorders, and the presence of multiple co-pathologies, however, have questioned the validity and status of many traditional disorders. If cases of neurodegenerative disease are not easily classifiable into distinct entities, but more continuously distributed, then a new descriptive framework may be required. This review proposes that there are four key neuropathological features of neurodegenerative disease (the 'primary determinants') that could be used to provide such a framework, viz., the anatomical pathways affected by the disease ('anatomy'), the cell populations affected ('cells'), the molecular pathology of 'signature' pathological lesions ('molecules'), and the morphological types of neurodegeneration ('morphology'). This review first discusses the limitations of existing classificatory systems and second provides evidence that the four primary determinants could be used as axes to define all cases of neurodegenerative disease. To illustrate the methodology, the primary determinants were applied to the study of a group of closely related tauopathy cases and to heterogeneity within frontotemporal lobar degeneration with TDP-43 proteinopathy (FTLD-TDP).