Alzheimer’s Disease Genetics: A Dampened Microglial Response?

Abstract

Alzheimer’s disease (AD) is a debilitating age-related neurodegenerative condition. Unbiased genetic studies have implicated a central role for microglia, the resident innate immune cells of the central nervous system, in AD pathogenesis. On-going efforts are clarifying the biology underlying these associations and the microglial pathways that are dysfunctional in AD. Several genetic risk factors converge to decrease the function of activating microglial receptors and increase the function of inhibitory receptors, resulting in a seemingly dampened microglial phenotype in AD. Moreover, many of these microglial proteins that are genetically associated with AD appear to interact and share pathways or regulatory mechanisms, presenting several points of convergence that may be strategic targets for therapeutic intervention. Here, we review some of these studies and their implications for microglial participation in AD pathogenesis.

Keywords

Alzheimer’s disease microglia genetics HLA CD33 TREM2 SPI1

Alzheimer’s Disease

Alzheimer’s disease (AD) is a debilitating age-related neurodegenerative condition. It is a progressive disorder that deteriorates memory and cognition and is ultimately fatal, with an immense personal and societal burden. Five million individuals in the United States have AD, and this number is projected to increase to about 13 million by 2050 (Brookmeyer and others 2011). Currently, there is no disease-modifying therapy for AD. Over the past few decades, the aging and dementia research community has made progress in identifying risk factors, particularly genetic loci, and defining the clinical and pathologic features of the disease (Morris and others 1993). These discoveries have led to the understanding that AD is a progressive disease with dementia only resulting in its terminal stage. Initially, AD has a long subclinical phase in which molecular changes occur, including the gradual accumulation of characteristic amyloid and tau pathologies, in the absence of overt cognitive impairment. This is followed by a phase of mild cognitive impairment and finally by AD dementia (Sperling and others 2011).

Late-onset Alzheimer’s disease (LOAD), in which disease presents after age 65, accounts for the majority of AD cases. Early-onset Alzheimer’s disease (EOAD) occurs in 5% of cases, in which disease presents before age 65 (Alzheimer’s Association 2015). These early presentations include autosomal dominant AD (ADAD) cases, a rare form of AD in which the disease is inherited in an autosomal dominant fashion at an early age (Bateman and others 2012). ADAD progresses rapidly and is usually associated with mutations in the APP or PSEN gene (Bateman and others 2012). Conversely, LOAD appears to be a sporadic disease, and its etiology and pathogenesis are poorly understood. While clinical presentations suggestive of Alzheimer’s dementia are classified by assessments of cognitive status, AD continues to be biologically defined by neuropathological hallmarks (Jack and others 2018). The known pathology of AD includes aberrant extracellular amyloid-beta (Aβ) aggregates which form diffuse and neuritic plaques and hyperphosphorylated tau aggregates which form intraneuronal neurofibrillary tangles (Duyckaerts and others 2009). These pathological hallmarks are accompanied by loss of synapses, neuronal death, and gross brain atrophy (Duyckaerts and others 2009). Interestingly, initial AD brains characterized by Alois Alzheimer revealed the presence of glial cells with abnormal morphology (Alzheimer 1907; Alzheimer and others 1995; Graeber and others 1997), which were later shown to be astrocytes and microglia (Maragakis and Rothstein 2006; Mosher and Wyss-Coray 2014). While chronic shifts in inflammation and immune signaling in AD were long considered to be a response to pathology, microglia are now regarded as playing key roles in both disease initiation and progression (Box 1). Here, we will review the current literature studying the impact of LOAD genetic risk-variants on the functional responses of microglia, and discuss the implications of these findings.

Box 1.

The percentage of Alzheimer’s disease publications that investigate microglia have been increasing over time. Publications listed in PubMed that respond to the search terms “Alzheimer’s disease” or “Alzheimer’s disease AND microglia” between the years of 1970 and 2019 are plotted above. While both search terms have increasing publications over time, the percentage of publications investigating microglia have especially increased in recent years, with a high of 5% in 2019. Open bars correspond to “Alzheimer’s disease” on the left y-axis and red bars indicate “Alzheimer’s disease AND microglia” on the right y-axis.

Microglia

Microglia are the resident innate immune cells of the central nervous system (CNS), where they mediate the local immune response. Like other tissue-resident macrophages, microglia originate from hematopoietic progenitors in the yolk sac, which emigrate to the CNS well before the development of the brain (embryonic days 8.5–10.5 in mice; Alliot and others 1999; Askew and Gomez-Nicola 2018; Ginhoux and others 2010; Ginhoux and others 2013; Hoeffel and others 2012; Schulz and others 2012). Microglia participate in a wide range of homeostatic processes including neurogenesis and synaptic pruning (Hammond and others 2018; Nimmerjahn and others 2005; Schafer and others 2012; Squarzoni and others 2014). They further contribute to maintaining brain homeostasis through constantly surveying the environment and performing innate immune functions such as responding to injury, tissue repair, and removing unwanted molecules or pathogens (Hammond and others 2018; Hanisch and Kettenmann 2007; Nimmerjahn and others 2005).

It has long been accepted and more recently shown that microglia are highly plastic, long-lived immune cells that respond to their environment and modulate their gene expression, phenotype, and function in certain CNS regions or disease states (Mastroeni and others 2017; McCarthy 2017; Olah and others 2018). They are thought to be a heterogeneous population with distinct subsets ranging in function. Indeed, specific microglial subpopulations have been observed in mouse models of Alzheimer’s disease, where microglia were profiled over time at the single-cell level (Keren-Shaul and others 2017; Mathys and others 2017). Microglial heterogeneity is now being explored in the human system by examining single-immune cell transcriptomics from autopsy tissue and biopsy material (Olah and others 2020; Sankowski and others 2019). In one study, the analysis of 21 samples identified nine microglial subsets with varying expressions of activation markers and immune response genes (Olah and others 2020). While the overwhelming majority of microglia belong to clusters that appear to represent homeostatic microglial populations, one subset enriched with markers of microglial activation and antigen presentation was negatively correlated with a clinical diagnosis of AD, AD pathology, and tau tangles. Conversely, a subset featuring genes involved in anti-inflammatory responses (interleukin [IL]-10, IL-4, IL-13) was found to be positively associated with amyloid pathology and clinical AD (Olah and others 2020).

Pre-GWAS Perspective

Prior to the age of transcriptomics and genome-wide association studies (GWAS), microglial profiling in AD was characterized by morphology, targeted phenotyping, and functional outcomes, with an emphasis on inflammatory cytokine production, phagocytosis, and ‎Aβ clearance. Inflammation had long been thought to be a driving force in AD. Several studies including those performed in AD mouse models led to the conclusion that activated microglia contribute to neuronal death in AD, and that they need to be suppressed with anti-inflammatory drugs (Akiyama and others 2000). Aligning with this inflammation theory, epidemiological observations suggested the use of nonsteroidal anti-inflammatory drugs (NSAIDs) to be protective against AD, although a meta-analysis of these findings suggests that the reported benefits of NSAIDs may largely be due to bias in analysis and reporting (de Craen and others 2005; in t’ Veld and others 2001). When tested for preventative effects against AD in a recent clinical trial no evidence was observed for improved outcomes with NSAID treatment in presymptomatic individuals including cognition, and the study raised safety concerns of significant cardiovascular and gastrotintestinal side effects (Meyer and others 2019). The failure of a beneficial effect in clinical trials has contributed to the inflammation dogma being challenged, and it is now suggested that microglial activation may play a beneficial role in AD, as discussed throughout this review. Another clinical trial with ibuprofen and cromolyn has recently been completed, the results of which once published may help clarify whether NSAIDs provide a preventative or therapeutic benefit against AD (Elmaleh 2020).

While innate immune cells have many roles in the CNS, impaired phagocytosis resulting in an inability to clear Aβ has long been thought to be critical to LOAD progression (Fiala and others 2005; Fiala and others 2007; Hickman and others 2008; Wildsmith and others 2013; Yoon and Jo 2012). ‎Aβ clearance is impaired in the AD brain, suggesting an imbalance between rates of Aβ production and clearance (Mawuenyega and others 2010; Wildsmith and others 2013). Interestingly, monocytes and monocyte-derived macrophages from AD subjects are ineffective at phagocytosis of Aβ compared to those from age-matched controls, independent of genetic background (Avagyan and others 2009; Fiala and others 2005; Zaghi and others 2009). As these cells are derived from the circulation, it suggests that there is an underlying defect in the cells themselves, rather than a response to the local AD brain environment. In addition, macrophages from AD subjects appear to undergo apoptosis following exposure to Aβ (Avagyan and others 2009; Fiala and others 2005). These findings suggest that altered phagocytic function of macrophages and microglia contribute to AD pathogenesis. However, the molecular mechanisms underlying impaired phagocytosis and the implications of this altered phenotype for microglial participation in disease is unclear. Genetic studies have begun to clarify risk factors and molecular pathways that may contribute to the mechanisms involved in altered microglial functioning in AD.

GWAS and Sequencing Identification of LOAD-Associated Genes

With the ongoing discovery and validation of LOAD susceptibility loci, genetic risk factors have been identified that give insights into the earliest phases of the pathophysiological processes leading to LOAD (Hollingworth and others 2011a). It is thought that genetics contribute 53% of the phenotypic variance observed in AD (Ridge and others 2016). In particular, unbiased genetic studies have highlighted the role of the innate immune system in LOAD susceptibility, implicating a central role for microglia in disease. Initially, large consortiums such as the International Genomics of Alzheimer’s Project (IGAP) identified several AD-associated genetic loci which implicate genes that are preferentially or exclusively expressed in microglia in the CNS, such as ABCA7, BIN1, CASS4, CD33, CLEF1, HLA, INPP5D, MEF2C, MS4A region, NME8, PICALM, PTK2B, RIN3, SORL1, and ZCWPW1 (Harold and others 2009; Hollingworth and others 2011a; Hollingworth and others 2011b; Lambert and others 2009; Naj and others 2011; Seshadri and others 2010). In the last 7 years, additional GWAS, GWAS by proxy, and sequencing studies have only enhanced the suggestion that genetically susceptible microglia play a central role in LOAD pathogenesis. These genetic findings, in combination with transcriptomic data, have allowed us to hypothesize which microglial functional programs may be impacted in AD. For example, the combined genetic association of CD33, TREM2, TYROBP, and CR1 with AD suggests that phagocytosis may be altered in microglia as a pathogenic mechanism of disease (Bradshaw and others 2013; Griciuc and others 2013; Konishi and Kiyama 2018; Zhang and others 2013).

It is very clear from human genetics and transcriptomics that a microglial-specific intervention could be a strategic therapeutic approach for LOAD (Box 2). However, it may be a major challenge to identify ways in which to positively modulate microglial phagocytic activities while dampening any detrimental properties in order to preserve neuronal function and plasticity. While findings strongly support a causal involvement of microglial cells in AD pathogenesis, the underlying mechanisms remain unclear. GWAS indicate the genetic loci that are associated with disease; they do not definitively identify the associated gene(s). Tools such as fine mapping, eQTL analysis, and other functional genomic studies are needed to elucidate the biology underlying these associations. Such efforts are ongoing and will continue to clarify pathways that are dysfunctional in LOAD (Andrews and others 2020).

Box 2.

Over the past decade, many genes have been identified as genetically associated with LOAD. Many efforts are underway to identify the cell type(s) affected and the functional outcomes of these associations with the goal of targeting them for treatment of LOAD.

Microglial Suppression in LOAD

The dogma that arose in the pre-GWAS era concluded that the immune system needs to be therapeutically suppressed in AD, and has been supported by studies in AD mouse models (Akiyama and others 2000). However, as reviewed below, the recent human genetic findings are in stark contrast to this hypothesis and conversely suggest that a less responsive microglial phenotype is associated with AD risk. Indeed, AD risk variants appear to converge to suppress immune signaling, while protective alleles enhance the immune response, resulting in a shared inhibitory microglial functional outcome that may contribute to AD pathogenesis (Bradshaw and others 2013; de Rojas and others 2020; Huang and others 2017; Kleinberger and others 2014; Magno and others 2019; Rathore and others 2018; Replogle and others 2015). Microglial suppression resulting in the inability to resolve inflammatory processes, mitigate pathology, and maintain brain homeostasis may heighten susceptibility to neuronal injury and contribute to AD progression. Importantly, this hypofunctional microglial phenotype may extend past genetic risk variants and functionally converge with age-associated senescence, and other lifestyle and environmental factors associated with AD. Below, we review some of these genetic findings and their implications for microglial participation in LOAD pathogenesis. As many of the innate immune proteins genetically associated with LOAD share pathways or regulatory mechanisms, here we additionally outline several points of convergence that may serve as advantageous microglial targets for therapeutic intervention.

CD33

GWAS established the CD33 locus as a genetic risk factor for AD in 2011 (Hollingworth and others 2011b; Naj and others 2011). The locus had been previously identified to be associated with AD in 2008 in a large familial study (Bertram and others 2008). CD33, also known as Siglec-3, is a 67-kDa transmembrane glycoprotein expressed on the surface of myeloid progenitor cells, including monocytes, macrophages, and microglia (Bradshaw and others 2013; Griciuc and others 2013; Malik and others 2013; Walker and others 2015). It functions as a lectin, a carbohydrate-binding protein, and specifically binds sialic acid. CD33 is an inhibitory molecule, limiting endocytosis and the activation of immune signaling pathways, and inducing apoptosis (Crocker and Redelinghuys 2008). Its intracellular component contains immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are implicated in the inhibition of cellular activity (Crocker and Redelinghuys 2008; Taylor and others 1999). In support of its putative inhibitory function, CD33 has been shown to constitutively repress monocyte-derived pro-inflammatory cytokines (Lajaunias and others 2005). Furthermore, reducing the surface expression of CD33 using siRNA or an anti-CD33 antibody leads to an increase in spontaneous pro-inflammatory cytokine production by monocytes (Lajaunias and others 2005).

There are two main protein isoforms of CD33, termed CD33M and CD33m (Pérez-Oliva and others 2011). The smaller isoform, CD33m, is generated through alternative splicing and lacks the Ig V-set domain which contains the sialic acid-binding motif (Malik and others 2013; Pérez-Oliva and others 2011). This domain is encoded in exon 2. The AD-associated risk allele rs3865444^C is strongly associated with increased expression of CD33 exon 2. This can be observed at the protein level with dramatically increased expression of CD33M (the exon 2-containing isoform) with the risk allele (Malik and others 2013; Raj and others 2014). While the function of the small CD33m isoform remains unclear, it has been suggested to be a protective isoform (Estus and others 2019). The causative AD-associated single nucleotide polymorphism (SNP) is most likely rs12459419, which is located in a putative SRSF2 splice site of exon 2 and is in high linkage disequilibrium with the AD-risk GWAS SNP, rs3865444 (Malik and others 2013; Raj and others 2014). CD33 rs3865444^C has also been associated with reduced uptake in monocytes (Bradshaw and others 2013). This finding is supported by experiments in which CD33 was expressed in an immortalized microglial cell line and phagocytosis was reduced (Griciuc and others 2013).

The ligands for CD33 include the sialic acid moiety on glycoproteins (Crocker and Redelinghuys 2008). However, few CD33 protein binding partners have been identified, with only CD14 being suggested as a binding partner in monocyte-derived dendritic cells and S100A9 in myeloid-derived suppressor cells (Chen and others 2013; Ishida and others 2014). The binding of CD14 by CD33 is suggested to lead to downregulation of Toll-Like Receptor 4 (TLR4)-mediated signaling (Ishida and others 2014). Upon CD33 signaling, its intracellular ITIM motifs become phosphorylated, which then attract SH2-domain-containing proteins. In the case of CD33, these are primarily thought to be the phosphatases SHP-1 and SHP-2 (Taylor and others 1999). Phosphorylated CD33 has also been reported to recruit the SH2-domain-containing proteins SOCS3 and Cbl (Orr and others 2007; Walter and others 2008). However, it remains unclear which recruited proteins and sialic acid-dependent binding partners are the most relevant in microglia and in the context of AD-associated stimuli such as amyloid pathology.

Of note, the murine ortholog of CD33 lacks the ITIM motif and is expressed only at low levels on microglia, presenting additional challenges in studying the role of CD33 in mouse models of AD (Bhattacherjee and others 2019; Crocker and others 2007). The murine ortholog additionally includes a unique transmembrane lysine residue, whereas the human CD33M contains a unique cytoplasmic tyrosine (Cao and Crocker 2011; Crocker and others 2007). These differences may functionally result in divergent roles for CD33 in regulating phagocytosis in humans compared to mice. While human CD33M impairs phagocytosis in monocytes and primary microglia, murine CD33 has conflicting reports. In one study, primary microglial cultures derived from CD33^−/− mice demonstrated increased uptake in vitro (Griciuc and others 2013). However, in a more recent study, no effect of CD33 genetic deletion was found on uptake in the BV-2 mouse microglial cell line, and only deletion of human CD33M, but not murine CD33, resulted in increased uptake in a monocytic cell line (Bhattacherjee and others 2019). Moreover, the protective CD33 rs3865444A variant associated with higher expression of the small CD33m isoform seems to be unique to humans (Schwarz and others 2016). These findings highlight the importance of specifically studying human CD33 in AD pathogenesis. However, in studies in 5xFAD mice, knockout of murine CD33 attenuated amyloid-beta pathology and improved cognition, suggesting the existence of additional mechanisms independent of its ITIM motif or repressed phagocytosis through which CD33 may contribute to AD pathogenesis (Griciuc and others 2019).

As our understanding of CD33 biology and its contribution to AD risk has grown, downregulating its expression or inhibiting its function has become a clear potential therapeutic approach. In fact, clinical trials are currently examining the safety and efficacy of anti-CD33 antibodies as a therapy for AD (Paul 2019).

PILRA

Like CD33, PILRA is a sialic acid-binding and ITIM-containing inhibitory receptor that is specifically expressed in innate immune cells. In the cortex, its expression is enriched 15-fold in microglia compared to bulk tissue (Olah and others 2018). Similar to CD33, its expression was also found to be increased in peripheral leukocytes of AD patients (Park and others 2020). PILRA was determined to be the most likely causal gene in the AD-associated locus originally described as the ZCWPW1 locus (Rathore and others 2018). The PILRA genetic variation leads to a missense amino acid change in the ligand-binding domain, in which the AD-associated risk variant glycine 78 is replaced by an arginine (G78R) in the protective variant. This protective amino acid change from glycine to arginine reduces ligand binding (Rathore and others 2018). As such, like CD33, PILRA is another example of an AD-genetic risk factor that results in the increased function of an inhibitory microglial receptor.

TREM2

Among all the identified AD-associated genetic risk factors involved in innate immune function, none have generated as much translational investigation as the mutations within the TREM2 gene (Guerreiro and others 2013; Jonsson and others 2013). As TREM2 is exclusively expressed in cells of the myeloid lineage, microglia remain a prime focus for TREM2-related AD research (Bouchon and others 2001; Colonna 2003). The identified TREM2 mutations associated with LOAD are primarily within exons and include H15Y, R62H, and R47H. Of these variants, R47H has been the most heavily studied and has been demonstrated to result in reduced ligand-binding affinity and phagocytosis (Kleinberger and others 2014; Yao and others 2019). Similarly, in vitro studies have demonstrated that these mutations reduce lipid–ligand affinity and subsequent signaling, indicating a loss of TREM2 function (Kober and others 2016; Sudom and others 2018; Wang and others 2015). Thus, it is suggested that impaired signaling due to a partial loss of TREM2 function may be a major contributor to AD pathogenesis in individuals with these mutations. Additionally, soluble TREM2 in the CSF of AD patients positively correlates with both total tau and phosphorylated tau, but not with Aβ levels (Piccio and others 2016). Individuals carrying the R47H TREM2 variant have been shown to have increased levels of total and phosphorylated tau, with no effect on amyloid (Lill and others 2015). These findings suggest that the participation of TREM2 in AD may involve tau dysfunction.

The TREM2 signaling partner, TYROBP (also known as DAP12), contains an immunoreceptor tyrosine-based activation motif (ITAM) and is an adaptor protein known to associate with many cell surface molecules (Lanier 2009). TYROBP was identified via an integrated systems approach as a key driver for developing LOAD (Zhang and others 2013). Upon ligand binding of TREM2, TYROBP becomes phosphorylated at its ITAM motif. Syk kinase is then recruited to further initiate downstream signaling, promoting cell survival, proliferation, inflammation, and phagocytosis (Konishi and Kiyama 2018). SHIP1, another AD genetically associated protein (gene name INPP5D), has been shown to inhibit this signaling pathway (Peng and others 2010). INPP5D expression was found to be increased in the brains of LOAD patients and to correlate with amyloid plaque density (Tsai and others 2021). However, further research must be done to uncover the functional outcome of the INPP5D AD-associated variant in microglia, and how its inhibitory effects on the TREM2 signaling pathway impact AD risk.

While TREM2 signaling results in cellular activation, TREM2 is thought to be a negative regulator of the inflammatory response. Unlike other TYROBP-associated activating receptors such as TREM1, antibody crosslinking-induced TREM2 signaling does not result in degradation of IkB-α and nuclear translocation of NF-κB (Bouchon and others 2001). Silencing TREM2 expression with shRNA increased TNF (tumor necrosis factor) production in response to diverse TLR ligands including zymosan and CpG in bone marrow- derived macrophages (BMDM; Hamerman and others 2006). Similar findings have also been demonstrated in BMDM from TREM2 deficient mice, in which increased IL-6 and TNF were observed in response to an LPS challenge compared to wild-type controls (Turnbull and others 2006). Conversely, TREM2 overexpression in microglia has been shown to decrease TNF transcription following exposure to apoptotic neurons (Takahashi and others 2005). These findings suggest a role for TREM2 as a regulator of cytokine production and inflammatory responses in a context-specific manner.

Many functional consequences of altered TREM2 expression have been reported, including impaired phagocytosis. Microglia in TREM2-deficient mice were shown to have reduced expression of CD68, a phagocytic marker, as well as to associate less with dying cells in a mouse model of demyelination (Kawabori and others 2015; Sieber and others 2013). Similarly, TREM2 deficiency in an AD mouse model was shown to reduce internalization of Aβ throughout disease (Jay and others 2017). Expression of a LOAD variant of TREM2 in the BV2 mouse microglial cell line resulted in impaired phagocytosis (Kleinberger and others 2014). A more recent study based on stem cell derived microglia-like cells also found TREM2-deficient microglia to be less effective in Aβ clearance (Claes and others 2019). In this regard, the TREM2/TYROBP signaling pathway presents a potential axis to modulate microglial protective functions, including Aβ clearance (Schlepckow and others 2020).

Mouse models of amyloidosis suggest a protective function for TREM2 in disease pathogenesis. In mouse models of AD crossed with TREM2 or TYROBP knockout mice, microglial activation and Aβ clearance are impaired (Wang and others 2015; Wang and others 2016; Yuan and others 2016). Additionally, recent studies in mouse models of AD demonstrate that overexpression of human TREM2 modulates microglial function and morphology, and improves memory and AD pathology (Lee and others 2018). In mice, Trem2-null microglia also fail to acquire an activated transcriptional signature upon various neuropathological insults (Keren-Shaul and others 2017). However, while loss of TREM2 expression results in a detrimental phenotype in mouse models of amyloidosis, in tau models TREM2 deletion as well as mutant TREM2 appears to be beneficial and protects against neurodegeneration at advanced stages of the disease (Gratuze and others 2020; Leyns and others 2017; Sayed and others 2018). This pathology-specific discrepancy highlights two shortcomings in the field: (1) that we are lacking a robust mouse model of LOAD that encapsulates both amyloid and tau pathologies and (2) that we must comprehensively phenotype individuals selected for clinical trials and treatment to better understand AD stage or pathology-specific effects of therapeutic manipulation.

In fact, clinical trials are currently examining the efficacy of a TREM2 agonistic antibody as a therapeutic approach for AD (Alector-Inc. 2020). This antibody binds TREM2 and activates its signaling, resulting in increased phosphorylation of the downstream effector Syk. In early clinical trial findings, soluble TREM2 was decreased in the cerebrospinal fluid (CSF) following treatment, while a colony-stimulating factor 1 receptor (CSF1R) fragment was increased in the CSF (Wang and others 2020). This therapy aims to increase microglial activation via TREM2 signaling, and may compensate for the reduced microglial responsiveness associated with the TREM2 risk variant, and perhaps the other AD variants discussed in this review. However, close examination for phase and pathology specific effects will be necessary to determine whether activating TREM2 may be beneficial at different stages of the disease.

Interestingly, the CD33 and TREM2 genetic associations converge such that CD33 modulates TREM2 expression, and TREM2 functions downstream of CD33. In a targeted study, the CD33 LOAD risk variant was found to be associated with increased TREM2 surface expression. This association was greatly reduced after adjusting for membrane-bound CD33 expression (Chan and others 2015). Importantly, the CD33 risk variant is associated with seven-fold increased CD33 surface expression (Bradshaw and others 2013). These findings suggest that the degree of membrane-bound CD33 expression mediates the interaction between the CD33 risk variant and TREM2 expression. Additionally, TREM2 has been demonstrated to function downstream of CD33. While knockout of CD33 in a mouse model of AD was found to mitigate pathology, this effect was abrogated upon subsequent knockout of TREM2 (Griciuc and others 2019). Similarly, changes in gene expression observed in CD33 knockout AD mice were found to be dependent on TREM2. The convergence of TREM2 and CD33 highlights the relevance of these two signaling pathways in facilitating LOAD risk, and offers a potentially advantageous point for therapeutic modulation.

TREM1

While not a GWAS association, an intronic variant of TREM1 (rs6910730) that confers AD risk was identified in a targeted analysis of the TREM locus in relation to features of AD (Replogle and others 2015). This variant has been found to decrease TREM1 surface expression in myeloid cells, and is associated with increased neuritic and diffuse plaques, Aβ density, and cognitive decline. TREM1 has been shown to facilitate Aβ phagocytosis (Jiang and others 2016). As this intronic variant is associated with amyloid pathology but not tau pathology, TREM1 may be important in the interaction between Aβ and the innate immune system.

In addition to decreasing TREM1 surface expression, the AD-associated TREM1 intronic variant also decreases the ratio of TREM1:TREM2 expression in myeloid cells (Chan and others 2015). In contrast, another TREM1 variant (rs2627567) that decreases TREM1 expression but does not alter the TREM1:TREM2 ratio was not associated with AD. This finding suggests that the TREM1 association with AD may be mediated not by a simple reduction in TREM1 expression but by influencing the balance of signaling molecules that affect myeloid activation (Chan and others 2015). This concept aligns with the recurring theme of an altered balance of activating and inhibitory signaling molecules in the AD-associated risk variants discussed thus far. Importantly, TREM1 and TREM2 both signal through the adaptor protein TYROBP (Arts and others 2012; Bouchon and others 2001). While crosstalk between TREM1 and TREM2 has not been thoroughly explored, TREM2 appears to negatively regulate the TREM1 proinflammatory response, such that TREM2 signaling reduces TREM1-induced TNFα production (Takahashi and others 2005). This finding highlights the importance of a balanced TREM1:TREM2 ratio for myeloid cell signaling.

The myeloid literature is clear in reporting TREM1 as an important receptor that participates in augmenting proinflammatory responses, and a null allele of TREM1 in mice results in diminished inflammatory reactions (Colonna 2003; Gibot and others 2007; Hommes and others 2014; Lagler and others 2009; Sharif and Knapp 2008). TREM1 signaling via TYROBP phosphorylation regulates NF-κB activation and expression of inflammatory genes (Arts and others 2012). TREM1 amplifies the innate immune response in part through synergistic crosstalk with pattern recognition receptors such as TLRs, which facilitate the response to a diverse range of pathogenic and inflammatory stimuli (Bleharski and others 2003; Dower and others 2008). In this way, TREM1 has been shown to play a regulatory role in influencing inflammatory disease outcomes, likely through its known role in activating antigen-presenting cells (APCs) and promoting their survival through several anti-apoptotic mechanisms (Cai and others 2013; Numata and others 2011; Prieto and others 2010; Yuan and others 2014). TREM1 may exert protective effects in AD by augmenting microglial activation and phagocytosis, as well as enhance microglial fitness through its antiapoptotic effects. TREM1 expression is decreased in the context of the rs6910730 risk variant, potentially impairing such effects and conferring AD risk (Chan and others 2015; Replogle and others 2015). However, few human studies have examined this topic, and the role of TREM1 in AD pathogenesis remains unclear.

SPI1

The SPI1 SNP rs1057233 has been established as a genetic risk factor for AD through GWAS. SPI1 encodes the protein PU.1, a well-known hematopoietic transcription factor that controls the expression of hundreds of myeloid genes (Huang and others 2017; Kunkle and others 2019). The SPI1 AD risk variant increases PU.1 expression (Huang and others 2017). PU.1 is involved in the early stages of hematopoietic progenitor cell development and in the later stages of myeloid cell maturation (Friedman 2007). Granulocyte-macrophage progenitor cells from mice deficient in PU.1 have impaired commitment to a macrophage fate, demonstrating that PU.1 tightly regulates monocyte and macrophage differentiation (Dakic and others 2005). Furthermore, PU.1 appears to be a nonclassical pioneer transcription factor, meaning it can reprogram cell-type fates (Minderjahn and others 2020). This capability is likely due in part to its ability to bind the chromatin remodeling complex SWI/SNF.

PU.1 regulates the activation and fitness of macrophages and monocytes, in part by regulating the expression of innate immune pattern recognition receptors including TLR1, TLR2, and TLR4. PU.1 binds elements within the proximal human TLR4 promoter and directly regulates its expression (Rehli and others 2000). In human monocytes, GM-CSF was found to downregulate TLR1, TLR2, and TLR4 in a PU.1-dependent manner (Sadeghi and others 2016). Reduced expression of these receptors resulted in decreased downstream p38 and ERK signaling and impaired proinflammatory cytokine production following exposure to lipopolysaccharide or lipoteichoic acid (Sadeghi and others 2016). In these ways, PU.1 acts as a critical regulator of innate immune responses to environmental triggers.

Interestingly, PU.1 also regulates the expression of several LOAD-associated genes (Fig. 1). This comes as little surprise, as PU.1 regulates many myeloid genes, and many LOAD-risk associated variants are enriched in myeloid cells. An overwhelming 60 genes that both lie in LOAD-associated loci and are expressed in microglia contain PU.1 binding sites (Huang and others 2017). Specifically, PU.1 regulates TREM2 expression, evidenced by studies in which silencing of SPI1 with siRNA led to reduced TREM2 expression (Rustenhoven and others 2018). PU.1 has also been established as a negative regulator of TREM1 (Zeng and others 2007). As the SPI1 risk allele increases PU.1 expression, the risk allele is associated with decreased TREM1 expression (Huang and others 2017). PU.1 also regulates TYROBP, through which TREM1 and TREM2 both signal, solidifying its regulation of the TREM1/TREM2/TYROBP axis at several points (Huang and others 2017). CD33 expression is also regulated by PU.1, and PU.1 is necessary for CD33 expression in myeloid cells (Bodger and Hart 1998). PU.1 additionally modulates PTPN6, the gene which encodes the protein SHP-1 (Wlodarski and others 2007). SHP-1 is the major downstream signaling molecule of CD33, demonstrating the importance of PU.1 in regulating several points in the CD33 signaling cascade (Taylor and others 1999). The PILRA gene also contains a PU.1 binding site, and PU.1 positively regulates its expression (Huang and others 2017). These findings clarify that PU.1 presents a point of convergence for regulating several LOAD genetic risk factors and their associated signaling cascades. Accordingly, PU.1 may serve as an advantageous therapeutic target for LOAD.

Figure 1.

The LOAD genetically associated transcription factor PU.1 modulates the expression of other genetically associated genes (TREM2, CD33, PILRA, TREM1, and IL34) and their key signaling partners.

PLCG2

PLCG2, which encodes an enzyme that catalyzes the hydrolysis of phospholipids, was found to have a rare protective variant that is genetically associated with LOAD (Sims and others 2017). An independent association of PLCG2 with LOAD was also identified in a GWAS by proxy (Marioni and others 2018). A recent preprint manuscript confirmed the protective variant and the GWAS by proxy association in the promoter region, as well as identified an additional independent association with LOAD (de Rojas and others 2020). The rare AD-protective variant leads to an amino acid change, proline 522 to an arginine (P522R). This alteration increases PLCγ2 function, and knockin mice carrying the P522R mutation exhibit an enhanced acute inflammatory response (Magno and others 2019; Takalo and others 2020). This direction of effect is consistent with those of other AD genetic variants in which decreased microglial activation is associated with disease risk, and activation is associated with protection.

Interestingly, there are other known hypermorphic mutations in PLCG2 that are associated with autoinflammatory disease (Zhou and others 2012). The AD-protective variant has no such association with autoinflammation; increased PLCγ2 activity in the context of this variant is protective for AD neurodegeneration and supportive of longevity (van der Lee and others 2019). A more comprehensive understanding of these mutations is needed to differentiate detrimental and protective effects of increased PLCγ2 signaling in disease risk. It is possible that the mutations preferentially influence PLCγ2 participation in different signaling pathways, or that they disrupt interactions with regulatory binding partners that are differentially expressed in the various innate immune cells. Differential regulation of signaling pathways is possible, as the S707Y autoinflammatory mutation is located in the c-terminal SH domain, while the AD P522R mutation is located between the catalytic phosphatidylinositol-specific phospholipase C X domain and the regulatory N-terminal SH2 domain. Similarly, these mutations might differentially disrupt interactions between various domains and disturb autoinhibition.

PLCγ2 is a downstream signaling molecule of TREM2, TREM1, and several TLRs (Andreone and others 2020; Arts and others 2012; Bae and others 2017; Chae and others 2015; Chiang and others 2012; Zhu and others 2018). Its participation in TLR and TREM signaling pathways likely occurs through distinct mechanisms, as its ability to induce proinflammatory responses downstream of TLR2 signaling occurs through a TREM2 independent mechanism (Andreone and others 2020). In iPSC-derived microglia, PLCγ2 was shown to mediate TREM2 regulation of cholesterol metabolism (Andreone and others 2020). Further understanding of these genetic associations will clarify the role of PLCγ2 in AD and whether it may be modulated as a potential therapeutic target (de Rojas and others 2020).

IL34

An IL34 variant resulting in an early stop codon (Tyr213Ter) has been genetically associated with increased risk for AD (de Rojas and others 2020). IL-34 is produced by neurons, and may act as a neuronal signal to microglia (Mizuno and others 2011). IL-34 signals through the CSF1R, as does colony stimulating factor 1 (CSF-1), although IL-34 binds CSF1R with a higher affinity than CSF-1 (Lin and others 2008; Liu and others 2012). Both IL-34 and CSF-1 induce proliferation and cell survival in monocytes and other myeloid cells (Chitu and Stanley 2006; Dai and others 2002; Lin and others 2008; Wei and others 2010). While knockout of CSF-1 did not affect microglial development in mice, knockout of CSF1R resulted in microglial depletion (Wang and others 2012). These findings suggest that CSF1R-mediated signaling is essential for microglial development independent of CSF-1, with IL-34 potentially mediating this effect.

Interestingly, IL34 mRNA was found to be decreased in postmortem inferior temporal gyrus samples from AD patients compared to samples from MCI and controls, whereas CSF1R and CSF1 mRNA were both found to be increased (Walker and others 2017). This finding is consistent with the genetically associated variant resulting in an early stop codon, suggesting that decreased IL-34 expression is associated with AD risk. Of note, IL-34 and CSF1R functionally converge with other AD genetically associated molecules. Both IL-34 and CSF1R expression were found to be regulated by SPI1/PU.1 in BV2 cells (Huang and others 2017). Additionally, CSF1R interacts with PLCγ2 as a binding partner (Bourette and others 1997). Treatment with IL-34 was found to promote microglial proliferation and uptake of soluble oligomeric Aβ in vitro, as well as to reduce neurotoxic effects of Aβ (Ma and others 2012). Additionally, IL-34 signaling decreases oxidative stress and facilitates neuronal survival (Luo and others 2013; Ma and others 2012). IL-34 may similarly provide cognitive benefits, as intracerebral administration of IL-34 reduced impairments in associative learning in AD mice (Mizuno and others 2011). These findings further support the protective effects of IL-34 in AD, as well as the AD genetic risk associated with the truncating Tyr213Ter mutation.

Despite these protective effects of IL-34 signaling via CSF1R, a clinical trial is currently underway for a CSF1R antagonist as a therapeutic approach for AD, to reduce the microglial population as well as to suppress their activity (Raymont 2020). As decreased IL-34 signaling via CSF1R is associated with AD risk as reviewed above, it may instead be therapeutically advantageous to activate CSF1R in AD. However, ablation of an immune cell type has been found to be a beneficial therapeutic approach in some autoimmune diseases, such as the use of B cell depletion therapy in the treatment of multiple sclerosis (Greenfield and Hauser 2018).

HLA

Somewhat surprisingly, the human leukocyte antigen (HLA) complex region is another genetically associated locus that implicates the innate immune system in LOAD pathogenesis (Lambert and others 2013). HLA is one of the strongest genetic associations for most autoimmune diseases, in which the adaptive immune system engages in dysregulated, self-reactive responses. HLA variants have now been identified as genetic risk factors for several neurodegenerative diseases (Barcellos and others 2006; Hamza and others 2010; Jansen and others 2019; Lambert and others 2009; Nalls and others 2014; Steele and others 2017; Wissemann and others 2013). Genetic variants in both major histocompatibility complex (MHC) class I and class II have been associated with LOAD (Steele and others 2017). MHC class II is almost exclusively expressed on innate immune cells, and therefore is predominantly expressed on microglia within the CNS. MHC class I has also been identified as a microglial enriched gene in brains from aged individuals, although it can be expressed by all cell types (Olah and others 2018). Antigen-presenting cells (APCs) including microglia, utilize MHC class II molecules (HLA-DR, HLA-DQ, HLA-DP) to present antigen to CD4+ helper T cells. Similarly, cells utilize MHC class I molecules (HLA-A, HLA-B, HLA-C) to present antigen to CD8+ cytotoxic T cells.

Both an MHC class I haplotype ((A*03:01~B*07:02) and a class II haplotype (DRB1*15:01- DQA1*01:02- DQB1*06:02 (DR15)) were found to be genetically associated with LOAD (Kunkle and others 2019; Steele and others 2017). This same class II haplotype, DR15, has also been associated with Parkinson’s disease and multiple sclerosis (Barcellos and others 2006; Wissemann and others 2013). Different HLA molecules have varying specificities with which they bind antigen peptides. The association of DR15 with autoimmune disease suggests its proclivity in presenting self-peptides, in some circumstances initiating a self-reactive adaptive immune response. It has also been suggested that a large hydrophobic pocket in DR15’s peptide-binding groove allows for metabolite binding in addition to peptides (Misra and others 2019). These small molecules are abundant and are often derived from the human microbiome. NAD⁺ is one such metabolite that is predicted to bind this pocket (Misra and others 2019). A number of studies, including mouse model studies of AD, have established decreased NAD⁺ as a hallmark of aging and have linked its decrease to the development of chronic diseases (Braidy and others 2018; Demarest and others 2019; Lautrup and others 2019; Schultz and Sinclair 2016).

These HLA genetic findings solidify the involvement of innate immune cells, presumably microglia, in LOAD susceptibility. They also suggest an important role for the interface between the innate and adaptive immune systems in disease pathogenesis and a role for microglia as APCs. This microglial T cell interface may be especially important in the hippocampus, where extravascular CD4+ and CD8+ T cells have been found to be increased in LOAD (Itagaki and others 1988; Merlini and others 2018; Rogers and others 1988). It has been reported that there are more CD8+ T cells than CD4+ T cells among the LOAD infiltrating T cells (Merlini and others 2018). However, no detailed characterization of the T cell phenotype has yet been performed. Future work may reveal the nature of their interaction with microglia and their role in brain neuropathology.

Convergence of GWAS Loci and Functional Outcomes

Many of the proteins genetically associated with LOAD reviewed here appear to interact and share pathways or regulatory mechanisms (Fig. 2). These findings present several points of convergence that may be strategic targets for therapeutic intervention. In summary, CD33 functions upstream of TREM2 and regulates TREM2 expression (Chan and others 2015; Griciuc and others 2019). The TREM2/TREM1/TYROBP axis described in this review is tightly interconnected and appears to be modulated by PU.1 (Arts and others 2012; Bouchon and others 2001; Chan and others 2015; Huang and others 2017). PLCγ2 functions downstream of TREM1, TREM2, and TYROBP, providing another point in this axis that may be advantageous to manipulate therapeutically (Andreone and others 2020; Arts and others 2012). PU.1 also regulates CD33, PILRA, IL-34, and CSF1R (the IL-34 receptor). Interestingly, the SPI1 AD risk variant appears to increase expression of CD33 and PILRA and to decrease expression of TREM1 and IL-34. These outcomes are consistent with the alterations in protein expression or function observed with the CD33, PILRA, TREM1, and IL34 risk variants, further supporting the association of these patterns of protein expression with LOAD risk. PU.1 may therefore present a particularly advantageous therapeutic target as it modulates the expression of several genetically associated LOAD risk genes.

Figure 2.

Many of the proteins that are coded for by LOAD-associated genes interact with each other and are in shared microglial signaling pathways; proteins in dark purple are encoded by genes in LOAD genetic loci, proteins in light purple are critical binding partners.

In addition to the convergence of molecular pathways, it is valuable to consider whether these risk variants converge on a shared microglial functional outcome that may contribute to AD pathogenesis. Of the reviewed genetically identified variants enriched in microglia for which a direction of effect has been established, the findings suggest that a less responsive microglial phenotype is associated with AD risk. Supporting this trend, the risk alleles in CD33, TREM1, TREM2, PILRA, SPI1, and IL34 converge to suppress immune signaling, while the protective allele in PLCG2 enhances the immune response (Fig. 3; Bradshaw and others 2013; de Rojas and others 2020; Huang and others 2017; Kleinberger and others 2014; Magno and others 2019; Rathore and others 2018; Replogle and others 2015). These findings are in stark contrast to the dogma that arose in the pre-GWAS era that concluded that the immune system needed to be suppressed (Akiyama and others 2000).

Figure 3.

AD genetic risk is associated with decreased signaling of activating microglial receptors, and increased signaling of inhibitory receptors, resulting in a dampened microglial response in AD.

The AD brain environment prominently features sterile inflammatory stimuli, such as highly inflammatory molecules released by deteriorating or dead cells and misfolded protein aggregates. Microglial suppression resulting in the inability to effectively resolve these sterile inflammatory processes, mitigate pathology, and maintain brain homeostasis may heighten susceptibility to neuronal injury. Indeed, the inability to clear pathology is thought to be critical to AD progression (Fiala and others 2005; Fiala and others 2007; Hickman and others 2008; Wildsmith and others 2013; Yoon and Jo 2012). This phenotype may be mirrored by senescent deterioration associated with aging, the primary risk factor for Alzheimer’s disease (Damani and others 2011; Hefendehl and others 2014; Kronenberg and others 2017; Mosher and Wyss-Coray 2014; Streit 2006; Streit and others 2004). Therefore, the genetic findings discussed in this review support the loss of microglial fitness in AD, and may converge with the functional outcomes of other known LOAD risk factors on an ineffective microglial phenotype. A number of ongoing clinical trials, as mentioned above, will shed more light on this topic.

Genetics, Microglial Immune Responses, and the Pathogen Hypothesis

In addition to the amyloid cascade and tau hypotheses, it has been suggested that LOAD pathogenesis may be triggered by an infectious agent, a controversial and debated concept referred to as the “pathogen hypothesis” (Li and others 2018). Viral, fungal, and bacterial pathogens have been associated with LOAD, including herpes simplex virus type 1 (HSV-1), human herpesvirus 6A (HHV-6A), and HHV-7, as well as the intracellular bacteria Chlamydophila pneumoniae and Porphyromonas gingivalis (Dominy and others 2019; Mawanda and Wallace 2013; Readhead and others 2018). Many of the pathogens that have been associated with LOAD are known to be neuroinvasive, or found in the CNS (Stevens 1993). However, it remains unclear which CNS cell types are susceptible to infection, and whether infection contributes to impaired cellular fitness in LOAD pathogenesis. While HHV6 was not found in higher abundance in AD brains compared to controls, measures of viral load would not rule out any slow pathogenic processes toward AD set in motion by infection decades in the past, or any interaction between AD genetic susceptibility and these environmental exposures(Allnutt and others 2020). A number of studies have also found an association between infectious agents and the accumulation of amyloid (Eimer and others 2018; Soscia and others 2010). Furthermore, it has been demonstrated that amyloid endogenously functions as an anti-microbial peptide. Amyloid-beta plaque deposition may represent a mechanism by which to sequester and eliminate pathogens (Eimer and others 2018; Kumar and others 2016; Soscia and others 2010; White and others 2014). Several outstanding questions for further research remain, including how microglia respond to these pathogens, whether microglia themselves are infected with these pathogens, whether there is an ongoing immune response to these pathogens in the CNS, as well as the degree to which pathogens are in complex with amyloid in the LOAD brain.

In human subjects, LOAD genetic risk variants appear to converge to dampen the innate immune response, as detailed in this review. Microglial enrichment for these risk variants is suggestive of a suppressed microglial phenotype as a pathological mechanism of disease. Importantly, age is the number one risk factor for LOAD. Age also profoundly impacts the immune system’s fitness, similarly converging with the functional outcomes of the reviewed LOAD genetic risk variants to suppress immune function (Del Giudice and others 2018). An aged immune system, as well as carriage of suppressive LOAD genetic variants, may alter the local microglial response to infectious and sterile inflammatory stressors. Some AD risk variants may also directly alter host-pathogen interactions. For example, glycoprotein B of herpes simplex virus 1 (HSV-1) has been identified as a ligand for PILRA (Satoh and others 2008). As the AD-protective variant leads to reduced PILRA ligand binding, viral entry into the host cell via glycoprotein B may be limited (Rathore and others 2018). The HLA DR15 association has been shown to lead to a higher viral load and attenuated immune control of EBV in a model of multiple sclerosis (Zdimerova and others 2021). Additionally, the HLA genetic associations with LOAD raise the question of what antigens are potentially being presented by microglia to CNS-infiltrating T cells in AD. Are these self-antigens modified by the disease process such as misfolded proteins, or peptides from pathogens that are residing in the CNS? Further study of microglial-T cell interactions will lead to a better understanding of the inflammatory stimuli present in the CNS that may be contributing to LOAD, and the immune system’s ability to respond to these challenges.

Rather than an active infection contributing to LOAD, it is possible that infectious processes over a lifetime may act as triggering events in AD pathogenesis, or exacerbate preexisting disease processes. Inflammatory challenges, both sterile and infectious, are known to result in long term alteration of immune dynamics and may facilitate immunosenescence (Coppé and others 2010; Dregan and others 2014; Effros 2016; Fairweather and others 2012; Furman and others 2019; Jose and others 2017; Pawelec and others 2005; Root-Bernstein and Fairweather 2014; Stack and others 2012). An inflammatory insult may prime microglial activation or dysfunction, dictated by factors including genetic susceptibility and the nature of the challenge. These challenges may alter functional programs of microglia, which may be potentiated long after the resolution of an infection or sterile inflammatory process. In this way, a lifetime burden of inflammatory challenges may alter microglial fitness and function. Such alterations, compounded in the context of immune aging and suppressive innate immune genetic risk variants, may contribute to disease. Importantly, inflammatory challenges need not be restricted locally within the CNS. Systemic inflammatory challenges have been shown to result in sustained alterations of microglial functional programs in mice (Wendeln and others 2018). In humans, systemic infections have been demonstrated to modify the late-stage immune response in AD, resulting in an immunosuppressive brain environment, a finding that converges with the impaired immune environment associated with AD risk as outlined in this review (Rakic and others 2018).

Concluding Thoughts

The genetic variants described in this review implicate a role for suppressed innate immune function, particularly in microglia, in LOAD pathogenesis. Innate immune dysregulation likely results in microglial inability to mitigate pathologies and resolve inflammatory processes, whether sterile or pathogenic, in AD. In the context of age and the accompanying immunosenescence, as well as immune alterations resulting from pathogenic and sterile inflammatory challenges, a dampened microglial phenotype may give way to microglial failure. This microglial failure may lead to a loss of homeostasis in the local brain environment and increased vulnerability to neuronal injury and toxicity. As AD is a progressive disease with age as its primary risk factor, and as microglia live on average for 4 years and up to two decades, it is particularly important to consider the dynamics of increased suppressive tone in microglia over several years of inflammatory challenges (Réu and others 2017). Future research must uncover the consequences of the dampened immune tone resulting from these genetic variants on microglial function, and how impairments in microglial capability participate in LOAD pathogenesis.

Footnotes

Acknowledgements

Figures 2 and 3 and Box 2 were created in BioRender. Figure 1 was made using String, Cytoscape, and BioRender.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: EMB is a founder of IMAD Therapeutics. ZKC has no conflicts of interest to disclose.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Elizabeth M. Bradshaw

References

Akiyama

Barger

Barnum

Bradt

Bauer

Cole

, and others. 2000. Inflammation and Alzheimer’s disease. Neurobiol Aging 21(3):383–421.

Alector-Inc. 2020. A phase 2 study to evaluate efficacy and safety of AL002 in participants with early Alzheimer’s disease (INVOKE-2). Available from: https://clinicaltrials.gov/ct2/show/NCT04592874

Alliot

Godin

Pessac

1999. Microglia derive from progenitors, originating from the yolk sac, and which proliferate in the brain. Brain Res Dev Brain Res 117(2):145–52.

Allnutt

Johnson

Bennett

Connor

Troncoso

Pletnikova

, and others. 2020. Human herpesvirus 6 detection in Alzheimer’s disease cases and controls across multiple cohorts. Neuron 105(6):1027–35.e1022.

Alzheimer

. 1907. Uber eine eigenartige erkrankung der hirnrinde. Zentralbl Nervenh Psych 18:177–9.

Alzheimer

Stelzmann

Schnitzlein

Murtagh

. 1995. An English translation of Alzheimer’s 1907 paper, “Uber eine eigenartige erkankung der hirnrinde”. Clin Anat 8(6):429–31.

Alzheimer’s Association. 2015. Alzheimer’s disease facts and figures. Alzheimers Dement. 11(3):332–84.

Andreone

Przybyla

Llapashtica

Rana

Davis

van Lengerich

, and others. 2020. Alzheimer’s-associated plcγ2 is a signaling node required for both trem2 function and the inflammatory response in human microglia. Nat Neurosci 23(8):927–38.

Andrews

Fulton-Howard

Goate

2020. Interpretation of risk loci from genome-wide association studies of Alzheimer’s disease. Lancet Neurol 19(4):326–35.

10.

Arts

RJW

Joosten

LAB

Meer

JWM

Netea

. 2012. Trem-1: intracellular signaling pathways and interaction with pattern recognition receptors. J Leukoc Biol 93(2):209–15.

11.

Askew

Gomez-Nicola

2018. A story of birth and death: insights into the formation and dynamics of the microglial population. Brain Behav Immun 69:9–17.

12.

Avagyan

Goldenson

Tse

Masoumi

Porter

Wiedau-Pazos

, and others. 2009. Immune blood biomarkers of Alzheimer disease patients. J Neuroimmunol 210(1–2):67–72.

13.

Bae

Lee

Jung

Lee

Suh

. 2017. Phospholipase cγ in toll-like receptor-mediated inflammation and innate immunity. Adv Biol Regul 63:92–7.

14.

Barcellos

Sawcer

Ramsay

Baranzini

Thomson

Briggs

, and others. 2006. Heterogeneity at the hla-drb1 locus and risk for multiple sclerosis. Hum Mol Genet 15(18):2813–24.

15.

Bateman

Xiong

Benzinger

TLS

Fagan

Goate

Fox

, and others. 2012. Clinical and biomarker changes in dominantly inherited Alzheimer’s disease. N Engl J Med 367(9):795–804.

16.

Bertram

Lange

Mullin

Parkinson

Hsiao

Hogan

, and others. 2008. Genome-wide association analysis reveals putative Alzheimer’s disease susceptibility loci in addition to apoe. Am J Hum Genet 83(5):623–32.

17.

Bhattacherjee

Rodrigues

Jung

Luzentales-Simpson

Enterina

Galleguillos

, and others. 2019. Repression of phagocytosis by human cd33 is not conserved with mouse cd33. Commun Biol 2:450.

18.

Bleharski

Kiessler

Buonsanti

Sieling

Stenger

Colonna

2003. A role for triggering receptor expressed on myeloid cells-1 in host defense during the early-induced and adaptive phases of the immune response. J Immunol 170(7):3812–8.

19.

Bodger

Hart

DNJ

. 1998. Molecular cloning and functional analysis of the cd33 promoter. Br J Haematol 102(4):986–95.

20.

Bouchon

Hernández-Munain

Cella

Colonna

2001. A dap12-mediated pathway regulates expression of cc chemokine receptor 7 and maturation of human dendritic cells. J Exp Med 194(8):1111–22.

21.

Bourette

Myles

Choi

Rohrschneider

. 1997. Sequential activation of phoshatidylinositol 3-kinase and phospholipase c-gamma2 by the M-CSF receptor is necessary for differentiation signaling. EMBO J. 16(19):5880–93.

22.

Bradshaw

Chibnik

Keenan

Ottoboni

Raj

Tang

, and others. 2013. CD33 Alzheimer’s disease locus: altered monocyte function and amyloid biology. Nat Neurosci 16(7):848–50.

23.

Braidy

Grant

Sachdev

. 2018. Nicotinamide adenine dinucleotide and its related precursors for the treatment of Alzheimer’s disease. Curr Opin Psychiatry 31(2):160–6.

24.

Brookmeyer

Evans

Hebert

Langa

Heeringa

Plassman

. 2011. National estimates of the prevalence of Alzheimer’s disease in the United States. Alzheimers Dement 7(1):61–73.

25.

Cai

Chen

Liu

Hou

Zeng

2013. Activation of triggering receptor expressed on myeloid cells-1 protects monocyte from apoptosis through regulation of myeloid cell leukemia-1. Anesthesiology 118(5):1140–9.

26.

Cao

Crocker

. 2011. Evolution of CD33-related siglecs: regulating host immune functions and escaping pathogen exploitation? Immunology. 132(1):18–26.

27.

Chae

Park

Hwang

Hoffmann

Kehrl

. 2015. Connecting two pathways through Ca²⁺ signaling: Nlrp3 inflammasome activation induced by a hypermorphic PLCG2 mutation. Arthritis Rheumatol 67(2):563–7.

28.

Chan

White

Winn

Cimpean

Replogle

Glick

, and others. 2015. CD33 modulates TREM2: convergence of Alzheimer loci. Nat Neurosci 18(11):1556–8.

29.

Chen

Eksioglu

Zhou

Zhang

Djeu

Fortenbery

, and others. 2013. Induction of myelodysplasia by myeloid-derived suppressor cells. J Clin Invest 123(11):4595–611.

30.

Chiang

Veckman

Limmer

David

2012. Phospholipase cγ-2 and intracellular calcium are required for lipopolysaccharide-induced toll-like receptor 4 (TLR4) endocytosis and interferon regulatory factor 3 (IRF3) activation. J Biol Chem 287(6):3704–9.

31.

Chitu

Stanley

. 2006. Colony-stimulating factor-1 in immunity and inflammation. Curr Opin Immunol 18(1):39–48.

32.

Claes

Van Den Daele

Boon

Schouteden

Colombo

Monasor

, and others. 2019. Human stem cell-derived monocytes and microglia-like cells reveal impaired amyloid plaque clearance upon heterozygous or homozygous loss of TREM2. Alzheimers Dement 15(3):453–64.

33.

Colonna

. 2003. Trems in the immune system and beyond. Nat Rev Immunol 3(6):445–53.

34.

Coppé

Desprez

Krtolica

Campisi

2010. The senescence-associated secretory phenotype: the dark side of tumor suppression. Annu Rev Pathol 5:99–118.

35.

Crocker

Paulson

Varki

2007. Siglecs and their roles in the immune system. Nat Rev Immunol 7(4):255–66.

36.

Crocker

Redelinghuys

2008. Siglecs as positive and negative regulators of the immune system. Biochem Soc Trans 36(6):1467–71.

37.

Dai

Ryan

Hapel

Dominguez

Russell

Kapp

, and others. 2002. Targeted disruption of the mouse colony-stimulating factor 1 receptor gene results in osteopetrosis, mononuclear phagocyte deficiency, increased primitive progenitor cell frequencies, and reproductive defects. Blood 99(1):111–20.

38.

Dakic

Metcalf

Di Rago

Mifsud

Nutt

. 2005. Pu.1 regulates the commitment of adult hematopoietic progenitors and restricts granulopoiesis. J Exp Med 201(9):1487–502.

39.

Damani

Zhao

Fontainhas

Amaral

Fariss

Wong

. 2011. Age-related alterations in the dynamic behavior of microglia. Aging Cell 10(2):263–76.

40.

de Craen

Gussekloo

Vrijsen

Westendorp

. 2005. Meta-analysis of nonsteroidal antiinflammatory drug use and risk of dementia. Am J Epidemiol 161(2):114–20.

41.

de Rojas

Moreno-Grau

Tesi

Grenier-Boley

Andrade

Jansen

, and others. 2020. Common variants in Alzheimer’s disease: novel association of six genetic variants with ad and risk stratification by polygenic risk scores. medRxiv. https://doi.org/10.1101/19012021

42.

Del Giudice

Goronzy

Grubeck-Loebenstein

Lambert

Mrkvan

Stoddard

, and others. 2018. Fighting against a protean enemy: immunosenescence, vaccines, and healthy aging. NPJ Aging Mech Dis 4:1.

43.

Demarest

Babbar

Okur

Dan

Croteau

Fakouri

, and others. 2019. NAD+ metabolism in aging and cancer. Annu Rev Cancer Biol 3(1):105–30.

44.

Dominy

Lynch

Ermini

Benedyk

Marczyk

Konradi

, and others. 2019. Porphyromonas gingivalis in Alzheimer’s disease brains: evidence for disease causation and treatment with small-molecule inhibitors. Sci Adv 5(1):eaau3333.

45.

Dower

Ellis

Saraf

Jelinsky

Lin

. 2008. Innate immune responses to trem-1 activation: overlap, divergence, and positive and negative cross-talk with bacterial lipopolysaccharide. J Immunol 180(5):3520–34.

46.

Dregan

Charlton

Chowienczyk

Gulliford

. 2014. Chronic inflammatory disorders and risk of type 2 diabetes mellitus, coronary heart disease, and stroke: a population-based cohort study. Circulation 130(10):837–44.

47.

Duyckaerts

Delatour

Potier

. 2009. Classification and basic pathology of Alzheimer disease. Acta Neuropathol 118(1):5–36.

48.

Effros

. 2016. The silent war of CMV in aging and HIV infection. Mech Ageing Dev 158:46–52.

49.

Eimer

Vijaya Kumar

Navalpur Shanmugam

Rodriguez

Mitchell

Washicosky

, and others. 2018. Alzheimer’s disease-associated β-amyloid is rapidly seeded by herpesviridae to protect against brain infection. Neuron 99(1):56–63.e53.

50.

Elmaleh

. 2020. Safety and efficacy study of ALZT-OP1 in subjects with evidence of early Alzheimer’s disease (COGNITE). Available from: https://www.clinicaltrials.gov/ct2/show/study/NCT02547818?term=COGNITE&rank=1

51.

Estus

Shaw

Devanney

Katsumata

Press

Fardo

. 2019. Evaluation of CD33 as a genetic risk factor for Alzheimer’s disease. Acta Neuropathol 138(2):187–99.

52.

Fairweather

Petri

Coronado

Cooper

. 2012. Autoimmune heart disease: role of sex hormones and autoantibodies in disease pathogenesis. Expert Rev Clin Immunol 8(3):269–84.

53.

Fiala

Cribbs

Rosenthal

Bernard

2007. Phagocytosis of amyloid-beta and inflammation: two faces of innate immunity in Alzheimer’s disease. J Alzheimers Dis 11(4):457–63.

54.

Fiala

Lin

Ringman

Kermani-Arab

Tsao

Patel

, and others. 2005. Ineffective phagocytosis of amyloid-beta by macrophages of Alzheimer’s disease patients. J Alzheimers Dis 7(3):221–32.

55.

Friedman

. 2007. Transcriptional control of granulocyte and monocyte development. Oncogene 26(47):6816–28.

56.

Furman

Campisi

Verdin

Carrera-Bastos

Targ

Franceschi

, and others. 2019. Chronic inflammation in the etiology of disease across the life span. Nat Med 25(12):1822–32.

57.

Gibot

Massin

Marcou

Taylor

Stidwill

Wilson

, and others. 2007. TREM-1 promotes survival during septic shock in mice. Eur J Immunol 37(2):456–66.

58.

Ginhoux

Greter

Leboeuf

Nandi

See

Gokhan

, and others. 2010. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science 330(6005):841–5.

59.

Ginhoux

Lim

Hoeffel

Low

Huber

2013. Origin and differentiation of microglia. Front Cell Neurosci 7:45.

60.

Graeber

Kösel

Egensperger

Banati

Müller

Bise

, and others. 1997. Rediscovery of the case described by Alois Alzheimer in 1911: historical, histological and molecular genetic analysis. Neurogenetics 1(1):73–80.

61.

Gratuze

Leyns

Sauerbeck

St-Pierre

Xiong

Kim

, and others. 2020. Impact of TREM2R47H variant on tau pathology-induced gliosis and neurodegeneration. J Clin Invest 130(9):4954–68.

62.

Greenfield

Hauser

. 2018. B-cell therapy for multiple sclerosis: entering an era. Ann Neurol 83(1):13–26.

63.

Griciuc

Patel

Federico

Choi

Innes

Oram

, and others. 2019. TREM2 acts downstream of CD33 in modulating microglial pathology in Alzheimer’s disease. Neuron 103(5):820–35.e827.

64.

Griciuc

Serrano-Pozo

Parrado Antonio

Lesinski Andrea

Asselin Caroline

Mullin

, and others. 2013. Alzheimer’s disease risk gene CD33 inhibits microglial uptake of amyloid beta. Neuron 78(4):631–43.

65.

Guerreiro

Wojtas

Bras

Carrasquillo

Rogaeva

Majounie

, and others. 2013. TREM2 variants in Alzheimer’s disease. N Engl J Med 368(2):117–27.

66.

Hamerman

Jarjoura

Humphrey

Nakamura

Seaman

Lanier

. 2006. Cutting edge: inhibition of TLR and FCR responses in macrophages by triggering receptor expressed on myeloid cells (TREM)-2 and DAP12. J Immunol 177(4):2051–5.

67.

Hammond

Robinton

Stevens

2018. Microglia and the brain: complementary partners in development and disease. Annu Rev Cell Dev Biol 34:523–44.

68.

Hamza

Zabetian

Tenesa

Laederach

Montimurro

Yearout

, and others. 2010. Common genetic variation in the HLA region is associated with late-onset sporadic Parkinson’s disease. Nat Genet 42(9):781–5.

69.

Hanisch

Kettenmann

2007. Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat Neurosci 10(11):1387–94.

70.

Harold

Abraham

Hollingworth

Sims

Gerrish

Hamshere

, and others. 2009. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer’s disease. Nat Genet 41(10):1088–93.

71.

Hefendehl

Neher

Sühs

Kohsaka

Skodras

Jucker

2014. Homeostatic and injury-induced microglia behavior in the aging brain. Aging Cell 13(1):60–9.

72.

Hickman

Allison

El Khoury

2008. Microglial dysfunction and defective β-amyloid clearance pathways in aging Alzheimer’s disease mice. J Neurosci 28(33):8354.

73.

Hoeffel

Wang

Greter

See

Teo

Malleret

, and others. 2012. Adult Langerhans cells derive predominantly from embryonic fetal liver monocytes with a minor contribution of yolk sac-derived macrophages. J Exp Med 209(6):1167–81.

74.

Hollingworth

Harold

Jones

Owen

Williams

2011a. Alzheimer’s disease genetics: current knowledge and future challenges. Int J Geriatr Psychiatry 26(8):793–802.

75.

Hollingworth

Harold

Sims

Gerrish

Lambert

Carrasquillo

, and others. 2011b. Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with Alzheimer’s disease. Nat Genet 43(5):429–35.

76.

Hommes

Hoogendijk

Dessing

Veer

CVT

Florquin

Colonna

, and others. 2014. Triggering receptor expressed on myeloid cells-1 (TREM-1) improves host defence in pneumococcal pneumonia. J Pathol 233(4):357–67.

77.

Huang

Marcora

Pimenova

Narzo

AFD

Kapoor

Jin

, and others. 2017. A common haplotype lowers PU.1 expression in myeloid cells and delays onset of Alzheimer’s disease. Nat Neurosci 20(8):1052–61.

78.

in t’ Veld

Ruitenberg

Hofman

Launer

van Duijn

Stijnen

, and others. 2001. Nonsteroidal antiinflammatory drugs and the risk of Alzheimer’s disease. N Engl J Med 345(21):1515–21.

79.

Ishida

Akita

Mori

Tanida

Toda

Inoue

, and others. 2014. Negative regulation of toll-like receptor-4 signaling through the binding of glycosylphosphatidylinositol-anchored glycoprotein, CD14, with the sialic acid-binding lectin, CD33. J Biol Chem 289(36):25341–50.

80.

Itagaki

McGeer

Akiyama

1988. Presence of T-cytotoxic suppressor and leucocyte common antigen positive cells in Alzheimer’s disease brain tissue. Neurosci Lett 91(3):259–64.

81.

Jack

Jr Bennett

Blennow

Carrillo

Dunn

Haeberlein

, and others. 2018. NIA-AA research framework: toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14(4):535–62.

82.

Jansen

Savage

Watanabe

Bryois

Williams

Steinberg

, and others. 2019. Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimer’s disease risk. Nat Genet 51(3):404–13.

83.

Jay

Hirsch

Broihier

Miller

Neilson

Ransohoff

, and others. 2017. Disease progression-dependent effects of TREM2 deficiency in a mouse model of Alzheimer’s disease. J Neurosci 37(3):637.

84.

Jiang

Zhang

Gao

Zhou

Zhu

, and others. 2016. TREM1 facilitates microglial phagocytosis of amyloid beta. Acta Neuropathol 132(5):667–83.

85.

Jonsson

Stefansson

Steinberg

Jonsdottir

Jonsson

Snaedal

, and others. 2013. Variant of TREM2 associated with the risk of Alzheimer’s disease. N Engl J Med 368(2):107–16.

86.

Jose

Bendickova

Kepak

Krenova

Fric

2017. Chronic inflammation in immune aging: role of pattern recognition receptor crosstalk with the telomere complex? Front Immunol 8:1078–8.

87.

Kawabori

Kacimi

Kauppinen

Calosing

Kim

Hsieh

, and others. 2015. Triggering receptor expressed on myeloid cells 2 (TREM2) deficiency attenuates phagocytic activities of microglia and exacerbates ischemic damage in experimental stroke. J Neurosci 35(8):3384–96.

88.

Keren-Shaul

Spinrad

Weiner

Matcovitch-Natan

Dvir-Szternfeld

Ulland

, and others. 2017. A unique microglia type associated with restricting development of Alzheimer’s disease. Cell 169(7):1276–90.e1217.

89.

Kleinberger

Yamanishi

Suárez-Calvet

Czirr

Lohmann

Cuyvers

, and others. 2014. TREM2 mutations implicated in neurodegeneration impair cell surface transport and phagocytosis. Sci Transl Med 6(243):243ra286.

90.

Kober

Alexander-Brett

Karch

Cruchaga

Colonna

Holtzman

, and others. 2016. Neurodegenerative disease mutations in Trem2 reveal a functional surface and distinct loss-of-function mechanisms. Elife 5.

91.

Konishi

Kiyama

2018. Microglial TREM2/DAP12 signaling: a double-edged sword in neural diseases. Front Cell Neurosci 12:206.

92.

Kronenberg

Uhlemann

Schöner

Wegner

Boujon

Deigendesch

, and others. 2017. Repression of telomere-associated genes by microglia activation in neuropsychiatric disease. Eur Arch Psychiatry Clin Neurosci 267(5):473–7.

93.

Kumar

Choi

Washicosky

Eimer

Tucker

Ghofrani

, and others. 2016. Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer’s disease. Sci Transl Med 8(340):340ra372.

94.

Kunkle

Grenier-Boley

Sims

Bis

Damotte

Naj

, and others. 2019. Genetic meta-analysis of diagnosed Alzheimer’s disease identifies new risk loci and implicates aβ, tau, immunity and lipid processing. Nat Genet 51(3):414–30.

95.

Lagler

Sharif

Haslinger

Matt

Stich

Furtner

, and others. 2009. TREM-1 activation alters the dynamics of pulmonary IRAK-M expression in vivo and improves host defense during pneumococcal pneumonia. J Immunol Baltim Md 1950 183(3):2027–36.

96.

Lajaunias

Dayer

Chizzolini

2005. Constitutive repressor activity of CD33 on human monocytes requires sialic acid recognition and phosphoinositide 3-kinase-mediated intracellular signaling. Eur J Immunol 35(1):243–51.

97.

Lambert

Heath

Even

Campion

Sleegers

Hiltunen

, and others. 2009. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer’s disease. Nat Genet 41(10):1094–9.

98.

Lambert

Ibrahim-Verbaas

Harold

Naj

Sims

Bellenguez

, and others. 2013. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 45(12):1452–8.

99.

Lanier

. 2009. DAP10- and DAP12-associated receptors in innate immunity. Immunol Rev 227(1):150–60.

100.

Lautrup

Sinclair

Mattson

Fang

. 2019. NAD(+) in brain aging and neurodegenerative disorders. Cell Metab 30(4):630–55.

101.

Lee

CYD

Daggett

Jiang

Langfelder

, and others. 2018. Elevated TREM2 gene dosage reprograms microglia responsivity and ameliorates pathological phenotypes in Alzheimer’s disease models. Neuron 97(5):1032–48.e1035.

102.

Leyns

CEG

Ulrich

Finn

Stewart

Koscal

Remolina Serrano

, and others. 2017. TREM2 deficiency attenuates neuroinflammation and protects against neurodegeneration in a mouse model of tauopathy. Proc Natl Acad Sci U S A 114(43):11524.

103.

Liu

Zheng

Huang

. 2018. Amyloid, tau, pathogen infection and antimicrobial protection in Alzheimer’s disease—conformist, nonconformist, and realistic prospects for ad pathogenesis. Transl Neurodegener 7:34.

104.

Lill

Rengmark

Pihlstrøm

Fogh

Shatunov

Sleiman

, and others. 2015. The role of TREM2 R47H as a risk factor for Alzheimer’s disease, frontotemporal lobar degeneration, amyotrophic lateral sclerosis, and Parkinson’s disease. Alzheimers Dement 11(12):1407–16.

105.

Lin

Lee

Hestir

Leo

Huang

Bosch

, and others. 2008. Discovery of a cytokine and its receptor by functional screening of the extracellular proteome. Science 320(5877):807–11.

106.

Liu

Leo

Chen

Wong

Williams

Lin

, and others. 2012. The mechanism of shared but distinct CSF-1R signaling by the non-homologous cytokines IL-34 and CSF-1. Biochim Biophys Acta 1824(7):938–45.

107.

Luo

Elwood

Britschgi

Villeda

Zhang

Ding

, and others. 2013. Colony-stimulating factor 1 receptor (CSF1R) signaling in injured neurons facilitates protection and survival. J Exp Med 210(1):157–72.

108.

Doi

Jin

Sonobe

Takeuchi

, and others. 2012. TGF-β induced by interleukin-34-stimulated microglia regulates microglial proliferation and attenuates oligomeric amyloid β neurotoxicity. Neurosci Lett 529(1):86–91.

109.

Magno

Lessard

Martins

Lang

Cruz

Asi

, and others. 2019. Alzheimer’s disease phospholipase C-gamma-2 (PLCG2) protective variant is a functional hypermorph. Alzheimers Res Ther 11(1):16.

110.

Malik

Simpson

Parikh

Wilfred

Fardo

Nelson

, and others. 2013. CD33 Alzheimer’s risk-altering polymorphism, CD33 expression, and exon 2 splicing. J Neurosci 33(33):13320–5.

111.

Maragakis

Rothstein

. 2006. Mechanisms of disease: astrocytes in neurodegenerative disease. Nat Clin Pract Neurol 2(12):679–89.

112.

Marioni

Harris

Zhang

McRae

Hagenaars

Hill

, and others. 2018. GWAS on family history of Alzheimer’s disease. Transl Psychiatry 8(1):99.

113.

Mastroeni

Nolz

Sekar

Delvaux

Serrano

Cuyugan

, and others. 2017. Laser-captured microglia in the Alzheimer’s and Parkinson’s brain reveal unique regional expression profiles and suggest a potential role for hepatitis B in the Alzheimer’s brain. Neurobiol Aging 63:12–21.

114.

Mathys

Adaikkan

Gao

Young

Manet

Hemberg

, and others. 2017. Temporal tracking of microglia activation in neurodegeneration at single-cell resolution. Cell Rep 21(2):366–80.

115.

Mawanda

Wallace

2013. Can infections cause Alzheimer’s disease? Epidemiol Rev 35(1):161–80.

116.

Mawuenyega

Sigurdson

Ovod

Munsell

Kasten

Morris

, and others. 2010. Decreased clearance of CNS beta-amyloid in Alzheimer’s disease. Science 330(6012):1774.

117.

McCarthy

. 2017. Location, location, location: Microglia are where they live. Neuron 95(2):233–5.

118.

Merlini

Kirabali

Kulic

Nitsch

Ferretti

. 2018. Extravascular CD3+ T cells in brains of Alzheimer disease patients correlate with tau but not with amyloid pathology: an immunohistochemical study. Neurodegen Dis 18(1):49–56.

119.

Meyer

Tremblay-Mercier

Leoutsakos

Madjar

Lafaille-Maignan

Savard

, and others. 2019. Intrepad: a randomized trial of naproxen to slow progress of presymptomatic Alzheimer disease. Neurology 92(18):e2070–e2080.

120.

Minderjahn

Schmidt

Fuchs

Schill

Raithel

Babina

, and others. 2020. Mechanisms governing the pioneering and redistribution capabilities of the non-classical pioneer PU.1. Nat Commun 11(1):402.

121.

Misra

Damotte

Hollenbach

. 2019. Structure-based selection of human metabolite binding P4 pocket of DRB1*15:01 and DRB1*15:03, with implications for multiple sclerosis. Genes Immun 20(1):46–55.

122.

Mizuno

Doi

Mizoguchi

Jin

Noda

Sonobe

, and others. 2011. Interleukin-34 selectively enhances the neuroprotective effects of microglia to attenuate oligomeric amyloid-β neurotoxicity. Am J Pathol 179(4):2016–27.

123.

Morris

Edland

Clark

Galasko

Koss

Mohs

, and others. 1993. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD): Part IV. Rates of cognitive change in the longitudinal assessment of probable Alzheimer’s disease. Neurology 43(12):2457.

124.

Mosher

Wyss-Coray

2014. Microglial dysfunction in brain aging and Alzheimer’s disease. Biochem Pharmacol 88(4):594–604.

125.

Naj

Jun

Beecham

Wang

Vardarajan

Buros

, and others. 2011. Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer’s disease. Nat Genet 43(5):436–41.

126.

Nalls

Pankratz

Lill

Hernandez

Saad

, and others. 2014. Large-scale meta-analysis of genome-wide association data identifies six new risk loci for Parkinson’s disease. Nat Genet 46(9):989–93.

127.

Nimmerjahn

Kirchhoff

Helmchen

2005. Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science 308(5726):1314–8.

128.

Numata

Araya

Fujii

Hara

Takasaka

Kojima

, and others. 2011. Insulin-dependent phosphatidylinositol 3-kinase/AKT and ERK signaling pathways inhibit TLR3-mediated human bronchial epithelial cell apoptosis. J Immunol Baltim Md 1950 187(1):510–9.

129.

Olah

Menon

Habib

Taga

Yung

, and others. 2020. Single cell RNA sequencing of human microglia uncovers a subset associated with Alzheimer’s disease. Nat Commun 11(1):6129.

130.

Olah

Patrick

Villani

White

Ryan

, and others. 2018. A transcriptomic atlas of aged human microglia. Nat Commun 9(1):539.

131.

Orr

Morgan

Elliott

Burrows

Scott

McVicar

, and others. 2007. CD33 responses are blocked by SOCS3 through accelerated proteasomal-mediated turnover. Blood 109(3):1061–8.

132.

Park

Hodges

Simmons

Lovestone

Weiner

Kim

, and others. 2020. Association of blood-based transcriptional risk scores with biomarkers for Alzheimer disease. Neurol Genet 6(6):e517.

133.

Paul

. 2019. First in human study for safety and tolerability of AL003. Available from: https://clinicaltrials.gov/ct2/show/NCT03822208

134.

Pawelec

Akbar

Caruso

Solana

Grubeck-Loebenstein

Wikby

2005. Human immunosenescence: is it infectious? Immunol Rev 205:257–68.

135.

Peng

Malhotra

Torchia

Kerr

Coggeshall

Humphrey

. 2010. TREM2- and DAP12-dependent activation of PI3K requires DAP10 and is inhibited by SHIP1. Sci Signal 3(122):ra38.

136.

Pérez-Oliva

Martínez-Esparza

Vicente-Fernández

Miguel

RCS

García-Peñarrubia

Hernández-Caselles

2011. Epitope mapping, expression and post-translational modifications of two isoforms of CD33 (CD33M and CD33M) on lymphoid and myeloid human cells. Glycobiology 21(6):757–70.

137.

Piccio

Deming

Del-Águila

Ghezzi

Holtzman

Fagan

, and others. 2016. Cerebrospinal fluid soluble TREM2 is higher in Alzheimer disease and associated with mutation status. Acta Neuropathol 131(6):925–33.

138.

Prieto

Cuenca

Través

Fernández-Velasco

Martín-Sanz

Boscá

2010. Lipoxin A4 impairment of apoptotic signaling in macrophages: implication of the PI3K/AKT and the ERK/NRF-2 defense pathways. Cell Death Differ 17(7):1179–88.

139.

Raj

Ryan

Replogle

Chibnik

Rosenkrantz

Tang

, and others. 2014. CD33: increased inclusion of exon 2 implicates the IG V-set domain in Alzheimer’s disease susceptibility. Hum Mol Genet 23(10):2729–36.

140.

Rakic

Hung

YMA

Smith

Tayler

Varney

, and others. 2018. Systemic infection modifies the neuroinflammatory response in late stage Alzheimer’s disease. Acta Neuropathol Commun 6(1):88.

141.

Rathore

Ramani

Pantua

Payandeh

Bhangale

Wuster

, and others. 2018. Paired immunoglobulin-like type 2 receptor alpha G78R variant alters ligand binding and confers protection to Alzheimer’s disease. PLoS Genet 14(11):e1007427.

142.

Raymont

. 2020. Microglial colony stimulating factor-1 receptor (CSF1R) in Alzheimer’s disease (MICAD). Available from: https://clinicaltrials.gov/ct2/show/NCT04121208

143.

Readhead

Haure-Mirande

Funk

Richards

Shannon

Haroutunian

, and others. 2018. Multiscale analysis of independent Alzheimer’s cohorts finds disruption of molecular, genetic, and clinical networks by human herpesvirus. Neuron 99(1):64–82.e67.

144.

Rehli

Poltorak

Schwarzfischer

Krause

Andreesen

Beutler

2000. PU.1 and interferon consensus sequence-binding protein regulate the myeloid expression of the human toll-like receptor 4 gene. J Biol Chem 275(13):9773–81.

145.

Replogle

Chan

White

Raj

Winn

Evans

, and others. 2015. A TREM 1 variant alters the accumulation of Alzheimer-related amyloid pathology. Ann Neurol 77(3):469–77.

146.

Réu

Khosravi

Bernard

Mold

Salehpour

Alkass

, and others. 2017. The lifespan and turnover of microglia in the human brain. Cell Rep 20(4):779–84.

147.

Ridge

Hoyt

Boehme

Mukherjee

Crane

Haines

, and others. 2016. Assessment of the genetic variance of late-onset Alzheimer’s disease. Neurobiol Aging 41:200.e213–200.e220.

148.

Rogers

Luber-Narod

Styren

Civin

. 1988. Expression of immune system-associated antigens by cells of the human central nervous system: relationship to the pathology of Alzheimer’s disease. Neurobiol Aging 9(4):339–49.

149.

Root-Bernstein

Fairweather

2014. Complexities in the relationship between infection and autoimmunity. Curr Allergy Asthma Rep 14(1):407.

150.

Rustenhoven

Smith

Smyth

Jansson

Scotter

Swanson

MEV

, and others. 2018. PU.1 regulates Alzheimer’s disease-associated genes in primary human microglia. Mol Neurodegener 13(1):44.

151.

Sadeghi

Wisgrill

Wessely

Diesner

Schüller

Dürr

, and others. 2016a. GM-CSF down-regulates TLR expression via the transcription factor PU.1 in human monocytes. PLoS One 11(10):e0162667.

152.

Sankowski

Böttcher

Masuda

Geirsdottir

Sagar, Sindram

, and others. 2019. Mapping microglia states in the human brain through the integration of high-dimensional techniques. Nat Neurosci 22(12):2098–110.

153.

Satoh

Arii

Suenaga

Wang

Kogure

Uehori

, and others. 2008. Pilralpha is a herpes simplex virus-1 entry coreceptor that associates with glycoprotein B. Cell 132(6):935–44.

154.

Sayed

Telpoukhovskaia

Kodama

Zhou

, and others. 2018. Differential effects of partial and complete loss of TREM2 on microglial injury response and tauopathy. Proc Natl Acad Sci U S A 115(40):10172–7.

155.

Schafer

Lehrman

Kautzman

Koyama

Mardinly

Yamasaki

, and others. 2012. Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron 74(4):691–705.

156.

Schlepckow

Monroe

Kleinberger

Cantuti-Castelvetri

Parhizkar

Xia

, and others. 2020. Enhancing protective microglial activities with a dual function TREM2 antibody to the stalk region. EMBO Mol Med 12(4):e11227.

157.

Schultz

Sinclair

. 2016. Why NAD(+) declines during aging: it’s destroyed. Cell Metab 23(6):965–6.

158.

Schulz

Gomez Perdiguero

Chorro

Szabo-Rogers

Cagnard

Kierdorf

, and others. 2012. A lineage of myeloid cells independent of MYB and hematopoietic stem cells. Science 336(6077):86–90.

159.

Schwarz

Springer

Altheide

Varki

Gagneux

Varki

2016. Human-specific derived alleles of CD33 and other genes protect against postreproductive cognitive decline. Proc Natl Acad Sci U S A 113(1):74–9.

160.

Seshadri

Fitzpatrick

Ikram

DeStefano

Gudnason

Boada

, and others. 2010. Genome-wide analysis of genetic loci associated with Alzheimer disease. JAMA 303(18):1832–40.

161.

Sharif

Knapp

2008. From expression to signaling: roles of TREM-1 and TREM-2 in innate immunity and bacterial infection. Immunobiology 213(9–10):701–13.

162.

Sieber

Jaenisch

Brehm

Guenther

Linnartz-Gerlach

Neumann

, and others. 2013. Attenuated inflammatory response in triggering receptor expressed on myeloid cells 2 (TREM2) knock-out mice following stroke. PLoS One 8(1):e52982.

163.

Sims

van der Lee

Naj

Bellenguez

Badarinarayan

Jakobsdottir

, and others. 2017. Rare coding variants in PLCG2, ABI3, and TREM2 implicate microglial-mediated innate immunity in Alzheimer’s disease. Nat Genet 49(9):1373–84.

164.

Soscia

Kirby

Washicosky

Tucker

Ingelsson

Hyman

, and others. 2010. The Alzheimer’s disease-associated amyloid beta-protein is an antimicrobial peptide. PLoS One 5(3):e9505.

165.

Sperling

Aisen

Beckett

Bennett

Craft

Fagan

, and others. 2011. Toward defining the preclinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging–Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7(3):280–92.

166.

Squarzoni

Oller

Hoeffel

Pont-Lezica

Rostaing

Low

, and others. 2014. Microglia modulate wiring of the embryonic forebrain. Cell Rep 8(5):1271–9.

167.

Stack

Stacey

Humphreys

. 2012. Herpesvirus exploitation of host immune inhibitory pathways. Viruses 4(8):1182–201.

168.

Steele

Carr

Bonham

Geier

Damotte

Miller

, and others. 2017. Fine-mapping of the human leukocyte antigen locus as a risk factor for Alzheimer disease: a case-control study. PLoS Med 14(3):e1002272.

169.

Stevens

. 1993. HSV-1 neuroinvasiveness. Intervirology 35(1–4):152–63.

170.

Streit

. 2006. Microglial senescence: does the brain’s immune system have an expiration date? Trends Neurosci 29(9):506–10.

171.

Streit

Sammons

Kuhns

Sparks

. 2004. Dystrophic microglia in the aging human brain. Glia 45(2):208–12.

172.

Sudom

Talreja

Danao

Bragg

Kegel

Min

, and others. 2018. Molecular basis for the loss-of-function effects of the Alzheimer’s disease-associated R47H variant of the immune receptor TREM2. J Biol Chem 293(32):12634–46.

173.

Takahashi

Rochford

CDP

Neumann

2005. Clearance of apoptotic neurons without inflammation by microglial triggering receptor expressed on myeloid cells-2. J Exp Med 201(4):647–57.

174.

Takalo

Wittrahm

Wefers

Parhizkar

Jokivarsi

Kuulasmaa

, and others. 2020. The Alzheimer’s disease-associated protective PLCΓ2-P522R variant promotes immune functions. Mol Neurodegener 15(1):52.

175.

Taylor

Buckley

Douglas

Cody

Simmons

Freeman

. 1999. The myeloid-specific sialic acid-binding receptor, CD33, associates with the protein-tyrosine phosphatases, SHP-1 and SHP-2. J Biol Chem 274(17):11505–12.

176.

Tsai

Lin

PBC

Dong

Moutinho

Casali

Liu

, and others. 2021. INPP5D expression is associated with risk for Alzheimer’s disease and induced by plaque-associated microglia. Neurobiol Dis 153:105303.

177.

Turnbull

Gilfillan

Cella

Aoshi

Miller

Piccio

, and others. 2006. Cutting edge: TREM-2 attenuates macrophage activation. J Immunol 177(6):3520–4.

178.

van der Lee

Conway

Jansen

Carrasquillo

Kleineidam

van den Akker

, and others. 2019. A nonsynonymous mutation in PLCG2 reduces the risk of Alzheimer’s disease, dementia with Lewy bodies and frontotemporal dementia, and increases the likelihood of longevity. Acta Neuropathol 138(2):237–50.

179.

Walker

Tang

Lue

. 2017. Studies on colony stimulating factor receptor-1 and ligands colony stimulating factor-1 and interleukin-34 in Alzheimer’s disease brains and human microglia. Front Aging Neurosci 9:244.

180.

Walker

Whetzel

Serrano

Sue

Beach

Lue

. 2015. Association of CD33 polymorphism RS3865444 with Alzheimer’s disease pathology and CD33 expression in human cerebral cortex. Neurobiol Aging 36(2):571–82.

181.

Walter

Häusermann

Raden

Teckchandani

Kamikura

Bernstein

, and others. 2008. Phosphorylated ITIMs enable ubiquitylation of an inhibitory cell surface receptor. Traffic 9(2):267–79.

182.

Wang

Mustafa

Yuede

Salazar

Kong

Long

, and others. 2020. Anti-human TREM2 induces microglia proliferation and reduces pathology in an Alzheimer’s disease model. J Exp Med 217(9):e20200785.

183.

Wang

Cella

Mallinson

Ulrich

Young

Robinette

, and others. 2015. TREM2 lipid sensing sustains the microglial response in an Alzheimer’s disease model. Cell 160(6):1061–71.

184.

Wang

Szretter

Vermi

Gilfillan

Rossini

Cella

, and others. 2012. IL-34 is a tissue-restricted ligand of CSF1R required for the development of Langerhans cells and microglia. Nat Immunol 13(8):753–60.

185.

Wang

Ulland

Ulrich

Song

Tzaferis

Hole

, and others. 2016. TREM2-mediated early microglial response limits diffusion and toxicity of amyloid plaques. J Exp Med 213(5):667–75.

186.

Wei

Nandi

Chitu

Yeung

Huang

, and others. 2010. Functional overlap but differential expression of CSF-1 and IL-34 in their CSF-1 receptor-mediated regulation of myeloid cells. J Leukoc Biol 88(3):495–505.

187.

Wendeln

Degenhardt

Kaurani

Gertig

Ulas

Jain

, and others. 2018. Innate immune memory in the brain shapes neurological disease hallmarks. Nature 556(7701):332–8.

188.

White

Kandel

Tripathi

Condon

Taubenberger

, and others. 2014. Alzheimer’s associated β-amyloid protein inhibits influenza a virus and modulates viral interactions with phagocytes. PLoS One 9(7):e101364.

189.

Wildsmith

Holley

Savage

Skerrett

Landreth

. 2013. Evidence for impaired amyloid β clearance in Alzheimer’s disease. Alzheimers Res Ther 5(4):33.

190.

Wissemann

Hill-Burns

Zabetian

Factor

Patsopoulos

Hoglund

, and others. 2013. Association of Parkinson disease with structural and regulatory variants in the HLA region. Am J Hum Genet 93(5):984–93.

191.

Wlodarski

Zhang

Liu

Kasprzycka

Marzec

Wasik

. 2007. PU.1 activates transcription of SHP-1 gene in hematopoietic cells. J Biol Chem 282(9):6316–23.

192.

Yao

Coppola

Schweig

Crawford

Mullan

Paris

2019. Distinct signaling pathways regulate TREM2 phagocytic and NFκB antagonistic activities. Front Cell Neurosci 13:457.

193.

Yoon

. 2012. Mechanisms of amyloid-β peptide clearance: potential therapeutic targets for Alzheimer’s disease. Biomol Ther (Seoul) 20(3):245–55.

194.

Yuan

Condello

Keene

Wang

Bird

Paul

, and others. 2016. TREM2 haplodeficiency in mice and humans impairs the microglia barrier function leading to decreased amyloid compaction and severe axonal dystrophy. Neuron 90(4):724–39.

195.

Yuan

Syed

Panchal

Joo

Colonna

Brantly

, and others. 2014. Triggering receptor expressed on myeloid cells 1 (TREM-1)-mediated BCL-2 induction prolongs macrophage survival. J Biol Chem 289(21):15118–29.

196.

Zaghi

Goldenson

Inayathullah

Lossinsky

Masoumi

Avagyan

, and others. 2009. Alzheimer disease macrophages shuttle amyloid-beta from neurons to vessels, contributing to amyloid angiopathy. Acta Neuropathol 117(2):111–24.

197.

Zdimerova

Murer

Engelmann

Raykova

Deng

Gujer

, and others. 2021. Attenuated immune control of Epstein-Barr virus in humanized mice is associated with the multiple sclerosis risk factor HLA-DR15. Eur J Immunol 51(1):64–75.

198.

Zeng

Ornatowska

Joo

Sadikot

. 2007. TREM-1 expression in macrophages is regulated at transcriptional level by NF-κB and PU.1. Eur J Immunol 37(8):2300–8.

199.

Zhang

Gaiteri

Bodea

Wang

McElwee

Podtelezhnikov

, and others. 2013. Integrated systems approach identifies genetic nodes and networks in late-onset Alzheimer’s disease. Cell 153(3):707–20.

200.

Zhou

Lee

Brady

Datta

Katan

Sheikh

, and others. 2012. A hypermorphic missense mutation in PLCG2, encoding phospholipase Cγ2, causes a dominantly inherited autoinflammatory disease with immunodeficiency. Am J Hum Genet 91(4):713–20.

201.

Zhu

Jones

Zhang

2018. The role of phospholipase c signaling in macrophage-mediated inflammatory response. J Immunol Res 2018:5201759.