Glial Cytokines in Alzheimer’s Disease

Abstract

The roles of activated glia and of glial cytokines in the pathogenesis of Alzheimer’s disease are reviewed. Interleukin-1 (IL-1), a microglia-derived acute phase cytokine, activates astrocytes and induces expression of the astrocyte-derived cytokine, S100β, which stimulates neurite growth (and thus has been implicated in neuritic plaque formation) and increases intracellular free calcium levels. Interleukin-1 also upregulates expression and processing of β-amyloid precursor proteins (β-APPs) (thus favoring β-amyloid deposition) and induces expression of α₁-antichymotrypsin, thromboplastin, the complement protein C3, and apolipoprotein E, all of which are present in neuritic plaques. These cytokines, and the molecular and cellular events that they engender, form a complex of interactions that may be capable of self-propagation, leading to chronic overexpression of glial cytokines with neurodegenerative consequences. Self-propagation may be facilitated by means of several reinforcing feedback loops. β-Amyloid, for instance, directly activates microglia, thus inducing further IL-1 production, and activates the complement system, which also leads to microglial activation with IL-1 expression. Self-propagation also could result when S100β-induced increases in intraneuronal free calcium levels lead to neuronal injury and death with consequent microglial activation. Such chronic, self-propagating, cytokine-mediated molecular and cellular reactions would explain the progressive neurodegeneration and dementia of Alzheimer’s disease.

Extracellular deposition of β-amyloid, initially in the form of diffuse amyloid deposits not associated with glial, neuronal, neuritic, or synaptic alterations, is widely believed to be an important early event in the progressive neurodegeneration and dementia of Alzheimer’s disease.¹ These diffuse deposits, however, neither establish the diagnosis of Alzheimer’s disease nor explain the progressive dementia characteristic of this disease. It is the appearance of the diagnostic neuritic plaques, with associated dystrophic neurites and “condensed” (congophilic) amyloid, and ultimately the loss of neurons and synapses,^2–4 that correlate with the intellectual decline of Alzheimer’s disease.^3,5,6 Neuritic plaques have been shown to contain, in addition to β-amyloid and dystrophic neurites, various nonamyloid proteins (including α₁-antichymotrypsin,⁷ thromboplastin,⁸ the complement protein C3,⁹ and apolipoprotein E¹⁰), and activated glial cells (astrocytes^11–13 and microglia¹⁴) overexpressing neurotrophic cytokines.^12,13

The cytokine concept developed from earlier work with leucocyte-derived polypeptide factors (lymphokines, monokines, etc) that orchestrate the multiple cellular interactions involved in immune responses. With the discovery that these regulatory molecules were not limited to cells of the immune system either in origin or in target spectrum, the more general term cytokine was introduced.¹⁵ Cytokines are secreted molecules, characteristically expressed constitutively in low amounts, which are promptly upregulated in response to various environmental stimuli (including increased concentrations of other cytokines). They may have effects on their cell of origin (autocrine effects), on adjacent cells of other types (paracrine effects), or systemically (endocrine effects).

There is increasing evidence that plaque-associated activated glia and glia-derived cytokines are important pathogenic factors in the progression of neuropathologic changes, and, by extension, in the progression of dementia in Alzheimer’s disease. This has led to the concept of a chronic, smoldering inflammatory process underlying the pathological progression of this disease. In this article we examine the roles of activated astrocytes and microglia, and of the cytokines they elaborate, in Alzheimer’s disease. We propose that these cytokines are key links in a complex of molecular and cellular interactions that, in the form of limited acute phase responses to various forms of central nervous system injury or insult, assist in repair and recovery from such injury. However, established functions of acute-phase-response cytokines suggest that chronic elevation (eg, in response to cumulative insults) may lead to the characteristic pathological features of Alzheimer’s disease. This concept suggests that Alzheimer’s disease has a multifactorial origin, and may be better conceived as a final common pathway of neurological decline rather than as a distinct disease entity with a single, identifiable origin.

MICROGLIA

Microglia are a resident population of brain cells of uncertain (but probably mesodermal) embryological origin.¹⁶ A variety of pathological insults result in microglial activation, a complex of morphological and biochemical alterations that include enlargement and proliferation, expression of novel cell-surface antigens (leukocyte common antigen, immunoglobulin Fc receptors, major histocompatibility complex [MHC] class I and class II glycoproteins, β2-integrins, and the vitronectin receptor), ^17,18 and release of various proteinases, cytokines (including interleukin-1 [IL-1], interleukin-6 [IL-6], and tumor necrosis factor-α [TNF-α]),^19,20 and reactive oxygen and nitrogen intermediates.²¹ Activated microglia are active in phagocytosis of necrotic material in a variety of pathological conditions and may be responsible for the alternative pathway cleavage of the β-amyloid precursor protein (β-APP) to yield amyloidogenic fragments in Alzheimer’s disease.²² Perineuronal microglia also may be involved in synaptic remodeling after neuronal injury.²³

Activated microglia are constant and prominent components of neuritic plaques in Alzheimer’s disease.²⁴ They also are found, although in lesser numbers, in the diffuse cerebral amyloid deposits that are thought to precede neuritic plaque formation in this disease²⁵ but are not found in significant numbers in the diffuse amyloid plaques of the cerebellum,²⁶ where progression to neuritic plaques is not observed.^27–30

ASTROCYTES

Astrocytes, like microglia, become activated in response to various pathological insults, and such astrocytic activation is prominent in Alzheimer’s disease and in many other neurodegenerative conditions. This activation is marked by hypertrophy³¹ and results in expression of structural proteins (eg, glial fibrillary acidic protein, [GFAP]), adhesion molecules; antigen presenting capabilities including MHC antigens (for reviews see Frohman et al³² and Malhotra et al³³) and cytokines, such as S100β.^12,34,35 Activated astrocytes respond to IL-1 by producing a variety of cytokines, including TNF-α,³⁶ IL-6, IL-8, colony stimulating factor-1 (CSF-1),³⁷ granulocyte-macrophage CSF, granulocyte-CSF³⁸ and S100β.³⁹ Interleukin-1–stimulated, activated astrocytes also produce the complement protein C3⁴⁰ and apolipoprotein E.⁴¹

In Alzheimer’s disease activated astrocytes form a halo around neuritic plaques, and extend processes into the cores of these plaques.^12,13,42 As is the case with activated microglia, these cells are found in fewer numbers in diffuse cerebral amyloid deposits (Mrak, Sheng, Griffith; unpublished data) and are not significantly increased in number in the cerebellum of Alzheimer patients.²⁶

INTERLEUKIN-1

Interleukin-1 is an acute phase cytokine that has numerous systemic effects.⁴³ The principal source of both the α and β isoforms are macrophages in the periphery and microglia in the central nervous system.^44–46 These two isoforms, encoded by genes on chromosome 2,⁴⁷ are synthesized as 33 kDa precursors that are cleaved to yield 17 kDa products. For the β isoform only the 17 kDa cleavage product is biologically active. For the α isoform both the 33 kDa and 17 kDa molecules are biologically active, but only the 17 kDa molecule is secreted (for recent reviews of IL-1 functions see Rothwell⁴⁸ and Dinarello and Wolffe⁴³).

Interleukin-1 also seems to function as an acute phase cytokine in the brain. For instance, IL-1 is expressed in microglia within hours following head injury in humans, and this expression is accompanied by dystrophic, swollen β-APP immunoreactive (β-APP⁺) neurites.⁴⁹ Interleukin-1 upregulates expression of β-APP^50–52 through stimulation of the β-APP promoter,⁵³ and it also stimulates β-APP processing.⁵⁴ Interleukin-1 has autocrine effects on microglia, with promotion of microglial proliferation in mixed neural cell cultures⁵⁵ and upregulation of microglial expression of IL-6, TNF-α, and IL-1 itself.^19,20 Interleukin-1 paracrine effects include activation of astrocytes^44,56 with upregulation of expression of astrocytic cytokines, including CSF-1,^57,58 IL-6, TNF-α,^20,36 and S100β³⁹ as well as expression of astrocyte-derived apolipoprotein E.⁴¹ In vitro, low levels of IL-1 promote survival of cultured neurons,⁵⁹ but higher levels are toxic.⁶⁰ Interleukin-1 also upregulates expression of tissue factor (thromboplastin) by vascular endothelial cells⁶¹ and of α₁-antichymotrypsin (an acute phase inflammatory protein⁶²) in cultured cells.⁶³ Both of these proteins, like β-amyloid¹ and apolipoprotein E,¹⁰ are components of neuritic plaque cores in Alzheimer’s disease.^7,8,64

Interleukin-1 is elevated in tissue homogenates of temporal lobe¹² and in cerebrospinal fluid⁶⁵ from patients with Alzheimer’s disease. The α isoform of IL-1 accounts for most of the excessive expression in temporal lobe.⁶⁶ In Alzheimer patients activated immunoreactive IL-1 microglia are increased sixfold over those present in age matched controls,¹² and these microglia show distinct patterns of distribution in these brains. Interleukin-1α immunoreactivity in various brain regions (hippocampus, cortex of the various cerebral lobes, and cerebellum) correlates both with generally recognized patterns of regional susceptibility in Alzheimer’s disease and with concomitant involvement of activated astrocytes in these regions.²⁶ Furthermore, IL-1α immunoreactivity is intimately associated with neuritic plaques in Alzheimer’s disease (Fig 1A and B).^25,66 This evidence suggests a seminal role for these cells and this cytokine in the progression of Alzheimer’s disease.

FIGURE 1.

Immunohistochemical demonstration of IL-1α⁺ activated microglia and S100β⁺ activated astrocytes in Alzheimer’s disease and Down’s syndrome, and following head trauma. (A) Alzheimer’s disease: activated IL-1α⁺ microglia (brown, arrowheads) associated with β-APP⁺ dystrophic neurites (red) in a neuritic plaque (*). (B) Alzheimer’s disease: activated IL-1α⁺ microglia (brown, arrowheads) associated with β-amyloid (β-AP) immunopositive deposits (red) in a neuritic plaque (*). (C) Alzheimer’s disease: activated S100β⁺ astrocytes (brown, arrows) associated with Tau 2⁺ dystrophic neurites (red) in a neuritic plaque (*). (D) Head trauma: activated IL-1α⁺ microglia (brown, arrowheads) associated with β-APP⁺ dystrophic neurites in a neuritic plaque-like structure in a patient surviving 48 hours after head trauma. (E) Down’s syndrome: activated S100β⁺ astrocytes (brown) in a trisomy 21 fetus at 23 weeks’ gestation; (F) Down’s syndrome: activated IL-1α⁺ microglia in a 2-year-old patient with Down’s syndrome. Bars represent 15 μm. Immunohistochemical technique¹²²: Paraffin blocks of formalin-fixed mesial temporal lobe tissue were sectioned at 10-μm thickness. Sections were deparaffinized in xylene, rehydrated in a graded series of ethanol solutions, and permeabilized in 0.05% Triton-X 100 for 10 minutes followed by 0.2 N HCl for 20 minutes. Endogenous peroxidase was blocked by incubation with 3% H₂O₂ in 97% methanol for 30 minutes. Sections for β-APP immunoreaction were pretreated with 99% formic acid for 5 minutes and washed with phosphate-buffered saline. Antibodies employed were polyclonal anti-IL-1α (Cistron, Pine Brook, NJ), diluted 1:20; polyclonal anti-β-amyloid (Boehringer-Mannheim Biochemica, Indianapolis, IN), diluted 1:10; monoclonal anti-β-amyloid (a gift from G.W. Roberts, SmithKline Beecham Pharmaceuticals, Essex, UK), diluted 1:1,000; monoclonal anti-β-APP (clone 22C11; Boehringer-Mannheim Biochemica), diluted 1:10; polyclonal antibovine S100β (a gift from L.J. Van Eldik, Northwestern University, Chicago, IL), diluted 1:300; and monoclonal Tau 2 antibody (Sigma Chemical Company, St Louis, MO), diluted 1:100. For dual immunoreaction sections were further processed according to the manufacturer’s protocol in DAKO’s (Glostrup, Denmark) double immunolabeling kit (K-665).

Further evidence of a role for IL-1 in Alzheimer’s disease is the differential distribution of activated IL-1α⁺ microglia among different plaque types.²⁵ This differential distribution suggests that activated IL-1α⁺ microglia are involved in the evolution of these plaques—from the clinically silent, diffuse plaque through the diagnostic neuritic plaque to the inert, end-stage dense core, nonneuritic plaque. Interleukin-1α⁺ microglia are found in small numbers in most cerebral cortical diffuse plaques in Alzheimer’s disease, suggesting an early involvement of these cytokine-expressing cells in plaque evolution.²⁵ This is in contrast to their absence in the diffuse plaques found in the cerebellum in Alzheimer’s disease,²⁶ where progression to neuritic forms is not observed.^27–30 Interleukin-1α⁺ microglia are larger and most abundant in neuritic plaques without dense β-amyloid cores, again suggesting an important role in plaque evolution. In contrast, fewer of these cells are found in “mature” neuritic plaques (those with a dense β-amyloid core), and no IL-1α⁺ microglia are found in “burnt-out” plaques (β-amyloid dense core plaques that do not have associated β-APP⁺ neurites),²⁵ suggesting a waning of microglia-associated inflammatory activity as the dystrophic neurites disappear and an end stage of plaque evolution is reached.

Patients with Down’s syndrome invariably develop striking Alzheimer’s-like neuropathological changes in middle age.⁶⁷ As Down’s syndrome patients (unlike Alzheimer’s patients) can be recognized at (or even before) birth, these patients provide a model for studying early (predementia) neuropathological changes in Alzheimer’s disease. In Down’s syndrome there is striking, early overexpression of IL-1¹² by activated microglia (Fig 1F), which increases progressively in fetal, young (0 to 9 years), and adult Down’s syndrome patients.⁶⁸ In addition to Down’s syndrome, head trauma recently has been recognized as a significant risk factor for the later development of Alzheimer’s disease.^69–76 Such acute cerebral trauma has been shown to elicit “acute phase” cellular and molecular responses: β-APP expression is increased within days of a single episode of head injury,⁷⁷ and this elevation is associated with elevated microglial IL-1α expression⁴⁹ (Fig 1D). These findings suggest that the mechanism underlying the elevated risk of Alzheimer’s disease observed in head injury patients may involve cytokine-mediated initiation of neurodegenerative processes, which, in concert with advancing age and additional risk factors, may become self-sustaining and terminate in Alzheimer’s disease.

S100β

S100 is most familiar to pathologists as an immunohistochemical marker for tumors of neuroectodermal origin. There are two 10 kDa isoforms, α and β, that are encoded by single genes on chromosomes 1 and 21,⁷⁸ respectively. In the brain the β isoform is most abundant and is synthesized by and released from astrocytes, whereas the α isoform is expressed in small quantities by neurons.^79,80 Activated (“reactive”) astrocytes in Alzheimer’s disease express greatly elevated levels of S100β,⁴² and both astrocyte activation⁵⁶ and increased expression of S100β³⁹ are induced in vivo by IL-1.

Several lines of evidence suggest an important intercellular regulatory (cytokine) function for S100β released by activated astrocytes. S100β increases cytoplasmic free calcium levels in neurons,⁸¹ stimulates neurite outgrowth in vitro,⁴² and promotes neuronal survival in vivo.⁸² During fetal neurological development S100β seems to be an important neurotrophic agent^82,83 with effects on neuroblasts and glia during this period.⁸³ S100β also has autocrine effects, including elevation of astrocyte intracellular free calcium levels⁸¹ and stimulation of astrocytic proliferation and hypertrophy.⁸⁴

Mapping of the S100β gene to the Down’s (q22) region of chromosome 21 suggests a signal pathogenic role for S100β in Down’s syndrome and, by analogy, in Alzheimer’s disease.⁸⁵ This suggestion is supported by the finding of elevated S100β expression (ie, in excess of the 1.5-fold increase expected from gene loading) in activated astrocytes in Down’s syndrome at all ages¹² and by the elevated levels of biologically active S100β and S100β mRNA in Alzheimer’s disease.⁴²

Biologically active (homodimeric) S100β levels⁴² and the numbers of activated S100β⁺ astrocytes¹² are dramatically increased in brains of patients with Alzheimer’s disease, and the distribution of this increase across different brain regions parallels the pattern of involvement of these regions in Alzheimer’s disease.⁸⁶ These observations, together with the spatial orientation of these activated cells to neuritic (Tau 2⁺) plaques¹³ (Fig 1C) and the ability of S100β to promote neurite extension,⁴² suggest an important participatory (and not merely reactive) role for these astrocytes and for this cytokine in the evolution of neuritic plaques in Alzheimer’s disease.

In Down’s syndrome overexpression of S100β is apparent as early as 18 to 19 weeks in utero,^12,68 an effect that may be interpreted as a direct genetic consequence of trisomy 21. This increase in S100β expression in Down’s syndrome fetuses (Fig 1E) is not accompanied by elevated expression of GFAP.⁶⁸ Like the early expression of IL-1 observed in Down’s syndrome, S100β expression increases progressively in fetal, young, and adult Down’s syndrome patients.⁶⁸ In contrast, deposition of extracellular β-amyloid and formation of neuritic plaques are not detectable before adolescence.^68,87 These findings strongly suggest that overexpression of IL-1 and S100β are important early events in the genesis of Alzheimer’s-like neuropathological changes in Down’s syndrome.

OTHER CYTOKINES

Interleukin-6 has been shown to be elevated in temporal lobe tissue from patients with Alzheimer’s disease,⁸⁸ and IL-6 immunoreactivity is demonstrable in and around some senile plaques in Alzheimer’s disease.^89,90 Unlike IL-1, IL-6 is not elevated in serum or cerebrospinal fluid from patients with Alzheimer’s disease.^89,91 The activated microglia in Alzheimer’s brain express interferon-α.⁹² Interleukin-2 immunoreactivity is profuse in brain tissue from both control and Alzheimer patients.⁹³ Serum TNF-α levels have been reported to be both elevated⁹⁴ and depressed⁹⁵ in Alzheimer’s disease.

CYTOKINE-β-AMYLOID RELATIONSHIPS

The involvement of β-amyloid and its precursor, β-APP, in Alzheimer’s disease is, of course, well established.¹ β-Amyloid is an important and prominent constituent of the characteristic plaques of Alzheimer’s disease, and β-APP was first identified in the course of a search for the origin of these amyloid deposits.⁹⁶ Dystrophic neurites within plaques react intensely with anti-βAPP antibodies^25,49 (Fig 1A). These structures are intimately associated with activated IL-1α⁺ microglia²⁵ and with classical complement pathway proteins,^9,97 suggesting a cytolytic attack by this system against neuronal membranes. Upregulation of a membrane inhibitor of reactive lysis (that protects host cells against complement-mediated lysis) in dystrophic neurites and neurofibrillary tangle-containing neurons in Alzheimer’s disease has been reported.⁹⁸

The β-amyloid precursor proteins are a closely related group of integral membrane proteins.^96,99 These proteins are synthesized in cell somas and undergo fast anterograde axonal transport.¹⁰⁰ They are present at synapses,^96,101 interact with second messenger systems,¹⁰² and appear to be important in memory formation.¹⁰³ Like S100β, these proteins are encoded by a gene mapped to chromosome 21 but, unlike S100β,⁸⁵ this gene lies outside the Down’s syndrome region.^104,105 The major degradation pathway for β-APP does not yield amyloidogenic fragments.⁹⁹ However, cleavage of these proteins within their hydrophobic intramembranous domains (normally a minor degradation pathway) yields an amyloidogenic peptide that accumulates in both the diffuse and neuritic plaques of Alzheimer’s disease.^1,106 β-Amyloid precursor protein is not a cytokine but does enhance survival of cultured neurons¹⁰⁷ and regulate the effects of nerve growth factor on neurite outgrowth¹⁰⁸ and cell adhesion.¹⁰⁹ Furthermore, two of the three β-APP isoforms (751 kDa and 770 kDa) contain a Kunitz type protease inhibitor^110,111 and have regulatory effects on neuronal growth.¹⁰² Neuronal β-APP expression also has been associated with neurofibrillary tangle formation in Alzheimer’s disease.¹¹²

β-Amyloid, the amyloidogenic cleavage product of β-APP, was once thought to be inert, but now seems to participate intimately in the neurodegenerative events of Alzheimer’s disease. β-Amyloid stimulates microglial activation and microglial expression of IL-1,¹¹³ and directly activates (ie, without the intervention of immunoglobulins) the classical, but not the alternative, complement pathway leading to cell lysis.⁹⁷ Purified β-amyloid also has been shown to form calcium channels in artificial phosphatidylserine bilayers¹¹⁴; in this way β-amyloid might act synergistically with S100β to effect neuronal cell injury and death through increases in intracellular calcium concentrations.

PATHOGENIC IMPLICATIONS

It is evident from the foregoing discussion that activated glia and their cytokines, β-amyloid and its precursor proteins, and the complement proteins are components of a complex system of cellular and molecular interactions in Alzheimer’s disease. The potential positive feedback and autocrine activities within such a system suggest that these interactions might become chronically self-propagating. Such a self-propagating system, shown diagrammatically in Fig 2, would be capable of explaining many of the characteristic neuropathological features of Alzheimer’s disease, including formation of neuritic plaques containing activated astrocytes, activated microglia, β-APP^{+ 115} and Tau 2^{+ 116} overgrown dystrophic neurites as well as extracellular deposits containing β-amyloid, α₁antichymotrypsin, thromboplastin, the complement protein C3, and apolipoprotein E.

FIGURE 2.

Diagram of proposed pathogenic interactions between glial cytokines and molecular and cellular events in Alzheimer’s disease. Micro, microglial; Astro, astroglial; PN-1, protease nexin 1; α₁-ACT, α₁ antichymotrypsin; (Ca)i, intracellular free calcium concentration; P-tau, excessively phosphorylated tau; PHF, paired helical filaments in neurofibrillary tangles; PolyA⁺ mRNA, polyadenylated mRNA.

This self-propagating pathogenic process might result from any of several events capable of initiating chronic elevation of IL-1 or from the cumulative effects of several such events. Interleukin-1, in turn, would stimulate neuronal and endothelial cell synthesis and the processing of β-APP; upregulate expression of plaque components, such as α₁-antichymotrypsin, thromboplastin, the complement protein C3, and apolipoprotein E; and activate microglia and astrocytes with induction of S100β expression.

S100β, in turn, would stimulate neurite growth, promote astrocyte activation, and increase intracellular free calcium levels in neurons and astrocytes. Chronic elevation of intracellular free calcium would favor calcium-mediated events, such as excessive phosphorylation, possibly including abnormal phosphorylation of the tau present in the neurofibrillary tangles¹¹⁷ in neurons and neurites, and could ultimately result in the neuronal cell death characteristic of Alzheimer’s disease.

The combined effects of β-APP upregulation and processing would favor increased β-amyloidogenic cleavage products. This, together with increased production of α₁-antichymotrypsin, thromboplastin, and apolipoprotein E, could result in extracellular deposits (plaques) containing these proteins. The β-amyloid in these plaques, in turn, could have effects that would feed back to propagate the system. These include (1) direct activation of microglia with microglial IL-1 production and consequent astrocyte activation with increased S100β expression, and (2) activation of the classical complement pathway that, in turn, leads to microglial activation with IL-1 production.

Neuronal cell injury and cell death might result from the combined effects of increased neuronal intra-cytoplasmic calcium levels (attributable to S100β-mediated increases in neuronal free calcium levels and possibly to β-amyloid-mediated formation of neuronal cell membrane calcium channels), accumulation of neurofibrillary tangles (caused by calcium-mediated phosphorylation of tau), and complement-mediated cell damage (resulting from β-amyloid binding to and activation of complement). Such cellular injury would, in turn, activate microglia and provide yet another feedback mechanism for propagation of the cytokine cycle.

CONCLUSIONS

Central nervous system injury provokes a limited acute phase cellular and molecular response, including elaboration of the glial cytokines IL-1 and S100β, which is important in healing and repair. However, chronic or repeated stimulation of this response may produce a self-sustaining cycle that gives rise to the characteristic neuropathological changes of Alzheimer’s disease. Thus, Alzheimer’s disease may not be a distinct disease entity but rather a syndrome with multiple origins, all leading to chronic self-propagation of a vicious circle of glial acute phase responses, formation of plaques, neuronal cell death, and synaptic loss. As such, it may be impossible to identify a single initiating molecular or cellular event as the cause of Alzheimer’s disease. Recognized risk factors for Alzheimer’s disease include age,¹¹⁸ trisomy 21,⁶⁷ certain mutations in the β-APP gene,¹¹⁹ the type 4 allele of apolipoprotein E,¹²⁰ and a history of head trauma.^69–76 However, regardless of initiating or promoting factors, chronic elevation of IL-1 would seem to be a key factor in the pathogenic progression observed in Alzheimer’s disease. This hypothesis, together with our finding of elevated levels of IL-1 in Alzheimer’s disease, has led to the suggestion of a novel treatment strategy based on pharmacological intervention in IL-1 production or actions.¹²¹

Abbreviations

IL

interleukin

APP

amyloid precursor protein

MHC

major histocompatibility complex

TNF

tumor necrosis factor

GFAP

glial fibrillary acidic protein

CSF

colony stimulating factor