Bringing Order out of Chaos: Establishing an Epistatic Relationship Between CD33 and TREM2
Timothy Y. Huang,Huaxi Xu
DOI: https://doi.org/10.1016/j.neuron.2019.08.019
IF: 16.2
2019-01-01
Neuron
Abstract:TREM2 and CD33 are microglial receptors associated with Alzheimer’s disease (AD) risk. In this issue of Neuron, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar demonstrate opposing effects of CD33 and TREM2 on AD phenotypes, where CD33 deletion promotes neuroprotection in a manner dependent on TREM2. TREM2 and CD33 are microglial receptors associated with Alzheimer’s disease (AD) risk. In this issue of Neuron, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar demonstrate opposing effects of CD33 and TREM2 on AD phenotypes, where CD33 deletion promotes neuroprotection in a manner dependent on TREM2. Alzheimer’s disease (AD) is the most prevalent form of neurodegeneration in the elderly and steadily becoming a significant societal concern. AD is pathologically characterized by the presence of amyloid β (Aβ)-enriched plaques and neurofibrillary tangles comprising hyperphosphorylated forms of the microtubule-stabilizing component, tau. Although our understanding of the basic biology underlying Aβ generation and toxicity has blossomed within the last three decades, pharmacologically targeting Aβ has seen little or no success so far clinically in AD. Indeed, Aβ is an obligately fundamental driver for familial forms of AD, where mutations in the amyloid precursor protein (APP) or presenilins (PS1/PS2) can enhance Aβ plaque pathology and trigger early cognitive decline. However, since the majority of AD cases (estimated to be over 90%) occur sporadically, it seems likely that genetic factors other than APP or PS1 contribute to sporadic AD onset. Interestingly, identification of new genetic components related to AD risk suggests that variations in genes enriched in microglia may have a significant role in enhancing AD onset. Recent results from genome-wide association studies (GWAS) and whole-exome sequencing have identified an R47H mutation in the microglial immune receptor TREM2 linked to sporadic AD (Guerreiro et al., 2013Guerreiro R. Wojtas A. Bras J. Carrasquillo M. Rogaeva E. Majounie E. Cruchaga C. Sassi C. Kauwe J.S. Younkin S. et al.Alzheimer Genetic Analysis GroupTREM2 variants in Alzheimer’s disease.N. Engl. J. Med. 2013; 368: 117-127Crossref PubMed Scopus (1881) Google Scholar, Jonsson et al., 2013Jonsson T. Stefansson H. Steinberg S. Jonsdottir I. Jonsson P.V. Snaedal J. Bjornsson S. Huttenlocher J. Levey A.I. Lah J.J. et al.Variant of TREM2 associated with the risk of Alzheimer’s disease.N. Engl. J. Med. 2013; 368: 107-116Crossref PubMed Scopus (1635) Google Scholar). TREM2 is almost exclusively expressed in microglia and binds various immune ligands (bacterial lipopolysaccharide, LPS) and signals through its association with its obligate co-receptor or adaptor, DAP12 (TYROBP), which transmits downstream signaling pathways through immunoreceptor tyrosine-based activation motifs (ITAMs) (Figure 1A). TREM2 deletion can impair microglial clustering to Aβ plaques, increases autophagy, impairs energetic metabolism, and increases Aβ plaque load in AD mouse models (Ulland et al., 2017Ulland T.K. Song W.M. Huang S.C. Ulrich J.D. Sergushichev A. Beatty W.L. Loboda A.A. Zhou Y. Cairns N.J. Kambal A. et al.TREM2 Maintains Microglial Metabolic Fitness in Alzheimer’s Disease.Cell. 2017; 170: 649-663.e613Abstract Full Text Full Text PDF PubMed Scopus (526) Google Scholar, Wang et al., 2015Wang Y. Cella M. Mallinson K. Ulrich J.D. Young K.L. Robinette M.L. Gilfillan S. Krishnan G.M. Sudhakar S. Zinselmeyer B.H. et al.TREM2 lipid sensing sustains the microglial response in an Alzheimer’s disease model.Cell. 2015; 160: 1061-1071Abstract Full Text Full Text PDF PubMed Scopus (911) Google Scholar). Conversely, overexpression of TREM2 can reduce Aβ plaques, attenuate AD-related transcriptomic signatures, and correct memory deficits in AD mice (Lee et al., 2018Lee C.Y.D. Daggett A. Gu X. Jiang L.L. Langfelder P. Li X. Wang N. Zhao Y. Park C.S. Cooper Y. et al.Elevated TREM2 Gene Dosage Reprograms Microglia Responsivity and Ameliorates Pathological Phenotypes in Alzheimer’s Disease Models.Neuron. 2018; 97: 1032-1048.e1035Abstract Full Text Full Text PDF PubMed Scopus (182) Google Scholar), together suggesting that TREM2 has a protective role in AD. In human AD, the R47H mutation that resides within the extracellular V-type Ig domain can confer AD risk ∼3-fold above TREM2 common variants and may impair interactions with various extracellular ligands, including ApoE, LDL, clusterin (Gratuze et al., 2018Gratuze M. Leyns C.E.G. Holtzman D.M. New insights into the role of TREM2 in Alzheimer’s disease.Mol. Neurodegener. 2018; 13: 66Crossref PubMed Scopus (203) Google Scholar), and Aβ (Zhao et al., 2018Zhao Y. Wu X. Li X. Jiang L.L. Gui X. Liu Y. Sun Y. Zhu B. Pina-Crespo J.C. Zhang M. et al.TREM2 Is a Receptor for beta-Amyloid that Mediates Microglial Function.Neuron. 2018; 97: 1023-1031.e1027Abstract Full Text Full Text PDF PubMed Scopus (308) Google Scholar). Thus, TREM2-associated microglial dysfunction is closely linked to AD pathogenesis. Polymorphisms in the microglia siglec-family receptor CD33 (Siglec-3) have also been identified by GWAS. Similar to TREM2, CD33 is also comprised of an IgV domain with an additional IgC2 domain in its ectodomain and, through engagement of extracellular sialic acid ligands, triggers inhibitory signals through immunoreceptor tyrosine-based inhibitory motifs (ITIMs) within its cytosolic tail. Human CD33 comprises both ITIM and ITIM-like motifs in the cytosolic tail, while murine CD33 features an ITIM-like motif only (Figure 1A). Interestingly, “protective” CD33 polymorphisms (SNP rs865444) are associated with decreased CD33 expression, and deletion of CD33 in mice can reduce Aβ plaque deposition in an APP/PS1 AD mouse model (Griciuc et al., 2013Griciuc A. Serrano-Pozo A. Parrado A.R. Lesinski A.N. Asselin C.N. Mullin K. Hooli B. Choi S.H. Hyman B.T. Tanzi R.E. Alzheimer’s disease risk gene CD33 inhibits microglial uptake of amyloid beta.Neuron. 2013; 78: 631-643Abstract Full Text Full Text PDF PubMed Scopus (636) Google Scholar). This sets the stage to address a fundamental question: deletions of two microglial receptors show converse effects on Aβ plaque pathology in AD mice, so in a case in which both receptors are deleted, which phenotype would prevail? In order to determine how opposing effects on CD33 deletion (protective) and TREM2 deletion (deleterious) interact epistatically, Griciuc et al. compared 5xFAD mouse lines expressing APP and PS1 familial AD mutations crossed with CD33 KO, TREM2 KO, or CD33/TREM2 double KO and characterized resulting behavior, pathological, and transcriptomic profiles in these animals (Figure 1B) (Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar). In line with previous results in brain tissue from CD33 KO mice expressing APP and PS1 mutations (APP/PS1 double Tg) demonstrating decreased Aβ levels and plaque pathology, CD33 KO/5xFAD mice showed improved spatial learning and memory function (retention memory) in Morris water maze assays. Given that TREM2 KO mice bear increased Aβ levels in aged AD mouse models with impaired association with plaques, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar now demonstrate that TREM2 KO/5xFAD mice feature cognitive deficits comparable to 5xFAD animals alone. Interestingly, TREM2 and CD33 deletion in combination together in 5xFAD mice similarly results in cognitive impairment comparable to TREM2 KO/5xFAD or 5xFAD animals, implicating epistatic dominance of the deleterious TREM2 KO state over protective effects with CD33 deletion. To further consider the impact of TREM2 and CD33 deletion in combination on other aspects of pathological dysfunction, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar evaluated and compared single and combined TREM2/CD33 deletion in 5xFAD animals for changes in Aβ levels and plaque pathology. In agreement with behavioral assays, CD33 deletion decreased Aβ levels as well as the formation of Aβ plaques. While TREM2 KO/5xFAD showed elevations in Aβ/Aβ plaques compared to 5xFAD alone, these effects predominated in TREM2 KO/CD33 KO/5xFAD animals; effects on reducing Aβ levels and Aβ plaques with CD33 deletion were masked by TREM2 deletion. Enhanced cell loss in 5xFAD cortex and hippocampus as indicated by active caspase-3 staining was attenuated by CD33 deletion in 5xFAD brain, while TREM2 KO alone or in combination with CD33 deletion in 5xFAD mice elevated active caspase-3 levels above 5xFAD alone. Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar also show reduced Iba1+ microglial clustering around Aβ plaques in TREM2 KO/5xFAD animals; while CD33 KO/5xFAD showed little to no difference in Iba1 cell density or microglial clustering to Aβ plaques compared to 5xFAD alone, CD33 KO/TREM2 KO/5xFAD in combination did not alter reductions in Iba1 staining or microglial Aβ homing as seen in TREM2 KO/5xFAD animals. Together, this suggests that while protective effects resulted from CD33 deletion are dependent on the presence of TREM2, deleterious effects seen in TREM2 KO brain are not altered by CD33 KO, defining a downstream role for TREM2 in neuroprotection with CD33 deletion. Lastly, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar analyzed transcriptomic profiles in sorted microglia isolated from 5xFAD mouse brain and compared these profiles to TREM2 KO, CD33 KO, and TREM2/CD33 KO animals in a 5xFAD background. Genes associated with both pro-inflammatory and anti-inflammatory microglial states, as well as genes associated with the microglial “sensome” to detect bacterial and/or fungal and apoptotic ligands such as TREM2 and DAP12, were seen to be upregulated in 5xFAD microglia. Microglial genes associated with homeostasis, cytokine and/or sialic acid binding, and signaling (such as CD33) were found to be downregulated in 5xFAD mice. While CD33 deletion induced robust changes in the transcriptome in 5xFAD animals compared to 5xFAD alone (299 differentially expressed genes [DEGs]), CD33 deletion failed to induce these transcriptomic changes in a TREM2 KO background (only 10 DEGs observed; CD33 KO/TREM2 KO/5xFAD compared with TREM2 KO/5xFAD), indicating that transcriptomic shifts associated with CD33 are dependent on TREM2. Conversely, a large number of DEGs (117 DEGs) were induced in 5xFAD microglia with TREM2 deletion in both the presence and the absence of CD33, indicating that transcriptomic shifts associated with TREM2 deletion largely occur independently of CD33. Assuming that CD33 deletion results in “protective” while TREM2 deletion results in “deleterious” microglial profiles, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar searched for DEGs with differential expression in CD33 and TREM2 KO backgrounds in 5xFAD microglia and validated Il1b, Il1m, Gpnmb, and Vegfa as components upregulated in CD33 KO but downregulated in TREM2 KO in AD mouse microglia. The question remains, how does CD33 deletion confer neuroprotection, and conversely, how does TREM2 deletion enhance neurodegeneration? Categorical profiles of genes enriched in CD33 or TREM2 KO/5xFAD animals differ quite significantly in these transgenic lines; while TREM2 deletion resulted in enrichment of categories more relevant to immune response (“cytokine production secretion,” “receptor activity chemokine”), CD33 deletion resulted in enrichment of other categories related to morphogenesis (“morphogenesis branching development”) and the extracellular matrix (“extracellular collagen matrix”). Although this does not provide a direct molecular model of how CD33 deletion is protective and TREM2 deletion is deleterious, the molecular signatures observed suggest that CD33- and TREM2-dependent pathways are quite dissimilar in a 5xFAD background. Given that CD33 and TREM2 deletion show differential expression of proinflammatory mediators such as Il1b in 5xFAD mouse lines, it is anticipated that future studies will characterize other potential microglial modulators involved in opposing roles for CD33 and TREM2 in AD. The age-dependent molecular signatures identified by Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar in 5xFAD mice are interesting, though difficult to interpret. More specifically, decreased CD33 and increased TREM2 levels are observed in 8-month-old 5xFAD brain; assuming that TREM2 is protective and CD33 is deleterious, could this be part of a reactionary program to cope with Aβ proteotoxicity? Given that in addition to AD, TREM2 is upregulated in other neurodegenerative disorders, including amyotrophic lateral sclerosis (ALS), Parkinson’s disease, and stroke (Gratuze et al., 2018Gratuze M. Leyns C.E.G. Holtzman D.M. New insights into the role of TREM2 in Alzheimer’s disease.Mol. Neurodegener. 2018; 13: 66Crossref PubMed Scopus (203) Google Scholar), it is possible that microglia may comprise a responsive system to potentially slow or counter pathogenic effects due to neurodegenerative insult. Because CD33 and TREM2 mouse models currently used in Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar and other studies comprise constitutive deletion of these genes, it may be of interest to determine whether recapitulation and/or attenuation of these responsive changes during pathogenic onset are indeed protective. It is currently unclear how elevated CD33 expression can inhibit Aβ uptake and aggravate Aβ pathology. It is interesting that both TREM2 and CD33 are comprised of Ig-like modules within their ectodomains, and as TREM2 has been shown to bind directly to Aβ (Zhao et al., 2018Zhao Y. Wu X. Li X. Jiang L.L. Gui X. Liu Y. Sun Y. Zhu B. Pina-Crespo J.C. Zhang M. et al.TREM2 Is a Receptor for beta-Amyloid that Mediates Microglial Function.Neuron. 2018; 97: 1023-1031.e1027Abstract Full Text Full Text PDF PubMed Scopus (308) Google Scholar), it is tempting to speculate that CD33 may also bind Aβ. Indeed, a mutant version of CD33 lacking the sialic acid-binding Ig-V domain was shown to distribute equivalently to wild-type CD33 in BV2 cells but failed to impair Aβ uptake (Griciuc et al., 2013Griciuc A. Serrano-Pozo A. Parrado A.R. Lesinski A.N. Asselin C.N. Mullin K. Hooli B. Choi S.H. Hyman B.T. Tanzi R.E. Alzheimer’s disease risk gene CD33 inhibits microglial uptake of amyloid beta.Neuron. 2013; 78: 631-643Abstract Full Text Full Text PDF PubMed Scopus (636) Google Scholar). Moreover, a CD33K7R mutant with enhanced cell surface distribution showed enhanced impairment of Aβ uptake in BV2 cells (Griciuc et al., 2013Griciuc A. Serrano-Pozo A. Parrado A.R. Lesinski A.N. Asselin C.N. Mullin K. Hooli B. Choi S.H. Hyman B.T. Tanzi R.E. Alzheimer’s disease risk gene CD33 inhibits microglial uptake of amyloid beta.Neuron. 2013; 78: 631-643Abstract Full Text Full Text PDF PubMed Scopus (636) Google Scholar). Together, these results suggest that CD33 Ig-V domain interactions at the cell surface are important in attenuating Aβ uptake. However, whether this is due to sialic acid binding or interaction with an alternate ectodomain ligand remains to be seen. Results from Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar consolidate an interesting premise that AD risk genes can influence and cross-regulate each other. Indeed, analysis of a “pathogenic” CD33 SNP in human monocytes identified a correlation between the pathogenic rs865444 CD33 allele (elevated CD33 expression) and increased TREM2 expression at the cell surface (Chan et al., 2015Chan G. White C.C. Winn P.A. Cimpean M. Replogle J.M. Glick L.R. Cuerdon N.E. Ryan K.J. Johnson K.A. Schneider J.A. et al.CD33 modulates TREM2: convergence of Alzheimer loci.Nat. Neurosci. 2015; 18: 1556-1558Crossref PubMed Scopus (110) Google Scholar). In contrast to these cross-regulatory interactions in humans, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar show no direct effect of CD33 deletion on TREM2 or Tyrobp (Dap12) expression. Despite differences in quantification methods (FACS versus mRNA expression), cell type (monocytes versus microglia), and genetic context (SNP versus homozygous deletion), the possibility remains that CD33, TREM2, and other AD-associated microglial genes may interact differently in mouse and humans. Additionally, TREM2 and CD33 orthologs differ quite significantly, where TREM2 and CD33 coding sequences are only loosely conserved in mouse and human. More importantly, murine CD33 lacks a membrane proximal ITIM motif present in human CD33 and comprises a single “ITIM-like” domain (Figure 1A). Although these results contribute greatly to our understanding of how these genes interact in a mouse AD context, we may need to further evaluate how CD33 and TREM2 interact in human primary microglia or human iPSC/ESC-derived microglial cell lines in the context of naturally occurring CD33 polymorphisms and the TREM2R47H risk variant. While study of AD risk variants in human microglia would be ideal, the rarity of these alleles, combined with the sporadic, age-dependent nature of AD onset, impedes our ability to fully understand how genes such as CD33 and TREM2 modulate microglial behavior in the human brain during AD pathogenesis. Although combining CD33, TREM2, and other AD-associated polymorphisms through arranged marriages is theoretically possible, the practicality of human genetics is additionally hampered by a small litter size (n = 1), non-isogenicity (as inbreeding has been commonly outlawed) and a 60-year incubation period. For the moment, AD mouse models, while imperfect, still may be one of the most useful systems to study orthologous gene function. By establishing how CD33 and TREM2 interact in a 5xFAD model, Griciuc et al., 2019Griciuc A. Patel S. Federico A.N. Choi S.H. Innes B.J. Oram M.K. Cereghetti G. McGinty D. Anselmo A. Sadreyev R.I. et al.TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s Disease.Neuron. 2019; 103 (this issue): 820-835Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar have made a significant step that may yet be an uphill climb in our understanding of glial biology in AD. TREM2 Acts Downstream of CD33 in Modulating Microglial Pathology in Alzheimer’s DiseaseGriciuc et al.NeuronJuly 10, 2019In BriefMicroglial receptors CD33 and TREM2 exhibit opposite effects on Aβ pathology and microglial activity in AD mice. In crosstalk between these receptors, TREM2 is required for differential gene expression in 5xFAD mice deficient for CD33 and acts downstream of CD33. Full-Text PDF Open Archive