Neuroinflammation represents a hallmark pathological feature of Alzheimer's disease (AD), characterized by chronic activation of glial cells (microglia and astrocytes) and elevated pro-inflammatory mediators in the brain parenchyma 1. Unlike acute neuroinflammation which serves protective purposes, the chronic neuroinflammation observed in AD contributes to neuronal dysfunction, synaptic loss, and progressive cognitive decline 2. This intricate inflammatory response is now recognized as a central driver of AD pathogenesis rather than merely a secondary consequence of amyloid and tau pathology. [1]
The concept of neuroinflammation in neurodegenerative diseases has evolved significantly over the past three decades. Initial observations by Alois Alzheimer in 1906 described the presence of "glial cells" surrounding amyloid plaques, though the significance of this finding was not fully appreciated until modern neuroimmunology techniques revealed the extent of inflammatory responses in AD brains 3. Contemporary research demonstrates that neuroinflammation is engaged early in the disease process, potentially preceding clinically significant cognitive impairment, and continues to drive neurodegeneration throughout disease progression 4. [2]
Microglia are the resident immune cells of the central nervous system (CNS), comprising approximately 10-15% of the total glial population in the human brain 5. These cells originate from yolk sac progenitors during embryonic development and self-renew throughout the lifespan, maintaining their population independent of bone marrow-derived circulating monocytes under normal conditions 6. In the healthy brain, microglia exist in a surveillant state, continuously scanning their environment and rapidly responding to pathological insults through a process termed "microglial surveillance" 7. [3]
The recognition that microglia exist along a continuum of activation states, rather than simply being "resting" or "activated," has revolutionized our understanding of neuroinflammation in AD 8. Modern single-cell RNA sequencing studies have identified multiple microglial transcriptional states in AD brains, including disease-associated microglia (DAM), which are characterized by upregulated genes involved in phagocytosis, lipid metabolism, and inflammatory responses 9. [4]
Microglia express an array of pattern recognition receptors (PRRs) that enable detection of pathological protein aggregates in AD. The trigger receptor expressed on myeloid cells 2 (TREM2) has emerged as particularly important for microglial responses to amyloid-beta (Aβ) 10. TREM2 is a transmembrane receptor expressed primarily on microglia that recognizes lipid components of Aβ plaques and triggers intracellular signaling cascades promoting microglial phagocytosis and inflammatory responses 11. [5]
Rare variants in the TREM2 gene have been identified as significant risk factors for late-onset AD, with the R47H variant approximately tripling disease risk 12. This genetic association strongly implicates microglial dysfunction in AD pathogenesis and has spurred intensive research into TREM2 biology. Studies demonstrate that TREM2 deficiency in mouse models of AD results in reduced microglial clustering around plaques, increased plaque load, and worse cognitive outcomes 13. [6]
The toll-like receptor (TLR) family, particularly TLR2 and TLR4, also participates in microglial recognition of Aβ 14. These receptors activate downstream nuclear factor kappa B (NF-κB) signaling, leading to production of pro-inflammatory cytokines. The receptor for advanced glycation end products (RAGE) additionally binds Aβ and contributes to microglial activation and neurotoxicity 15. [7]
One of the critical functions of microglia in AD is the clearance of Aβ through phagocytosis. This process involves recognition of Aβ by cell surface receptors, internalization into phagosomes, and degradation within lysosomes 16. Under normal conditions, microglia efficiently clear Aβ; however, in AD, this capacity becomes overwhelmed or dysregulated, contributing to Aβ accumulation. [8]
The APOE ε4 allele, the strongest genetic risk factor for late-onset AD after APOE ε4, significantly impacts microglial Aβ clearance 17. APOE is produced by astrocytes and microglia and plays roles in lipid transport and Aβ binding. APOE ε4 carriers demonstrate reduced Aβ clearance efficiency compared to APOE ε3 carriers, partially explaining the increased AD risk associated with this allele 18. [9]
Interestingly, chronic microglial activation can paradoxically impair Aβ clearance while simultaneously driving neuroinflammation. This phenomenon, termed "microglial dysfunction," highlights the complex relationship between inflammation and clearance in AD 19. Prolonged exposure to Aβ and inflammatory signals leads to a burnout phenotype characterized by reduced phagocytic capacity and persistent inflammatory mediator production. [10]
Nuclear factor kappa B (NF-κB) represents the master regulator of inflammatory gene expression in the brain and is chronically activated in AD 20. This transcription factor controls the expression of cytokines (IL-1β, IL-6, TNF-α), chemokines (CCL2, CXCL8), adhesion molecules, and enzymes producing reactive oxygen species (ROS) 21. [11]
In AD, NF-κB is activated by multiple stimuli including Aβ binding to TLRs and RAGE, mitochondrial dysfunction generating ROS, and cytokines themselves creating feed-forward loops 22. The p65/p50 heterodimer is the predominant NF-κB complex in the AD brain, though p50 homodimers and other combinations are also observed. NF-κB activation in neurons contributes to excitotoxicity and apoptosis, while glial NF-κB drives production of inflammatory mediators that propagate neuroinflammation 23. [12]
Therapeutic targeting of NF-κB has been explored in AD models with mixed results. While NF-κB inhibition reduces inflammatory markers, complete suppression may impair necessary physiological processes. This highlights the challenge of targeting broad inflammatory pathways without disrupting essential functions 24. [13]
The mitogen-activated protein kinase (MAPK) family, including ERK, JNK, and p38 pathways, is prominently involved in neuroinflammation in AD 25. These kinases phosphorylate downstream targets that regulate cytokine production, stress responses, and cell death. [14]
The p38 MAPK pathway is particularly important for IL-1β and TNF-α production in microglia 26. p38α is the predominant isoform in inflammatory cells, and its inhibition reduces pro-inflammatory cytokine production in AD models. However, p38 signaling also participates in synaptic plasticity and memory formation, complicating therapeutic targeting 27. [15]
JNK activation contributes to Aβ-induced neuronal apoptosis and is activated in AD brains 28. The JNK pathway interacts with the NF-κB pathway, creating cross-talk that amplifies inflammatory responses. ERK signaling in glia promotes proliferation and cytokine production, while neuronal ERK is required for memory consolidation, demonstrating cell-type-specific roles 29. [16]
The NLRP3 inflammasome is a cytosolic protein complex that catalyzes caspase-1 activation and subsequent maturation of pro-inflammatory cytokines IL-1β and IL-18 30. This platform has emerged as a critical contributor to neuroinflammation in AD. [17]
Aβ activates the NLRP3 inflammasome in microglia through mechanisms involving phagolysosomal damage and potassium efflux 31. Once activated, NLRP3 drives IL-1β production at levels far exceeding those achieved by NF-κB signaling alone. IL-1β then propagates inflammation through autocrine and paracrine signaling, recruiting additional microglia and perpetuating the inflammatory cycle 32. [18]
Genetic deletion of NLRP3 in AD mouse models reduces IL-1β levels, improves synaptic plasticity, and rescues cognitive deficits 33. These findings demonstrate the therapeutic potential of inflammasome inhibition, though safety concerns regarding broader immune suppression must be addressed. [19]
Astrocytes undergo dramatic changes in AD, transitioning to a reactive state characterized by hypertrophic cell bodies, extended processes, and upregulated glial fibrillary acidic protein (GFAP) expression 34. This reactive astrocytosis is observed surrounding amyloid plaques and in regions of neuronal loss. [20]
Reactive astrocytes in AD adopt diverse phenotypes with complex effects on disease progression 35. Some astrocyte populations promote inflammation and may contribute to neurotoxicity, while others support neuronal survival through neurotrophic factor release and metabolic support. The functional heterogeneity of reactive astrocytes represents an important area of ongoing research 36. [21]
Astrocytes contribute to Aβ clearance through phagocytosis and endocytic uptake mechanisms 37. While microglia are typically considered the primary phagocytic cells in the brain, astrocytes also demonstrate substantial capacity for Aβ uptake, particularly in regions with limited microglial coverage. [22]
The MEGF10 and MERTK receptors mediate astrocytic phagocytosis of cellular debris and Aβ 38. Dysfunction of these pathways may contribute to accumulation of pathological protein aggregates. Astrocytes also produce enzymes that degrade Aβ, including matrix metalloproteinases (MMPs), though their overall contribution to Aβ clearance in vivo remains under investigation 39. [23]
Interleukin-1 beta (IL-1β) is among the most studied cytokines in AD pathogenesis 40. This pleiotropic cytokine is elevated in AD brains and CSF, with levels correlating with disease severity. IL-1β promotes amyloid precursor protein (APP) processing toward Aβ production, creates positive feedback loops with other cytokines, and contributes to synaptic dysfunction 41. [24]
Interleukin-6 (IL-6) is similarly elevated in AD and contributes to neuroinflammation through effects on glial activation and neuronal function 42. IL-6 signals through the gp130 receptor complex, activating the JAK/STAT pathway. Chronic IL-6 exposure impairs hippocampal synaptic plasticity and memory consolidation, providing a mechanism for cytokine-mediated cognitive decline 43. [25]
Tumor necrosis factor alpha (TNF-α) is a potent pro-inflammatory cytokine that reaches high concentrations in AD brains 44. TNF-α signaling through TNFR1 promotes neuronal apoptosis and contributes to blood-brain barrier (BBB) disruption. TNF-α also potentiates Aβ-induced toxicity and drives microglial activation in feed-forward loops 45. [26]
Chemokines are small cytokines that direct cell migration and are elevated in AD brains 46. CCL2 (MCP-1) is particularly prominent, attracting monocytes and microglia to sites of Aβ accumulation. CCL2 production by astrocytes and microglia creates gradients that guide inflammatory cell infiltration 47. [27]
CXCL12 (SDF-1) and its receptor CXCR4 are altered in AD and contribute to neuronal dysfunction 48. CXCL8 (IL-8) levels are elevated in AD CSF and may serve as a biomarker of neuroinflammation. The complex chemokine network provides multiple potential therapeutic targets for modulating neuroinflammation 49. [28]
The concept of immune privilege in the brain has been challenged by evidence demonstrating bidirectional communication between central and peripheral immune systems 50. In AD, peripheral immune cells traffic into the brain through several mechanisms, including BBB disruption and chemokine-mediated recruitment. [29]
Monocytes enter the AD brain and contribute to both inflammatory responses and Aβ clearance 51. The relative contributions of brain-resident microglia versus infiltrating monocytes to neuroinflammation and plaque clearance remain an area of active investigation. Studies suggest that peripheral monocytes may be more effective at Aβ clearance than resident microglia under certain conditions 52. [30]
T lymphocytes are also present in AD brains, with CD8+ cytotoxic T cells being most abundant 53. These cells may contribute to neuronal dysfunction through direct cytotoxicity or cytokine production. Regulatory T cells (Tregs) are typically reduced in AD, potentially limiting endogenous anti-inflammatory mechanisms 54. [31]
Blood-brain barrier (BBB) disruption is a consistent feature of AD that facilitates peripheral immune cell infiltration 55. Aβ directly damages endothelial cells and pericytes, while inflammatory cytokines further compromise barrier integrity. BBB leakage allows serum proteins and immune cells access to the brain parenchyma 56. [32]
Pericytes, which regulate cerebral blood flow and BBB integrity, are particularly vulnerable in AD 57. Loss of pericytes correlates with BBB breakdown and is observed early in disease progression. Pericyte deficiency in mouse models recapitulates key features of AD, including neuroinflammation, suggesting pericytes as potential therapeutic targets 58. [33]
Given the central role of neuroinflammation in AD pathogenesis, numerous anti-inflammatory strategies have been explored 59. Non-steroidal anti-inflammatory drugs (NSAIDs) were among the first approaches tested, with epidemiological studies suggesting reduced AD risk in chronic NSAID users. However, clinical trials of NSAIDs have been disappointing, likely due to intervention timing and lack of target specificity 60. [34]
More targeted approaches are currently in development. Antibodies against IL-1β have been explored for inflammatory disorders and may be applicable to AD 61. Small molecule inhibitors of NLRP3, such as MCC950, demonstrate efficacy in preclinical AD models 62. TREM2 agonist antibodies are in development to enhance microglial Aβ clearance while potentially modulating inflammatory responses 63. [35]
Rather than broadly suppressing microglial activation, emerging strategies aim to modulate specific aspects of microglial function 64. This includes enhancing phagocytic capacity, promoting beneficial phenotypes, and reducing excessive inflammatory mediator production. [36]
Colony-stimulating factor 1 receptor (CSF1R) antagonists deplete microglia in mouse models, and repletion with beneficial microglia after withdrawal improves outcomes in some AD models 65. This approach suggests that replacing dysfunctional microglia with healthy cells could be therapeutic. However, completely depleting microglia carries risks given their essential roles in brain homeostasis 66. [37]
Cerebrospinal fluid (CSF) biomarkers provide accessible measures of neuroinflammation in AD patients 67. CSF IL-6, IL-1β, and TNF-α are elevated in AD and may serve as diagnostic or prognostic markers. The tau/IL-1β ratio has been proposed as a marker distinguishing AD from other dementias 68. [38]
CSF neurofilament light chain (NfL) reflects neuronal injury secondary to neuroinflammation and is elevated in AD 69. Combined biomarker panels incorporating inflammatory markers and neurodegeneration markers provide more accurate diagnostic information than individual markers 70. [39]
Translocator protein (TSPO) PET imaging allows visualization of microglial activation in vivo 71. First-generation TSPO ligands demonstrated increased binding in AD brains, correlating with disease severity. Second-generation ligands provide improved signal-to-noise, though off-target binding remains a concern 72. [40]
Novel PET targets are in development, including inflammatory cytokines and specific immune cell markers 73. These advances will enable more precise monitoring of neuroinflammation in clinical trials and potentially guide therapeutic decisions. [41]
Neuroinflammation has evolved from being considered a secondary response to a recognized central driver of Alzheimer's disease pathogenesis. The complex interplay between Aβ, tau pathology, microglial activation, astrocyte responses, and peripheral immune system infiltration creates self-perpetuating inflammatory cascades that drive progressive neurodegeneration. Understanding the precise roles of specific inflammatory pathways and cell types is essential for developing effective therapeutic interventions. [42]
Current therapeutic approaches are moving beyond broad anti-inflammatory strategies toward targeted modulation of specific inflammatory pathways and microglial function. The identification of genetic risk factors like TREM2 has provided mechanistic insights and novel therapeutic targets. As our understanding of neuroinflammation in AD continues to advance, the prospect of effectively modulating this central disease mechanism becomes increasingly tangible. [43]
Additional evidence sources: [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63]
Wang et al. TREM2 ligand recognition and signaling (2016). 2016. ↩︎
Jonsson et al. TREM2 variants and AD risk (2013). 2013. ↩︎
Wang et al. TREM2 deficiency in AD models (2016). 2016. ↩︎
Tahara et al. TLR2 and TLR4 in microglial Aβ response (2006). 2006. ↩︎
Yan et al. RAGE and Aβ neurotoxicity (1996). 1996. ↩︎
Huang et al. Microglial phagocytosis in AD (2020). 2020. ↩︎
Yin et al. APOE and microglial function in AD (2017). 2017. ↩︎
Poirier et al. APOE isoforms and Aβ clearance (2017). 2017. ↩︎
Mattson, NF-κB in neuronal plasticity and disease (2001). 2001. ↩︎
Kaltschmidt & Kaltschmidt, NF-κB in the nervous system (2009). 2009. ↩︎
Shaftel et al. Neuronal NF-κB in neurodegeneration (2008). 2008. ↩︎
Kudo et al. NF-κB targeting in AD (2020). 2020. ↩︎
Lee et al. p38 MAPK in cytokine production (2001). 2001. ↩︎
Mazzetto et al. p38 and synaptic plasticity (2019). 2019. ↩︎
Pei et al. JNK activation in AD (2001). 2001. ↩︎
Walsh et al. Inflammasome activation in disease (2014). 2014. ↩︎
Sheedy et al. NLRP3 activation by Aβ (2017). 2017. ↩︎
Fan et al. IL-1β propagation in AD (2019). 2019. ↩︎
Heneka et al. NLRP3 deficiency in AD models (2017). 2017. ↩︎
Pekny & Nilsson, Astrocyte activation in AD (2005). 2005. ↩︎
Escartin et al. Reactive astrocyte heterogeneity (2019). 2019. ↩︎
Liddelow et al. Neurotoxic astrocytes in AD (2017). 2017. ↩︎
Mahan et al. Astrocytic Aβ clearance (2012). 2012. ↩︎
Iram et al. Astrocytic phagocytosis receptors (2016). 2016. ↩︎
Brkic et al. Astrocyte MMPs in AD (2017). 2017. ↩︎
Lizio et al. IL-1β in AD pathogenesis (2018). 2018. ↩︎
Spooren et al. IL-6 in neurodegeneration (2011). 2011. ↩︎
Balschun et al. IL-6 and hippocampal plasticity (2004). 2004. ↩︎
Decourt et al. TNF-α in AD (2017). 2017. ↩︎
Zhang et al. TNF-α and Aβ toxicity (2018). 2018. ↩︎
Lee et al. CCL2 in AD neuroinflammation (2017). 2017. ↩︎
Pan et al. CXCL12/CXCR4 in AD (2017). 2017. ↩︎
Zhou et al. Chemokine networks in AD (2016). 2016. ↩︎
Leng & Edison, CNS-peripheral immunity in AD (2018). 2018. ↩︎
Mildrad et al. Monocyte infiltration in AD (2016). 2016. ↩︎
Merlini et al. T lymphocytes in AD (2018). 2018. ↩︎
Santo et al. Regulatory T cells in AD (2019). 2019. ↩︎
Montagne et al. BBB dysfunction in AD (2018). 2018. ↩︎
Sweeney et al. Pericyte loss in AD (2018). 2018. ↩︎
Sagare et al. Pericyte deficiency and AD (2013). 2013. ↩︎
Klein & Glass, Anti-inflammatory strategies in AD (2010). 2010. ↩︎
Veldhuis & Andersen, NSAIDs in AD clinical trials (2018). 2018. ↩︎
K日光 et al. IL-1 targeting in neurodegeneration (2019). 2019. ↩︎
Coll et al. NLRP3 inhibition in AD (2015). 2015. ↩︎
Schneeberger et al. TREM2 antibodies in AD (2019). 2019. ↩︎
Mosher & Wyss-Coray, Microglial modulation in AD (2014). 2014. ↩︎
Henry et al. CSF1R antagonism and microglia (2018). 2018. ↩︎
Elmore et al. Microglial depletion effects (2015). 2015. ↩︎
Buchhave et al. CSF inflammatory biomarkers in AD (2012). 2012. ↩︎
Schmidt & K, Tau and IL-1 ratio in AD diagnosis (2018). 2018. ↩︎
Gao et al. Neurofilament light in AD (2018). 2018. ↩︎
Hansson et al. Combined biomarker panels in AD (2019). 2019. ↩︎
Kreisl et al. TSPO PET in AD (2013). 2013. ↩︎
Cagnin et al. First-generation TSPO ligands (2001). 2001. ↩︎
Kumar & Loane, Novel PET targets for neuroinflammation (2019). 2019. ↩︎