The complement cascade in Alzheimer’s disease: modern implications of an ancient immune protagonist Authors: Maria-Tzousi Papavergi, Praveen Bathini, Brijendra Singh, Cynthia A. Lemere Published: 2025 ABSTRACT The complement system, a critical arm of innate immunity, has emerged as a key contributor to the pathogenesis of Alzheimer’s disease (AD). While complement-mediated synaptic pruning is essential during brain development, its reactivation in the aging and diseased brain can promote neurodegeneration. In AD and cerebral amyloid angiopathy (CAA), aberrant complement activity contributes to amyloid-β (Aβ) accumulation, synapse loss, neuroinflammation, vascular dysfunction and blood-brain barrier (BBB) disruption. This review traces the evolving understanding of complement dysregulation in AD, from foundational findings in human studies to mechanistic discoveries in animal models and emerging induced pluripotent stem cell (iPSC)-derived cellular systems. It describes how genetic and epigenetic factors, including risk variants, chromatin modifications, and microRNAs, modulate complement pathways in the AD brain. Systemic influences, such as gut-brain axis disruption, are also considered for their potential to exacerbate complement-mediated neuroinflammation in AD. We further highlight efforts to identify biomarkers of complement activation and review the therapeutic potential of targeting complement components and regulators. By integrating molecular, experimental, and translational perspectives, this review outlines the multifaceted involvement of complement in AD and discusses key directions for future research and therapeutic interventions. Background Neurodegeneration involves the progressive loss of neuronal structure and function and is a core feature of diverse neurological disorders including Alzheimer’s disease (AD), Parkinson’s disease (PD), Huntington’s disease (HD), and amyotrophic lateral sclerosis (ALS) [1]. Neurodegeneration is also linked to other neurological conditions such as ischemic stroke, traumatic brain injury (TBI) and vascular diseases. These conditions, while distinct in their aetiologies, can all lead to neuronal damage, cerebrovascular changes, functional decline, and risk of developing dementia [2]. AD, the most prevalent type of dementia, is characterized by extracellular deposition of amyloid-β (Aβ), intracellular neurofibrillary tangles (NFTs) and progressive synapse loss. It is a complex disease with a pathological continuum of degenerative events starting decades before the onset of clinical symptoms [3]. Despite decades of research into key proteins like Aβ and tau and strategies to combat neuroinflammation, AD remains difficult to treat effectively due to its complex and heterogeneous biology. Importantly, diverse combinations of risk or resilience factors affect distinct biological domains implicated in the development and course of AD pathogenesis [4]. Indeed the aetiology of AD is shaped by interplay of genetic risk factors, ageing and other, modifiable, risk factors including cardiometabolic disorders, environmental factors and lifestyle [5]. It is becoming more apparent that inflammatory processes represent a key driver of neurodegenerative diseases rather than acting as a bystander, with confirmation from genetic studies indicating many risk genes to be associated with inflammation and the innate immune response [6] including complement regulators like CLU, which encodes for clusterin, a weak negative regulator of membrane attack complex (MAC) assembly, and CR1, encoding for complement receptor 1 [7–9]. The complement system, a crucial component of innate immunity, has emerged as a key player in both central nervous system (CNS) development and neurological disease [10]. Originally characterized for its role in host defense and inflammation [11], recent research has revealed increasingly complex functions of complement proteins in the brain, ranging from synaptic pruning during development to involvement in various neurodegenerative conditions [12, 13]. Complement has well-established physiological functions in the CNS, with C1q and downstream components guiding synaptic pruning during development, ensuring refinement of neural circuits [14, 15]. Complement components also exert direct neuroprotective effects; C1q enhances neuronal viability and neurite outgrowth, modulates gene and microRNA expression, regulates neurotrophin levels and protects neurons from Aβ toxicity [16]. Understanding these diverse roles has become crucial for developing new therapeutic strategies for neurological disorders [17]. The complement cascade operates through three main activation pathways: classical, alternative, and lectin which are described in more detail in the next section. This cascade system provides a carefully regulated mechanism for immune response and tissue homeostasis, but its dysregulation can contribute to pathological conditions. Beyond the classical, alternative and lectin pathways, recent work has identified non-canonical routes of complement activation. Notably, granzyme K (GZMK), a serine protease released by cytotoxic lymphocytes, can cleave C2 and C4 to generate active convertases, thereby initiating full complement activation [18]. The role of complement dysregulation in neurological disorders has been extensively documented [19–23]. In AD, complement components participate in multiple pathological processes, including Aβ opsonization, chronic inflammation, and aberrant synaptic pruning [24, 25], driving neurodegeneration and cognitive decline. The clinical implications of complement research extend to both diagnostic and therapeutic applications, with some biomarkers like CLU, Factor I and terminal complement complex (TCC) able to predict disease progression in AD [26]. Understanding the precise mechanisms of complement dysregulation in different neurological conditions will be crucial for developing effective therapeutic strategies. As research continues to uncover new aspects of complement function in the CNS, the potential for therapeutic intervention grows. Future directions include the development of more specific inhibitors, improved delivery systems for crossing the blood-brain barrier (BBB), and a better understanding of the temporal aspects of complement activation in different disease states [12, 19]. The continued investigation of the role of the complement system in neurological disorders promises to yield new therapeutic strategies for these challenging conditions. In this review, we report on the multifaceted role of complement activation in the pathogenesis and progression of AD. We discuss its evolving role in amyloid and tau pathology, synaptic loss, and neuroinflammation, as well as its contribution to cerebral amyloid angiopathy (CAA). We also highlight the genetic and epigenetic mechanisms regulating complement activity in the AD brain and consider peripheral influences, including gut–brain interactions. Finally, we review ongoing efforts to identify biomarkers of complement activation and evaluate the therapeutic potential of targeting complement components and regulators in AD. The complement system Overview of the complement system It all started in the late 19th century when Jules Bordet was working in Élie Metchnikoff’s lab and discovered that serum contained a heat-labile component (inactivated at 56 °C for 30 min) that could lyse bacteria in the presence of specific antibodies [27, 28]. To characterize this bactericidal component, Bordet adopted the word ‘alexin’, introduced by Buchner in 1891 [29], which comes from the ancient Greek word αλέξειν, meaning ‘to defend’ [28, 30, 31]. In 1899, Paul Ehrlich was the first to use the now widely-known term complement [32] to describe how these newly-found components ‘complement’ the action of antibodies to fight pathogens. Over a century later, the complement system has been established as an evolutionarily conserved and sophisticated system [33] that plays a key role in innate immunity [34–38]. The complement system is recognized as one of the major effector mechanisms of the innate immune system and is composed of a complex network of plasma and membrane-associated serum proteins, which can robustly induce inflammatory and cytolytic immune responses against infection and injury [39]. Pepy’s seminal experiments in the 1970s revealed for the first time that complement also affects adaptive immunity [40, 41]. Researchers now define complement as a bridge between innate and adaptive immunity via regulating B- and T-cell responses, shaping the host’s ability to combat invasion from pathogens [39, 42]. More specifically, complement components such as C3a and C5a anaphylatoxins may interact with complement receptors on B- and T-lymphocytes, influencing processes like clonal expansion, memory cell formation, and antigen presentation [43, 44]. Delving deeper, the complement cascade includes more than 50 soluble and membrane-bound proteins, regulators and receptors [45, 46]. Complement proteins comprise ~ 5% of total plasma protein [39], and are constitutively present in circulation in their inactive, soluble forms under homeostatic conditions [19]. Upon activation of the complement cascade, these proteins are proteolytically cleaved into fragments (such as iC3b and C3d), which are considered activation products and are not constitutively found in the blood [19]. Complement can be activated via three different pathways: the classical, lectin, and alternative complement pathways (Fig. 1). All three pathways converge on the cleavage of the C3 component, triggering downstream events leading to inflammation, opsonization, and target cell lysis. The classical pathway, the first to be discovered, is initiated when the C1 complex, composed of C1q and the serine proteases C1r and C1s, binds to an activator surface, such as antigen-antibody complexes, fibrillar Aβ, neurofibrillary tangles, or neuronal blebs [19]. This binding induces the cleavage of C4 and C2, resulting in the formation of the C3 convertase C4bC2b (historically known as C4bC2a) [47]. The lectin pathway is initiated by binding carbohydrates such as mannose-binding lectin (MBL), ficolins, and collectins on the surface of pathogens, activating two proteases, MASP1 and MASP2 (mannose-binding serine proteases 1 and 2), causing cleavage of C4 and then C2, whose fragments combine to form the same C3 convertase as in the classical pathway [48]. The alternative pathway is constitutively active at low levels in the blood via spontaneous C3 hydrolysis [49]. Upon pathogen recognition -in the case of e.g. viruses, fungi, bacteria, lipopolysaccharides (LPS), etc.- C3b binds Factor B, which is cleaved by Factor D, forming the alternative C3 convertase (C3bBb) [49]. Both C3 convertases cleave C3 into C3a and C3b. C3b can combine with the two different C3 convertases to form C5 convertase (C4bC2bC3b in the classical and lectin pathways and C3bBbC3b in the alternative pathway), which in turn cleaves C5 into C5a and C5b. Finally, C5b recruits C6, C7, C8, and multiple copies of C9 to form the membrane attack complex (MAC or C5b-9), which perforates target cell membranes [50]. Notably, a putative fourth pathway of complement activation has recently been described [18]. In this pathway, lymphocyte-derived Granzyme K acts as the initiator protease by directly cleaving C4 and C2, thereby propagating the entire cascade [18]. As mentioned above, complement activation exerts a range of effector functions. C3b and its inactivated Factor I-mediated cleavage product, iC3b, act as opsonins, tagging pathogens, apoptotic cells, or debris for clearance by phagocytes bearing complement receptors [44]. Specifically, C3b binds to complement receptor 1 (CR1) (Fig. 1), whereas iC3b engages complement receptor 3 (CR3) [44]. C3a and C5a are anaphylatoxins that bind their respective receptors, C3aR and C5aR, promoting chemotaxis, cytokine release, and immune cell activation [44]. Importantly, there are two forms of C5aR. C5aR1 (also known as CD88) is pro-inflammatory, while C5aR2 (formerly known as C5L2 or GPR77) may have regulatory or anti-inflammatory roles [51]. Beyond these core functions, complement also interacts with Toll-like receptor (TLR) signaling [52], Nod-like receptor family pyrin domain containing 3 (NLRP3) inflammasome activation [46], and coagulation pathways [53], amplifying inflammatory cascades and bridging innate and adaptive responses. The role of these pathways in the CNS is discussed later. The tight regulation of complement activation is crucial for maintaining immune homeostasis, which is why the complement system is often described as “a double-edged sword”. Excessive or misdirected activation can result in severe host tissue damage. A complex network of fluid phase and membrane-bound regulatory proteins, including Factor H, CD55 (decay-accelerating factor), and CD59 (protectin) (Fig. 1), ensures precise control of complement attack, and adjusts its propagation and endpoints to the cellular target [44, 54]. More specifically, these regulators enable discrimination between self and non-self surfaces, thereby preventing the formation or stabilization of C3 and C5 convertases and inhibiting MAC assembly on host cells [54] (Fig. 1). Nevertheless, genetic variants, autoantibodies, or insufficient expression of complement regulators may cause dysregulation of the complement cascade, resulting in chronic inflammation, autoimmune disease, and tissue injury [54, 55]. These mechanisms play a central role in the pathogenesis of disorders such as systemic lupus erythematosus, age-related macular degeneration, cancer, and a range of neurodegenerative diseases, including AD, in which sustained or mislocalized complement activation exacerbates tissue damage and disease progression [56]. It should be emphasized that, over the past decade, there has been a fundamental change in our understanding of complement biology with the discovery of its crucial role in the intracellular space of a broad range of cellular populations and tissues [57, 58]. This intracellularly active complement system was defined as ‘the complosome’ [58]. From an evolutionary standpoint, the discovery of intracellular complement activity has formed the hypothesis that, in its most ancestral form, C3 may have been retained within the cell as a host defense protein, functioning to tag and neutralize microorganisms engulfed by primitive unicellular eukaryotes such as choanoflagellates [11, 45]. The role of the complosome has been shown to extend beyond traditional immune functions, serving as a central regulator of immune cell metabolism [59–61], autophagy [62–65], and gene transcription [66–70]. Its function is not limited within immune cells [57, 67, 68, 70–76], but expands across a variety of non-immune cell types and tissues [65, 77–84], orchestrating cellular homeostasis, modulating responses to both infectious and non-infectious stimuli, and controlling cell turnover and survival [46]. Altered complosome function has been implicated in the pathogenesis of various diseases, ranging from infection-related complications [66, 72, 85] to chronic inflammatory conditions [86, 87], cancer [88–90], and renal dysfunction [91, 92]. Although its role in the brain is only beginning to be understood, growing evidence suggests that its dysregulation may also contribute to neurodegenerative processes [93]. Thus, selectively modulating the complosome, potentially in combination with extracellular complement inhibition, could represent a promising therapeutic strategy for complement-mediated disorders, including those affecting the nervous system. Fig. 1Overview of the complement cascade. (A) The complement system can be activated by three different pathways: the classical, lectin and alternative pathways. The classical pathway is initiated when complement component 1 (C1) complex binds to an activator surface, such as antibody complexes, fibrillar amyloid-β (Aβ), and neurofibrillary tangles (NFTs), and can be inhibited by SRPX2 (Sushi Repeat Containing Protein X-linked 2) via binding to C1q. The lectin pathway is activated when microbial carbohydrates bind mannan binding lectin (MBL) in complex with MASP1/2. C1-INH (C1 inhibitor) is the central regulator of the classical and lectin pathways, acting by inactivating C1r, C1s, MASP-1, and MASP-2. The alternative pathway is activated by spontaneous hydrolysis of C3 and formation of a C3 convertase. All three pathways converge at C3, which is cleaved by the C3 convertases into C3a and C3b. At the level of the classical/lectin C3 convertase (C4bC2b; historically known as C4bC2a), regulators including CR1 (Complement Receptor 1), C4bp (C4 binding protein), CSMD1 (CUB and Sushi Multiple Domains 1), CD46 (membrane cofactor protein), CD55 (decay-accelerating factor), FI (Factor I) and SUSD4 (Sushi Domain Containing 4) either accelerate convertase decay or act as cofactors for FI-mediated cleavage of C3b and/or C4b. C3a promotes chemotaxis and cell activation via C3a receptor (C3aR). C3b can be converted to iC3b, which opsonizes targets for microglial phagocytosis via complement receptor 3 (CR3). Conversion of C3b to iC3b requires FI together with cofactors including CR1, CD46, and FH (Factor H). C3b can also bind to C4bC2b or C3bBb to form the C5 convertases. C5 is then cleaved into C5a, which promotes chemotaxis and cell activation via C5aR1, and C5b, which binds to C6, C7, C8 and C9 to form the membrane attack complex (MAC; C5b–9), ultimately resulting in membrane disruption and cell lysis. As illustrated, C5 convertases are negatively regulated by CR1, CD55, and FH through decay-acceleration, as well as by CR1, CD46, and FH acting as cofactors for FI-mediated inactivation of C3b. C5a, as well as C3a, can be inactivated by CPN (carboxypeptidase N), which removes their C-terminal arginine, thereby reducing their ability to bind and signal through their respective receptors. Finally, terminal pathway regulators including clusterin (Clu), vitronectin (Vn), CD59 (protectin), and CSMD1, prevent MAC formation. MAC structure adapted from PDB ID: 6H04. Figure created with Biorender.com Overview of the complement cascade. (A) The complement system can be activated by three different pathways: the classical, lectin and alternative pathways. The classical pathway is initiated when complement component 1 (C1) complex binds to an activator surface, such as antibody complexes, fibrillar amyloid-β (Aβ), and neurofibrillary tangles (NFTs), and can be inhibited by SRPX2 (Sushi Repeat Containing Protein X-linked 2) via binding to C1q. The lectin pathway is activated when microbial carbohydrates bind mannan binding lectin (MBL) in complex with MASP1/2. C1-INH (C1 inhibitor) is the central regulator of the classical and lectin pathways, acting by inactivating C1r, C1s, MASP-1, and MASP-2. The alternative pathway is activated by spontaneous hydrolysis of C3 and formation of a C3 convertase. All three pathways converge at C3, which is cleaved by the C3 convertases into C3a and C3b. At the level of the classical/lectin C3 convertase (C4bC2b; historically known as C4bC2a), regulators including CR1 (Complement Receptor 1), C4bp (C4 binding protein), CSMD1 (CUB and Sushi Multiple Domains 1), CD46 (membrane cofactor protein), CD55 (decay-accelerating factor), FI (Factor I) and SUSD4 (Sushi Domain Containing 4) either accelerate convertase decay or act as cofactors for FI-mediated cleavage of C3b and/or C4b. C3a promotes chemotaxis and cell activation via C3a receptor (C3aR). C3b can be converted to iC3b, which opsonizes targets for microglial phagocytosis via complement receptor 3 (CR3). Conversion of C3b to iC3b requires FI together with cofactors including CR1, CD46, and FH (Factor H). C3b can also bind to C4bC2b or C3bBb to form the C5 convertases. C5 is then cleaved into C5a, which promotes chemotaxis and cell activation via C5aR1, and C5b, which binds to C6, C7, C8 and C9 to form the membrane attack complex (MAC; C5b–9), ultimately resulting in membrane disruption and cell lysis. As illustrated, C5 convertases are negatively regulated by CR1, CD55, and FH through decay-acceleration, as well as by CR1, CD46, and FH acting as cofactors for FI-mediated inactivation of C3b. C5a, as well as C3a, can be inactivated by CPN (carboxypeptidase N), which removes their C-terminal arginine, thereby reducing their ability to bind and signal through their respective receptors. Finally, terminal pathway regulators including clusterin (Clu), vitronectin (Vn), CD59 (protectin), and CSMD1, prevent MAC formation. MAC structure adapted from PDB ID: 6H04. Figure created with Biorender.com Complement in the CNS The hepatic origin of most complement components contributed to the long-standing assumption that the complement system operated exclusively in systemic immune responses. Over the years, it became progressively evident that complement proteins, as well as their receptors and regulators, are expressed throughout the CNS [94]. Currently, it is well established that CNS-resident cells, including neurons, astrocytes, microglia, and oligodendrocytes express complement proteins [95–97]. Notably, astrocytes are the primary source of C3 in the brain, while microglia are the dominant producers of C1q, particularly under inflammatory or neurodegenerative conditions [98]. Furthermore, astrocytes together with oligodendrocytes secrete C4, and neurons express C1q and C3, indicative of cell-intrinsic functions in synaptic regulation, beyond their classical immune roles [96]. Additionally, under stress or neurodegenerative conditions, astrocytes and mast cells have been shown to release C3a via connexin- and pannexin-based hemichannels, representing a non-canonical route of complement-mediated neuroinflammatory signaling [99]. The pivotal work of Stevens and colleagues [14] elucidated the fundamental role of the complement system in synaptic refinement during brain development. Following birth, excessive or weak synaptic terminals are eliminated by a process defined in 1982 as synaptic pruning [100], thereby effectively refining neural circuits. Early components of the classical complement pathway, such as C1q and C3, localize to synapses tagging them for removal by microglia expressing complement receptors, e.g. CR3/CD11b [14, 15, 101, 102]. During both development and adulthood, microglia prune synapses in an activity-dependent manner via CR3 recognition of synapse-bound iC3b [101]. In addition to CR3, CR1 has also been shown to modulate microglial reactivity. In vitro, pharmacological blockade of CR1 suppressed Aβ-induced microglial activation in primary rodent microglia, highlighting a functional role for CR1 in glial complement signaling [103]. In parallel, astrocytes contribute to synaptic engulfment through the C1q binding protein MEGF10 (multiple EGF-like domains 10) [104], as well as via MERTK (MER proto-oncogene, tyrosine kinase) and apolipoprotein E (APOE)-dependent phagocytic pathways, both of which operate in a C1q-dependent manner [105–107]. Notably, recent findings by Brott et al. revealed that C4-derived fragments engage human Leukocyte immunoglobulin-like receptor type B2 (LilrB2)/murine Paired immunoglobulin receptor B (PirB) with high affinity, linking complement activation to synaptic pruning in the AD brain [108]. Although C1q plays a well-established central role in synaptic pruning during both development and later in life, recent findings have revealed that it is dispensable for proper synaptic circuit refinement in the binocular zone of the primary visual cortex (V1b), pointing to region-specific mechanisms underlying synaptic pruning [109]. To this end, evidence regarding neuronally expressed complement regulators, including transmembrane proteins such as CSMD1 (CUB and Sushi Multiple Domains 1) (Fig. 1) [110–112], CSMD3 (CUB and Sushi Multiple Domains 3) [113, 114], and SUSD4 (Sushi Domain Containing 4) (Fig. 1) [115, 116], and the secreted protein SRPX2 (Sushi Repeat Containing Protein X-linked 2) (Fig. 1) [117, 118], has demonstrated their region-specific involvement in synaptic pruning [19, 119]. Importantly, aberrant synaptic pruning can lead to inappropriate synaptic retention or elimination, ultimately disrupting brain connectivity. Altered connectivity driven by complement dysregulation has been linked to neurodevelopmental and neurological disorders, including autism spectrum disorder [120], schizophrenia [121, 122], multiple sclerosis [123–125], TBI [126], stroke [127], and AD [128]. Taking a more detailed look at the implications of complement factors in the CNS, alterations in the expression or genetic regulation of C1q, C3, and C4 have been linked to aberrant synaptic pruning and dysfunctional connectivity in autism spectrum disorder and schizophrenia [120–122, 129, 130]. Complement components C3 and C3aR have been identified as key regulators of adult hippocampal neurogenesis [131, 132]. In multiple sclerosis, complement contributes to demyelination and lesion formation, with C1q, C3, and MAC deposition detected in both active and chronic plaques [22, 133, 134]. Following TBI or ischemic stroke, complement is rapidly activated at the site of injury, amplifying tissue damage through anaphylatoxin signaling and MAC-induced cell lysis [126, 135–137]. Over the past decades, data from human studies, in vivo and in vitro models, and omics approaches have elucidated the central and multifaceted role of complement in AD initiation, progression and exacerbation. This role spans from synaptic loss and glial activation to integration with genetic and epigenetic risk factors. Emerging evidence also implicates complement in anti-amyloid antibody-induced Amyloid-related Imaging Abnormalities (ARIA), the primary side effect of anti-amyloid immunotherapy in AD, in both nonclinical and clinical studies [138, 139]. Importantly, a recent retrospective case-control study of aducanumab-treated participants identified complement activation in regions corresponding to ARIA on magnetic resonance imaging (MRI) [140]. In light of these findings, the bidirectional crosstalk between peripheral and CNS complement players requires further investigation to better understand the progression of neuroinflammatory diseases. The BBB is a highly selective structure composed of tightly joined endothelial cells, vascular pericytes, and perivascular glia [141, 142]. Under physiological conditions, the BBB compartmentalizes the peripheral and central immune environments, effectively restricting circulating complement proteins from entering the brain parenchyma and enabling the CNS to regulate its own complement activity via local synthesis [13]. Nevertheless, BBB breakdown can occur with aging and sustained chronic inflammation [143], and is a hallmark of several neurological disorders, including multiple sclerosis, TBI, stroke, and AD [144–148]. This breach allows the infiltration of the brain by peripheral complement components, including C3, C5, and anaphylatoxins, amplifying neuroinflammation via the introduction of pre-activated or excessive complement proteins into an already vulnerable microenvironment [149, 150]. At the same time, brain-derived Damage-Associated Molecular Patterns (DAMPs), cytokines, or complement fragments may leak into the circulation and modulate systemic complement activation, creating a feedback loop that exacerbates immune responses on both sides of the barrier [151, 152]. Importantly, the choroid plexus has also been shown to undergo early complement-related proteomic changes in AD [153]. In the APPNL−G−F mouse model, Delvenne and colleagues identified dysregulation of complement and immune signaling pathways in the choroid plexus as early as 7 weeks of age, with corresponding alterations in human cerebrospinal fluid (CSF) proteomes [153]. These findings reinforce the notion that peripheral-central immune crosstalk at barrier sites may drive early complement activation in AD. This crosstalk has significant implications for therapeutic strategies: complement inhibitors targeting systemic components may have limited CNS efficacy when the BBB remains intact, whereas CNS-penetrant agents or dual-compartment approaches may be required in pathologies with documented barrier disruption [13]. Thus, it is of great significance to understand how complement flows and signals across this compartmental boundary in order to develop targeted and effective neuroimmune therapies. This is particularly relevant in AD, where both peripheral and central complement dysregulation converge, and where BBB dysfunction may further amplify complement-driven neurodegeneration. These insights provide a foundation for exploring the evolving understanding of complement in the context of AD. The evolution of complement in AD Evidence in human AD Insights into complement involvement in AD began in the 1980s, when Eikelenboom and Stam reported the presence of complement proteins within amyloid plaques in late-stage AD postmortem brain tissue for the first time [154]. In the following years, these findings were further confirmed [155–159], sparking scientific interest towards the role of complement in AD pathophysiology. These studies revealed that early components of the classical pathway, such as C1q, C3, and C4, were particularly abundant in amyloid plaques and NFTs, whereas detection of terminal complement proteins was initially inconsistent [158, 160, 161]. By the late 1990s, however, the presence of MAC depositions was confirmed in neurons and plaques [162, 163], and complement component genes were shown to be upregulated in the AD brain [164]. The fact that complement activation progresses to the MAC stage indicates that the regulatory mechanisms of the complement system have been unable to halt the complement activation process [165]. Altogether, these findings led to the hypothesis that complement activation is not merely a consequence of neurodegeneration, but may actively contribute to disease progression. Nonetheless, several caveats should be noted when interpreting these early immunohistochemical studies. Although some antibodies, such as commonly used C1q reagents, have been validated in knockout mouse tissue, such validation cannot be performed in humans as complement knockout samples are not available. Moreover, the sticky nature of amyloid plaques and NFTs can lead to non-specific binding of both primary and secondary antibodies, raising the possibility of false-positive signals. Importantly, independent confirmation of these findings comes from proteomic and transcriptomic studies, which consistently identify complement changes in the human AD brain. With the turn of the millennium, a new wave of studies sought to unravel the ‘mystery’ of complement and neuroinflammation in AD [165–167]. For instance, the use of Down’s syndrome brains as a temporal model for AD revealed that the classical complement activation follows the compaction of Aβ42 deposits and it may culminate in neuronal MAC expression in response to Aβ plaque maturation [168]. Seminal studies demonstrated that complement component C1q binds fibrillar Aβ, triggering activation of the C1 complex and inhibiting Aβ uptake by microglia [169–171]. Similarly, tau, as another antibody-independent activator of C1q, was shown to activate complement through C1q binding [172]. Notably, in patients with CAA, cerebrovascular Aβ colocalized with C3d and C9 [173]. Rogers and colleagues showed that in the periphery, Aβ is cleared by C3b-dependent adherence to CR1 on erythrocytes [174], highlighting the involvement of complement in Aβ trafficking and clearance beyond the CNS. These initial findings laid the foundation for a series of studies over the past two decades exploring the involvement of complement in human AD. Studies examining complement proteins in CSF have reported increased levels of C3 and complement regulators such as Factor H and CR1 in patients with AD compared to controls [175, 176]. In line with these fluid biomarker findings, CR1 expression is now well documented in human brain tissue and induced pluripotent stem cell (iPSC)-derived microglia [177], and recent work has demonstrated a functional role for CR1 in mediating phagocytosis [178]. Furthermore, Wu et al. demonstrated increased levels of C3 in postmortem AD brains, especially at synapses, as well as in CSF, where it correlated with tau pathology [179]. In the aging human brain, C1q levels rise substantially and are predominantly localized near hippocampal synapses, suggesting a potential role in age-related synaptic vulnerability [180]. Similarly, C1q has been found to increase and associate with synapses in human AD brains [181]. Additionally, C1q-ApoE complexes were found in plaques and the choroid plexus, suggesting a role for APOE genotype in modulating C1q function at key neuroimmune interfaces [182, 183]. Chatterjee and colleagues reported elevated C1q in extracellular vesicles derived from the CSF of AD patients, further supporting its potential as a biomarker for AD [184]. These data reinforce the clinical relevance of complement activation in AD and suggest its potential utility as a biomarker of disease progression and synaptic dysfunction. Complement dysregulation in CAA in AD CAA is characterized by the deposition of Aβ protein in the vascular walls of the meninges and vessels, predominantly affecting the cortical and meningeal vessels, with involvement of cortical structures like white matter, thalamus, basal ganglia and brainstem [185]. CAA is commonly linked to AD and Lewy body dementia, and, to a lesser extent, to frontotemporal lobar degeneration, corticobasal degeneration and supranuclear palsy [186]. Notably, approximately 90% of AD cases exhibit comorbid CAA pathology [187] with 25% having severe vascular amyloidosis [188], indicating overlap between both pathologies [189]. While extracellular Aβ deposition promotes tau pathology, CAA vasculopathy is associated with vascular inflammation, bleeds and stroke risk [190]. The deposition of complement fragments in vascular amyloid deposits has been reported [191–193]. These vascular amyloid deposits have also been associated with accumulation of many inflammatory mediators [194], including CLU (clusterin; Apo J), a potent complement regulator found at significantly higher levels in CAA patients than in those with AD or healthy controls, based on laser-captured cortical tissue proteome analysis [194]. Recent proteomic studies identified multiple proteins involved in CAA pathology. Proteome data from the leptomeningeal and large cortical vessels isolated from cryopreserved postmortem brain specimens of CAA cases identified an enrichment for apoE and CLU in the vessels of CAA patients compared to non-CAA patients. Also, in silico analysis identified enrichment for classical complement and extracellular matrix remodelling pathways in the differentially expressed vascular proteome [195, 196]. Furthermore, a recent large scale proteomics study on capillary microvessels in patients with confirmed capillary CAA, AD and healthy controls showed distinct proteomic profile in CAA cases independent of AD, with microvascular proteome from CAA brain showing enrichment for several proteins as reported in above studies including accumulation of complement C1qb, C1qc, C3 and complement regulator CLU, indicating the role of complement cascade in CAA-associated pathology [197]. This could overall influence Aβ aggregation or CAA clearance through a peri-arterial drainage process. In another proteomic study, comparison between CAA(+) vessels with neighbouring CAA(-) vessels in patients suffering from mild cognitive impairment (MCI) and AD showed proteins that are associated with vascular matrix reorganization and BBB breakdown [198]. CAA proteomics meta-analysis from some of the above studies indicated vascular upregulation of proteins like C3 and CLU in CAA cases with plaque amyloid pathology compared to vascular proteome of plaque-positive AD patients without any vascular amyloid [199]. Additionally, immunohistochemical studies of AD brains revealed expression of complement proteins and activated complement protein fragments C1q, C3c, C4d, C5b-9 in plaques [154] as well as in CAA vessels [191–193]. Altogether, these discoveries redefined the role of complement in AD and CAA expanding its function from a participant in inflammation to a critical regulator of synaptic and vascular health. Complement modulation in AD mouse models Mouse models have emerged as a valuable tool in understanding the molecular mechanisms underlying the role of the complement cascade in AD. Thus, numerous investigations have leveraged AD mouse models to dissect the complex, context-dependent functions of complement. As mentioned previously, the discovery of the critical role of complement in mediating synaptic pruning during development [14], utilizing mouse models, constituted a cornerstone in our understanding of complement’s function in the pathogenesis of AD. This led scientists to hypothesise that this physiological mechanism could be aberrantly reactivated in neurodegenerative diseases, including AD. The importance of this hypothesis is reflected in the fact that there are no currently approved treatments for synaptic loss or neuronal death and dysfunction, which is the cause of cognitive loss in AD and the most impactful clinical presentation of this disease. Evidence from mouse models has strongly supported this hypothesis. Hong et al. observed elevated C1q levels in the brains of J20 AD model mice as early as one month of age [200], mirroring findings of C1q mRNA upregulation at two months in the 3xTg AD model [201]. Notably, this increase occurred prior to detectable amyloid plaque deposition [200]. Early synapse loss, preceding plaque formation, was documented in both J20 and APP/PS1 mice at 3–4 months of age, particularly within the CA1, CA3, and dentate gyrus regions [200]. This loss was associated with increased C1q and C3 deposition at synapses and was absent in APP/PS1 mice genetically deficient for C3. In the same study, injection of oligomeric Aβ into wild-type mice resulted in synapse loss and enhanced microglial engulfment of synaptic material within 72 h [200]. These effects were not observed in C1q-deficient mice, in mice co-injected with anti-C1q antibodies, or in those lacking the C3b/iC3b receptor CR3 [200]. Together, these findings established a direct mechanistic link between complement tagging and early synapse loss in AD. They also identified CR3, the microglial receptor for iC3b and to a lesser degree C3b, as a key effector of synaptic phagocytosis in AD. Others have explored the role of C1q in Alzheimer’s pathology using genetic deletion strategies. A foundational study by Fonseca et al. in APP transgenic mice, revealed that C1q deficiency significantly reduced microgliosis and neuritic damage, suggesting a detrimental role of the classical complement pathway in AD progression [202]. In Tg2576 mice, C1q deficiency reduced C4 deposition but unexpectedly increased astrocytic C3 expression and overall C3 deposition, likely via alternative pathway activation [203]. C1q-deficient mice showed reduced gliosis and neuronal loss despite elevated C3, suggesting that alternative pathway activation alone is insufficient to reproduce the detrimental effects associated with classical pathway activity [203]. In line with these findings, genetic deletion of C1q or C4 in PS2APP mice rescued synapse loss without altering cortical amyloid plaque burden, further implicating classical pathway activation as a direct driver of synaptic vulnerability [204]. In Arctic APP transgenic mice with microglia-specific C1qa deletion (via Cx3cr1-CreERT2), microglia were shown to be the primary source of brain C1q, while deletion reduced C1q levels and altered downstream complement activation [205]. In a tauopathy context, global C1q knockout in TauP301S (PS19) mice rescued synapse density by preventing both microglial and astrocytic engulfment of synapses [107]. More recently, Petrisko and colleagues reported that global C1q deficiency in Arctic and Tg2576 models altered gut microbiota composition, suggesting a potential role for C1q in mediating peripheral-central interactions during neurodegeneration [206]. In addition to genetic deletion, pharmacological inhibition of C1q has also been investigated. Administration of a C1q-blocking antibody in TauP301S tauopathy mice preserved synaptic density by preventing microglial-mediated synapse engulfment [181]. Furthermore, Wang et al. showed that antisense oligonucleotide (ASO)-mediated knockdown of C1r, C1s, or C4 protected synapses in PS2APP mice, demonstrating that targeted inhibition of the classical pathway can dose-dependently rescue synapse loss [204]. Beyond genetic and pharmacological targeting of C1q, recent studies have uncovered a diverse set of upstream and downstream mechanisms that govern its synaptic actions in AD mouse models. In APP/PS1 mice, inhibition of metabotropic glutamate receptor 1 (mGluR1) signaling mitigated microglial phagocytosis of the glutamatergic synapses and restored synaptic strength and cognition [207], while mGluR5 inhibition prevented C1q synaptic tagging and reversed synapse loss in both APP/PS1 and APPNL−G−F/hMAPT mice [208]. Other work has linked synaptic vulnerability to broader cellular and systemic changes. In APP/PS1 mice, synaptic mitochondrial dysfunction and septin accumulation promoted C1q-mediated pruning [209], while periodontal infection in APPNL−G−F mice increased microglial C1q expression and exacerbated synapse loss [210]. Importantly, several endogenous regulators of C1q have also been identified. In aged TauP301S mice, neuronal pentraxin 2 (Nptx2) overexpression ameliorated synapse loss, potentially via C1q inhibition [211]. In 5xFAD mice, triggering receptor expressed on myeloid cells 2 (TREM2) was shown to bind C1q and prevent excessive pruning, suggesting a microglial-mediated protective role of TREM2-C1q complex formation in the pathogenesis of AD [212]. Recent transcriptomic and proteomic analyses in APPNL−G−F mice further revealed that TREM2 deficiency reduces microglial expression of C1qa, C1qb, and C1qc, disrupts plaque engagement, and exacerbates dystrophic neurite formation, reinforcing TREM2’s regulatory role in complement-mediated synaptic vulnerability [213]. ApoE was also identified as a key endogenous regulator of classical complement activation through its ability to bind C1q and return complement signaling to homeostasis [182]. This interaction was reduced in mice expressing the human APOE ε4 allele, resulting in sustained complement activity [182]. Tyrosine kinase binding protein (TYROBP) deficiency markedly reduced C1q expression in both APP/PS1 and TauP301S mice, indicating shared upstream regulation across AD models [214, 215]. Moreover, in the FAD4T model, microglial upregulation of C1qA was associated with synapse loss and cognitive decline [216]. Finally, in 5xFAD mice, microglial CD2-associated protein (CD2AP) deficiency reduced C1q-mediated pruning [217], while physical exercise inhibited C1q-mediated microglial engulfment through downregulation of transmembrane protein 9 (Tmem9), ultimately improving cognitive function [218]. Extending beyond C1q-driven synaptic tagging, C3 and its downstream signaling components have emerged as central effectors of complement-mediated dysfunction in AD mouse models. Shi and colleagues reported a region- and age-dependent decline in hippocampal synaptic puncta density, dendritic spine density, neurons, long-term potentiation (LTP), and cognitive performance in wild-type mice, deficits that were markedly attenuated or absent in age-matched C3 knockout mice [219]. This was further supported by a second study by Shi et al., who demonstrated that C3 deficiency in APP/PS1 mice not only preserved synaptic integrity and cognitive performance in aged mice but also attenuated microgliosis, astrogliosis, and proinflammatory cytokine expression, despite ongoing amyloid pathology [220]. Wu et al. showed that C3 deletion protected against neurodegeneration in both PS2APP and TauP301S mouse models, establishing C3 as a central pathological driver [179]. In addition, Carpanini and colleagues reported elevated levels of C3b/iC3b activation fragments in the brains of aged APPNL−G−F and 3xTg mice [221]. Furthermore, C3aR deletion in TauP301S mice attenuated tau pathology and normalized immune gene networks [222], while C3aR knockout in APPNL−G−F mice enhanced microglial Aβ clearance and improved cognition [223]. Importantly, the findings in TauP301S mice were not replicated in a more recent study, which reported that C3aR deletion did not attenuate neurodegeneration or alter acute inflammation-induced gene expression changes in the brain [224]. Beyond neuronal pathology, endothelial C3aR has been shown to mediate vascular inflammation and BBB permeability during aging [225], as well as hippocampal pathology and cognitive impairment in VCID (vascular contributions to cognitive impairment and dementia) mouse models [226]. Pharmacological C3aR inhibition reduced tau pathology and improved cognition in TauP301S mice [227]. Nonetheless, significant caveats accompany these findings as the compound SB290157 has consistently been shown to display partial agonist activity, off-target effects, and poor pharmacokinetic properties, which limits its suitability as a tool for probing C3aR inhibition in vivo [228, 229]. Notably, overexpression of a soluble variant of the rodent-specific complement regulator Crry (complement receptor 1-related gene/protein y), which inhibits C3 activation by acting as a Factor I cofactor and by accelerating the decay of classical and alternative pathway C3 convertases, reduced hyperphosphorylated tau levels in aged P301L/sCrry double-transgenic mice [230]. However, it was also reported that local silencing of membrane-bound Crry by lentiviral injection in the cortex and hippocampi of TauP301S mice reduced neuroinflammatory cytokines and complement C3 components, leading to improved neuronal integrity and attenuated disease progression [231]. Taken together, these studies underscore that C3 and its downstream signaling components orchestrate key pathological processes beyond initiation, contributing to both neurodegeneration and impaired vascular integrity in AD. Several recent studies have highlighted “complementary” mechanisms that modulate complement activity and microglial responses in AD rodent models. For example, Factor H, known as a key regulator of the alternative pathway, has emerged as a promising therapeutic target, with evidence suggesting that enhancing its function may restrain complement overactivation in APP/PS1 mice [232]. At the level of microglial checkpoints, loss of SIRPα (Signal Regulatory Protein alpha), a ‘don’t eat me’ signal, led to exaggerated synaptic pruning in APP/PS1 mice [233]. Vascular dysfunction has also emerged as a contributor to complement-mediated pathology. Fibrinogen leakage into the brain parenchyma induces microglia-mediated dendritic spine elimination via CD11b (part of CR3), leading to synapse loss and cognitive impairment in 5XFAD mice [234]. Similarly, chronic cerebral hypoperfusion was shown to activate both coagulation and complement cascades in APP23 mice, linking vascular stress to innate immune activation [235]. Complement receptors have additionally been implicated in the clearance of pathological proteins: complement receptor 4 (CR4) was recently shown to mediate the uptake of extracellular tau fibrils by microglia, suggesting that complement receptor signaling may facilitate aggregate removal downstream of iC3b opsonization [236]. Such studies expand our understanding of how complement activity is fine-tuned in the AD brain, bridging upstream activation with cellular responses. The C5a–C5aR1 signaling axis has emerged as a critical driver of neuroinflammation and synaptic dysfunction in AD mouse models. Pharmacological inhibition of C5aR1 using the antagonist PMX205 has consistently demonstrated therapeutic benefit in transgenic AD models such as Tg2576 and Arctic48, where it reduced amyloid burden, gliosis, and cognitive decline [111, 237]. Similarly, active immunization against C5a improved memory and reduced inflammatory markers in Tg2576 mice [238]. In contrast, a study using intermittent administration of EP67, a conformationally biased analog of C5a (C5aR1 agonist) engineered to preferentially activate mononuclear phagocytes over neutrophils [239], reported enhanced microglial phagocytosis of Aβ and preserved memory in 5xFAD mice [240]. However, EP67 is not a selective C5aR1 agonist, as it can also activate C3aR and C5aR2, the latter of which has been associated with anti-inflammatory signaling [241]. Therefore, the beneficial effects observed in [240] may have resulted from this broader receptor activity or from the specific intermittent dosing schedule, rather than direct, sustained C5aR1 activation. Together, these considerations highlight that the outcomes of complement receptor agonism depend not only on dose but also on ligand specificity and dosing context [239, 242]. Supporting a pathogenic role for sustained C5a–C5aR1 signaling, Carvalho and colleagues showed that C5a overexpression in Arctic mice accelerated cognitive decline and inflammatory gene expression, whereas C5aR1 deletion delayed these changes [243]. Additional studies demonstrated that C5aR1 antagonism promotes a neuroprotective microglial phenotype [244], dampens inflammatory glial signaling [111], and protects against region- and age-dependent synaptic loss in Tg2576 and Arctic48 models [245]. Notably, C5aR1 deletion has also been associated with age-dependent shifts in gut microbiota composition in both wild-type and AD mouse models, suggesting a potential role for peripheral immune signaling in broader neuroimmune dynamics [206]. Collectively, these findings establish C5aR1 as a critical effector of complement-mediated neurotoxicity and a promising target for disease modification in AD. Mounting evidence implicates the terminal complement pathway, and in particular, MAC formation, as a critical effector of neurodegeneration in AD. Genetic deletion of CD59a, the endogenous inhibitor of MAC, in tauopathy models exacerbated tau pathology and neuronal loss, underscoring the detrimental effects of unrestrained MAC activation [230]. In 3xTg-AD mice, genetic deletion of C6 significantly reduced synaptic loss, supporting a pathogenic role for MAC formation in synapse degeneration [221]. In the APPNL−G−F model, pharmacological inhibition of the terminal pathway using an anti-C7 antibody preserved dendritic spine density near amyloid plaques, further implicating MAC in amyloid-associated synaptic vulnerability [221]. More recent studies by Zelek et al. showed that both genetic deletion of C7 and systemic administration of an anti-C7 antibody in APPNL−G−F mice led to reduced complement activation, decreased amyloid burden, preserved synaptic density, and improved cognitive function, demonstrating that targeting the terminal pathway can mitigate multiple AD-related pathologies [246]. Moreover, a brain-penetrant anti-C7 antibody was reported to cross the BBB, inhibit MAC formation, and protect against neurodegeneration in APPNL−G−F mice [247]. Extending the relevance of MAC activation beyond parenchymal pathology, Hu et al. demonstrated that macrophage-derived migrasomes (large extracellular vesicles formed during cell migration) accumulate along cerebral vessels and induce complement-dependent cytotoxicity and BBB disruption in a model of CAA (Tg-SwDI/B mice), highlighting the role of terminal complement activation in cerebrovascular injury [248]. Together, the data support terminal complement inhibition as a promising strategy to counteract synaptic, cognitive, and vascular decline in AD. Recent advances in spatial transcriptomics have provided a powerful new perspective on how complement activation unfolds in a neuroanatomical context. In APPNL−G−F mice, Chen et al. used in situ sequencing to identify a plaque-induced gene (PIG) network enriched for complement components, including C1q and C4, that were specifically upregulated in microenvironments surrounding amyloid plaques [249]. This spatially restricted expression pattern was also observed in human AD brain tissue, reinforcing the concept that complement activation in AD is not diffuse but instead confined to local inflammatory niches [249]. These findings underscore the importance of spatial resolution in dissecting complement-mediated neuroinflammation and synaptic vulnerability. While the studies discussed above emphasize the crucial roles of complement activation in AD, it is important to recognize that these effects can vary depending on the disease model. Fonseca and colleagues investigated the impact of C1q, C3, and C5 deficiency in 3xTg-AD mice and found that the contribution of individual complement components to amyloid pathology and neuroinflammation differed in other transgenic models (Tg2576 and Arc48) [250]. Similarly, Maier et al., found detrimental effects of germline C3 deficiency in J20 APP transgenic mice [251], whereas Shi et al. found neuroprotective effects of germline C3 deficiency in APP/PS1 mice [220]. These findings highlight that the consequences of complement modulation are not universally consistent and may be influenced by factors such as the underlying transgenic background, disease stage, and pathological burden. This variability reinforces the importance of context when interpreting nonclinical data and suggests that complement-targeted therapies should be tailored to model- and disease stage-specific dynamics. Nevertheless, a mega-analysis of transcriptomic data across 10 AD mouse studies identified C1qa, C1qb, and C1qc as among the most consistently upregulated genes during the early disease stages [252]. These transcriptional changes preceded major pathological shifts and were not solely attributable to gliosis, suggesting that classical complement activation may represent a common and early hallmark of AD pathogenesis across models [252]. Proteomic and lipidomic analyses in APP/PS1 mice further supported this, revealing early upregulation of complement and coagulation components, consistent with complement activation as an early molecular event in AD [253]. Furthermore, computational analyses have reinforced the cross-species relevance of complement involvement in AD. Lee et al. identified conserved molecular signatures between AD mouse models and human brain tissue, including complement cascade activation, TYROBP-associated innate immune responses, and Tyro3/Axl/MerTK (TAM) receptor agonists, distinct from changes attributable to aging [254]. Varma et al. found age- and sex-specific changes in CSF complement proteins (C1qa, C3, C9, Serping1, CFI) in Caribbean vervet non-human primates that develop cerebral amyloidosis with aging, noting elevations in males and reductions in females [255]. At the cellular level, complement-dependent signaling between microglia and astrocytes plays a pivotal role in shaping Aβ dynamics and driving AD progression. This bidirectional communication is increasingly recognized as a core component of AD-related neuroinflammation. The role of microglia as complement-dependent phagocytes in the diseased brain marked a major advance in our understanding of AD pathophysiology. Nonetheless, their role expands beyond phagocytosis. As AD pathology progresses, microglia produce pro-inflammatory cytokines, including interleukin 1 alpha (IL-1α), tumor necrosis factor alpha (TNF-α), and C1q, which act on nearby astrocytes to induce a neurotoxic phenotype [256]. These neurotoxic, reactive astrocytes are unable to promote neuroprotective functions, synaptogenesis, and phagocytosis, and lead to neuronal and oligodendrocyte death [256]. Moreover, neuronal Aβ activates nuclear factor-κΒ (NF-κB) signaling in astrocytes, leading to the extracellular release of complement C3a [257, 258]. Released C3a binds to C3aR expressed on microglia and neurons, amplifying Aβ pathology through a ‘feedforward’ loop. Pharmacological blockade of this signaling cascade using a C3aR antagonist attenuates microglial activation and reduces Aβ accumulation. These findings highlight the NF-κB/C3/C3aR axis as a key mediator of astrocyte-microglia crosstalk in the context of Aβ-driven pathology [257, 258]. Of particular note, C1q has also been suggested to act as an AD-specific modulator of the cellular crosstalk between microglia and astrocytes [259], as well as a potential regulator of astrocyte-mediated Aβ clearance [260]. Adding to this multicellular complement-signaling axis, recent work has shown that perivascular macrophages can induce a complement-expressing, phagocytic state in microglia via secretion of SPP1 (secreted phosphoprotein 1/osteopontin) [261]. In APPNL−F mice, SPP1-driven upregulation of C1qa in microglia led to enhanced synaptic engulfment, while genetic ablation of Spp1 prevented synapse loss despite ongoing amyloid deposition [261]. Emerging cellular models of complement in AD Translating insights from animal models to human disease remains a key challenge in AD and complement research. Human iPSC-derived models, particularly when paired with Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-based gene editing and co-culture approaches, now provide a platform to investigate complement-driven processes such as synaptic pruning and neuroinflammation. Although still limited in maturity and complexity, these systems represent a promising tool to bridge complement findings from animal models to human disease. Protocols for generating microglia-like cells from human iPSCs (iMGLs) now allow for functional studies of innate immune signaling in a human context [262]. These iMGLs express complement-related genes and have been used to investigate AD risk variants such as TREM2, which modulates microglial activation, and may shape complement responses [263]. TREM2 variants have been shown to alter the functional properties of iPSC-derived microglia, including their response to chemotactic cues such as complement component C5a [264]. Importantly, CR1 expression has been demonstrated in iPSC-derived microglia, providing a platform to connect human complement genetics with cellular mechanisms relevant to AD [177]. In parallel, iPSC-derived astrocytes have been used to model inflammatory activation and neurotoxicity, with reactive astrocytes serving as a major source of C3 and potentially contributing to synapse elimination [265]. Collectively, these studies highlight how stem cell–derived systems can be leveraged to probe complement mechanisms relevant to AD. CRISPR-based strategies are emerging as powerful approaches to dissect complement mechanisms in human cellular models. For example, a CRISPR interference and activation (CRISPRi/a) platform in iPSC-derived microglia was used to identify regulators of inflammatory activation, survival, and phagocytosis, processes that intersect with complement signaling in neurodegeneration [266]. In parallel, CRISPR/Cas9 editing has been employed to generate human cell lines with targeted knockouts of complement regulatory proteins. Thielen and colleagues established human cell lines lacking CD46, CD55, and CD59 (Fig. 1), providing robust tools to study complement activation and regulation directly in a human context [267]. These innovative approaches highlight how gene editing can be applied to uncover complement-specific pathways and identify potential therapeutic targets in AD. Beyond monocultures, co-culture and organoid systems offer more physiologically relevant platforms to study complement across interacting human brain cell types. For instance, cellular specificity within the complement system could be explored by selectively manipulating complement genes in defined cell types using CRISPR, and examining the resulting effects within co-culture systems. Notably, a tri-culture model comprising iPSC-derived neurons, astrocytes, and microglia demonstrated that microglia-astrocyte communication induces robust astrocytic C3 expression, especially in the presence of APP mutations, modeling complement-driven inflammatory signaling in an AD-relevant context [259]. In three-dimensional systems, brain organoids containing innately developed microglia responded more effectively to β-amyloid challenge than those with externally added microglia, suggesting that microglial ontogeny may influence complement-related responses [268]. In addition, long-term adhesion organoids with integrated iPSC-derived microglia showed reduced tau pathology, preserved synaptic density, and neuronal protection, underscoring the potential of these systems to recapitulate complement-associated neurodegeneration [269]. These models broaden the capacity of iPSC-based systems to investigate complement signaling within complex multicellular environments relevant to AD. While iPSC-derived systems provide powerful tools to study complement in a human context, they face important limitations, including developmental immaturity, lack of vasculature and aging, and limited capacity to model chronic, progressive inflammation [270, 271]. Nevertheless, emerging approaches based on direct differentiation (transdifferentiation or direct reprogramming) are beginning to address the limitation of developmental immaturity observed in iPSC-derived systems [272, 273]. Differences in complement gene expression between iPSC-derived cells and adult human brain tissue further constrain translational interpretation, particularly in the context of aging and AD [274]. Even so, these systems remain uniquely genetically tractable and, when combined with genome-editing and co-culture approaches, are poised to yield critical insights into complement dysregulation and to accelerate therapeutic discovery in AD. Non-canonical roles of complement Beyond its well-established extracellular functions, complement activity has also been detected inside cells [60, 275]. Complement proteins C3 and C5 have been found intracellularly where they regulate fundamental processes including metabolic reprogramming [57, 61, 67, 276], mitochondrial ROS production and stress signaling [275, 277], as well as autophagy regulation [278]. For example, intracellular C3 sustains the mechanistic target of rapamycin (mTOR)-dependent metabolic pathways in T cells [57], while C3 and C5 signaling within subcellular compartments cross-talks with innate sensors such as the NLRP3 inflammasome and the mitochondrial antiviral signaling protein (MAVS), linking complement activity to mitochondrial stress responses and inflammasome activation [275, 277]. Although the existence of a functional intracellular convertase remains unproven and is biochemically implausible under endosomal conditions, these findings suggest that complement proteins can exert non-canonical, cell-intrinsic roles independent of the traditional cascade [279]. Within the CNS, complement proteins have also been implicated in non-canonical intracellular and synaptic functions. In particular, C1q and C3 have been suggested to modulate synaptic communication and neurotransmission (reviewed in [280]). Additionally, microglia-derived C1q has been detected within neuronal ribosomes in the aging mouse brain, raising the possibility of intracellular complement activity relevant to neuronal homeostasis [281]. Together, these findings point to a growing body of evidence that complement functions extend beyond classical immunity to encompass synaptic modulation, proteostasis, and cellular stress responses. Although direct evidence in AD models is limited, non-canonical complement functions could plausibly intersect with pathways central to neurodegeneration, including energy metabolism, mitochondrial dysfunction, inflammasome activation, and neuronal stress responses. As tools for high-resolution subcellular profiling and genetic manipulation advance, systematic investigation of these alternative roles may uncover novel mechanisms linking innate immunity to AD pathology. Decades of research have reshaped our view of complement in AD from an amyloid-associated bystander to a central regulator of synapse loss, inflammation, and vascular dysfunction. Animal models and human-iPSC systems have uncovered diverse, context-dependent roles for key components such as C1q, C3, and C5aR1, while emerging work on intracellular complement points to additional, underexplored mechanisms. These findings establish complement as a dynamic contributor to AD pathogenesis. The next section explores how genetic and epigenetic factors shape this system and modulate disease risk. Genetic & epigenetic implications of complement in AD Genetic associations between complement and AD risk For over 16 years (1993–2009), APOE remained the only major genetic risk factor identified in sporadic AD. Numerous studies consistently confirmed that the ɛ4 allele increases AD risk, while the ɛ2 allele appears to confer a protective effect [282, 283]. Early genome-wide association studies (GWAS) in AD, conducted between 2007 and 2009, were constrained by limited cohort sizes, often fewer than 1,000 individuals per group, which reduced their ability to detect risk variants for such a genetically complex disorder [284–289]. This changed in 2009, when two landmark GWAS studies identified novel susceptibility loci for late-onset AD (LOAD), for the first time since the discovery of APOE. Lambert and colleagues reported genome-wide significant associations for CLU and CR1 [9], while Harold et al. independently confirmed the CLU signal and identified PICALM (phosphatidylinositol binding clathrin assembly protein) as an additional risk gene [7]. Unlike rare mutations in APP, PSEN1, or PSEN2 that cause early-onset familial AD, the variants identified in these studies contribute to risk for the more common, polygenic form of late-onset, sporadic AD. The strength of GWAS lies in their ability to scan hundreds of thousands of variants in an unbiased, hypothesis-free manner. Therefore, the fact that both studies independently identified CLU in a genome-wide context, further reinforced its association with AD [290]. These findings demonstrated that sufficiently powered GWAS could uncover biologically meaningful variants, many of which implicate immune pathways, endocytic trafficking, and lipid metabolism in AD pathogenesis. The CLU gene encodes clusterin which is also known as apolipoprotein J (ApoJ) [291], and has long been regarded as a risk gene in AD, leading to the accumulation of substantial functional evidence over the years (Table 1). For instance, in 2004, DeMattos and colleagues observed that ApoE and clusterin cooperatively regulate extracellular Aβ metabolism in vivo [292]. Genetic deletion of either protein led to increased Aβ accumulation, while combined deletion resulted in an even greater burden, supporting their complementary roles in Aβ clearance and deposition suppression [292]. It is now known that CLU, a ~ 75 kDa extracellular chaperone protein, regulates the complement system by inhibiting MAC formation and facilitating the clearance of cellular debris and misfolded proteins, including Aβ [293, 294]. It is highly expressed in astrocytes and upregulated in the AD brain, where it may exert both protective and potentially detrimental effects, depending on the disease state and molecular context [294–296]. CLU has been shown to reduce Aβ aggregation, gliosis, and neurotoxicity, and to support synaptic integrity by modulating astrocyte and microglial activity [295, 296]. Notably, recent work also demonstrated that clusterin, while protective against Aβ aggregation, can paradoxically enhance tau seeding and propagation in vivo [297]. Earlier studies further suggest that CLU may stabilize toxic Aβ oligomers, promote plaque formation, or exacerbate neuritic damage under certain pathological conditions [298, 299]. Although the precise mechanisms by which CLU variants influence AD risk remain unresolved, studies suggest that genetic variation near the CLU locus affects both its expression and alternative splicing, potentially modifying its neuroprotective functions and complement-modulating capacity [300, 301]. The CR1 gene encodes a membrane-bound receptor that primarily binds C3b and C4b to mediate immune complex clearance [8], and has consistently emerged as a significant genetic risk factor in AD (Table 1). GWAS and meta-analyses have identified robust associations between LOAD and single nucleotide polymorphisms (SNPs) within CR1, particularly rs6656401 and rs3818361, across European and Chinese populations [9, 302–307]. In Han Chinese individuals, additional missense variants have been linked to increased AD risk, underscoring ethnic-specific genetic susceptibility [308]. Structural polymorphisms, including intragenic duplications that generate the CR1-S isoform, increase C3b/C4b binding sites but may impair Aβ clearance due to altered receptor trafficking and reduced peripheral clearance, especially via erythrocytes [8, 309–311]. Although minimally expressed under normal conditions, CR1 is upregulated in AD microglia and astrocytes and has been linked to greater amyloid burden, CAA, cognitive decline, and depression [177, 312–315]. Recent functional work further implicates CR1 in the regulation of microglial phagocytosis [178]. CR1 also appears to interact with APOE-ε4 to accelerate cognitive decline [316]. Notably, the rs3818361 CR1 variant has been shown to modulate the relationship between plasma ApoE levels and cortical amyloid deposition, indicating that CR1 genotype may influence amyloid accumulation through ApoE-dependent mechanisms [317]. Functional assays have shown that AD-associated CR1 variants modestly increase C3b and C1q binding, suggesting altered complement clearance dynamics in risk carriers [315]. Altogether, these studies suggest that CR1 contributes to AD pathogenesis through combined effects on immune complex handling, amyloid processing, and neuroinflammatory responses. While CR1 and CLU represent the most robust and consistently replicated complement-related genetic risk loci in AD, several additional complement genes have been suggested by candidate gene studies, rare variant analyses, or pathway-level enrichment analyses. These include C1S, a serine protease of the classical pathway [204], and complement Factor H (CFH), a regulator of the alternative pathway more strongly associated with age-related macular degeneration but functionally relevant to neuroinflammation [318–321] (Table 1). Another interesting candidate is CSMD1, encoding a transmembrane protein that inhibits classical complement activation and is highly expressed in the brain, where it may regulate synaptic refinement and modulate neuroimmune responses [112, 322, 323]. Though associations involving these genes have not consistently reached genome-wide significance or have shown limited replication, they highlight the broader involvement of the complement cascade in AD susceptibility (Table 1). In summary, common variants in complement genes, particularly CR1 and CLU, act as risk modifiers in sporadic AD, showcasing complement’s relevance in the development of LOAD. Beyond the genome: epigenetic mechanisms governing complement in AD Genetic variation alone cannot fully account for complement dysregulation in AD. Over the past 20 years, increasing evidence points to epigenetic regulation as a key mechanism shaping complement gene expression in the aging and diseased brain [324, 325]. Processes such as DNA methylation, histone modifications, chromatin remodeling, and non-coding RNA activity dynamically influence gene accessibility and transcription in response to intrinsic and extrinsic cues. These layers of regulation may unravel how complement activity varies by age, sex, brain region, or inflammatory state, and how it contributes to disease progression even in the absence of strong genetic drivers. This subsection examines the emerging evidence linking epigenetic mechanisms to complement pathway dysregulation in AD. DNA methylation is a fundamental epigenetic mechanism that involves the addition of methyl groups to cytosine residues predominantly occurring in cytosine-phosphate‐guanine (CpG) dinucleotides [326]. It plays a critical role in regulating gene expression across development, aging, and disease. Over the past decade, major technological advances have enabled the first epigenome‐wide association studies (EWAS) of DNA methylation in AD brain tissue [327, 328]. Such studies have reported altered methylation at complement-related loci, suggesting that genes such as CLU and CR1 may be epigenetically modulated in AD (Table 1). For instance, hypomethylation at CpG island shores (regions flanking CpG-rich promoter sequences that are particularly susceptible to age- and disease-related methylation changes) has been observed at CLU, CR1 and PICALM in the blood of Japanese AD patients [329]. In the same study, CLU methylation levels, in combination with APOE genotype, could distinguish AD patients from cognitively healthy elderly individuals with high sensitivity and specificity [329]. Thus, methylation differences at the CpG island shores of AD-associated genes may have diagnostic value in AD, however, further studies are required to exclude ethnic-specificity. In postmortem AD brain tissue, DNA methylation at specific CpG sites within CLU and CR1 has also been associated with neuritic plaque burden, independently of genetic variation [330]. Importantly, at the CR1 locus, the direction of this association was modified by rs6656401 genotype, highlighting a gene-epigenome interaction that may influence complement-related pathology in AD [330]. Additional evidence suggests that CLU methylation may also relate to cognitive function independently of disease status. In a cohort of healthy and schizophrenia subjects, Alfimova et al. found that DNA methylation within a regulatory region of the CLU gene predicted performance on episodic verbal memory tasks, suggesting that epigenetic variation at this AD-associated locus may influence cognition beyond genetic background [331]. Consistent with this, a recent multi-omics Mendelian randomization study identified CLU as an AD-causal gene whose brain expression is influenced by DNA methylation in the context of mitochondrial dysfunction and inflammation, further supporting its epigenetic sensitivity in AD [332]. Supporting its epigenetic plasticity, the CLU promoter contains several binding sites for stress-responsive transcription factors, including AP-1 (activator protein 1) and HSF (heat-shock factor) [333], and lies within a CpG-rich domain susceptible to methylation [334–336]. Epigenetic regulation of CLU has also been observed in multiple cell types, including prostate cancer cells, neural cells, and retinal pigment epithelial cells [334–336], suggesting that CLU expression is epigenetically modulated by factors such as aging, stress, and inflammation, known contributors to AD pathology. Further supporting this, Gasparoni et al. performed a cell type-specific DNA methylation analysis in purified human neurons and glia, identifying differential CLU methylation in neurons, reinforcing the idea that complement gene regulation in AD may differ across cell types [337]. Methylation patterns may also differ significantly by sex. Recent analyses have revealed sex-specific methylation changes in AD brains, potentially affecting immune and complement-related genes [338]. In particular, Zhang and colleagues observed that in male samples, differentially methylated genes were significantly enriched for complement system-related processes, suggesting sex-dependent epigenetic regulation of complement activity [338]. Collectively, these findings support DNA methylation as a mechanism linking complement risk loci to disease-specific gene expression patterns in AD. In addition to DNA methylation, histone modifications are central to epigenetic regulation, influencing gene expression by altering chromatin structure and transcriptional accessibility. Post-translational modifications such as histone acetylation (e.g., H3K27ac) and methylation (e.g., H3K27me3) act as dynamic markers of gene activity or repression [339, 340]. In AD, aberrant patterns of histone modifications have been observed, particularly in glial cells, and are believed to contribute to neuroinflammation and immune dysregulation [341]. Direct evidence of histone-mediated regulation of complement genes in AD is limited but growing. Notably, a histone acetylome-wide association study by Marzi and colleagues identified differential H3K27ac enrichment in the entorhinal cortex of AD patients, including at the CR1 locus [342]. These acetylation changes were associated with altered gene expression, supporting a role for histone modifications in modulating complement activity in AD. CLU has also been shown to be responsive to chromatin state. In retinal pigment epithelial cells, histone deacetylase (HDAC) inhibition upregulated CLU expression, indicating that increased histone acetylation can promote CLU transcription [336]. Similarly, Nuutinen et al. demonstrated that treatment with various HDAC inhibitors induced CLU mRNA and protein expression in neural cells, and that this effect was further enhanced by co-treatment with a DNA methylation inhibitor, highlighting a synergistic relationship between histone acetylation and DNA methylation in the transcriptional regulation of CLU [335]. More broadly, enhancer remodeling has been implicated in immune gene regulation in AD. Gjoneska et al. reported conserved enhancer signatures marked by H3K27ac and H3K4me1 in microglia from mouse models of neurodegeneration, with overlapping patterns in human AD cortex [341]. These enhancer landscapes were enriched for immune and phagocytic genes, providing indirect support for complement-associated regulation [341]. Extending this work, Nott et al. demonstrated elevated H3K27ac at microglia-specific enhancers in AD brain tissue [343]. Transcriptional regulation plays a central role in the context-dependent expression of complement components in the brain, influencing the epigenetic signature. Transcription factors such as PU.1 (SPI1) and IRF8 are essential for microglial development and immune-related gene expression [344]. While complement genes like C1qa are highly expressed in microglia, and PU.1 is known to shape the microglial enhancer landscape [345], direct transcriptional regulation of specific complement components remains to be fully elucidated. Nonetheless, emerging in vivo evidence suggests that PU.1 modulation alters AD-related phenotypes and immune transcriptomes, including complement pathways, further supporting its role in disease-context regulation [346]. MicroRNAs (miRNAs), a class of non-coding RNAs, are considered epigenetic regulators owing to their ability to post-transcriptionally silence gene expression without altering the underlying DNA sequence [347]. They are increasingly recognized as modulators of immune and complement-related pathways in AD. Among them, miR-146a is one of the most consistently upregulated in AD brain and cellular models, acting in an NF-κB-dependent manner to suppress inflammatory signaling and directly target CFH [348–352] (Table 1). This suppression of CFH appears particularly prominent in microglia, where miRNA-146a expression is robustly induced by Aβ42 and TNF-α exposure, as shown in primary human neural and glial cell cultures [353]. In the hippocampal CA1 region of sporadic AD cases, miR-146a and miR-34a were both significantly upregulated and have been shown to downregulate a number of target mRNAs involved in immune and synaptic regulation, including CFH, SHANK3, and TREM2, further implicating miRNA dysregulation in both complement activation and synaptic vulnerability in AD [354]. miR-155, a pro-inflammatory miRNA also implicated in neurodegeneration, has been associated with decreased CFH expression and altered glial activation in both AD and Down syndrome brain tissues [355, 356]. Other miRNAs, such as miR-132 and miR-124, have been linked to synaptic regulation and complement-mediated neuroinflammation, including reported interactions with C1q family members and related signaling pathways [357–359]. A recent systematic review and meta-analysis further confirmed the dysregulation of miRNAs including miR-146a, miR-132, and miR-155 across multiple AD brain regions, strengthening their relevance as epigenetic regulators and candidate biomarkers [360]. Recent network-based and pathway analyses have identified additional miRNAs with potential roles in AD-related neuroinflammation and immune regulation, though direct links to complement components remain to be clarified [361]. Broad alterations in miRNA expression profiles have been reported in both AD brain tissue and CSF, further supporting their relevance as upstream regulators and potential biomarkers [362]. Extracellular vesicles may further contribute to miRNA-based complement regulation by facilitating intercellular RNA transfer, including between astrocytes and neurons [363, 364]. These vesicles, particularly exosomes, not only mediate neuroinflammatory signaling but are also emerging as potential therapeutic vehicles and biomarkers in AD [365]. Although less characterized, long non-coding RNAs (lncRNAs) and circular RNAs (circRNAs) are also emerging as potential regulators of neuroimmune signaling and complement gene expression in the aging brain [366]. Intriguingly, studies have also suggested a potential reverse signaling effect, whereby complement proteins such as C1q may influence miRNA expression in neurons, hinting at complex bidirectional communication between complement and non-coding RNA networks [16]. These findings suggest that non-coding RNAs form an epigenetic regulatory layer that may fine-tune complement activity in AD in a context- and cell type–specific manner. Integrative models and environmental triggers in complement regulation Complement activation in AD reflects the convergence of genetic predisposition, epigenetic modulation, and environmental influences. As discussed above, while GWAS-identified variants in complement-related genes such as CR1 and CLU establish foundational susceptibility, these effects are dynamically shaped by epigenetic mechanisms including DNA methylation, histone modifications, and non-coding RNAs. Environmental exposures, such as viral infection, chronic stress, and aging, may trigger or amplify epigenetic remodeling, thereby shifting glial activation thresholds and altering complement expression profiles. For instance, Licastro and colleagues proposed that AD-associated genetic risk loci, including CR1 and CLU, may represent components of an immune gene network that influences the brain’s ability to respond to neurotropic viral infections, particularly herpesviruses, suggesting that viral exposure may act as an environmental trigger interacting with complement-related genetic susceptibility [367]. Recent multi-omics studies have further highlighted how mitochondrial function, epigenetic modifications, and inflammation converge in AD pathogenesis. Zhang et al. employed Mendelian randomization to identify mitochondrial dysfunction-related genes such as CLU, MAPT (Microtubule Associated Protein Tau), and ACE (Angiotensin I Converting Enzyme), whose expression levels are influenced by DNA methylation and are genetically linked to inflammatory cytokines, including IL-17 C and IL-18 [332]. These findings suggest that epigenetic regulation of mitochondrial and nuclear gene expression may jointly influence the inflammatory environment in AD, potentially interacting with complement and immune signaling cascades. In support of this, proteomic analyses have also implicated complement dysregulation in AD progression. Bai et al. identified upregulation of complement components in postmortem brain tissue, with supporting evidence from the 5xFAD mouse model [368]. Sex-based differences in neuroinflammatory processes have also emerged as key modulators of complement activity in AD. Chromosomal and hormonal factors, such as X chromosome inactivation and sex hormone–dependent epigenetic modulation, can shape immune signaling trajectories [369]. Consistent with this, Casali et al. showed that both sex chromosomes and gonadal hormones independently modulate microglial activation and plaque burden in 5xFAD mice, underscoring the complex genetic and hormonal influences on neuroinflammation [370]. Recent single-cell transcriptomic analysis in the THY-Tau22 mouse model revealed sex-dependent dysregulation of complement genes such as C1qa, C1qb, and C1qc, with more pronounced alterations in females [371]. Supporting this, a longitudinal study in 3xTg-AD mice reported male-specific upregulation of complement genes, including C1qa-c and C5ar1, mirroring findings in postmortem male AD brains [372]. These differences may underlie observed disparities in disease prevalence, progression, and response to treatment between sexes. In addition to genetic and hormonal factors, lifestyle-related environmental influences such as diet and exercise have been proposed to affect brain health via epigenetic mechanisms. Nutrients like polyphenols, B vitamins, and omega 3 fatty acids have been shown to modulate DNA methylation and histone modifications, potentially impacting neuroimmune signaling and glial responses [373]. Although their effects on complement regulation remain to be clarified, these findings support the broader view that environmental exposures can shape the epigenetic landscape in AD. Diet may also directly impact complement-mediated neuroinflammation. A recent study demonstrated that even a short-term high-fat diet impaired memory and enhanced synaptic degradation in 3xTg-AD mice via a complement-dependent mechanism, highlighting the rapid sensitivity of complement pathways to environmental triggers [374]. Similarly, Poxleitner and colleagues found that a Western diet disrupted glucose and fatty acid metabolism and induced adaptive immune responses in the brain of APP/PS1 mice, emphasizing the role of diet in modulating brain homeostasis [375]. Epidemiological studies have also supported gene-diet interactions in AD risk, with evidence suggesting that protective effects of omega 3 fatty acid intake may be modulated by APOE genotype and complement-related genes such as CLU and CR1 [376]. Beyond dietary influences, physical exercise appears to exert protective effects against complement-mediated neuroinflammation. Remarkably, plasma from exercising mice was shown to reduce neuroinflammation and improve cognition in AD models, an effect linked to increased levels of CLU [377]. Yang et al. reported that aerobic exercise regulates the GPR81 signaling pathway (a lactate-sensitive GPCR pathway) and maintains complement-microglia axis homeostasis, promoting synaptic protection in early-stage AD [378]. Li et al. demonstrated that physical exercise decreases complement-mediated synaptic loss and protects against cognitive impairment by inhibiting microglial Tmem9-ATP6V0D1 signaling [218]. Moreover, combining exercise with postbiotic treatment resulted in synergistic improvements in mitochondrial function and reduced AD-related gene expression, suggesting a multifaceted approach to mitigating neuroinflammation [379]. Together, these studies reinforce the need for integrative models of complement dysregulation that extend beyond static genetic risk to include dynamic, context-sensitive modifiers. Multi-omic approaches combining genetics, epigenomics, transcriptomics, proteomics, and exposome data will be critical for unraveling the complex regulatory networks underlying complement activity in the aging and diseased brain. Understanding how factors such as age, sex, lifestyle, and systemic inflammation converge on the regulation of key complement genes, including CR1, CLU, and others, will be essential to define risk trajectories and identify precision-targeted interventions. As research moves forward, emphasis should be placed on cell type-specific regulatory dynamics and longitudinal modeling to fully capture the evolving landscape of complement biology in AD. Table 1Genetic and epigenetic regulation of complement in ADGeneType of RegulationMechanismProposed Role in ADCNS ExpressionCLU [291, 333, 335](Clusterin or ApoJ)Genetic, Epigenetic,Environmental Factors [7, 9, 290, 329–332, 337, 367, 376, 377]GWAS risk locus [7,9,290]; methylation [329–332, 337]Chaperone for Aβ [292]; regulator of MAC formation [293, 294]; modulates glial inflammation and BBB function [294–296]; context-dependent neuroprotection [294–301]Astrocytes (primary), neurons, microglia, endothelial cells [294] CR1 (Complement Receptor 1)Genetic, Epigenetic, Environmental Factors [8, 9, 302–311, 317, 329, 330, 367, 376]GWAS risk locus (multiple SNPs and structural variants identified) [8, 9, 302–311, 316, 317]; methylation [329, 330]; histone acetylation [342]Mediates immune complex clearance via C3b/C4b binding [8, 309, 311]; modulates amyloid burden and neuroinflammation [312, 313]; regulates microglial phagocytosis [178]; associated with depression [314]Minimal in CNS parenchyma, primarily expressed in peripheral immune cells [177, 310, 311, 315] CFH (Complement Factor H)Genetic, Epigenetic [319–321, 348–351, 353, 354, 356]Suggestive genetic association [319–321]; suppression by miR-125b, miR-146a, miR-155 [348–351, 353, 354, 356]Downregulated by miRNAs [348–351, 353, 354, 356]; impaired regulation exacerbates complement-mediated neuroinflammation [349, 353, 356]Microglia (primary), astrocytes (moderate, inducible in AD), neurons (low) [349, 350] C1S (Complement component 1 subcomponent S)Limited genetic evidence [204]Classical pathway serine proteaseMediator of classical pathway activation; AD relevance under investigationLow in brain, primarily expressed in liver Table 1Genetic and epigenetic regulation of complement in ADGeneType of RegulationMechanismProposed Role in ADCNS ExpressionCLU [291, 333, 335](Clusterin or ApoJ)Genetic, Epigenetic,Environmental Factors [7, 9, 290, 329–332, 337, 367, 376, 377]GWAS risk locus [7,9,290]; methylation [329–332, 337]Chaperone for Aβ [292]; regulator of MAC formation [293, 294]; modulates glial inflammation and BBB function [294–296]; context-dependent neuroprotection [294–301]Astrocytes (primary), neurons, microglia, endothelial cells [294] CR1 (Complement Receptor 1)Genetic, Epigenetic, Environmental Factors [8, 9, 302–311, 317, 329, 330, 367, 376]GWAS risk locus (multiple SNPs and structural variants identified) [8, 9, 302–311, 316, 317]; methylation [329, 330]; histone acetylation [342]Mediates immune complex clearance via C3b/C4b binding [8, 309, 311]; modulates amyloid burden and neuroinflammation [312, 313]; regulates microglial phagocytosis [178]; associated with depression [314]Minimal in CNS parenchyma, primarily expressed in peripheral immune cells [177, 310, 311, 315] CFH (Complement Factor H)Genetic, Epigenetic [319–321, 348–351, 353, 354, 356]Suggestive genetic association [319–321]; suppression by miR-125b, miR-146a, miR-155 [348–351, 353, 354, 356]Downregulated by miRNAs [348–351, 353, 354, 356]; impaired regulation exacerbates complement-mediated neuroinflammation [349, 353, 356]Microglia (primary), astrocytes (moderate, inducible in AD), neurons (low) [349, 350] C1S (Complement component 1 subcomponent S)Limited genetic evidence [204]Classical pathway serine proteaseMediator of classical pathway activation; AD relevance under investigationLow in brain, primarily expressed in liver Genetic and epigenetic regulation of complement in AD CLU [291, 333, 335] (Clusterin or ApoJ) Genetic, Epigenetic, Environmental Factors [7, 9, 290, 329–332, 337, 367, 376, 377] CR1 (Complement Receptor 1) CFH (Complement Factor H) C1S (Complement component 1 subcomponent S) Gut-complement link in AD neuroinflammation The gut microbiome is a complex ecology composed of billions of microorganisms, predominantly bacteria, that live in the intestinal lining [380]. Emerging evidence suggests that the microbiome in the intestinal lining plays a major role in modulating the immune response in health and disease and, as such, can also be involved in neuroinflammatory and neurodegenerative conditions such as AD [381]. Complement proteins such as C3 and factor B, along with receptors like C3aR and C5aR, are present in the mucosal and epithelial tissues of the intestine under normal conditions, and also influence the immune response and contribute to the brain-gut axis communication relevant to AD [86, 382]. This intestinal homeostasis is partially maintained by the presence of complement regulators CD46, CD59, and CD55 on the apical and basolateral sides of the intestinal epithelial cells, secreted by the mucosal cells [383, 384], and their dysregulation can contribute to systemic inflammation, a known risk factor of AD. The exact role of intestinal complement activity in AD is not yet fully understood. However, these findings highlight the necessity of exploring gut–complement–brain interactions within the framework of neurodegeneration. Emerging evidence suggests that bidirectional communication between the gut microbiome and CNS through various pathways, including immune, entero-endocrine, vagus nerve, and microbial metabolic pathways [385–387], is influenced by microbial dysbiosis caused by factors like LPS, aging, diet, drug intake, and diseases [388, 389]. Disruption of gut immune homeostasis caused by microbial dysbiosis can result in increased levels of complement proteins and TLRs, leading to heightened proinflammatory cytokines in the systemic circulation [390–392]. The integrity of the BBB can be disrupted by these elevated proinflammatory cytokines and facilitate neuroinflammation, which is one of the hallmarks of AD [390–392]. In addition, recent studies have highlighted that specific interventions, such as dietary intake, prebiotics/probiotics, fecal microbiota transplantation, as well as factors like age and sex, can modulate the composition of the microbiome, which can influence neuroinflammation and the pathogenesis of AD [393–395]. In 5xFAD mice, fecal transplants from healthy mice (C57BL/6) dramatically improved cognitive function, lower neuroinflammation, and lessened the pathology of AD. On the other hand, in healthy mice, fecal microbiota transplants from both 5xFAD and healthy mice worsened systemic inflammation and cognitive impairments [396]. Hao and his team found that germ-free mice exhibited elevated serum C3 and C3/C3R signaling after receiving a microbiome from the feces of mice with chronic unpredictable mild stress (CUMS), which resulted in abnormal microglia-dependent synapse pruning and signs of depression in those mice [370]. This result signifies that the intestinal microbial components can influence both peripheral and central complement systems, which play a major role in the pathogenesis of neurological diseases like AD. Research indicates that the maturation, regulation, and mitochondrial function of microglia are regulated by short-chain fatty acids (SCFAs) secreted by the gut microbiome under homeostatic conditions [397–399]. Prebiotic and probiotic intake increases SCFAs in the gut microbiome and plays an important role in neuroprotective effects [400]. It has been shown that early intervention with sodium butyrate in 5xFAD mice inhibits Aβ deposition [401, 402]. However, it is very intriguing to note that in the germ-free AD mouse model, SCFAs have detrimental effects by inducing glial activation and worsening Aβ pathology [403]. These mixed findings indicate the more beneficial and physiological balance effect of endogenous SCFAs produced by gut microbiome than the external administered SCFAs. Furthermore, after LPS treatment, germ-free mice showed deficits in their microglial cell maturation and functioning, and after SCFAs supplements, their microglia were able to recover and mature [398, 404]. Altogether, these studies emphasize how crucial the gut microbiome is for keeping balance and ensuring glial cells function properly in the brain, considering that dysfunctional microglia and aberrant complement activation contribute to AD pathology [405]. In the brain, astrocytes and microglia are the main sources of C3 and C1q components of the complement system [388, 389]. The sustained presence of dysbiotic bacteria has the potential to disrupt the homeostasis of the gut microbiome, leading to gut leakiness (Fig. 2). This condition makes the metabolite LPS enter into blood circulation, thereby triggering systemic inflammation, eventually resulting in neuroinflammation within the brain [390–392]. The activation of the complement pathway, particularly the release of C3 from activated astrocytes, may further stimulate microglial activity, promoting the release of C1q and various cytokines [406, 407]. Such processes can exacerbate Aβ toxicity and plaque formation while also activating macrophages, which may contribute to synaptic loss and neurodegeneration, characteristic of neurodegenerative disorders such as AD [251, 408]. A growing body of studies recently reported the alteration of the gut microbiome in AD patients compared with healthy non-AD individuals. For example, the bacteria Alistipes, Actinomyces viscosus, Bacteroides, Bilophila, Clostridium_XlVa, Flavobacterium, Lactobacillus, Moheibacter, Neorhizobium, Phascolarctobacterium, Solobacterium, and Weissella were found to be more common in people with AD and less common in healthy individuals [409, 410]. In addition, recent studies also highlighted the age- and sex-specific difference in microbiome modulating the neuroimmune response towards AD pathology [395, 411]. Beneficial gut microbiota such as Firmicutes and Actinobacteria, including Bifidobacterium spp., Blautia spp., and members of Lachnospirales, were predominant in people below 75 years of age, while people above 75 years of age displayed a microbial community enriched in Bacteroidota, particularly Bacteroidia and Bacteroides spp., known to be associated with proinflammatory cytokines [412]. LPS and similar molecules can damage the gut barrier and enter the bloodstream. These microbial products are detected by pattern recognition receptors (PRRs), like TLRs, and can trigger the complement system, especially the alternative and lectin pathways, resulting in inflammation throughout the body and in the brain. Alterations in the gut microbiome associated with AD have also been observed in AD mouse models, including 5XFAD mice, which exhibit distinct fecal microbiome profiles after 9 months [412] and in 3xTg mice at 12 weeks of age [413] in comparison to normal wild-type mice. These distinct microbial profiles in two different studies might be due to differences in strain, diet and environmental exposure. Another important longitudinal study also highlighted the significant influence of distinct microbiome composition including reduced Turicibacter abundance (a potential mediator of gut-brain axis) and altered serotonin metabolism in 5xFAD mice indicative of gut-driven dysregulation of signaling molecules and their potential influence on AD [414]. Liang et al. demonstrated that different AD mouse models at different ages have different microbiome compositions, which shows how these microbiomes have the potential role in exacerbating amyloid burden in these models [415]. Certain dysbiotic Gram-negative bacteria generate extracellular proteins resembling amyloid, and elicit inflammatory reactions similar to those caused by amyloid aggregates in the CNS [416–418]. These secreted amyloid-like proteins can contribute to activation of the immune response and complement pathways [416, 417]. Long-term and ongoing disruption of gut microbiome integrity results in the continuous activation of the complement pathway, which may influence amyloid aggregation and contribute to the pathogenesis of AD. Fig. 2Schematic representation of gut-brain axis-mediated complement dysregulation and neuroinflammation in Alzheimer’s disease. Gut dysbiosis results in the disruption of the intestinal barrier, which is linked to altered complement C3 signaling and an increased release of bacterial lipopolysaccharides (LPS) and metabolites into the bloodstream. These factors cause widespread inflammation by activating immune cells and increasing proinflammatory cytokines (like IL-6, TNF-α, IL-1β, IL-12), which may compromise the blood-brain barrier (BBB). Simultaneously, the compromised BBB permits peripheral inflammatory signals to infiltrate the brain, resulting in the activation of microglia and increasing the production of C1q, which subsequently activates the astrocytes and thereby exacerbates neuroinflammation. The inflammatory environment facilitates or hastens the accumulation of amyloid-β (Aβ), a defining characteristic of Alzheimer’s disease pathology. Figure created with Biorender.com Schematic representation of gut-brain axis-mediated complement dysregulation and neuroinflammation in Alzheimer’s disease. Gut dysbiosis results in the disruption of the intestinal barrier, which is linked to altered complement C3 signaling and an increased release of bacterial lipopolysaccharides (LPS) and metabolites into the bloodstream. These factors cause widespread inflammation by activating immune cells and increasing proinflammatory cytokines (like IL-6, TNF-α, IL-1β, IL-12), which may compromise the blood-brain barrier (BBB). Simultaneously, the compromised BBB permits peripheral inflammatory signals to infiltrate the brain, resulting in the activation of microglia and increasing the production of C1q, which subsequently activates the astrocytes and thereby exacerbates neuroinflammation. The inflammatory environment facilitates or hastens the accumulation of amyloid-β (Aβ), a defining characteristic of Alzheimer’s disease pathology. Figure created with Biorender.com Biomarkers indicate complement system dysregulation in AD AD is defined by pathological features including the accumulation of amyloid plaques, tau protein aggregation, synaptic degeneration, and cognitive decline [419, 420]. According to the Alzheimer’s Association, 7.2 million Americans over 65 have AD dementia in 2025, and this number is expected to grow to more than 13.8 million individuals by 2060 [421]. A study published in Lancet in 2018 suggested that amyloid plaques begin to form at least 22 years before AD symptoms show up, and tau protein tangles start to develop 20 years after the amyloid plaques accumulate, before mature tau tangles form in AD [422]. Additionally, glucose hypometabolism and brain atrophy start 18 and 13 years before the onset of clinical symptoms becomes evident, respectively [423, 424]. Before 2012, postmortem imaging of Aβ buildup and tau tangles was considered a criterion for diagnosing AD. In 2012, the National Institute on Aging (NIA) - Reagan Institute formulated the diagnostic criteria for AD, which involves a substantial accumulation of amyloid and tau in the brain [425]. The AT(N) system, proposed by the NIA–Alzheimer’s Association (NIA-AA), categorizes AD biomarkers into three primary groups: A (amyloid deposition, e.g., CSF/plasma Aβ42, amyloid PET), T (tau pathology, e.g., CSF/plasma p-tau, tau PET), and N (neurodegeneration, e.g., total tau, NfL, GFAP, FDG-PET, MRI) [426–428]. Currently, CSF and blood-based biomarkers are gaining attention as accessible and minimally invasive tools for the detection and monitoring of AD [429, 430]. Aβ42/Aβ40, p-tau, and NfL can be detected in blood samples of AD patients and are now considered potent blood-based biomarkers [431, 432]. Recent evidence suggests that the biomarkers of the complement system may have added value in diagnosing AD, by revealing neuroinflammatory processes that A, T, or N biomarkers do not completely show. Specifically, complement proteins such as C1q, C3, C4, and their activated forms, exhibit alterations in AD and are detectable in blood plasma, offering a convenient method for assessing neuroinflammation [433, 434]. The identification of significant complement components in blood may be beneficial for early screening and large-scale population research in AD. Higher levels of C1q, C3a, C4d, and factor B in the blood are also associated with memory issues, amyloid accumulation, and the progression from MCI to AD [184, 435, 436]. Complement dysregulation in AD is further supported by human biomarker studies of CSF and plasma, with consistent changes in C3, C4, C1q, TCC, and Factor H confirming the translational significance of complement-targeted mechanisms [176, 321, 437]. Studies have found a significant elevation of plasma clusterin in patients with AD compared to the control group [438–440]. This signifies that clusterin, one of the complement analytes, may have potential as an AD biomarker. Overall, complement activation products and inhibitors show potential as biomarkers that could complement the AT(N) profile and serve as indicators of disease progression and treatment response. Complement and complement regulatory proteins as potential therapeutic targets As discussed in the previous sections, complement activation plays a central role in vascular disorders and neurodegenerative diseases such as AD and CAA. While the underlying pathology and triggers vary, inflammation is a common factor, and particularly dysregulated activation of complement cascade is a common driver of cellular damage [441]. Therefore, targeting complement cascade either to upregulate or downregulate the effector function presents a therapeutic opportunity. In early disease stages of AD or CAA, the complement pathway can exert protective effects by clearing debris, removing antigen-immune complexes, facilitating immune cell trafficking and supporting cell signalling. However, as the disease worsens, excessive complement activation can contribute to brain and vascular injury. This duality underscores the need for stage-specific therapeutic strategies based on disease progression as no single complement-targeted therapy is likely to be effective across all disease stages. Given the chronic nature linked to their progressive pathology, AD and CAA may require long-term modulation of complement activity due to sustained low-grade activation over many years [19]. Importantly, because the complement pathway plays a significant role in the defense against microbes and infections, blocking the complement cascade carries risks of infection or weakened immunity. Histochemical and genetic data [20, 442] have confirmed the role of complement in driving inflammation during the development and course of AD. Therefore, identifying reliable complement biomarkers could aid in tracking disease progression from the early diagnosis of MCI to AD, supporting disease stratification and patient selection [440, 442]. This may help to determine the optimal time window for complement-targeted interventions [442]. Multiplex assays have identified increased levels of clusterin in plasma, differentiating AD from controls, and modeling with relevant co-variables indicated clusterin, Complement Factor I, and MAC to be different in MCI-converters from non-converters, while plasma Complement Factor I levels were shown to be associated with brain atrophy [440]. These markers underscore the importance of predicting disease progression, identifying the right complement target and optimizing the therapeutic window. Also, understanding which complement proteins to target, including upstream (C1q, C1s, C4), central (C3, C3a/C3aR), and downstream (C5a/C5aR, C5-C9) or their regulators is also crucial [443, 444]. In the context of the complex pathology of AD, thus far, there are no complement-based therapeutics available. Most of the data investigating potential therapeutic interventions comes from animal models of amyloidosis and tauopathy, where modulating the complement cascade yielded neuroprotective effects. A sampling of the studies categorized as per the complement target are referred to in Table 2. In AD mouse models, intracranial administration of antisense oligonucleotides directed against C1r, C1s and C4 showed neuroprotection and rescued spine loss in the PS2APP amyloid mouse model [204], highlighting the use of nucleic acid-based medicine as a potential therapeutic approach for neurodegenerative diseases. C5aR1 antagonism with a cyclic hexapeptide, PMX205, also decreased amyloid load by 49–62%, and glial activation by 42–68% in AD mouse models [237]. In another AD mouse model, i.e. making use of APPNL−G−F mice, both C7 deficiency and C3 deletion were shown to protect against spine loss and improve cognition [179, 246]. In models of tauopathy, deletion of C3aR and C3 inhibition ameliorated tau pathology and lowered inflammation and neuronal loss [179, 222]. A clone 73D1 recombinant monoclonal antibody, a strong inhibitor for C7, fused with nanobody Nb62 targeting the transferrin receptor, was shown to penetrate the BBB and further reduce complement activation, synapse loss and amyloid levels in the APPNL−G−F mouse model [247]. Overall, these studies suggest that complement and its regulatory proteins serve as promising therapeutic targets in AD. Notably, there are several commercially available complement therapeutics that have demonstrated efficacy in other human diseases as shown in Table 3. Drug repurposing to lower complement activation in neurodegenerative diseases may hold potential, but, further studies are required to assess their efficacy and potential side effects. Table 2Drug development targeting complement in AD mouse modelsTherapeutic TargetDisease ModelInterventionKey OutcomesC1r, C1s, C4Amyloidosis (PS2APP)Antisense Oligonucleotides [204]Significant rescue of synapse lossC5aRAmyloidosis (Tg2576); Amyloid and Tau (3xTg)PMX205 (C5a receptor peptide antagonist) [111, 237, 244]Reduced amyloid deposits by 49–62%, glia by 42–68% and improved cognitive functionC3aRAmyloidosis (APP/TTA)SB290157 (C3aR antagonist) [258]Reduced amyloid load, microgliosisMembrane Attack ComplexAmyloidosis (APPNL−G−F)mAb 73D1 (C7 blocker) [247]Reduced synapse loss, amyloid burden, improved cognition Table 2Drug development targeting complement in AD mouse modelsTherapeutic TargetDisease ModelInterventionKey OutcomesC1r, C1s, C4Amyloidosis (PS2APP)Antisense Oligonucleotides [204]Significant rescue of synapse lossC5aRAmyloidosis (Tg2576); Amyloid and Tau (3xTg)PMX205 (C5a receptor peptide antagonist) [111, 237, 244]Reduced amyloid deposits by 49–62%, glia by 42–68% and improved cognitive functionC3aRAmyloidosis (APP/TTA)SB290157 (C3aR antagonist) [258]Reduced amyloid load, microgliosisMembrane Attack ComplexAmyloidosis (APPNL−G−F)mAb 73D1 (C7 blocker) [247]Reduced synapse loss, amyloid burden, improved cognition Drug development targeting complement in AD mouse models Table 3Complement therapeutics currently available for other indicationsComplement drugsApproveddrugs/classDiseasePathologySymptomsPlasma-derived C1 esterase inhibitors (C1INH)Cinryze™;Berinert™(Biologics)Hereditary angioedema [445]Deficiency or dysfunction of complement C1 inhibitorEpisodes of swelling in hands, feet, face, airways, intestinesC1s inhibitorSutimlimab(Monoclonal antibody)Cold agglutinin disease [446]IgM auto antibodies targeting RBCs causing hemolysisCold induced symptoms including anemia, pallor, muscle weaknessC3 inhibitorPegcetacoplan(small peptide inhibitor)PNH-Paroxysmal nocturnal hemoglobinuria (intra & extravascular hemolysis) [447]Age-related macular degeneration, [448]Lack of regulatory proteins CD55 and CD59 on the surface of red blood cells (RBCs) causes complement activation and hemolysisOveractivation of complement system int the eye causing subretinal deposits, death of retinal pigment epithelial cellsHemoglobinuria (hemoglobin in the urine)Central vision lossC5 inhibitorEculizumab™(Monoclonal antibody)PNH-Paroxysmal nocturnal hemoglobinuria (intravascular hemolysis) [449]Aquaporin-4–positive Neuromyelitis optica spectrum disorder (NMOSD) [450]HemolysisComplement (MAC) deposition around blood vesselsHemoglobinuriaSevere optic neuritis and transverse myelitis causing vision loss, paralysis, and bladder/bowel dysfunctionC5 inhibitorULTOMIRIS® (ravulizumab-cwvz); anti-C5 IgG2/4 (Monoclonal antibody)generalised myasthenia gravis (gMG) [451]atypical haemolytic uraemic syndrome (aHUS) [452]aquaporin-4–positiveNeuromyelitis optica spectrum disorder (NMOSD) [449, 451]Activation of the complement system at the neuromuscular junctionDysregulation of the complement leading to blood vessel damageComplement (MAC) deposition around blood vesselsMuscle weakness, fatiguekidney damage, blood clots, and anemiaSevere optic neuritis and transverse myelitis causing vision loss, paralysis, and bladder/bowel dysfunctionC5 inhibitorPozelimab (Monoclonal antibody)Crovalimab (Monoclonal antibody)Avacinacptad pegol (RNA aptamer)CD55-deficient protein-losing enteropathy (PLE) or CHAPLE disease [453]PNH [454]Geographic atrophy (advanced age related macular degeneration)Mutations in CD55 gene causing overactivation of complement system causing damage to blood, lymph vessels along upper digestive tractHemolysisExcessive complement activation in the eyeEnteropathyHemoglobinuriaCentral vision loss, reduced color perceptionC5 inhibitorZilucoplan (macrocylic peptide)generalised myasthenia gravis (gMG) in patients who are anti-AChR antibody positivity [455]Activation of the complement system at the neuromuscular junctionMuscle weakness, fatigueC5a inhibitorVilobelimab (Monoclonal antibody)COVID-19, acute respiratory distress syndrome [456]Swelling in the lungsDifficulty in breathingC5aR1 inhibitorAvacopan(small molecule)Polyangiitis [457]Autoimmune disease with anti-neutrophil cytoplasmic autoantibody associated vasculitisSwelling and irritation in blood vesselsFactor B inhibitorIptacopan(small molecule)PNH [458]Intravascular & extravascular hemolysisHemoglobinuriaFactor D inhiborDanicopan(small molecule)PNH [459]Extravascular hemolysisHemoglobinuria Table 3Complement therapeutics currently available for other indicationsComplement drugsApproveddrugs/classDiseasePathologySymptomsPlasma-derived C1 esterase inhibitors (C1INH)Cinryze™;Berinert™(Biologics)Hereditary angioedema [445]Deficiency or dysfunction of complement C1 inhibitorEpisodes of swelling in hands, feet, face, airways, intestinesC1s inhibitorSutimlimab(Monoclonal antibody)Cold agglutinin disease [446]IgM auto antibodies targeting RBCs causing hemolysisCold induced symptoms including anemia, pallor, muscle weaknessC3 inhibitorPegcetacoplan(small peptide inhibitor)PNH-Paroxysmal nocturnal hemoglobinuria (intra & extravascular hemolysis) [447]Age-related macular degeneration, [448]Lack of regulatory proteins CD55 and CD59 on the surface of red blood cells (RBCs) causes complement activation and hemolysisOveractivation of complement system int the eye causing subretinal deposits, death of retinal pigment epithelial cellsHemoglobinuria (hemoglobin in the urine)Central vision lossC5 inhibitorEculizumab™(Monoclonal antibody)PNH-Paroxysmal nocturnal hemoglobinuria (intravascular hemolysis) [449]Aquaporin-4–positive Neuromyelitis optica spectrum disorder (NMOSD) [450]HemolysisComplement (MAC) deposition around blood vesselsHemoglobinuriaSevere optic neuritis and transverse myelitis causing vision loss, paralysis, and bladder/bowel dysfunctionC5 inhibitorULTOMIRIS® (ravulizumab-cwvz); anti-C5 IgG2/4 (Monoclonal antibody)generalised myasthenia gravis (gMG) [451]atypical haemolytic uraemic syndrome (aHUS) [452]aquaporin-4–positiveNeuromyelitis optica spectrum disorder (NMOSD) [449, 451]Activation of the complement system at the neuromuscular junctionDysregulation of the complement leading to blood vessel damageComplement (MAC) deposition around blood vesselsMuscle weakness, fatiguekidney damage, blood clots, and anemiaSevere optic neuritis and transverse myelitis causing vision loss, paralysis, and bladder/bowel dysfunctionC5 inhibitorPozelimab (Monoclonal antibody)Crovalimab (Monoclonal antibody)Avacinacptad pegol (RNA aptamer)CD55-deficient protein-losing enteropathy (PLE) or CHAPLE disease [453]PNH [454]Geographic atrophy (advanced age related macular degeneration)Mutations in CD55 gene causing overactivation of complement system causing damage to blood, lymph vessels along upper digestive tractHemolysisExcessive complement activation in the eyeEnteropathyHemoglobinuriaCentral vision loss, reduced color perceptionC5 inhibitorZilucoplan (macrocylic peptide)generalised myasthenia gravis (gMG) in patients who are anti-AChR antibody positivity [455]Activation of the complement system at the neuromuscular junctionMuscle weakness, fatigueC5a inhibitorVilobelimab (Monoclonal antibody)COVID-19, acute respiratory distress syndrome [456]Swelling in the lungsDifficulty in breathingC5aR1 inhibitorAvacopan(small molecule)Polyangiitis [457]Autoimmune disease with anti-neutrophil cytoplasmic autoantibody associated vasculitisSwelling and irritation in blood vesselsFactor B inhibitorIptacopan(small molecule)PNH [458]Intravascular & extravascular hemolysisHemoglobinuriaFactor D inhiborDanicopan(small molecule)PNH [459]Extravascular hemolysisHemoglobinuria Complement therapeutics currently available for other indications Cinryze™; Berinert™ (Biologics) Sutimlimab (Monoclonal antibody) Pegcetacoplan (small peptide inhibitor) PNH-Paroxysmal nocturnal hemoglobinuria (intra & extravascular hemolysis) [447] Age-related macular degeneration, [448] Lack of regulatory proteins CD55 and CD59 on the surface of red blood cells (RBCs) causes complement activation and hemolysis Overactivation of complement system int the eye causing subretinal deposits, death of retinal pigment epithelial cells Hemoglobinuria (hemoglobin in the urine) Central vision loss Eculizumab™ (Monoclonal antibody) PNH-Paroxysmal nocturnal hemoglobinuria (intravascular hemolysis) [449] Aquaporin-4–positive Neuromyelitis optica spectrum disorder (NMOSD) [450] Hemolysis Complement (MAC) deposition around blood vessels Hemoglobinuria Severe optic neuritis and transverse myelitis causing vision loss, paralysis, and bladder/bowel dysfunction ULTOMIRIS® (ravulizumab-cwvz) ; anti-C5 IgG2/4 (Monoclonal antibody) generalised myasthenia gravis (gMG) [451] atypical haemolytic uraemic syndrome (aHUS) [452] aquaporin-4–positiveNeuromyelitis optica spectrum disorder (NMOSD) [449, 451] Activation of the complement system at the neuromuscular junction Dysregulation of the complement leading to blood vessel damage Complement (MAC) deposition around blood vessels Muscle weakness, fatigue kidney damage, blood clots, and anemia Severe optic neuritis and transverse myelitis causing vision loss, paralysis, and bladder/bowel dysfunction Pozelimab (Monoclonal antibody) Crovalimab (Monoclonal antibody) Avacinacptad pegol (RNA aptamer) CD55-deficient protein-losing enteropathy (PLE) or CHAPLE disease [453] PNH [454] Geographic atrophy (advanced age related macular degeneration) Mutations in CD55 gene causing overactivation of complement system causing damage to blood, lymph vessels along upper digestive tract Hemolysis Excessive complement activation in the eye Enteropathy Hemoglobinuria Central vision loss, reduced color perception Avacopan (small molecule) Iptacopan (small molecule) Danicopan (small molecule) Future directions As our understanding of complement biology in AD deepens, future research should prioritize clarifying the cell type- and brain region-specific roles of complement components across disease stages. Longitudinal multi-omic studies integrating genomics, epigenomics, proteomics, spatial transcriptomics, and the exposome will be essential for mapping temporal dynamics and functional outcomes. Further efforts should investigate sex-specific regulatory mechanisms, the interplay between peripheral and CNS complement systems, especially at barrier sites like the BBB and choroid plexus, and the influence of systemic factors such as microbiota and metabolism. Functional validation of GWAS and EWAS findings using iPSC-based model systems and CRISPR-based tools remains a key priority. Moreover, expanding our knowledge of intracellular complement signaling may uncover novel immunometabolic interactions. Therapeutically, identifying context-sensitive complement targets that can be engaged in combination with strategies supporting synaptic resilience and modulating microglial responses in late-stage AD models [460], may offer synergistic benefits. Finally, the discovery of reliable AD subtype- and stage-specific diagnostic and prognostic biomarkers of complement activity will be critical for advancing precision complement-modulating interventions in AD. Conclusion The complement system has emerged as a central player in the pathogenesis of AD, exerting diverse and context-dependent effects on synaptic function, neuroinflammation, and neurodegeneration. While genetic studies have highlighted key susceptibility loci such as CR1 and CLU, epigenetic and environmental factors increasingly appear to modulate complement activity in cell type-, sex-, and disease stage-specific ways. From developmental synaptic pruning to maladaptive activation in the aging and diseased brain, complement signaling represents a dynamic and multi-layered process. Advances in spatial transcriptomics, iPSC modeling, and systems-level analyses are now revealing a more nuanced view of complement as a regulator of both CNS and peripheral immune networks. Unraveling these complexities will be essential for the development of precision therapies aimed at harnessing the beneficial roles of complement while curbing its deleterious consequences in AD.