Single genetic factors, mostly alterations in the coding exons of genes, can directly cause neurodegenerative diseases. In rare Mendelian disorders such single-gene defects are responsible for most of the disease mechanisms and phenotypes. This is the most simple and straightforward way of modeling the influence of genetic factors on neurodegeneration. A good example of such modeling is Huntington’s disease (HD). HD is an autosomal-dominant neuropsychiatric disease characterized by progressive movement disorder, most commonly presenting as chorea, progressive cognitive decline leading to dementia, and various psychiatric symptoms that often precede diagnosis (Paulson and Albin 2011). The mutation causing HD is a CAG repeat expansion in the HTT gene (Group THsDCR 1993). This unique mutation is highly specific; no other type of mutation in HTT has been found to cause HD and all affected individuals with HD carry the CAG repeat expansion in HTT. It has been well recognized that the number of repeats is inversely correlated with age of onset (Andrew et al. 1993). Similarly, expanded CAG repeats commonly lead to a lengthy polyglutamine (polyQ) stretch in other autosomal-dominant neurodegenerative disorders including spinocerebellar ataxias, spinobulbar muscular atrophy, and dentatorubral–pallidoluysian atrophy (Zoghbi and Orr 2000). The strong correlation between the number of CAG repeats and age of onset has also been identified in these diseases, suggesting that the disease-associated polymorphic genotype also determines disease onset. Although the presence of a common pathology influencing the age of onset (and presumably disease progression) among these diseases has been expected, molecular mechanisms underlying these polyQ expansions appear to be quite complex, as reviewed elsewhere (La Spada et al. 2011). Nevertheless, CAG expansion involving multiple genes is a unique genetic factor that is specifically associated with neurodegeneration.

Sometimes, a single clinical phenotype can be caused by an alteration in many different genes. This condition is called “genetic heterogeneity.” Charcot–Marie–Tooth (CMT) disease is one of the most common neurogenetic disorders, with a prevalence of 1 in 2500. CMT is characterized by bilateral distal wasting, weakness, and sensory loss that begins in the lower limbs and slowly progresses in a length-dependent manner. CMT is further divided into two forms, CMT1 (demyelinating neuropathy) and CMT2 (axonal neuropathy), based on electrophysiological findings. CMT can be dominant, recessive, or X-linked; more than 60 CMT-associated genes have been identified and the list is still growing (Timmerman et al. 2014). Mutations in each gene likely affect the diverse functions of Schwann cells (associated with CMT1) and neurons (associated with CMT2) in the peripheral nervous system. The most common CMT-causing mutation is the genomic duplication of 17p11.2 that harbors PMP22, which encodes a myelin membrane protein. Another myelin-associated gene, MPZ, is also commonly mutated in patients with CMT1. Mutations in GJB1, encoding connexin32, affect the gap junctions of the myelin membrane. Genes involved in transcription and mRNA processing (EGR2 and CTDP1), cytoskeletal structure (PRX, IFN2, FGD4, and FBLN5), endosomal sorting and cell signaling (LITAF, SH3TC2, MTMR2, MTMR13, SBF1, FIG4, DNM2, and NDRG1), and mitochondria (GDAP1 and HK1) in Schwann cells can be affected in CMT1 (reviewed in Rossor et al. 2013). In CMT2, abnormalities that are even more diverse occur in the subcellular components of the neuronal axon and cell body, including defective nuclear envelope and mRNA processing, endoplasmic reticulum and Golgi apparatus, endosomal sorting and cell signaling, proteasome and protein aggregation, mitochondria, channels, axonal transport, and synaptic transmission (Rossor et al. 2013). These findings suggest that a single neurodegenerative phenotype can occur as a consequence of multi-layered pathological steps affecting tissues, cells, subcellular functional units, and genes, each of which can be responsible for the disease phenotype when mutated.

Single-gene alterations may only explain the common forms of neurodegenerative diseases in a small number of patients, while the vast majority of patients with the same disease may carry no such alterations. Dissecting the molecular pathology of such mutations not only explains the disease in rare cases, but also sheds light on the pathogenesis in most patients, because the rare genetic defect may affect the pathway followed by the common form of the disease. Alzheimer’s disease serves as an example of such a case. Analyses of a small number of early-onset Alzheimer disease families showing an autosomal-dominant pattern of inheritance uncovered highly penetrant mutations in three genes: APP, PSEN1, and PSEN2 (Goate et al. 1991; Sherrington et al. 1995; Rogaev et al. 1995; Levy-Lahad et al. 1995). APP encodes a single transmembrane protein from which amyloid β (Aβ) polypeptides are cleaved by β- and γ-secretases. Aβ polypeptides polymerize to form insoluble Aβ aggregates, the main component of senile plaque, which is the pathological hallmark in patients’ brains. Most APP mutations are heterozygous missense alterations located in or near Aβ-coding exons (Goate et al. 1991). Rare whole-gene duplications (Rovelet-Lecrux et al. 2006) as well as small recessive deletions and missense alterations have also been reported (Di Fede et al. 2009). These APP mutations either change Aβ production, increase the Aβ₄₂ to Aβ₄₀ ratio (which differs in length according to cleavage sites; Aβ₄₂ is more toxic), or enhance fibril formation. Interestingly, one particular APP single-nucleotide polymorphism (SNP), which leads to a change in Ala673Thr, was reported to strongly protect carriers from Alzheimer’s disease (1/OR = 5.29) (Jonsson et al. 2012). This SNP is located close to the cleavage site of BACE1 (β-site APP cleaving enzyme 1) and functionally inhibits BACE1 cleavage to produce Aβ. PSEN1 and PSEN2 are both critical components of γ-secretase, and mutations in either gene result in an increased ratio of Aβ₄₂ to Aβ₄₀ (De Strooper et al. 1998; Scheuner et al. 1996). Findings from these early-onset familial Alzheimer disease cases provided the solid genetic basis for the central role of Aβ in the pathogenesis of the common form of Alzheimer’s disease.

In most patients suffering from neurodegenerative diseases, genetic factors play a more complex role in the pathogenesis, onset, phenotype, and prognosis of the disease than in patients suffering from rare Mendelian forms of diseases (discussed in the previous section). Multiple genes appear to orchestrate together with environmental factors to formulate disease status. Recent advances in genome-wide association studies (GWAS) have been successfully delineating the susceptibility genes that contribute to each part of pathogenesis in the complex process of neurodegenerative diseases.

The best known example of such susceptibility genes with the largest effect size is the APOE gene in Alzheimer’s disease (Corder et al. 1993). In fact, the Manhattan Plot (an overview of GWAS results encompassing the entire human genome) demonstrated a single incomparable high peak at the APOE locus, while other associated loci across the entire genome revealed only modest effects. The APOE gene harbors three polymorphic alleles, ε2, ε3, and ε4, of which the ε4 allele is a risk allele for Alzheimer’s disease (Farrer et al. 1997). A meta-analysis in individuals of European descent showed that the risk of Alzheimer’s disease is increased in people carrying one copy of the ε4 allele (ε2/ε4, odds ratio 2.6; ε3/ε4, odds ratio 3.2) or two copies (ε4/ε4, odds ratio 14.9), compared with those carrying an ε3/ε3 genotype (Farrer et al. 1997). This association is also observed in other populations with a weaker (African American ε4/ε4, odds ratio 5.7; Hispanic ε4/ε4, odds ratio 2.2) or a stronger (Japanese ε4/ε4, odds ratio 33.1) effect (Farrer et al. 1997). APOE genotypes strongly affect Aβ deposition as a result of haplotype-specific differences in metabolizing Aβ. Findings from multiple studies in humans and mice suggest that APOE ε4 increases the risk by initiating and accelerating the accumulation, aggregation, and deposition of Aβ in the brain (Liu et al. 2013). APOE ε4 is less efficient at clearing Aβ from the brain. APOE genotypes are also associated with mild cognitive impairment (MCI), which is now considered a pre–Alzheimer disease condition. Individuals with MCI who carry the ε4 genotype are at a higher risk of progression from MCI to Alzheimer-type dementia (Petersen et al. 1995) and suffer a faster cognitive decline than non-carriers (Cosentino et al. 2008). Thus, APOE serves as a major susceptibility gene for Alzheimer’s disease with the ε4 allele increasing the risk by direct enhancement of Aβ pathogenesis.

By contrast, other susceptibility genes in at least nine loci (CLU, CR1, PICALM, BIN1, EPHA1, ABCA7, MS4A cluster, CD33, and CD2AP), identified and confirmed by multiple GWAS, have shown a much smaller effect size (reviewed in Bettens et al. 2013). Pathogenic mechanisms of variations in these nine susceptibility genes have not yet been established; however, these susceptibility genes may be related to specific functions including lipid processing (CLU and ABCA7), functioning of the complement system, inflammation, immune system (CLU, CR1, ABCA7, CD33, and EPHA1), and synaptic cell functions such as endocytosis (PICALM, BIN1, CD33, and CD2AP). However, the mechanisms by which these gene variations affect the pathological process of Alzheimer’s disease remains largely undetermined. Interestingly, common gene variants that have a predominant effect when mutated (i.e., APP, PSEN1, and PSEN2) are not risk factors for the complex forms of Alzheimer’s disease (Harold et al. 2009; Lambert et al. 2009). This contrasts with the findings in Parkinson’s disease, wherein common variants in the genes that are responsible for the Mendelian form of the disease (i.e., MAPT, SNCA, and LRRK2) also serve as risk alleles in the common form of the disease (Simon-Sanchez et al. 2009; Satake et al. 2009).

Genetic exploration and identification of the genes contributing to neurodegenerative disorders have begun to uncover the fact that clinically distinct neurodegenerative disorders share a common genetic basis to a significant degree. A single gene can contribute to the occurrence of different diseases. There are striking examples of this in amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD).

ALS is a devastating degenerative disorder of motoneurons. It is characterized by the onset of focal weakness, typically in the limbs but sometimes in bulbar muscles. This progresses to paralysis of almost all skeletal muscles, leading to death from respiratory failure typically within 5 years (Sreedharan and Brown 2013). FTD is a distinct form of dementia characterized by progressive degeneration of the frontal or temporal lobe, or both (Loy et al. 2014). FTD is clinically characterized by deterioration in behavior and speech, but memory and visuospatial function are spared to a certain degree. ALS and FTD have been thought of as distinct clinical entities, although rare overlapping cases have been reported. The seminal molecular pathology discovery connecting ALS and FTD was the identification of TDP-43 as the major component of ubiquitinated neuronal cytoplasmic inclusions in both diseases (Neumann et al. 2006), which also led to the identification of mutations in the TARDBP gene (which encodes TDP-43) in autosomal-dominant ALS and FTD families (Sreedharan et al. 2008; Chio et al. 2010). More recently, a massive expansion of a hexanucleotide repeat in an intron of C9ORF72 linked to chromosome 9q21 was found to cause ALS and FTD (Renton et al. 2011; DeJesus-Hernandez et al. 2011). Subsequently, it became apparent that this locus was responsible for a major part of both familial ALS (~40 %) and FTD (~25 %) (Majounie et al. 2012). In addition, at least three other genes are known to link ALS and FTD. FUS encodes an RNA-binding protein that has functional homology with TDP-43. FUS mutations have been found in about 4 % of familial ALS cases without dementia (Kwiatkowski et al. 2009; Vance et al. 2009); however, in rare families, FUS mutations also cause FTD (Blair et al. 2010). Note that FTD exhibiting a FUS-positive histopathological subtype is commonly observed; however, most of these cases do not carry FUS mutations (Snowden et al. 2011). Mutations in the VCP gene, which encodes valosin-containing protein, were initially found to cause inclusion body myopathy, an unusual clinical syndrome characterized by FTD, and Paget’s disease of the bone (Watts et al. 2004); however, they were later found to be responsible for 1–2 % of familial ALS cases too (Johnson et al. 2010). Ubiquitin 2 regulates the proteasome degradation of ubiquitinated proteins. The gene encoding this protein, UBQLN2, is responsible for rare familial cases of ALS and ALS/FTD syndrome (Deng et al. 2011). Since UBQLN2 pathology has been observed in patients with ALS who do not carry mutations in this gene, this protein may play an important role in the final common pathway associated with motor neuron degradation. These cumulative genetic findings have drastically changed our understanding of unexpectedly common molecular pathologies between ALS and FTD.

Astrocytes constitute the largest volume of cells in the central nervous system (CNS). Named for its star-like shape, astrocytes are distributed throughout the CNS in both the gray and white matter. Once considered little more than scaffolds for neurons, astrocytes play multiple roles by interacting with many types of cells and structural components in the CNS – including neurons, oligodendrocytes, other astrocytes, blood vessels, pial membranes, and the ependymal layer that lines ventricles – to maintain tissue homeostasis (Lanciotti et al. 2013).

In their interaction with neurons, astrocytes extend many fine processes to contact synapses and establish highly regulated bidirectional communication. By secreting trophic factors – such as brain-derived nerve growth factor (BDNF), glial cell–derived neurotrophic factor (GDNF), nerve growth factor (NGF), insulin-like growth factor (IGF), and thrombospondin – astrocytes contribute to the formation, maintenance, and remodeling of synapses. Astrocytes also modulate neuronal function by secreting cholesterol, apolipoprotein E, glutathione, and hydrogen sulfide. They express receptors to take up neurotransmitters – including glutamate, gamma-aminobutyric acid (GABA), norepinephrine, dopamine, serotonin, acetylcholine, and glycine – to remove them from the synaptic cleft and maintain their low extracellular concentrations. The clearance of glutamate, mainly mediated by a high-affinity excitatory amino acid transporter 1 (EAAT1, also known as “GLT1”), is critical to protecting neurons from excitotoxicity (Rothstein et al. 1996). Dysfunction of EAAT1 and accumulation of excessive extracellular glutamate have been implicated in several neurodegenerative processes and associated diseases, such as epilepsy, stroke, ALS, HD, and Alzheimer’s disease (Coulter and Eid 2012; Rothstein 2009; Faideau et al. 2010; Simpson et al. 2010). Astrocytes also supply neurons with energy substrates. Both glucose and lactate are transported to neurons to maintain the energy source for intense neuronal activity. Astrocytes also control the homeostasis of water and ion fluxes in the synaptic interstitial space through numerous ion channels and transporters.

Astrocytes also provide growth factors – such as platelet-derived growth factor alpha (PDGF-α), fibroblast growth factor 2 (FGF2), neurotrophin-3 (NT-3), IGF-1, and ciliary neurotrophic factor (CNTF) – which are necessary for the survival of oligodendrocytes (Lanciotti et al. 2013). Astrocytes are thought to regulate CNS myelination probably by secreting growth factors; elucidation of the underlying mechanism is important to understanding how astrocytes contribute to CNS repair and remyelination, which could be useful in the treatment of demyelinating diseases.

Astrocytes not only play essential roles to maintain healthy CNS homeostasis, they are also involved in many pathological processes. Specifically, genetic deficits in astrocyte-specific genes can cause neurodegenerative disorders, characterized as cystic leukodystrophies including Alexander disease (AxD), megalencephalic leukodystrophy with subcortical cysts (MLC), and vanishing white matter (VWM). These are all rare genetic diseases that commonly show the characteristic features of demyelinating leukodystrophy with progressive cystic or spongy (vacuolating) degeneration of myelin.

AxD is an autosomal-dominant disease caused by heterozygous mutations in the astrocyte-specific type III intermediate filament GFAP gene (Brenner et al. 2001). Clinical manifestations of AxD vary in severity and onset from more severe infantile and juvenile forms to milder adult onset forms (Messing et al. 2012). Patients with the infantile form show progressive features with seizures, bulbar dysfunction, psychomotor regression, and short lifespan. Pathologically, the brains of patients with severe AxD show loss of oligodendrocytes and myelin; cystic degeneration; neuronal loss most commonly in the hippocampus, striatum, and neocortex; and the presence of Rosenthal fibers in the cytoplasm of astrocytes (Messing et al. 2012). The exact mechanisms for neurodegeneration in AxD remain unknown, but both toxic gain-of-function and impairment of normal astrocyte supportive function have been implicated. Specifically, a marked downregulation of GLT-1 in astrocytes expressing mutant GFAP, and the vulnerability of hippocampal neurons to glutamate-induced cytotoxicity in co-cultures with astrocytes expressing mutant GFAP have been reported (Mignot et al. 2007). These findings may link GFAP mutations to glial glutamate transporter-1 (GLT-1) dysfunction and impairment of the neuron–astrocyte interaction, suggestive of a potential mechanism underlying neuronal loss in AxD. Similarly, dysregulation of the homeostatic functions of astrocytes – including extracellular K⁺ buffering by Kir4.1 and Na⁺/K⁺–ATPase activity – have been implicated in the demyelination of AxD (Hagemann et al. 2005).

Mature oligodendrocytes extend cell processes that are modified into thin sheaths of plasma membrane wrapping around a portion of an axon in a spiral fashion (so-called “myelin”). Each oligodendrocyte interacts with as many as 50 different axons by extending multiple processes. Each axon is ensheathed by multiple continuous segments of myelin, separated by small areas of bare axons exposed to interstitial spaces called “nodes of Ranvier.” Myelin functions as an insulator of electronic circuits in neuronal networks, where it enables rapid, efficient, and distant transmission without attenuation. In addition to these unique features of myelin as a structural component for highly efficient axonal conduction, oligodendrocytes communicate with axons through several signaling processes. Axonal signals that support myelin include neuregulins, neurotrophins, and the electrical activity of axons, while oligodendrocytes support axons through the function of some myelin membrane–bound proteins, such as PLP1 and CNP1. Mice in which either PLP1 or CNP1 has been disrupted develop normal myelin, suggesting that these myelin proteins are dispensable for normal myelination. On the other hand, these mice develop broad axonal swelling, indicating that PLP1 and CNP1 support axons independent of myelin (Nave and Trapp 2008). Of these two myelin proteins only PLP1 is directly associated with human disease.

Leukodystrophy comprises a group of rare genetic disorders that primarily target white matter in which myelinating oligodendrocytes and axons predominate. Pelizaeus–Merzbacher disease (PMD) is a hypomyelinating leukodystrophy characterized by deficit in CNS myelination (Inoue 2005). PMD is caused by mutations in PLP1, which encodes a major myelin membrane protein. PMD-causing PLP1 mutations include point mutations that mostly result in amino acid substitutions, genomic duplications that lead to an additional copy of the gene, and null mutations caused by deletions or early-truncating mutations (Inoue 2005). In PMD the structure of the neuronal network is essentially intact. However, slowly progressive axonal degeneration may occur later in the disease probably because of the loss of axonal support by myelin and the resulting metabolic and electronic insufficiency in the axons. Activated microglia have also been found at the site of hypomyelinated axons; thus, microglia may additionally contribute to hypomyelination-associated axonopathy (Ip et al. 2006). In addition to these features (generally observed in all types of mutations), patients with null mutations show an apparent length-dependent progressive axonopathy characterized by axonal swelling (Griffiths et al. 1998; Garbern et al. 2002). This is despite the fact that clinical severity and associated leukodystrophy observed by magnetic resonance imaging (MRI) in patients with null mutations are surprisingly milder than in patients with either point mutations or duplications. It is not completely understood why a milder condition with almost normal myelinating oligodendrocytes with no PLP1 expression is more susceptible to axonal degeneration than severe conditions with bare axons and almost no mature oligodendrocytes caused by other types of mutations. Nevertheless, the expression of myelin-specific genes such as PLP1 appears to play a role in preventing the degeneration of axons.