Genetic defects in the sphingolipid degradation pathway and their effects on microglia in neurodegenerative disease

Abstract

Sphingolipids, which function as plasma membrane lipids and signaling molecules, are highly enriched in neuronal and myelin membranes in the nervous system. They are degraded in lysosomes by a defined sequence of enzymatic steps. In the related group of disorders, the sphingolipidoses, mutations in the genes that encode the individual degradative enzymes cause lysosomal accumulation of sphingolipids and often result in severe neurodegenerative disease. Here we review the information indicating that microglia, which actively clear sphingolipid-rich membranes in the brain during development and homeostasis, are directly affected by these mutations and promote neurodegeneration in the sphingolipidoses. We also identify parallels between the sphingolipidoses and more common forms of neurodegeneration, which both exhibit evidence of defective sphingolipid clearance in the nervous system.

1. Introduction

Sphingolipids are a diverse family of lipids that are characterized by a sphingosine backbone. They are one of the three major categories of plasma membrane lipids, with the other two being cholesterol and glycerol backbone-based phospholipids [1]. Sphingolipids also have important functions as signaling molecules [2]. The biosynthesis of sphingolipids begins with the condensation of an amino acid, usually serine, with a fatty acid, usually 16 carbons in length, to produce an 18-carbon “sphingoid” base. Derivatization with another fatty acid produces ceramide, a membrane anchor that can be further derivatized with hydrophilic head groups to give rise to plasma membrane sphingolipids, which include sphingomyelin and the glycosphingolipids. Signaling lipids, ceramide, sphingosine, and sphingosine-1-phosphate are produced by the degradation of plasma membrane sphingolipids [2].

Sphingolipids are highly expressed in the brain. Gangliosides–glycosphingolipids containing one or more sialic acid residues–are prominent in the gray matter and are major components of neuronal membranes, where they make up to 10–12% of the total lipid content of neurons [3–5]. Glycosphingolipids in the form of cerebrosides (glucosylceramide and galactosylceramide) and sulfatides (sulfated galactosylceramide) are highly enriched in the white matter [6]. These sphingolipids represent up to 80% of the non-sterol lipid of the highly specialized myelin membranes, which are produced by oligodendrocytes and Schwann cells, and wrap axons to facilitate the neuronal transmission of electrical impulses [6].

The major sphingolipid degradation pathway occurs in lysosomes and utilizes an orchestrated series of enzymatic reactions (Fig. 1). After the sphingolipids are internalized into lysosomes, hydrolases and activator proteins coordinate the removal of the hydrophilic head groups. For sphingomyelin, the phosphorylcholine head group is removed by acid sphingomyelinase to generate ceramide. For the glycosphingolipids, each sugar residue is removed by an individual reaction in a defined sequence until ceramide is produced. The ceramide backbone is then split by acid ceramidase into sphingosine and a fatty acid, both of which are transferred into the cytoplasm and recycled to serve as substrates in other biosynthetic pathways. If a single sphingolipid degradative step fails to occur due to gene mutation, the lipid substrate that is normally degraded by the missing hydrolase accumulates, giving rise to a lysosomal storage disease. These generally result in severe neurodegenerative diseases. Because the lipids that accumulate are sphingolipids, these disorders are termed sphingolipidoses. In these disorders, an enzyme-deficient cell type will “store” sphingolipid whether it is synthesized endogenously or comes from an exogenous source. The storage process can contribute to the pathogenic pathways of the sphingolipidoses in a number of ways (e.g., formation of toxic lipids, aberrant activation of signaling pathways, blockage in the endosomal-lysosomal pathway [which precludes autophagic function], and an increase in synaptic spines and formation of meganeurites) [7–12]. Inflammatory processes also contribute to disease progression [7, 8].

Fig. 1.

Disorders resulting from mutational defects in the lysosomal sphingolipid degradation pathway. Oligosaccharide structures are illustrated by colored symbols. Substrate names are presented in gray rounded boxes, genes are presented in green text, and disorders are presented in boxed red arrows. An asterisk (*) identifies a disease predisposition associated with a gene variant.

Microglia are the resident macrophages of the brain and have critical functions during development and homeostasis [13, 14]. They mediate injury responses and pathogen defense, and are a primary source of pro-inflammatory cytokines [14]. They also have important roles in the pathogenesis of neurodegenerative diseases [14].

Normally microglia constantly surveil the brain parenchyma and eliminate dying neurons, remove neuronal synapses, and clean up myelin debris [13, 15–17] (Fig. 2). Sphingolipid-rich neuronal and myelin membranes captured through these processes undergo lysosomal degradation within microglia. This degradative process is facilitated by a lipid-sensing receptor, TREM2, that is activated by various lipids (including sphingolipids, sphingomyelin, and sulfatide) (Fig. 2). TREM2 signals via two receptor tyrosine kinases, DNAX-activation protein 10 (DAP10) and DAP12, to modulate several important microglial functions, including lipid uptake, transport, and processing [18, 19].

Fig. 2.

The sphingolipid degradation pathway and microglia in neurodegenerative disease. Microglia function during both development and homeostasis to eliminate dying neurons, prune neuronal synapses, and clear myelin debris. To accomplish these functions, sphingolipid-rich membranes are acquired from exogenous sources through the lipid-sensing receptor TREM2 and then undergo lysosomal degradation within microglia. During aging and the diseases indicated, these microglial processes are accelerated. Gene mutations in the sphingolipid degradation pathway genes reduce catabolic capacity and lead to lysosomal storage of sphingolipids, causing intrinsic microglial defects (such as a dysfunctional endosome-lysosome system), as well as compromising homeostatic functions and inducing a pro-inflammatory phenotype that promotes neurodegeneration.

To provide an understanding of how the sphingolipid degradation pathway influences microglial function, we review the relevant findings for the sphingolipid storage diseases in which the individual steps of the sphingolipid degradation pathway are disrupted. Microglia are directly affected by these mutations and promote neurodegeneration in the sphingolipidoses by impairment of their normal homeostatic functions and the acquisition of disease promoting characteristics. We also present evidence indicating that sphingolipid degradation pathway defects are a risk factor in common forms of neurodegeneration

2. Farber disease

2.1. Disease description

Farber disease is a rare autosomal recessive, progressive neurovisceral disorder that is characterized by macrophages ladened with lipid storage and inflammation throughout the body. Accumulation of ceramide in the visceral organs and cerebral white matter of severely affected Farber disease patients was first reported by Mosher and colleagues [20, 21]. The severity and symptoms vary among patients depending on the degree and location of the lipid build-up. The most common symptoms are subcutaneous nodules, joint deformities, and laryngeal hoarseness associated with neurological deficit. Patients with significant neurological involvement usually die early in infancy. Farber disease is a lysosomal storage disorder caused by mutations in the ASAH1 gene, which encodes acid ceramidase. Acid ceramidase catalyzes the degradation of ceramides to sphingosine and fatty acids (Fig. 1). Therefore, mutations of this gene cause ceramide accumulation in the lysosome [22].

2.2. Microglia in Farber disease

In a limited study of a Farber disease patient’s brain, the neurons were distended with storage materials and neuronal loss was accompanied by marked proliferation and activation of microglia and gliosis [21]. The ASAH1^{P361R/ P361R} mouse model exhibits features of human neurovisceral Farber disease [23]. In ASAH1^{P361R/ P361R} mouse brain, ceramide, hydroxyceramides, dihydroceramides, sphingosine, dihexosylceramides, and GM3 ganglioside were elevated [24]. Numerous microglial and/or macrophage abnormalities were noted as pathological hallmarks [24]. At 3 weeks of age, expanded foamy cells with varying levels of lipid storage appeared in the white matter. These cells formed fused multi-nucleated granuloma-like structures and frequently organized into perivascular cuffs at later ages.

Abnormal neuronophagic clusters appeared later only in specific gray matter regions, such as the hippocampal CA1 region, some thalamic nuclei, and layer II of the primary somatosensory cortex. At later ages, granuloma-like clustering was observed in the cerebellum. The morphologic alteration of neurons was not as pronounced as in the microglia. These findings indicate that microglial and/or macrophage abnormalities are a dominant neuropathological feature in Farber disease.

3. Gaucher disease

3.1. Disease description

Gaucher disease is an autosomal recessive lysosomal storage disorder characterized by the deficiency of lysosomal acid β-glucosylceramidase, which is caused by mutations in the GBA1 gene [7]. Specific GBA1 mutations correlate in part with disease severity and residual enzyme activity. Gaucher disease occurs as a neuropathic form with an early (type 2) or late onset (type 3), or as a systemic form (type 1) without central nervous system (CNS) involvement [25]. Neuropathic Gaucher disease is characterized by progressive neurodegeneration and brain inflammation [26].

β-glucosylceramidase is responsible for the removal of β-linked glucose from glucosylceramide to yield ceramide (Fig. 1) [27]. β-glucosylceramidase activity is dependent on the activator protein saposin C, which facilitates accessibility of the glycosphingolipid substrate [28]. Reduced β-glucosylceramidase activity due to mutations in GBA1 leads to the accumulation of glucosylceramide. Some accumulated glucosylceramide is converted by the action of acid ceramidase to its lyso-form, glucosylsphingosine, which also accumulates in Gaucher disease [29]. Glucosylceramide and glucosylsphingosine storage are particularly predominant in blood, liver, and spleen macrophages, which are enlarged with glycosphingolipid-loaded lysosomes and referred to as Gaucher cells [7]. Neuropathic forms of Gaucher disease are characterized by perivascular and parenchymal accumulation of Gaucher cells, accompanied by severe neuronal loss in several areas of the brain and brain stem, with extensive astrogliosis and microglial activation [30, 31].

3.2. Microglia in Gaucher disease

Accumulation of glucosylceramide and glucosylsphingosine in neuronopathic Gaucher disease leads to massive neuronal loss and inflammation in the brain [30, 32, 33]. Elevated levels of pro-inflammatory cytokines, nitric oxide, and reactive oxygen species were detected in fetal Gaucher disease brain [34]. An increase in inflammatory markers correlated with disease progression and severity in Gaucher disease mice [32]. In addition, a gene expression profile of a Gaucher disease type 2 patient brain revealed a pattern associated with activated astrocytes, a cell type that mediates injury responses and inflammation in the CNS [35].

Activated microglia are a common finding in the brain of Gaucher disease patients [35], and are the only myeloid population detected in significant numbers within the brain parenchyma in Gaucher disease mice [32, 34, 36]. Studies in mouse models indicate that while the neuronal deficiency of GBA1 is a primary determinant of CNS pathogenesis, GBA1-deficient microglia may also contribute by influencing the onset and progression of the disease [32, 37, 38]. Microglial activation together with astrogliosis are spatially and temporally correlated with neuronal loss, but it is not known if these inflammatory events precede or are concomitant with neuronal cell death [33, 37]. However, in a Gaucher disease zebrafish model, glycosphingolipid accumulation induced activated microglia and marked inflammation prior to neuronal loss [39]. Together, these findings support the idea that the activation of microglia in Gaucher disease stemming from the defective sphingolipid degradation pathway contributes to the process of neuronal cell death.

4. Krabbe disease

4.1. Disease description

Krabbe disease, also known as globoid cell leukodystrophy, is a rare autosomal recessive lysosomal storage disorder [40]. Krabbe disease is caused by mutations in the GALC gene resulting in a deficiency of the lysosomal enzyme β-galactocerebrosidase (Fig. 1). Functioning with activator saposin A, β-galactocerebrosidase removes β-linked galactose mainly from galactosylceramide and galactosylsphingosine (also known as psychosine) [41]. Psychosine, a bioactive lipid, accumulates at several hundred-fold levels above normal in Krabbe disease, leading to progressive demyelination and neurodegeneration in both the CNS and the peripheral nervous system (PNS) [42].

More than 90% of Krabbe disease cases have an onset before 6 months. Initial symptoms typically include hyperirritability, hypersensitivity, unknown episodic fever, and limb stiffness. The neurological dysfunction soon progresses into hypertonicity, as well as loss of vision and hearing. The affected patients have difficulty moving, eating, and breathing, which leads to death by the age of 2 [43].

4.2. Microglia in Krabbe disease

Multi-nucleated giant phagocytic globoid cell formation is the pathological hallmark for Krabbe disease [42]. In the brain, globoid cells may originate from residential microglia or infiltrating macrophages [44]. In vivo studies in twitcher mice, a naturally occurring mouse model of Krabbe disease, found that neither inflammation nor the globoid cell population was altered in the brain when macrophage infiltration is blocked by deficiency of the chemokine receptor CXCR2, suggesting that microglia, not infiltrating macrophages, are the major contributor to globoid cell formation in the CNS in Krabbe disease [45]. Moreover, in vitro, primary-cultured microglia formed multi-nucleated globoid cells after psychosine treatment [46], and became cytotoxic for oligodendrocytes [47].

In Krabbe disease, microglia play important roles in disease progression. They may induce astrogliosis via the inflammation mediator prostaglandin D₂ and promote myelin loss and neurodegeneration [48]. Microglia are also needed for maintaining myelin homeostasis, and compromised microglia lead to increases in myelin debris and more deleterious demyelination in twitcher mice [49].

Activation of microglia and globoid cell formation are early events in Krabbe disease. Evidence of globoid cell formation was found in 21-week-old Krabbe disease human fetal spinal cord [50]. Activated microglia were observed in 2-week-old twitcher mouse brain before demyelination onset [51]. In mice carrying the E130K missense Krabbe disease mutation (which have a shorter lifespan than twitcher mice), gliosis, globoid cells, and psychosine accumulation were present without significant demyelination in the CNS [52], suggesting that neuroinflammation plays a critical role in disease progression independent of myelin loss.

Galc-deficient microglia in twitcher mice may also have intrinsic defects that contribute to Krabbe disease progression. Scott-Hewitt et al. [53] showed that Galc^+/− mouse microglia had reduced upregulation of the TREM2 lipid-sensing receptor (Fig. 2), as well as decreased phagocytic clearance of myelin debris. In addition, PNS macrophages (microglia are not present in the PNS) require β-galactocerebrosidase for phagocytosis and myelin turnover [54]. In this paradigm, hematopoietic stem cell transplantation for Krabbe disease is beneficial by providing β-galactocerebrosidase-competent macrophages/microglia to restore the myelin phagocytosis function, thereby reducing the highly inflammatory globoid cell reaction [54].

5. GM1 gangliosidosis

5.1. Disease description

GM1 gangliosidosis is an autosomal recessive lysosomal storage disorder characterized by the deficiency of lysosomal acid β-galactosidase encoded by GLB1 [55]. β-Galactosidase cleaves β-linked terminal galactosyl residues from GM1 ganglioside (Fig. 1), as well as from glycoproteins and glycosaminoglycans [7]. GM1 ganglioside degradation does not require an activator protein, although it is enhanced by saposin B and GM2 activator protein [56].

GM1 gangliosidosis presents as a fatal infantile form, characterized by developmental arrest and rapid, progressive neurodegeneration. Less severe juvenile and chronic, adult forms of GM1 gangliosidosis also exist. In GM1 gangliosidosis, GM1 ganglioside accumulation occurs in the CNS and other organs, and neurons become filled with membranous cytoplasmic bodies and vacuoles as a result of the GM1 ganglioside storage [7, 57].

5.2. Microglia in GM1 gangliosidosis

Neuroinflammation plays an important role in disease progression for GM1 gangliosidosis, which has been most extensively studied in Glb1-deficient mice, a model that recapitulates features of the infantile form of GM1 gangliosidosis [58]. Brains of GM1 gangliosidosis mice exhibited elevated levels of inflammatory markers and cytokines [59, 60]. Microglial activation was noted in pre-symptomatic GM1 gangliosidosis mouse brains, starting in areas with the highest GM1 ganglioside accumulation and increasing gradually with disease progression, suggesting that microglial activation may contribute to neuronal death in this disease [59]. In addition, infiltration of monocytes and macrophages contributed to the inflammatory response, which was suppressed by bone marrow transplantation [60]. These results are consistent with a block in GM1 ganglioside degradation causing microglial activation and macrophage infiltration that is damaging to the nervous system.

6. Metachromatic leukodystrophy

6.1. Disease description

Metachromatic leukodystrophy (MLD) is a rare autosomal recessive lysosomal storage disease. MLD is caused by deficiency of lysosomal enzyme arylsulfatase A, which is encoded by the ARSA gene [61]. It cleaves 3-sulfate from the galactosyl moiety of sulfatide (Fig. 1), and requires an activator protein, saposin B [62]. In MLD, sulfatide storage was found in neurons, astrocytes, and activated macrophages/microglia, as well as in Schwann cells and oligodendrocytes [63], and leads to progressive demyelination and neurodegenerative manifestations [64].

Based on the disease onset, MLD is classified into three types: late-infantile (age of onset before 30 months); juvenile (2.5–16 years); and adult (after 16 years) [65]. Late-infantile onset accounts for nearly 50% of all cases worldwide, and is generally characterized by rapidly progressive psychomotor regression, including muscle weakness, areflexia, and ataxia. In the late stage of late-infantile onset disease, patients suffer from dysphagia, neuropathic pain, and severe foot deformities. Death usually occurs within a few years after disease onset.

6.2. Microglia in MLD

In human MLD, intense gliosis is marked in demyelinating areas [66, 67]. In an autopsy study, Bergner et al. [68] reported that in human MLD brains, activated amoeboid microglia could be found clustered in “prelesional” areas with normal white matter morphology. Microglia in prelesional areas showed evidence of apoptosis. Closer to the early gliotic scar areas, the percentage of amoeboid microglial cells increased and microglial decay/death became apparent, accompanied by lysosomal breakdown and cell membrane lysis. In contrast, in advanced gliotic scarring centers with complete demyelination, only myelin and oligodendrocyte remnants were observed. In mouse models, introduction of ARSA-competent macrophages/microglia into the MLD nervous system by transplantation with genetically engineered hematopoietic stem cells has been shown to be effective in reducing the severity of the disease [69].

Activation of microglia is an early event in MLD mouse models. ARSA knockout (KO) mice developed activated microglia with no obvious demyelination. In a demyelinating ARSA KO mouse model engineered to exhibit excess sulfatide synthesis, robust microglial activation and inflammatory cytokine (Ccl-2, Ccl-3, Ccl-4, and interleukins) elevation occurred before obvious demyelination appeared [70]. Correspondingly, several inflammatory cytokines (Ccl-2, Ccl-4, IL-1Ra, IL-8, and VEGF) were found to be significantly elevated in the cerebrospinal fluid of MLD patients [71]. In vitro, exposure of cultured primary microglia to sulfatide led to activation and production of inflammatory molecules [72].

These results point to a central pathophysiologic role of activated, inflammatory microglia in MLD stemming from their intrinsic inability to degrade the sphingolipid sulfatide within myelin membrane fragments.

7. Niemann-Pick disease type A

7.1. Disease description

Niemann-Pick disease type A (NPA) is a rare autosomal recessive, neurodegenerative disorder characterized by brain atrophy, hypomyelination, and Purkinje cell degeneration [73–76]. These defects lead to tremors and ataxia, and ultimately result in death in early childhood [7]. NPA is a lysosomal storage disorder caused by mutations in the SMPD1 gene, which encodes acid sphingomyelinase (ASM). ASM catalyzes the degradation of sphingomyelin to ceramide and phosphorylcholine (Fig. 1) [77, 78]. Mutations of this gene therefore cause build-up of sphingomyelin within lysosomes. In the nervous system, extensive sphingomyelin accumulation occurs in macrophages, vascular endothelial cells, and neurons. Schwann cells and, to a minor extent, oligodendrocytes also store sphingomyelin in NPA [79–82].

7.2. Microglia in NPA

Activated and expanded microglia were found in regions of the brain–cerebellum, cortex, and hippocampus–associated with neuronal death in NPA [83, 84]. Further, both inflammatory and anti-inflammatory activated microglia were detected in a mouse model of NPA. The anti-inflammatory microglia initially provided a beneficial function by clearing myelin debris, but became neurotoxic because of lipid overload caused by their intrinsic inability to degrade sphingomyelin. The sphingomyelin accumulation in microglia induced lysosomal damage, resulting in cathepsin B extracellular release, which was toxic to neurons [84]. The findings directly implicate the ASM-deficient microglia in the neurodegenerative process in NPA.

8. Niemann-Pick disease type C

8.1. Disease description

Niemann-Pick disease type C (NPC) is an autosomal recessive, neurodegenerative sphingolipid storage disorder that results in ataxia, dysarthria, dementia, dysphagia, and vertical supranuclear gaze palsy [85]. Disease presentation and symptom onset are variable, but most patients are diagnosed in late childhood and live until 10 to 25 years of age [86]. NPC is predominantly caused by mutations in the NPC1 gene (95% of cases), and very infrequently by mutations in the NPC2 gene [85]. NPC1 and NPC2 both reside in the late endosome/lysosome (LE/L), where NPC1 is an integral membrane protein and NPC2 is a soluble lumenal protein [87]. Both proteins have cholesterol-binding pockets (CBPs) and play a crucial role in regulating intracellular cholesterol levels [88].

In lysosomes, NPC2 binds to free cholesterol and transfers it to the NPC1 N-terminal CBP [87–89]. CBP-bound cholesterol is then moved to the NPC1 transmembrane sterol-sensing domain, from where the lipid is subsequently exported from the LE/L (Fig. 1) [89, 90]. Mutations in either NPC gene cause faulty lipid trafficking, resulting in cholesterol build-up in the LE/L [91]. Although cholesterol accumulation is the hallmark NPC cellular phenotype, mutations in NPC1 and NPC2 also result in sphingolipid accumulation in the LE/L [91, 92].

8.2. Microglia in NPC

Accumulation of cholesterol and sphingolipids were found in brain and other organs in NPC [91]. Because multiple lipids are stored, it is difficult to know the contribution of specific lipids to NPC disease progression. However, it is known that this sphingolipid accumulation occurs in axons of neurons in motor and sensory pathways, with pronounced effect on Purkinje cells in the cerebellum [93]. In fact, Lloyd-Evans et al. [94] found that sphingosine is the first lipid to build up in NPC-mutant cells, while the other lipids, including cholesterol, take longer to accumulate. The accumulation of sphingosine in NPC has been proposed to induce an imbalance in calcium homeostasis and affect lysosomal trafficking [94].

Microglial activation is a crucial CNS event in NPC that occurs before any neurodegeneration is evident [95]. Microglia are profoundly affected in NPC, with enhanced phagocytic uptake and impaired lipid processing that compromises their functions during development [95]. Changes in microglial morphology, gene expression, inflammatory markers, and phagocytic function provide evidence for a highly activated microglial state in NPC [95–97].

In an NPC mouse model, activated microglia showed significant reduction in expression of microglial lineage markers (CX₃CR1 and CD11b) and increased expression of chemokines and innate immune receptors [96]. Cologna et al. [97] found an increase in inflammatory markers in cerebrospinal fluid of NPC patients, providing further evidence that neuroinflammation is associated with microglial activation and disease progression.

Loss of NPC1 enhanced phagocytic uptake and impaired lipid trafficking in microglia [95]. This resulted in accumulation of phagocytosed myelin in LE/L and multi-vesicular bodies, possibly resulting from impaired myelin degradation [95] due to impaired vesicular fusion. The increase in phagocytic activity, in addition to impaired lipid turnover, creates a severely dysfunctional microglial phenotype in NPC. Cougnoux et al. [96] found that deletion of IRF8, a transcription factor that enables microglial activation, slowed disease progression and prolonged lifespan in NPC1 KO mice. These findings support the notion that intrinsic changes in microglial function, caused by lipid storage, drive NPC disease progression in the CNS.

9. Sandhoff disease and Tay-Sachs disease

9.1. Disease description

Sandhoff and Tay-Sachs diseases are rare autosomal recessive, lysosomal storage disorders caused by the deficiency of the lysosomal hydrolase β-hexosaminidase. β-Hexosaminidases A and B are dimers composed of two subunits, α and β, encoded by the HEXA gene and HEXB gene, respectively. In Tay-Sachs disease, mutations in the HEXA gene result in a deficiency of β-hexosaminidase A (αβ); in Sandhoff disease, mutations in the HEXB gene result in a deficiency of both β-hexosaminidase A (αβ) and B (ββ) isozymes [98].

β-hexosaminidase A is the only isozyme that removes terminal β-linked N-acetyl-D-galactosamine residues from GM2 ganglioside to form GM3 ganglioside (Fig. 1). The enzyme requires the GM2 activator protein for its ganglioside-degrading activity. Deficiency of β-hexosaminidase A activity results in the accumulation of GM2 ganglioside, as well as GA2 glycolipid, in the lysosomes, forming lamellar membranous inclusions in neurons and glial cells in the CNS in both Sandhoff and Tay-Sachs diseases [7, 98, 99].

Sandhoff and Tay-Sachs diseases present as severe forms in childhood and adolescence, or as chronic, late-onset forms, characterized by progressive neurodegeneration. Symptoms include progressive motor deterioration, hypotonia, blindness, seizures, and macrocephaly in the infantile forms, with slower progression and milder symptoms in later onset forms [98].

9.2. Microglia in Sandhoff and Tay-Sachs diseases

Neuroinflammation has been shown to have a crucial role in the pathogenesis of Sandhoff and Tay-Sachs diseases, and the mechanisms involved have been most extensively studied in the Sandhoff disease (Hexb KO) mouse model, which recapitulates central aspects of both of these human diseases. In this model, increased pro-inflammatory cytokine expression in the brain correlates with disease manifestations [59, 100]. Moreover, deletion of the cytokine tumor necrosis factor-α decreased levels of astrogliosis and reduced neuronal cell death, with no alterations in neuronal storage of gangliosides, and slightly improved lifespan in Sandhoff disease mice [101].

Microglial activation plays a significant role in disease progression in Sandhoff and Tay-Sachs diseases. Evidence for intense inflammatory responses in the brain attributed to macrophage/microglial activation was found in Sandhoff and Tay-Sachs disease patients, as well as in Sandhoff disease mice [59, 100, 102, 103]. Activated microglial expansion and other inflammatory processes attributed to reactive microglia were found to precede the substantial neuronal cell death observed in Sandhoff disease mouse brain, suggesting that microglia drive the inflammatory response and contribute to disease progression, including neuronal cell death [59, 101, 102]. When β-hexosaminidase expression was induced specifically in neurons of Sandhoff disease mice, their lifespan was partially extended, with a substantial improvement in neuropathology [104, 105]. This incomplete rescue may point to an additional requirement for β-hexosaminidase expression in other CNS cell types, such as microglia. Interestingly, HEXB is highly expressed by microglia compared with other cell types in the CNS and is considered a microglial signature gene in mice [106]. Notably, deletion of hexb in zebrafish caused abnormalities in radial glia, which are neuronal and glial progenitors, and within microglia themselves [107].

In addition to microglia functioning as potentially damaging inflammatory mediators, peripheral monocytes and macrophages can infiltrate the brain in Sandhoff disease mice and contribute to inflammation [59, 100, 108, 109]. Supplying normal monocytes and macrophages through the transplantation of normal bone marrow into Sandhoff disease mice prolonged lifespan by suppression of both the explosive expansion of activated microglia and neuronal cell death, without detectable decreases in GM2 ganglioside storage [102, 110]. This effect was thought to be due to enzyme-competent monocytes/macrophages that had infiltrated the CNS. The suppression of enzyme-deficient monocyte infiltration from the peripheral circulation by deletion of monocyte chemoattractant protein-1α in Sandhoff disease mice retarded neurodegeneration [100, 109]. These studies support an important role for inflammation responses mediated by microglia and infiltrating monocytes/macrophages in the neurodegeneration that characterizes Sandhoff and Tay-Sachs diseases.

10. The sphingolipid degradation pathway and microglia in common forms of neurodegeneration

10.1. Aging

Aging is considered the primary risk factor for neurodegenerative disease [111]. During aging, microglia develop defective homeostatic functions and elevated reactive oxygen species and secrete pro-inflammatory cytokines–processes that may contribute to age-related neurodegeneration. Aged microglia have a defective lysosomal degradative pathway and accumulate inclusions and lipofucin (lipid-containing pigment) in their lysosomes [112, 113]. In mouse brain, debris from myelin fragmentation was found to be increased with age. These myelin fragments were cleared by microglia, but led to lysosomal accumulation of myelin fragments and lipofucin-like material. The accumulation of large amounts of myelin debris caused subsequent activation of microglia [16]. Thus, age-related myelin fragmentation is substantial, leading to lysosomal myelin storage and contributing to microglial senescence and immune dysfunction in aging. These results suggest that the microglial lysosomal degradative pathway for sphingolipid-rich myelin debris generated during aging is sensitive to overloading [16].

10.2. Alzheimer’s disease

Alzheimer’s disease, the most prevalent neurodegenerative disorder, is also the most common cause of age-related dementia [114]. Its pathological features include brain atrophy, neurodegeneration, extracellular amyloid plaque accumulation (amyloid β peptide aggregates), and tau protein neurofibrillary tangles. The accumulation of lipids in microglial cells of patients with Alzheimer’s disease was first observed by Alois Alzheimer in 1907 when he described glial cells with numerous fibers and adipose saccules [115]. Microglia are crucial for the preliminary defense against the progression of Alzheimer’s disease, in part because they surround initial fibrillar amyloid depositions and act as a stable barrier inhibiting polymerization [116]. They are also responsible for phagocytotic clearance of amyloid β [117]. Importantly, the majority of risk genes for Alzheimer’s disease are expressed in microglia, with many being microglia-specific [118].

Lipid sensing by microglia appears to reduce the risk for Alzheimer’s disease. A loss of function mutation in the TREM2 gene is associated with Alzheimer’s disease [119]. TREM2 is a transmembrane glycoprotein expressed on microglia that associates with transmembrane adapter DAP12 for downstream signaling to promote microglial proliferation and survival. TREM2 binds to and is activated by various membrane lipids, including the sphingolipids, sphingomyelin, and sulfatide. In the absence of TREM2, microglia accumulate sphingolipids after myelin fragmentation, suggesting that activation of TREM2 promotes the clearance of myelin fragments via the enhancement of the microglial lysosomal degradation pathway for sphingolipids [18, 19, 120].

HEXB (see Section 9) has been identified as a potential Alzheimer’s disease risk gene (Fig. 1) that is specifically expressed in microglia and responsive to amyloid β [121]. In the mouse model of Sandhoff disease (Hexb KO), aggregated proteins associated with Alzheimer’s disease, amyloid β peptide-like, and phospho-tau-like all accumulated in the brain [122]. These findings indicate that impaired degradation of GM2 ganglioside, which accumulates in Sandhoff disease, hinders the clearance of proteins connected with Alzheimer’s disease, possibly by impairment of lysosomal function [123, 124].

10.3. Parkinson’s disease

Parkinson’s disease is the second-most common neurodegenerative disorder. The clinical symptoms of this disease center around motor dysfunction, resulting in a resting tremor, rigid face muscles, and bradykinesia [125]. These symptoms are driven by the loss of dopaminergic neurons within the substantia nigra pars compacta. A hallmark of Parkinson’s disease is that the remaining neurons contain α-synuclein-positive inclusions, referred to as Lewy bodies [125]. Neuroinflammatory processes, including activated microglia, have been reported to play an important role in the pathophysiology of Parkinson’s disease [126].

Whereas biallelic mutations in GBA1 cause Gaucher disease (see Section 3), heterozygous mutations in GBA1 are the most important risk factor for developing Parkinson’s disease [127, 128]. Other damaging sphingolipid storage disease gene variants have also been identified in association with Parkinson’s disease risk. Variants of the SMPD1 gene, whose deficiency is responsible for NPA, the ASAH1 gene, whose deficiency causes Farber disease, GALC, whose deficiency causes Krabbe disease, and ARSA, whose deficiency is responsible for MLD, have been identified as candidate Parkinson’s disease susceptibility genes (Fig. 1) [129–131]. In addition, reduced activities of lysosomal sphingolipid hydrolases have been observed in Parkinson’s disease [132]. These findings link defects in the sphingolipid degradation pathway with increased risk for Parkinson’s disease. However, whether impaired sphingolipid degradation in microglia is a factor in Parkinson’s disease onset or progression remains to be determined.

11. Conclusions

Sphingolipids are widely expressed, and are relatively highly expressed in the brain. As a consequence, genetic disruption of the sphingolipid degradation pathway often causes catastrophic neurodegeneration. While the sphingolipid accumulation that occurs in the sphingolipidoses likely has a direct impact on most nervous system cell types, microglia are often affected. Evidence has emerged that sphingolipidosis microglia have intrinsic defects that can contribute to the disease process by the corruption of their normal homeostatic functions and by the acquisition of damaging pro-inflammatory phenotypes (Fig. 2).

The sphingolipid degradation pathway has been linked to more common forms of neurodegeneration. During aging, a major risk factor for neurodegeneration, sphingolipid-degrading enzyme activities are reduced [132], and microglial lipid storage is increased. In the more common neurodegenerative diseases, Alzheimer’s disease and Parkinson’s disease, increased disease risk has been associated with genetic variants in the pathway of sphingolipid uptake and degradation (Fig. 1).

Normally, cells have essentially no lysosomal sphingolipid accumulation because the maximum turnover rate of the sphingolipid degradation pathway is much higher than the influx rate of the substrate entering into the lysosomal compartment [133]. Microglia must continually take up and degrade cell membranes containing high levels of sphingolipids during their developmental and homeostatic activities of synaptic pruning, removal of dying neurons, and clearance of myelin debris [13]. During aging and neurodegenerative disease, these processes are accelerated, increasing the sphingolipid substrate influx into the lysosomal compartment, which would be predicted to raise the critical threshold value for catabolic activity required to prevent sphingolipid storage [133]. Reductions in the catabolic activity of the pathway below this critical threshold value would therefore result in sphingolipid accumulation. When sphingolipid degradation pathway genes are mutated, cells with excessive sphingolipid burden, such as microglia and other phagocytic cells, may be affected early in the pathogenic process and contribute to neurodegeneration. Although this appears to be the case for some of the sphingolipidoses, further studies are needed to define the link between microglial sphingolipid degradation and the more common forms of neurodegeneration.

Highlights.

• Sphingolipids are highly expressed in the brain

• Sphingolipids are degraded in lysosomes by a sequence of enzymatic steps

• Sphingolipidoses are caused by defects in sphingolipid degradation pathway genes

• Sphingolipid degradation pathway genetic defects affect microglial functions

• Altered microglial functions promote neurodegeneration in the sphingolipidoses

• Gene variants in the sphingolipid degradation pathway may increase risk for more common neurodegenerative diseases

Abbreviations:

ASM

acid sphingomyelinase

CBP

cholesterol-binding pocket

CNS

central nervous system

DAP

DNAX-activation protein

KO

knockout

LE/L

late endosome/lysosome

MLD

Metachromatic leukodystrophy

NPA

Niemann-Pick disease type A

NPC

Niemann-Pick disease type C

PNS

peripheral nervous system