Neurodegenerative disorders (NDs) are a heterogeneous group of pathologies that is characterized by the progressive degeneration of the structure and function of the central or peripheral nervous system1,2,3. Severity and propensity to progression not only vary across different NDs, but also across patients presenting with the same pathology, which considerably complicates the effectiveness of therapeutic interventions4. In fact, no disease-modifying therapeutic strategies are currently available for most NDs5,6,7, perhaps with the exception of lecanemab, a controversial agent recently licensed by the US Food and Drug Administration (FDA) for the therapy of early Alzheimer’s disease (AD)8.

Common NDs affecting the central nervous system (CNS) including but not limited to AD, Parkinson’s disease (PD), and Huntington’s disease (HD) are generally attributed to intracellular defects in specific neuronal populations that over time compromise cellular homeostasis (often along with the progressive accumulation of specific proteins within neurons or in their microenvironment), ultimately culminating with widespread or localized neuronal loss coupled with clinical cognitive, behavioral or motor symptoms1,3,9. In line with this notion, while juvenile variants of central NDs exist10, the prevalence and severity of most NDs increases with age11. Moreover, a number of genetic mutations affecting intracellular processes have been associated with an increased risk for central NDs12,13. For instance, mutations in amyloid beta precursor protein (APP), which encodes the precursor of the AD pathognomonic amyloid beta (Aβ) peptide, is associated with an increased risk for AD1. Similarly, most cases of HD are associated with mutations in huntingtin (HTT), which results in the generation of an altered HTT variant with pathogenic features9. That said, some NDs are not necessarily associated with genetic predisposition14,15, suggesting that additional factors may be involved in their pathogenesis.

A number of non-genetic factors have been proposed as contributors to the pathogenesis of NDs, including (but not limited to): (1) the abundance and composition of the gut microbiome16,17, (2) a history of infection with viral or bacterial pathogens18, (3) the existence of subjacent or overt cardiovascular and metabolic disorders19, as well as (4) dysfunctions of the innate or adaptive immune system20,21 (Box 1). Recent data from other biomedical disciplines point indeed to an underappreciated contribution of dysregulated immunity to the pathogenesis of disorders that have classically been attributed to cell-intrinsic mechanisms, including various cardiovascular conditions22,23 and cancer24. Moreover, microglial cells, which play a major role in the development of multiple NDs (see below), are brain-resident cells of hematopoietic origin with macrophage-like functions, which render them preferential interactors of other potentially pathogenic immune cells25.

Here, we discuss immunological mechanisms that promote central neurodegeneration in the context of human NDs and propose potential strategies to correct neurodegeneration-associated immunological dysfunctions with therapeutic purposes. Conversely, immunological aspects of peripheric NDs including multiple sclerosis and amyotrophic lateral sclerosis have been extensively covered elsewhere26,27, and hence will not be covered in this review.

Alzheimer’s disease

AD, which is the most common cause of dementia amongst the elderly, is characterized by a progressive loss of cognitive functions and memory associated with the accumulation of extracellular Aβ plaques and intracellular microtubule associated protein tau (MAPT, best known as tau) neurofibrillary tangles1.

AD has been consistently linked to genetic alterations in apolipoprotein E (APOE), encoding a protein involved in lipid metabolism28, as well as in a number of genes involved in innate immunity and microglia activation, notably triggering receptor expressed on myeloid cells 2 (TREM2), encoding a plasma membrane receptor that promotes phagocytosis29,30,31; complement C3b/C4b receptor 1 (CR1), encoding a complement component32; CD3333,34; major histocompatibility complex, class II, DR beta 1 (HLA-DRB1), HLA-DRB5 and HLA-DR15, all encoding MHC molecules35,36; protein tyrosine kinase 2 beta (PTK2B), inositol polyphosphate-5-phosphatase D (INPP5D), and phospholipase C gamma 2 (PLCG2), all encoding intracellular signal transducers31,35; and ABI family member 3 (ABI3), encoding an adaptor protein31, pointing to microglia-driven neuroinflammation as a major pathogenic determinant of this ND.

Abundant preclinical data suggest that microglial activation may have a context- and disease stage-dependent effect on the progression of AD. On the one hand, inhibition of TREM2 by genetic or pharmacological strategies has been associated with limited microglial activity and neuroprotection in various mouse models of AD and tau pathology, including mice expressing five protein variants associated with familial AD (i.e., 5XFAD mice)37,38,39, mice expressing a human pathogenic variant of tau (so-called P301S mice)40,41, P301S mice expressing human AD-linked variant of APOE (i.e., APOE4)42, as well as 5XFAD mice intracerebrally administered with sarkosyl-insoluble tau aggregates isolated from the frontal cortex of human AD brain tissue43,44,45. However, TREM2 overexpression as well as increased TREM2 activation upon inhibition of membrane shedding have also been linked with improved biochemical and cognitive manifestations of AD in rodent models of the disease46,47. Thus, TREM2 appears to influence AD progression in a complex manner that may involve an initial beneficial impact related to the degradation of amyloid plaques and a subsequent detrimental impact linked to neuroinflammation.

To add yet another layer of complexity, in at least some AD models, while Trem2 haploinsufficiency appears to aggravate tau pathology in mice, complete Trem2 loss reportedly limits tau-driven microglial activation and atrophy48. Whether these apparently contradictory findings may relate to the differential activation of inflammatory responses in cells other than the microglia, such as oligodendrocytes49, remains to be further investigated. The existence of different microglial clusters as documented by modern single-cell sequencing technologies may also explain, at least in part, the apparently context-dependent impact of TREM2 on AD progression in mice. For instance, a novel type of neurodegenerative disease-associated microglia (DAM) has been shown to mediate neuroprotective effects in 5XFAD mice via a mechanism that at least initially involves TREM2 signalling50. Similar results have been obtained in mice expressing pathogenic APP variants (namely, AppNL-G-F mice), a scenario in which APOE expression by a DAM-like microglia was positively associated with an improved clearance of Aβ plaques51. In line with this notion, APOE has been shown to promote AD progression in mice bearing pathogenic APP and presenilin 1 (PSEN1, best known as PS1) variants (namely, APP-PS1 mice)52 and P301S mice53, a neurodegenerative mechanism mapping to the subset of microglia that exhibit a common disease-associated phenotype in mice and humans52,54.

Importantly, at least part of these APOE-associated mechanisms leading to neurodegeneration have been mapped to the inability of AD-linked APOE variants to preserve homeostatic (tolerogenic) microglial functions, culminating with the expression of multiple pro-inflammatory transcription factors52. In line with this possibility, APOE expression in astrocytes favors the polarization of microglia towards a DAM state in P301S mice54. Moreover, neuronal expression of APOE4 drives a specific DAM subset with potent neurodegenerative effects in an APOE4-expressing tauopathy mouse model, an effect that can be circumvented via neuron-specific ApoE deletion55. That said, human AD-associated microglia (HAM) as characterized by RNA sequencing from frozen samples of frontal cortex from AD-affected individuals appears to exhibit little transcriptional resemblance with the DAM as identified in mouse AD models56, with the notable exception of a common APOE overexpression57. Moreover, the HAM signature appears to be detectable also in patients with non-AD NDs, which is not the case for the DAM transcriptional profile56.

Thus, while initial microglial engagement has beneficial effects on AD progression at least partly emerging from the clearance of Aβ deposits, the overactivation of microglia has consistently been attributed a neurotoxic activity related to accrued oxidative stress and inflammation. Interestingly, such overactivation has been reported to culminate with microglial dystrophy58,59,60, a phenotype that appears to be common to multiple NDs beyond AD61,62,63. Additional work is required to identify strategies to finely modulate microglial functions for the treatment of AD.

Importantly, pro-inflammatory molecules secreted by the activated microglia including interleukin 1A (IL1A), tumor necrosis factor (TNF) and complement C1q A chain (C1QA) have been shown to promote astrocyte activation, resulting not only in a loss of phagocytic and synapse-promoting activity, but also in the secretion of neurotoxins that promotes neuronal and oligodendrocyte death64. In line with a pathogenic role for astrocytes in AD, astrocytic tau accumulation in the dentate gyrus has been shown to promote neuronal dysfunction and memory deficits in mice65. Moreover, reactive astrocytes have been identified in early stages of human AD and appear to be present ubiquitously throughout disease progression66. Moreover, while mildly reactive astrocytes have limited neurotoxic potential, severe astrocyte activation has been linked to neurotoxic and ultimately pathogenic hydrogen peroxide production via monoamine oxidase B (MAOB), at least in mice67. That said, astrocytes resemble microglia in being highly heterogenous and exhibiting an age-dependent decline in neuroprotective activity68,69. In line with this notion, a disease-associated astrocyte (DAA) transcriptional profile as identified in multiple rodent models of AD appears to emerge early during disease progression and exacerbate over time, a pathogenic progression also observed in aged wild-type mice and humans70.

While transcriptionally different from their murine counterparts64, DAAs from individuals with AD have been subclassified into 8 different clusters, one of which exhibits signatures of immune signaling including transforming growth factor beta 1 (TGFB1) activation57. Of note, human DAAs overexpress APOE as well as glypican 4 (GPC4), a secreted factor that has been detected in post-mortem brains from patients with APOE4-associated AD70,71 and may contribute to disease progression upon direct interaction with APOE4 and consequent tau hyperphosphorylation57. Supporting the pathogenic role of this mechanism, Apoe4 deletion from astrocytes limits disease progression in P301S mice54. Thus, pro-inflammatory signaling elicited by microglia may promote AD progression also via astrocytes.

Yet another non-neuronal cell type involved in AD are so-called disease-associated oligodendrocytes (DOLs). A study integrating multiple datasets from mouse models of NDs and post-mortem data from ND patients identified three different oligodendrocyte activation states: disease-associated 1 (DA1), DA2 and interferon (IFN)-associated72. Intriguingly, not only IFN-associated DOLs but also DA1 DOLs exhibit upregulation of multiple genes involved in innate and adaptive immunity, including multiple cytokine- and complement-encoding genes as well as genes coding for MHC Class I and Class II molecules72. At least some of these DOL signatures are not restricted to AD, but can also be documented in other NDs and neuroinflammatory disorders, suggesting a common response to severe pathological conditions73. Of note, DOLs have TREM2-independent transcriptional responses to neurodegenerative conditions that resembles those of DAAs49, including the overexpression of stress-responsive proteins such as serpin family A member 3 (SERPINA3)73. That said the transcriptional profile of DOLs from individuals with AD considerably differs from that of mouse AD models72.

Importantly, not only brain-resident cells engaging in innate immune signaling, but also newly recruited immune cells have been shown to contribute to AD pathogenesis. For instance, neutrophil infiltration has been documented in individuals with various NDs including AD, a finding that has been mechanistically linked with reduced cerebral blood flow in APP-PS1 mice, 5XFAD mice, as well as mice expressing pathogenic variants of APP, PS1 and tau (namely, 3xTg mice). In these models, blood flow and short-term memory function rapidly improve when cerebral perfusion is restored by preventing neutrophil adhesion74,75. Along similar lines, several studies have demonstrated infiltration of the brain parenchyma by CD4+ and CD8+ T cells (which orchestrate and execute antigen-specific immune responses, respectively) in patients with AD76,77 and in animal models of the disease49,78. In this setting, extravascular cytotoxic CD8+ T cell abundance appears to correlate with disease stage, indicating a potential role for adaptive immunity in the pathogenesis of AD79,80,81. Similar findings have been obtained in APP-PS1 and 5XFAD mice81. Importantly, in this latter setting CD8+ T cell infiltration of the brain parenchyma could be mechanistically linked to the microglia, and not only CD8+ T cell depletion but also interference with CD8+ functions by interferon gamma (IFNG) or programmed cell death 1 (PDCD1, best known as PD-1) blockage mediated considerable neuroprotective effects81, potentially linked to the reversal of CD8+ T cell exhaustion82. Further supporting a link between CD8+ T cell activity and the pathogenesis of AD, peripheral blood mononuclear cells (PBMCs) from patients with AD are enriched in CD8+CD45RA+ T effector memory (TEMRA) cells displaying transcriptional signatures of activation, and their abundance negatively correlated with cognition83.

Conversely, the actual impact of helper CD4+ T cells on AD progression remains to be formally established. Indeed, administration of Aβ-specific type 1 (TH1) and type 17 (TH17) CD4+ T cells (two specialized populations of CD4+ T cells characterized by specific secretory profiles) reportedly exacerbates memory impairment and amyloid deposition in APP-PS1 mice84. Moreover, CD4+ T cell depletion appears to improve Aβ clearance and cognitive performance in 5XFAD mice85. However, CD4+ T cell depletion has also been associated with accelerated cognitive decline with no impact on amyloid pathology in APP-PS1 mice86. At least theoretically, these apparently contrasting findings may relate to the considerable phenotypic and functional diversity of CD4+ T cells, encompassing a highly reactive compartment as well as immunosuppressive subpopulations such as CD4+CD25+FOXP3+ regulatory T (TREG) cells87,88. Indeed, TREG cells have consistently been shown to limit AD progression in mouse models of the disease89,90,91.

Interestingly, multiple microglia–T cell interactions have been documented in patients with AD and/or mouse models thereof, including (but not limited to) TCR responses to MHC Class II-restricted peptides92. That said, single-cell analyses of T cells infiltrating the brain of mice affected by amyloid, tau or combined (amyloid and tau) pathology suggest that T cell reactivity is not influenced by TREM2 expression in the DAM49. Taken together, these observations suggest that while the DAM may drive pathogenic T cell responses during AD progression, such responses (at least initially) may not impinge on the APOE-TREM2 signalling axis. Recently, B cells (a population of lymphoid cells specialized in antigen presentation and antibody production) have also been mechanistically implicated in the pathogenesis of AD93. Specifically, 3×Tg mice have been shown to exhibit not only an expanded B cell comportment in the periphery, but also accrued B cell accumulation in the brain parenchyma associated with immunoglobulin deposition at amyloid plaques93. In the same setting, B cell depletion at disease onset appears to reduce amyloid accumulation, limit hippocampal microglial activation and overall decelerate disease progression93.

In summary, the pathogenesis of AD involves a complex neuroinflammatory reaction involving brain-resident cells as well as newly recruited immune cells that ultimately promote neurotoxicity coupled to cognitive disorders (Fig. 1).

Fig. 1: Immunological aspects of Alzheimer’s disease.
figure 1

Alzheimer’s disease (AD) develops in the context of complex immunological alterations that involve not only microglial cells, with a major role for altered apolipoprotein E (APOE) and triggering receptor expressed on myeloid cells 2 (TREM2) signalling, but also astrocytes and oligodendrocytes, culminating with a neuroinflammatory state associated with immune cell infiltration from the periphery. Aβ amyloid beta, BBB blood-brain barrier, C1QA complement C1q A chain, DA disease-associated, IFN interferon, IL1A interleukin 1A, SERPINA3 serpin family A member 3, tau (microtubule-associated protein tau, MAPT), TGFB1 transforming growth factor beta 1, TNF tumor necrosis factor. Created with

Parkinson’s disease

PD is the most prevalent ND that results in disordered movement, affecting 6–7 million individuals worldwide3. While most PD cases are idiopathic, a familial form of the disease has been associated with mutations in > 20 genes, including parkin RBR E3 ubiquitin protein ligase (PRKN), PTEN-induced kinase 1 (PINK1), both of which encode component of the molecular apparatus that removes dysfunctional mitochondria (so-called mitophagy)94, leucine-rich repeat kinase 2 (LRRK2), encoding a multifunctional kinase, and synuclein alpha (SNCA)95. The main pathological feature of PD is the degeneration of dopaminergic neurons in the substantia nigra (SN) involving the intraneuronal accumulation of SNCA aggregates called Lewy bodies3. Of note, SNCA aggregates have also been documented in the gastrointestinal tract of PD patients up to 20 years prior to their diagnosis96, and the administration of preformed SNCA fibrils into the duodenal and pyloric muscularis layer promotes PD development in mice97. These observations point to the existence of a gut-to-brain axis that contributes to the spread of pathogenic Lewy bodies to the central nervous system.

While neuronal dysfunction coupled with oxidative stress has a major role in the pathogenesis of PD, variations in numerous genes encoding key components of the innate and adaptive immune system have been associated with an increased risk for PD, including an MHC Class II haplotype that is displayed by ~15% of the population (namely, HLA-DRB1)98. Moreover, multiple genetic loci associated with an increased risk for PD appear to also predispose to some autoimmune and inflammatory diseases, such as Crohn’s disease99. Of note, PD patients often exhibit elevations in the circulating or cerebrospinal levels of cytokines such as TNF, interleukin 1 beta (IL1B), IL2 and IL10100,101. In line with this notion, microglial activation is a characteristic finding in the SN of post-mortem brains from patients with PD feature of the substantia nigra in post-mortem human brains with PD77,102, and the active microglia has been shown to actively engage in pro-inflammatory signalling via various molecular platform including (but not limited to) the NF-κB, inflammasome, JAK/STAT and Toll-like receptor (TLR) signaling103,104,105, at least in some settings as a direct consequence of tau accumulation106. Importantly, multiple pharmacological strategies aimed at interrupting these signal transduction cascades have been shown to decelerate PD progression in animal models, including rats intracerebrally administered with an SNCA-encoding adenovirus or treated with the PD-inducer rotenone103,104,105. Together with the fact that significant microgliosis has been documented in areas not showing significant neuronal death in post-mortem brains from patients with PD107,108,109 and with kinetic data from rodent models of PD110, these findings suggest that inflammatory microglial activation precedes and promotes the demise of dopaminergic neurons that characterize PD. At least in part, such a neurotoxic response involves the microglia-driven conversion of astrocytes to a pathogenic state, as mechanistically demonstrated with a small molecule that prevents this conversion (i.e., NLY01) in mice expressing a pathogenic variant of SNCA or administered intracerebrally with preformed SNCA fibrils111.

Further supporting a link between inflammation and the pathogenesis of PD, SNCA has been shown to promote microglial activation in an MHC Class II-dependent manner, culminating with the initiation of a pathogenic CD4+ T cell response112. In line with this notion, SNCA overexpression in the mouse midbrain results in the upregulation of MHC Class II molecules on myeloid cells coupled with abundant infiltration of IFNG-producing CD4+ and CD8+ T cells113, a process that at least in part involves so-called border associated macrophages (BAMs)114. However, while the absence of CD3+ T cells or CD4+ T cells reportedly decelerates PD progression and ameliorate behavioral symptoms in mice treated with the PD driver 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), the same does not hold true for the selective absence of CD8+ T cells102. That said, CD8+ T cell infiltration has been documented in the SN from pre-symptomatic PD patients, de facto preceding dopaminergic neuron loss and SNCA pathology115. Moreover, patients with overt PD exhibit: (1) a reduction in the levels of circulating naïve T cells and TREG cells116, (2) an increase in the ratio of IFNG- over IL4-producing CD4+ T cells in the periphery117, and (3) circulating T cells responding to SNCA-derived peptides118, an autoreactivity that appears to develop even prior to clinical manifestations of the disease119. Finally, results from Snca−/− mice indicate that SNCA is also required for the development of normal inflammatory and antigen-specific responses to intraperitoneal bacteria120, further strengthening the links between PD and immunity.

In summary, PD appears to involve a variety of innate and adaptive immune processes that culminate with the loss of dopaminergic neurons in the SN and the consequent motor symptoms (Fig. 2).

Fig. 2: Immunological aspects of Parkinson’s disease.
figure 2

Patients with Parkinson’s disease (PD) exhibit synuclein alpha (SNCA)-related microglial activation coupled with the initiation of multiple signalling pathways that result in the abundant secretion of pro-inflammatory cytokines in the cerebrospinal fluid (CSF). Such cytokines promote the recruitment of immune cells that contribute to neuroinflammation by secreting proinflammatory mediators such as interferon gamma (IFNG). BBB blood brain barrier, IL interleukin, TNF tumor necrosis factor, TREG regulatory T. Created with

Huntington’s disease

HD is an autosomal dominant neurological disorder caused by an aberrant expansion of CAG triplets in exon 1 of HTT, resulting in a long polyglutamine tract that disrupts HTT cellular functions and causes widespread neurotoxicity starting from the neostriatum3. Clinical HD manifestations include progressive cognitive, motor and behavioral impairments3.

Similar to AD and PD, HD is also characterized by microgliosis, both in humans121,122,123 and in mouse models of the disease such as R6/2 mice124,125,126. Moreover, HD resembles AD and PD in that inflammatory microglial activation constitutes an early event in the pathogenesis of disease as it has been documented in post-mortem brain samples from patients with pre-symptomatic HD126,127. Of note, HD-associated microgliosis has also been linked with alterations in circulating myeloid cells and cytokines, and at least some of these changes could be detected in mutant HTT carriers prior to clinical manifestations of the disease128,129.

Several lines of evidence implicate HTT-driven inflammatory microglial activation in the pathogenesis of HD. First, microglial depletion using an inhibitor of colony-stimulating factor 1 receptor (CSF1R) reduces mutant HTT accumulation and prevents striatal atrophy in R6/2 mice130. Second, the establishment of mouse chimeras incorporating human microglial cells expressing mutant HTT promotes motor impairment and neuronal dysfunction in the striatum, while the contrary is true when the human microglia expresses wild-type HTT131,132. Third, mutant HTT expression in microglial cells is sufficient to drive microgliosis and elicit neurodegeneration133.

HD-associated HTT mutations also affect astrocytic and oligodendrocytic cells. For instance, mutant HTT expression in microglia has been shown to deregulate the expression of multiple cell lineage-specific genes in astrocytes from R6/2 mice and zQ175 mice (another mouse model of HD)134. Moreover, data from multiple HD rodent models as well as from post-mortem HD brain tissues suggest that the accumulation of mutant HTT in the nucleus is way more frequent in astrocytes and oligodendrocytes than in the microglia135,136. Corroborating the pathogenic effect of mutant HTT accumulation in cells other than the microglia, selective inactivation of mutant HTT in NG2+ oligodendrocyte progenitors prevents myelin abnormalities and certain behavioral deficits in HD mice137. Moreover, the cell-specific downregulation of mutant HTT from astrocytes or neurons coupled with modern sequencing technologies reveal that astrocyte dysfunction has a smaller impact on the neuronal transcriptome than neuron dysfunction has on the astrocytic one138. In line with this notion, at least three different transcriptional clusters of disease-associated astrocytes have been documented in post-mortem cingulate cortex samples from patients with HD, including a cluster with prominent stress-responsive and reactive transcriptional profile139. Transcriptional data from two distinct mouse models of HD and post-mortem HD brains also delineate the existence of shared transcriptional alterations linked to astrocytic dysfunction that can be corrected by limiting mutant HTT expression140. Taken together, these observations highlight the prominent role of inflammatory microglial activation as a driver of cellular dysfunction in the context of HD.

Of note, mutant HTT levels in circulating leukocytes have been shown not only to positively correlate with disease burden in patients with HD, but also to elicit immunological dysfunction coupled with increased TNF and IL8 secretion downstream of altered NF-κB functions141. In this setting, HTT silencing by RNA interference was sufficient to reverse secretory and transcriptional alterations141. Whether such an intervention would modify disease course in mouse models of HD, though, remains to be further investigated. That said, signs of both central and peripheral immune activation have been documented for both dendritic cells (DCs)142 and macrophages143 in R6/2 mice and zQ175 mice. Moreover, mutant HTT carriers exhibit a cerebrospinal T cell compartment characterized by increased IL17 expression coupled with the acquisition of a TH17 polarization and elevated IL7 consumption prior to symptom onset, and the abundance of cerebrospinal TH17 cells negatively correlates with disease progression144. These latter observations suggest that T cells may be involved in the early pathogenesis of HD and hence that limiting T cell responses may delay the onset of symptomatic HD.

Thus, similar to AD and PD, HD appears to develop in the context of complex innate and adaptive immune responses that manifest both centrally and in the periphery (Fig. 3).

Fig. 3: Immunological aspects of Huntington’s disease.
figure 3

Huntingtin (HTT) defects as caused by the mutations that drive Huntington’s disease (HD) foster a robust neuroinflammatory process involving microglial cells as well as astrocytes and oligodendrocytes that is often associated with a TH17-polarized CD4+ T cell response dominated by the secretion of pro-inflammatory cytokines like interleukin 8 (IL8), IL17A and tumor necrosis factor (TNF). BBB blood brain barrier, mHTT mutant huntingtin. Created with


Dementia with Lewy bodies

Dementia with Lewy bodies (DLB) is a prevalent form of dementia in the elderly that is defined by cognitive impairment coupled with visual hallucinations, parkinsonism, sleep behavior disorders, as well as autonomic and psychiatric dysfunction145. Of note, both AD and PD not only share familiar risk factors with DLB, notably genetic variants of APOE and SNCA145, but also exhibit similar clinical manifestations, making differential diagnosis problematic146,147. Moreover, at least a fraction of DLB cases share with AD the deposit of Aβ plaques as well as tau hyperphosphorylation, and with PD the accumulation of SNCA aggregates (Lewy bodies), although this does not often involve the SN, but instead affects basal ganglia and the cerebral cortex145. Finally, DLB has been associated with a history of traumatic brain injury (TBI), at least potentially linked to neuroinflammation148,149. However, the pathogenesis of DLB remains poorly understood, partly due to a lack of precise cellular and animal models of the disease.

Analysis of post-mortem brains from patients with DLB revealed diffuse cerebral inflammation and an increased number of microglial cells in the proximity of Lewy bodies150, as well as increased levels of proinflammatory cytokines such as IL6 coupled to the downregulation of neurotrophic factors151. Similar to the case of AD, microgliosis as associated with DLB appears to occur early during disease pathogenesis152 and to decline over time as cognitive impairment emerges153. Moreover, DLB has been associated with an increase in the circulating levels of several proinflammatory cytokines including IL2, IL17A and C-C motif chemokine ligand 20 (CCL20) along with decreased IL8 concentrations153, pointing to an involvement of both innate and adaptive immune mechanisms in the pathogenesis of the disease.

Further supporting a pathogenic role for immune effectors in the progression of DLB, CD4+ (but not CD8+) T cells have been shown to infiltrate the brain parenchyma of patients with DLB and SNCA-expressing mice, a rodent model of the disease that also manifests intracranial accumulation of natural killer T (NKT) cells154, a small lymphoid cell population with potent reactive traits155,156. Intriguingly, DLB-associated CD4+ T cell brain infiltration appears to be maximal in the proximity of blood vessels154, and these cells appear to acquire a TH17 phenotype coupled with pathogenic IL17A secretion in the cerebrospinal fluid downstream of C-X-C motif chemokine ligand 12 (CXCL12)-driven, C-X-C motif chemokine receptor 4 (CXCR4)-mediated recruitment157, a potent chemotactic signalling axis158.

These findings suggest that, while the precise etiology of DLB remains to be clarified, adaptive immune effectors including IL17A-secreting CD4+ T cells may contribute to DLB establishment and progression.

Frontotemporal dementia

Frontotemporal dementia (FTD) is an early-onset neurodegenerative disorder driven by a progressive atrophy of the frontal and temporal lobes and characterized by alterations in behavior, impulse control, personality, and language159. Pathologically, FTD is characterized by abnormal accumulations of TAR DNA binding protein (TARDBP), hyperphosphorylated tau or FET proteins, which encompass EWS RNA binding protein 1 (EWSR1), TATA-box binding protein associated factor 15 (TAF15) and FUS RNA binding protein (FUS)159. Mutations in C9orf72-SMCR8 complex subunit (C9Orf72), MAPT, TREM2 and granulin precursor (GRN) have been associated with familial variants of the disease160,161,162,163, but sporadic FTD accounts for > 60% of FTD cases164. Interestingly, FTD-related mutations have also been associated with an increased risk for autoimmune conditions165,166, pointing to a potential role for neuroinflammation in the pathogenesis of this ND.

In line with this possibility, FTD has been consistently associated with microgliosis and neuroinflammation in patients167,168,169. Moreover, individuals affected by FTD exhibit increased cerebrospinal levels of multiple pro-inflammatory cytokines including (but not limited to) IL2, IL12, IL17A, TNF, TGFB1 and CXCL1, at least in some patient cohorts positively correlated with disease severity170,171,172. Interestingly, cerebrospinal alterations in cytokine levels appear to vary in sporadic vs familial FTD cases173, but the causes underlying the observations remain to be determined. One study also identified a decrease in circulating B cells in patients with FTD174, but these findings await validation in larger patient cohorts.

Of note, Grn−/− mice exhibit pro-inflammatory microglial activation downstream of NF-κB signaling175,176 coupled with the acquisition of an FTD-associated microglial state that: (1) is different from the AD- and ALS-associated DAMs, and (2) actively supports neurotoxic TARDBP granule deposition177, at least in part as a consequence of lysosomal dysfunction178. Along similar lines, mice expressing an FTD-associated TREM2 mutant exhibit brain-wide alterations including a delayed resolution of neuroinflammatory responses and a reduction of cerebral blood flow that may support disease progression179.

Vascular dysfunction and astrocytosis have also been observed in the frontal and temporal lobes of patients with FTD180. Specifically, a highly conserved astrocytic phenotype promoting synaptic degeneration and TARDBP neuropathy has been identified in the thalamus and frontal cortex of patients with GRN-associated FTD and Grn−/− mice181, pointing to a central role for these cells in FTD progression.

In summary, FTD also involves an immunological component, although it remains poorly characterized. Additional studies are required to elucidate innate and potentially adaptive immune mechanisms involved in the pathogenesis of FTD.


While most central NDs appear to originate from genetic or environmental alterations of cellular homeostasis in the brain parenchyma, it is now clear that such perturbations are accompanied by the activation of innate and (at least in some cases) adaptive immune effector mechanisms that contribute to disease pathogenesis. As abundantly discussed herein, multiple NDs are associated with mutations in genes encoding components of the innate or adaptive immune system, such as TREM229,30,31 or HLA-DRB135. Moreover, hitherto unrecognized connections are emerging between central ND susceptibility genes, such as SNCA, and core immunological functions, such as the development of normal innate and adaptive immune reactions to bacterial challenges120. Finally, patients affected by numerous NDs including AD, PD, HD, DLB and FTD exhibit shifts in the circulating levels of pro-inflammatory cytokines or peripheral immune populations, further supporting a pathogenic role for altered immune responses in the central nervous system in the progression of NDs. With a few exceptions including the robust implication of CD4+ in disease pathogenesis in mouse models of DLB157, most of the current links between immunological mechanisms and ND pathogenesis rely on observational and correlative rather than mechanistic experimental setups. While at least partially this reflects the limited number of rodent models that recapitulate the emergence and progression of NDs in humans, it will be important to harness currently available models to implement antibody-mediated depletion, pharmacological inhibition or genetic deletion/downregulation experiments to mechanistically link altered immune functions to ND pathogenesis and potentially identify novel targets for therapeutic interventions.

In the era of cancer immunotherapy (Box 2), the data summarized herein point indeed to the possibility of harnessing immunomodulatory agents beyond general anti-inflammatory and immunosuppressive drugs such as corticosteroids for the management of multiple NDs. While as mentioned above preclinical data in support of this possibility suffer from an overall observational nature, it is still tempting to postulate that currently approved therapeutics affecting immune functions may be beneficial for at least some patients with NDs. For instance, circulating IL17A elevations and/or polarization of the CD4+ T cell compartment toward a TH17 profile have been detected in patients with PD144, DLB153 and FTD172, and no less than three distinct IL17A blockers are currently available for the treatment of inflammatory conditions such as psoriasis182. Along similar lines, TREG cells have been demonstrated to limit disease progression in multiple mouse models of AD89,90,91, pointing to adoptive TREG transfer as an intriguing possibility to control AD progression in humans. Finally, PD-1 blockage has been associated with neuroprotective effects in mouse models of AD81, and multiple immune checkpoint inhibitors targeting PD-1 or its main ligand CD274 (PD-L1) are currently licensed for use in patients with various malignancies24.

Importantly, the use of immunomodulatory agents for the management of NDs has begun to be explored in the clinic. Specifically, a TREM2-targeting monoclonal antibody (AL002) is currently being assessed in patients with AD (NCT05744401)39, while an inflammasome inhibitor (RO7486967) is under investigation in individuals with PD (NCT05924243). Whether these or other immunomodulators are effective and will ultimately be approved for use in humans, however, remains to be established.

In summary, while additional work is required to elucidate the actual therapeutic potential of immunotherapy for patients with central NDs, both innate and immune dysfunctions have been documented during the progression of AD, PD, HD, DLB and FTD. It will be important to obtain further mechanistic insights into the immunological aspects of human degeneration in existing and newly developed rodent ND models to develop disease-modifying treatment options for these patient populations.