Researchers know that ApoE4 and Trem2 can worsen neurodegeneration in mouse models of tauopathy. Do these proteins conspire to exacerbate pathology? It turns out that ApoE4 causes much of the damage alone, according to scientists led by David Holtzman, Washington University School of Medicine in St. Louis. In the November 10 Neuron, they reported that despite knocking out TREM2, APOE4 drove a type of microgliosis that worsened tau pathology and brain atrophy. The microglia ramped up phagocytosis and production of lysosomal proteins, and their lysosomes swelled with lipids.

  • In mice, APOE4 exacerbates tau tangles and brain atrophy.
  • Knocking out TREM2 made these pathologies worse.
  • ApoE4 causes microglial lysosomes to swell with lipids.

“This study emphasizes the importance of TREM2-independent microglial mechanisms underpinning AD risk and pathogenesis,” wrote Michelle Ehrlich and Sam Gandy of Mount Sinai School of Medicine in New York. To Jessica Rexach of the University of California, Los Angeles, the study pointed to the complex role microglia play in disease pathogenesis. “The findings are consistent with a more nuanced view of microglia in neurodegeneration that is beginning to emerge,” she wrote.

Holtzman and colleagues previously reported that in the P301S model of tauopathy, human ApoE4 drove the accumulation of phosphorylated tau, lysosome dysfunction in microglia, and brain atrophy (Sep 2017 news). On the other hand, knocking out TREM2 quelled neurodegeneration in mice carrying the same tau mutation and their own endogenous ApoE (Oct 2017 news). Given these proteins’ prominent roles in microglial responses to disease, the scientists predicted that knocking out TREM2 would also protect mice expressing human APOE4. To the authors’ surprise, it did not.

First author Maud Gratuze crossed P301S mice with Apoe knockouts or a knock-in strain carrying the human APOE4 gene, and then she knocked out TREM2 in the crosses. This created four strains: TREM2 mice expressing either no Apoe (TREM2/Apoe-KO) or APOE4 (TREM2/E4), and their respective TREM2 knockouts (TREM2-KO/Apoe-KO and TREM2-KO/E4). She examined brain tissue from 9.5-month-old animals because, at that age, P301S mice have extensive tangles, gliosis, and neuronal loss in the hippocampus and cortex.

Despite knocking out TREM2, pathology worsened in APOE4 mice. In the hippocampus and cortex, TREM2-KO/E4 animals accumulated more phospho-tau, as measured by the AT8 and AT180 antibodies, and more tangles, as measured by MC1, than all other mouse strains. Likewise, in the TREM2-KO/E4 mice, these regions shrank 15 percent more than in TREM2/E4 mice and 30 percent more than in Apoe knockouts that did or did not have TREM2 (see image below). The authors concluded that APOE4 worsens tauopathy and atrophy even when TREM2 is absent.

Tangles and Atrophy. Compared to P301S tauopathy mice lacking Apoe (left), P301S mice carrying APOE4 (middle) had more neurofibrillary tangles in the hippocampus and cortex (top), and worse atrophy in the same regions (bottom). Knocking out TREM2 worsened pathology further (right). [Courtesy of Gratuze et al., Neuron, 2022.]

Could faulty microglia be to blame for the damage? The researchers analyzed microglial transcriptomes from the four mouse strains after RNA-Seq of PU.1-positive single nuclei isolated from hippocampal tissue. The 25,000 total microglial nuclei clustered into 10 subgroups. Three subclusters expressed homeostatic genes, such as CX3CR1 and CSF1R. Compared to Apoe knockouts, APOE4 tauopathy mice had fewer of these homeostatic microglia, and their TREM2 knockout counterparts had even less.

One of the clusters caught Gratuze’s eye because APOE4-positive P301S mice mustered twice as many of them as did APOE knockouts, regardless of TREM2 expression. These cells, dubbed Tau/ApoE4 reactive microglia (TERM), overexpressed some disease-associated microglia (DAM) genes typically upregulated in the presence of amyloid. Among these, LPL and CTSB, encoding a lipoprotein lipase and a lysosomal protease, respectively, hinted that something might be awry with their lysosomes. TERMs expressed more of the lysosomal damage marker galectin-3, and their lysosomes were stuffed with lipids. These results suggest that, in the context of tau pathology, ApoE4 activates DAM-like microglia yet hobbles their lysosomes, and it does so independently of Trem2. “This work supports an important shift away from considering microglia only with respect to amyloid plaques, or as a single ‘disease-associated’ state, to instead considering the multiple interactions between microglia and other aspects of [neurodegenerative] disease,” Rexach wrote.

Holtzman was struck by how severe the lysosomal abnormalities were in TREM2-KO/E4 mice. “They are more pronounced than those in amyloidosis mice,” he told Alzforum. “Trem2 can’t prevent the lysosomal damage caused by ApoE4.”

APOE4 microglia also ramped up phagocytosis. They engulfed four times as many PSD-95-positive synapses as microglia lacking APOE, independent of TREM2 expression. Such damage mirrors the thinned synapses seen in E4 mice, suggesting that their dysfunctional microglia drive neurodegeneration. “ApoE4 plays a more substantial role than Trem2 in tau-driven degeneration, at least in this tauopathy mouse model,” concluded David Hansen, Brigham Young University, Provo, Utah. William Meilandt of Genentech, South San Francisco, California, found the results compelling, noting how clearly damaging ApoE4 is in the presence of tau and absence of TREM2 (comments below).

Might similar lysosome changes happen in people carrying APOE4 and loss-of-function TREM2 mutations? The researchers analyzed published snRNA-Seq data from cortical tissue of 15 volunteers from Wash U’s Knight Alzheimer’s Disease Research Center and nine from the ROSMAP cohort (Brase et al., 2021). All had had AD and carried at least one copy of APOE4. Fourteen carried two copies of the common variant of TREM2; 10 had one copy of either R47H or R62H variants that associate with AD. Reactive microglia in the risk variant carriers expressed higher levels of genes involved in lysosome function. In the face of poor TREM2 function, it appears that ApoE4 drives similar lysosome dysfunction in microglia from people.

All told, knocking out TREM2 in tauopathy mice carrying human APOE4 exacerbated tau pathology, microgliosis, and neurodegeneration—a direct contrast with knocking out TREM2 in the P301S mice expressing normal mouse Apoe. “The question of why TREM2 deletion would be significantly protective in Apoe wild-type mice, inconsequential in Apoe knockout mice, and slightly detrimental in APOE4 knock-in mice remains unanswered,” Hansen wrote. Rexach believes the findings behoove further analysis. “This work clearly establishes a relationship between APOE4, tau pathology, and the lysosome/phagosome that opens the door for continued study,” she wrote. Holtzman and colleagues are looking for ways to reverse the lipid accumulation in these microglia.—Chelsea Weidman Burke

Comments

  1. In this excellent paper, Gratuze et al. report that Trem2 deletion exacerbates neurodegeneration in PS19 mice expressing human APOE4 but does not affect neurodegeneration in PS19 mice lacking Apoe. These results contrast with the same group’s previous report that Trem2 deletion greatly reduced neurodegeneration in PS19 mice with wild-type Apoe (Leyns et al., 2017).

    My main takeaway from this study is that Apoe plays a more substantial role than does Trem2 in tau-driven degeneration (at least in the PS19 model). Throughout the study, substantial differences were observed between tau+ Apoe knockout (TEKO) and tau+ APOE-e4 knock-in (TE4) mice, with TE4 mice repeatedly showing more severe pathological phenotypes than TEKO mice. These findings are consistent with earlier findings from the same group (Shi et al., 2017). In contrast, the presence or absence of Trem2 had relatively little impact on most disease measurements, though a notable exception was slightly worse degeneration in TE4 mice that lacked Trem2.

    The microglial snRNA-Seq data corroborated the pathological findings, since the transcriptional responses observed were driven primarily by tau and ApoE status, not by Trem2. All tau+ mice had increased numbers of interferon-responsive microglia, regardless of ApoE or Trem2 status. The newly defined “TERM” response (tau/E4 reactive microglia) was also induced in all tau+ mice, but more strongly in TE4 mice compared to TEKO mice. Thus, the TERM response was closely related to ApoE status and the severity of pathology and was Trem2-independent. Certain disease-associated microglia (DAM) response genes, including Lpl and Cd9, were induced in a tau-dependent and Trem2-dependent manner—independent of ApoE status—but such Trem2-responsive features did not yield distinct clusters of nuclei in the UMAP transcriptome data.

    A third area in which ApoE status was more influential than Trem2 status was in microglial lysosomal phenotypes. CD68 staining, PSD-95 puncta, Galectin-3 staining, and LipidTox staining were all elevated in an APOE4-dependent manner. Therefore, compared to ApoE status, Trem2 status showed relatively little impact on disease pathology, microglial gene expression, or lysosomal function.

    Although we sometimes tend to think of Trem2-deficient microglia as happily remaining in the unperturbed or “homeostatic” state, we should not be surprised by the Trem2-independent microglial activation described in this paper. As Keren-Shaul et al. reported in the 5xFAD model, the DAM response can be subdivided into two components, a Trem2-independent "stage 1” and a Trem2-dependent “stage 2” (Keren-Shaul et al., 2017). Similarly, we observed in the TauPS2APP model that the “DAM1” response was Trem2-independent, while the “DAM2” response was Trem2-dependent (Lee et al., 2021). Meilandt et al. (2020) analyzed whether microglial gene modules, as we had defined (Friedman et al., 2018), exhibited Trem2 dependence in the PS2APP model and reported that, while the “neurodegeneration-related” gene module was clearly Trem2-dependent, other modules within the DAM response were largely Trem2-independent (Meilandt et al., 2020). 

    An important question now is whether the Trem2-independent DAM1/TERM responses are causal, protective, or inconsequential with regard to neurodegeneration. To me, they appear mostly inconsequential. The overall lack of impact of Trem2 deletion we see in this paper by Gratuze et al. is reminiscent of the findings we reported in the pR5-183 (tau P301L) model: Trem2 deletion had little impact on most of the disease phenotypes measured (Lee et al., 2021). In contrast, we found that Trem2 deletion in the TauPS2APP model of combined amyloid and tau pathologies accelerated the spreading of tau pathology beyond the hippocampus and led to widespread degeneration. The importance of Trem2 (and, by extension, the Trem2-dependent DAM2 response) in mitigating Aβ toxicity is unmistakable. By comparison, the role of Trem2 in primary tauopathy models is equivocal, and the Trem2-independent DAM1/TERM responses appear to neither cause, nor strongly prevent, neurodegeneration.

    As for the PS19 tauopathy model, the question of why Trem2 deletion would be significantly protective in Apoe wild-type mice, inconsequential in Apoe KO mice, and slightly detrimental in APOE-e4 knock-in mice, remains unanswered. Whatever the reason, we should recognize that the PS19 tauopathy model appears to be an outlier, since Trem2 deletion did not reduce tau pathogenesis in the pR5-183 model, in the hTau model (Bemiller et al., 2017), or in the THY-Tau22 model (Vautheny et al., 2021). 

    To further clarify the role of Trem2 in the PS19 model, a more targeted breeding approach may be helpful. If I understood the authors’ description correctly, the breeding mice used were triple heterozygotes (Tau +/– ; Apoe –/E4 ; Trem2 +/–), yielding 27 possible genotypes in the offspring, and only males were analyzed. A breeding scheme in which all offspring are (Tau +/– ; Apoe E4/E4) and only Trem2 is variable would enable more direct comparison of littermates with different Trem2 genotypes.

    References:

    . TREM2 deficiency attenuates neuroinflammation and protects against neurodegeneration in a mouse model of tauopathy. Proc Natl Acad Sci U S A. 2017 Oct 24;114(43):11524-11529. Epub 2017 Oct 9 PubMed.

    . ApoE4 markedly exacerbates tau-mediated neurodegeneration in a mouse model of tauopathy. Nature. 2017 Sep 28;549(7673):523-527. Epub 2017 Sep 20 PubMed.

    . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

    . Trem2 restrains the enhancement of tau accumulation and neurodegeneration by β-amyloid pathology. Neuron. 2021 Apr 21;109(8):1283-1301.e6. Epub 2021 Mar 5 PubMed.

    . Trem2 Deletion Reduces Late-Stage Amyloid Plaque Accumulation, Elevates the Aβ42:Aβ40 Ratio, and Exacerbates Axonal Dystrophy and Dendritic Spine Loss in the PS2APP Alzheimer's Mouse Model. J Neurosci. 2020 Feb 26;40(9):1956-1974. Epub 2020 Jan 24 PubMed.

    . Diverse Brain Myeloid Expression Profiles Reveal Distinct Microglial Activation States and Aspects of Alzheimer's Disease Not Evident in Mouse Models. Cell Rep. 2018 Jan 16;22(3):832-847. PubMed.

    . TREM2 deficiency exacerbates tau pathology through dysregulated kinase signaling in a mouse model of tauopathy. Mol Neurodegener. 2017 Oct 16;12(1):74. PubMed.

    . THY-Tau22 mouse model accumulates more tauopathy at late stage of the disease in response to microglia deactivation through TREM2 deficiency. Neurobiol Dis. 2021 Jul;155:105398. Epub 2021 May 18 PubMed.

  2. Gratuze et. al. used a complicated set of genetic crosses to investigate the opposing interactions of ApoE4 and Trem2 in tau-mediated pathology, neuroinflammation, and brain atrophy (neurodegeneration). Previously, it was reported that the presence of human ApoE4 (in a mouse ApoE knockout background) in TauP301S mice (TE4) exacerbated tau-mediated neurodegeneration and neuroinflammation, as seen by increased CD68 and GFAP (Shi et al., 2017), whereas Trem2 deletion in the same TauP301S mouse model conferred neuroprotection against tau-mediated brain atrophy and neuroinflammation, as seen by reduced CD68 and GFAP (Leyns et al., 2017). However, deleting Trem2 in the presence of E4 (TE4-T2KO) does not prevent tau-mediated pathology, neurodegeneration, or neuroinflammation, but instead partially worsens these outcomes in these mice. Trem2 deletion increased hippocampal brain atrophy and elevated levels of p-tau (AT8 and AT180) in TE4 mice.

    It's difficult to explain these findings given the previous research showing that Trem2 knockout is protective in TauP301S mice. We found that Trem2 deletion had no effect on tau pathology or brain atrophy in a pure tauopathy model (TauP301L), but did exacerbate Aβ-induced acceleration of tau pathology and neurodegeneration in a mouse model with both Aβ and tau (Lee et al., 2021; Meilandt et al., 2020). Deletion or inhibition of Trem2 was also unable to ameliorate aberrant microglial activity and associated lysosomal dysfunction in progranulin knockout mice (GRN-/-), indicating the therapeutic challenges and complex nature of Trem2 activity in several neurodegenerative models (Reifschneider et al., 2022). One thing that is clear in Gratuze et al. is that mouse ApoE knockout is protective in the presence of tau and absence of Trem2. This suggests that mouse ApoE may have differential activity in the presence of amyloid or tau pathology.

    References:

    . ApoE4 markedly exacerbates tau-mediated neurodegeneration in a mouse model of tauopathy. Nature. 2017 Sep 28;549(7673):523-527. Epub 2017 Sep 20 PubMed.

    . TREM2 deficiency attenuates neuroinflammation and protects against neurodegeneration in a mouse model of tauopathy. Proc Natl Acad Sci U S A. 2017 Oct 24;114(43):11524-11529. Epub 2017 Oct 9 PubMed.

    . Trem2 restrains the enhancement of tau accumulation and neurodegeneration by β-amyloid pathology. Neuron. 2021 Apr 21;109(8):1283-1301.e6. Epub 2021 Mar 5 PubMed.

    . Trem2 Deletion Reduces Late-Stage Amyloid Plaque Accumulation, Elevates the Aβ42:Aβ40 Ratio, and Exacerbates Axonal Dystrophy and Dendritic Spine Loss in the PS2APP Alzheimer's Mouse Model. J Neurosci. 2020 Feb 26;40(9):1956-1974. Epub 2020 Jan 24 PubMed.

    . Loss of TREM2 rescues hyperactivation of microglia, but not lysosomal deficits and neurotoxicity in models of progranulin deficiency. EMBO J. 2022 Feb 15;41(4):e109108. Epub 2022 Jan 12 PubMed.

  3. The study by Gratuze et al. used P301S tau mice expressing human APOE4 isoform with deleted (TE4-T2KO) or intact Trem2 (TE4). The most unexpected result is the effect on neurodegeneration. The authors show that the TREM2 deletion significantly increased neurodegeneration, and that TE4-T2KO mice exhibited smaller hippocampus and entorhinal/piriform cortex volumes compared with TE4 mice with intact Trem2. Also unexpected is the finding that Trem2 deficiency increased hippocampal atrophy without a change in tau phosphorylation, implying that the overall effect on the brain may not be directly associated with tau pathology.

    On the one hand, the results are surprising because prior studies found that loss of Trem2 in the same tau model expressing mouse APOE either led to the preservation of cortical volume in the entorhinal and piriform areas, accompanied by ventricular enlargement (Leyns et al., 2017), or protected against hippocampal atrophy (Sayed et al., 2018). On the other hand, from a purely biological point of view, it is not surprising that the combination of two risk factors for Alzheimer’s disease (TREM2 dysfunction and AppE4) should have a negative effect on the brain.

    Without more experiments, it is difficult to say what is the reason for these discrepancies. The most probable explanation for the observed differences with previous reports is that the mutated tau interacts differentially with human and mouse APOE. From this point of view, it will be more appropriate to compare the phenotype of TE4-T2KO to TE3-T2KO because the effect might be entirely related to human APOE, not necessarily to its isoform. The authors correctly conclude that “Further studies will be required to confirm these results are specific to ApoE4 as opposed to ApoE3 or ApoE2.” The authors also confirmed that loss of Trem2 decreased the expression of DAM genes, as reported previously (Keren-Shaul et al., 2017; Fitz et al., 2021). Another interesting finding is the presence of TREM2-independent reactive microglial in TE4-T2KO, which was shown before in APP/PS1 mice expressing human APOE isoforms (Fitz et al., 2020). 

    References:

    . TREM2 deficiency attenuates neuroinflammation and protects against neurodegeneration in a mouse model of tauopathy. Proc Natl Acad Sci U S A. 2017 Oct 24;114(43):11524-11529. Epub 2017 Oct 9 PubMed.

    . Differential effects of partial and complete loss of TREM2 on microglial injury response and tauopathy. Proc Natl Acad Sci U S A. 2018 Oct 2;115(40):10172-10177. Epub 2018 Sep 19 PubMed.

    . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

    . Phospholipids of APOE lipoproteins activate microglia in an isoform-specific manner in preclinical models of Alzheimer's disease. Nat Commun. 2021 Jun 7;12(1):3416. PubMed.

    . Trem2 deficiency differentially affects phenotype and transcriptome of human APOE3 and APOE4 mice. Mol Neurodegener. 2020 Jul 23;15(1):41. PubMed.

  4. TREM2-independent effects of APOE4 on microgliosis and neurodegeneration in tauopathy

    Early, or presymptomatic, reduction of brain Aβ has, so far, met with only modest success in terms of the clinical benefit. The identification of neuroimmune targets, and when and how they might be therapeutically modulated, is an area of recent interest. TREM2 was one of the first neuroimmune molecules linked to the risk for late-onset Alzheimer’s disease (LOAD) and is among the most studied (Jones, 2013). The optimal therapeutic manipulation of TREM2 remains unclear since the timing and polarity of its effects are disease-stage dependent. Early enhancement of TREM2 can apparently provide beneficial anti-inflammatory effects (Li et al., 2019). Once amyloidosis is established, Trem2 deletion can restrain the Aβ-induced tauopathy (Lee et al., 2021). Yet, chronic TREM2 activation with one anti-TREM2 antibody exacerbated Aβ-associated tauopathy seeding and spreading (Jain et al., 2023), while a similar experiment conducted using a different anti-TREM2 antibody had no effect on Aβ plaque burden or microgliosis, yet reduced neurodegeneration and preserved learning behavior (Wang et al., 2020), raising the question of whether activating or silencing TREM2 would be the best therapeutic approach.  

    Although not initially recognized as such, ApoE4 is also a potent inflammogen (Serrano-Pozo et al., 2021). This new paper from Holtzman and colleagues reports that TREM2 deficiency can exacerbate microgliosis and neurodegeneration in APOE4 x MAPT P301S mice, indicating a TREM2-independent effect. The question arises as to which APOE-isotype-dependent, TREM2-independent signaling pathway(s) is/are likely to underpin this observation. TREM2 is part of a microglial cell-surface receptor complex, wherein ligands for TREM2 and other receptors trigger signal transduction via TYROBP and its cytoplasmic ITAM motifs. These motifs, in turn, regulate the activity of the protein kinase, SYK, recently implicated as a key intracellular integrator linked to microglial phenotype (Ennerfelt et al., 2022; Wang et al., 2022). 

    TREM2 deletion is associated with elevated levels of Toll-like receptor-4 (Zhou et al., 2020). This brings to mind an unusual pathway linking APOE isotype to TLR4 signaling through microRNA-146a (Teter et al., 2016). Inflammation-associated activation and translocation of the transcription factor NFκB require increased miR146a expression to resolve the inflammatory response. MiR146a is also a target of the inflammation hub regulator, the transcription factor PU.1, identified as a microglial hub of AD-modified gene expression (Gjoneska et al., 2015). AD patients show increased miR146a in the brain (Lukiw et al., 2008) and, similarly, several mouse models of AD, including 5xFAD mice, also show increased brain miR146a associated with pathology (Li et al., 2011). 

    Another pathway linking APOE4 and inflammation was described by Wong et al. (2020). They reported that cholesterol-25-hydroxylase (CH25H) expression is upregulated in human AD brain and in mouse models of cerebral amyloidosis or tauopathy. Treatment with LPS markedly upregulated CH25H expression in the mouse brain and stimulated CH25H expression and 25-hydroxycholesterol (25-HC) secretion in mouse primary microglia. LPS-induced microglial production of the pro-inflammatory cytokine IL-1β was markedly potentiated by 25-HC and attenuated by the deletion of CH25H. They also reported that microglia expressing APOE4 produce greater amounts of 25-HC than APOE3-expressing microglia following treatment with LPS. Remarkably, 25-HC treatment results in a greater level of IL-1β secretion in LPS-activated APOE4-expressing microglia than in APOE3-expressing microglia.

    We recently compared the effects of genetic manipulation of TREM2 and TYROBP on specific aspects of microglial phenotype (Audrain et al., 2021). Using a novel transgenic mouse overexpressing TYROBP in microglia, we observed a decrease of the amyloid burden and an increase of tau phosphorylation when crossed with APP/PSEN1 or MAPTP301S mice, respectively. Characterization of these mice revealed TYROBP-related modulation of APOE transcription. We also showed that TYROBP and APOE mRNAs were increased in TREM2-null microglia recruited around either Aβ deposits or a cortical stab injury. Conversely, microglial APOE transcription was dramatically diminished when TYROBP was absent. These results provided evidence that TYROBP-APOE signaling does not require TREM2 and could be an initiating step in establishment of the disease-associated microglia (DAM) phenotype. While we have not assessed the APOE-isotype effects in the absence of TREM2, other groups have reported APOE-isotype effects acting via TREM2 (Fitz et al., 2020, 2021). 

    In the absence of TREM2, however, we are left with the possibility that some, as-yet unrecognized, TYROBP-binding receptor is responsible for the new effects reported by Gratuze et al. We recently reviewed this topic, and therein listed over a dozen known TYROBP-binding receptor ectodomains, the binding of which may be altered in the absence of TREM2 (Haure-Mirande et al., 2022). While all of these are potential candidates for TREM2-independent effects of APOE4, not all are expressed in microglia. However, CD33 is one particularly attractive possibility. CD33 is genetically associated with AD and is known to show APOE-isotype-specific effects on AD risk (Bertram et al., 2008; Javor et al., 2020). While direct experiments will be required to test this, and other possible explanations, Gratuze et al. emphasize the importance of TREM2-independent microglial mechanisms underpinning AD risk and pathogenesis, some of which may be TYROBP-dependent and utilize the same signal transduction pathway. It is worth noting that TYROBP-dependent signaling has now been implicated in Nasu-Hakola Disease, AD, tauopathies, and, as shown in a recent preprint, in a mouse model of Huntington’s disease (Creus-Muncunill et al., 2022). As such, targeting TYROBP may be beneficial in a wide range of neurodegenerative proteinopathies.

    References:

    . Reactive or transgenic increase in microglial TYROBP reveals a TREM2-independent TYROBP-APOE link in wild-type and Alzheimer's-related mice. Alzheimers Dement. 2021 Feb;17(2):149-163. Epub 2020 Dec 12 PubMed.

    . Genome-wide association analysis reveals putative Alzheimer's disease susceptibility loci in addition to APOE. Am J Hum Genet. 2008 Nov;83(5):623-32. PubMed.

    . Deletion of the microglial transmembrane immune signaling adaptor TYROBP ameliorates Huntington’s disease mouse phenotype. bioRxiv 2022.02.18.480944. bioRxiv

    . SYK coordinates neuroprotective microglial responses in neurodegenerative disease. Cell. 2022 Oct 27;185(22):4135-4152.e22. Epub 2022 Oct 17 PubMed.

    . Trem2 deficiency differentially affects phenotype and transcriptome of human APOE3 and APOE4 mice. Mol Neurodegener. 2020 Jul 23;15(1):41. PubMed.

    . Phospholipids of APOE lipoproteins activate microglia in an isoform-specific manner in preclinical models of Alzheimer's disease. Nat Commun. 2021 Jun 7;12(1):3416. PubMed.

    . Conserved epigenomic signals in mice and humans reveal immune basis of Alzheimer's disease. Nature. 2015 Feb 19;518(7539):365-9. PubMed.

    . Microglial TYROBP/DAP12 in Alzheimer's disease: Transduction of physiological and pathological signals across TREM2. Mol Neurodegener. 2022 Aug 24;17(1):55. PubMed.

    . Chronic TREM2 activation exacerbates Aβ-associated tau seeding and spreading. J Exp Med. 2023 Jan 2;220(1) Epub 2022 Oct 11 PubMed.

    . Association of CD33 rs3865444:C˃A polymorphism with a reduced risk of late-onset Alzheimer's disease in Slovaks is limited to subjects carrying the APOE ε4 allele. Int J Immunogenet. 2020 Oct;47(5):397-405. Epub 2020 Apr 24 PubMed.

    . Alzheimer disease: TREM2 linked to late-onset AD. Nat Rev Neurol. 2013 Jan;9(1):5. PubMed.

    . Trem2 restrains the enhancement of tau accumulation and neurodegeneration by β-amyloid pathology. Neuron. 2021 Apr 21;109(8):1283-1301.e6. Epub 2021 Mar 5 PubMed.

    . TREM2 inhibits inflammatory responses in mouse microglia by suppressing the PI3K/NF-κB signaling. Cell Biol Int. 2019 Apr;43(4):360-372. Epub 2018 May 10 PubMed.

    . Increased expression of miRNA-146a in Alzheimer's disease transgenic mouse models. Neurosci Lett. 2011 Jan 3;487(1):94-8. PubMed.

    . An NF-kappaB-sensitive micro RNA-146a-mediated inflammatory circuit in Alzheimer disease and in stressed human brain cells. J Biol Chem. 2008 Nov 14;283(46):31315-22. PubMed.

    . Effect of APOE alleles on the glial transcriptome in normal aging and Alzheimer's disease. Nat Aging. 2021 Oct;1(10):919-931. Epub 2021 Oct 11 PubMed.

    . Apolipoprotein E isotype-dependent modulation of microRNA-146a in plasma and brain. Neuroreport. 2016 Aug 3;27(11):791-5. PubMed.

    . Anti-human TREM2 induces microglia proliferation and reduces pathology in an Alzheimer's disease model. J Exp Med. 2020 Sep 7;217(9) PubMed.

    . TREM2 drives microglia response to amyloid-β via SYK-dependent and -independent pathways. Cell. 2022 Oct 27;185(22):4153-4169.e19. PubMed.

    . 25-Hydroxycholesterol amplifies microglial IL-1β production in an apoE isoform-dependent manner. J Neuroinflammation. 2020 Jun 17;17(1):192. PubMed.

    . TLR4 Targeting as a Promising Therapeutic Strategy for Alzheimer Disease Treatment. Front Neurosci. 2020;14:602508. Epub 2020 Dec 18 PubMed.

  5. Growing evidence, including genetic studies, implicate innate immunity in the modulation of disease risk for late-onset Alzheimer’s disease (LOAD). Central to CNS innate immune responses are microglia, which can exhibit either neuroprotective or deleterious responses, depending on context and pathology types. Increasingly recognized as a key regulator of microglial responses, Trem2, which is one of the strongest genetic risk factors for LOAD, has been at the center of the “good versus bad” debate, as preclinical studies in various AD mouse models involving Trem2 ablation or disease variants have yielded some conflicting data. In this elegant new paper by Holtzman and colleagues, Gratuze et al. provide striking evidence that continues to fuel this debate by showing that Trem2 loss-of-function (LoF) exacerbates neurodegeneration in a tauopathy model in an ApoE4 background, whereas previous work from the same group showed Trem2 LoF was neuroprotective in the same tauopathy model expressing murine, endogenous ApoE. Through deep phenotyping of microglia, the authors identify unique microglial subpopulations with activation of lysosomal gene networks that may contribute to enhanced neurodegeneration in the tauopathy model expressing ApoE4, but lacking Trem2. These new findings, which demonstrate Trem2 LoF promotes tau pathology and neurodegeneration, generally agree with the human genetics showing TREM2 LoF coding variants increase AD risk and the therapeutic hypothesis that enhancing Trem2 function may be beneficial in AD.

    There is generally good evidence that Trem2 ablation in mice affects amyloid plaques in a disease-state-specific fashion, and, perhaps more specifically and strikingly, decreases amyloid plaque compaction and enhances neuritic dystrophy around diffuse plaques, suggesting that lack of Trem2 is deleterious to neurons (Yuan et al., 2016; Wang et al., 2016; Jay et al., 2017; Parhizkar et al., 2019Meilandt et al., 2020). In contrast, Trem2 ablation or expression of AD-linked coding variants in tau models, such as the MAPT P301S (PS19) model, is strongly neuroprotective, at least in the presence of murine ApoE (Gratuze et al., 2020; Leyns et al., 2017; Sayed et al., 2018). In separate studies, Holtzman and colleagues have demonstrated that expression of human ApoE4 exacerbates amyloid pathology in various models and that ApoE2 and ApoE KO are protective. Similarly, ApoE4 expression has been shown to exacerbate neuroinflammation-induced neurodegeneration in mutant tau models, such as the MAPT P301S model, with a key contribution of microgliosis to neurodegeneration (Shi and Holtzman, 2018; Shi et al., 2019Shi et al., 2017). Here, ApoE2 and ApoE KO are also protective and significantly attenuate neuroinflammation and neurodegeneration, with significant impacts on tau pathology. Since Trem2 ablation profoundly affects microglial states detected by single-cell RNA sequencing (scRNA-Seq) with a reduced ability of microglia to fully express “disease-associated microglia (DAM)” states in amyloid and demyelination models (see for instance, Keren-Shaul et al., 2017Nugent et al., 2020), a similar question was asked by Gratuze et al. in the context of a tauopathy model, particularly with human ApoE4 expression.

    Surprisingly, Holtzman and colleagues observed that unlike in the murine ApoE background, Trem2 ablation exacerbates neuropathology and neurodegeneration in MAPT P301S mice expressing human ApoE4 (TE4) in an age-dependent manner. Indeed, they found that Trem2 ablation increases atrophy in the entorhinal and piriform cortices, as well as in the hippocampi of TE4 mouse brains at 9.5 months but not at 3 months of age. Interestingly, despite the increased brain atrophy in TE4 mice lacking Trem2, plasma NfL and synapses, measured by PSD-95 immunohistochemistry (IHC), were not affected by Trem2 genotypes. Perhaps measurements of cerebrospinal fluid (CSF) levels of NfL could offer more sensitive measurements of CNS-associated neurodegeneration. Trem2 ablation also caused a mild, but significant increase in phosphorylated tau epitopes detected by AT8 and AT180 antibodies in IHC relative to TE4 mice, consistent with exacerbated tauopathy in TE4 brains lacking Trem2. Single-nucleus RNA-Seq (snRNA-Seq) analysis of microglia after isolation of PU.1-positive nuclei by FACS was performed on mouse hippocampi and revealed interesting shifts in microglial subpopulations. First, TE4 brains showed the lowest proportion of homeostatic microglia (60-64 percent) in the presence or absence of Trem2, corroborating the notion that TE4 brains have more reactive microglia. Unlike in some other models, Trem2 ablation did not appear to generally “lock” microglia in a more homeostatic state in TE4 brains. Second, expression of mutant MAPT generally caused an increase in interferon-responsive microglia (IRM), independently of Trem2 expression. Third, TE4 brains showed a unique DAM-like subpopulation named “TERM” (Tau/ApoE4 reactive microglia) with lower expression of homeostatic genes (e.g., P2ry12), high expression of DAM genes Lpl and Spp1, as well as many lysosomal genes, including Ctsb, Cstd, Cd63, and Cd68. Trem2 ablation attenuated the drop in homeostatic marker P2ry12 found in TE4 microglia. Importantly, as assessed in amyloid models (Keren-Shaul et al., 2017), Trem2 dependency for gene expression was evaluated in these DAM-like populations in TE4 brains. While the Trem2 dependency was confirmed for some genes (e.g., Lpl, Spp1, Cd9, Clec7a), higher expression of other genes, such as Ctsb, Cstd, Fth1, and Lyz2, was largely Trem2-independent, as they were in amyloid models (Keren-Shaul et al., 2017). The cathepsin genes Ctsb and Cstd were expressed at even higher levels in TE4 microglia lacking Trem2 relative to TE4 microglia. Generally, the CLEAR (Coordinated Lysosomal Expression and Regulation) network was upregulated in TE4 microglia and this upregulation was stronger in the Trem2 KO. Since expression of lysosomal genes from the CLEAR network often occurs as a result of lysosomal stress, the authors speculated that TE4 microglia may exhibit lysosomal defects and that the latter may be exacerbated in the Trem2 KO.

    As predicted, TE4 microglia exhibited an expansion of the CD68 compartment, which labels phagolysosomes, and this expansion was greater in TE4 microglia lacking Trem2. Since this phenotype is consistent with a lysosomal storage disorder, the authors examined potential phagocytic cargoes that may cause the CD68 phenotype. They found that TE4 microglia exhibit higher levels of synaptic markers, such as PSD-95 (suggesting that synapse phagocytosis has been enhanced), but also neutral lipids, as labeled with LipidTox. In extreme cases, lysosomal storage can lead to lysosomal membrane permeabilization (LMP), which can be detected with markers such as galectin-3. The authors found that TE4 microglia exhibit higher levels of galectin-3 irrespective of Trem2 expression, suggesting that LMP occurs. Finally, upregulation of some of the CLEAR genes was also observed in AD patient brain microglia expressing ApoE4 and the rare TREM2 coding variants associated with increased AD risk, such as the p.R47H or p.R62H variants, suggesting that the mouse data may be in part replicated in humans.

    What are the key takeaways from this study? First, it provides strong evidence for mutant tau (at least the P301S mutant) and ApoE4 causing lysosomal defects in microglia, and likely astrocytes as well. Given that neurodegeneration is dramatically increased in TE4 mice compared to simple E4 or MAPT P301S mice, it is tempting to interpret the ApoE4/mutant tau interaction as “synthetically lethal,” at least for CNS neurons. While MAPT is primarily expressed in neurons and APOE is primarily expressed in glia, ApoE4 is a secreted protein and tau can be unconventionally secreted, suggesting that these two proteins may theoretically exert their deleterious effects in the CNS in part via non-cell autonomous mechanisms. In fact, extracellular tau is believed to contribute to tau pathology spreading and disease progression in AD (Vaquer-Alicea and Diamond, 2019). ApoE4 and tau may interact physically in the endolysosomal system of microglia, perhaps even as a result of dead neuron efferocytosis, somehow reducing the degradative capacity of lysosomes.

    Second, considering other disease models as well, Trem2 ablation appears to be deleterious in mouse models associated with clear lysosomal dysfunction in microglia. This was clearly shown in Grn KO mice, a model for frontotemporal lobar degeneration (FTLD-GRN), where lack of Trem2 corrected a large fraction of transcriptional anomalies that are characteristic of Grn KO microglia (Reifschneider et al., 2022) Loss of TREM2 rescues hyperactivation of microglia, but not lysosomal deficits and neurotoxicity in models of progranulin deficiency. However, the lysosomal lipid phenotypes, such as the deficiency of bis(monoacylglycerol)phosphate (BMP) deficiency and accumulation of glucocerebrosidase substrate glucosylsphingosine (Logan et al., 2021), were not corrected or worsened by lack of Trem2. Importantly, NfL levels were increased in the Grn/Trem2 double KOs, suggesting that Trem2 ablation exacerbates neurodegeneration, albeit via unknown neurotoxic pathways (Reifschneider et al., 2022). Similarly, Gratuze et al. show that lack of Trem2 exacerbates neurodegeneration in TE4 brains, which exhibit lysosomal defects in microglia. It will be interesting to determine what LipidTox+ neutral lipids accumulate in the lysosomes of TE4 microglia in the presence or absence of Trem2. Generally neutral lipids include glycerolipids, such as triacylglycerol or diacylglycerol, as well as cholesteryl esters, and those have been found to accumulate in microglia from various disease models, including AD model (Nugent et al., 2020; Andreone et al., 2020; Marschallinger et al., 2020; Xia et al., 2022). Interestingly, neutral lipids are generally found in cytoplasmic lipid droplets, whereas in the TE4 microglia, they appear to accumulate in lysosomes. When the latter happens, lysosomal lipid accumulation may increase the risk of LMP, as observed for instance in lysosomal storage models (Logan et al., 2021; Cantuti-Castelvetri et al., 2018). 

    Third and finally, the study by Gratuze et al. raises critical issues around molecular and cellular mechanisms underlying increased neurodegeneration in TE4 brains lacking Trem2. Since lack of Trem2 exacerbates neurodegeneration, one could argue that the enhanced neurotoxicity is normally suppressed by Trem2 signaling and in that sense, is Trem2-dependent. There are multiple potential mechanisms that can be envisaged, including increased complement-derived synaptic loss and neurotoxicity, as seen for instance in FTLD-MAPT (Dejanovic et al., 2018; Wu et al., 2019) and Grn KO models ((Lui et al., 2016; Zhang et al., 2020); upregulation of the NLRP3 inflammasome pathway (Andreone et al., 2020); upregulation of type I IRM (Roy et al., 2020); and upregulation of neurotoxic oxidized lipids, which are normally cleared by TREM2+ microglia (Dong et al., 2021). Additionally, neurotoxic astrocytes (Han et al., 2021) or even adaptive immunity cannot be ruled out. However, some of the data generated by Gratuze et al. do not directly support involvement of complement or NLRP3 pathway, although additional work is required to test this experimentally with genetic or pharmacological approaches. Intriguingly, the snRNA-Seq data strongly suggest that a microglial subpopulation expressing higher levels of CLEAR network lysosomal genes, such as Ctsb and Ctsd, may be more prominent in TE4 brains lacking Trem2. Following Occam’s razor, one could speculate that this population may be the culprit neurotoxic microglia exacerbating neurodegeneration. Lysosomal hydrolases are critical for lysosomal function, but they can also have deleterious actions once secreted or released into the cytosol in large amounts. Additionally, the CLEAR network appears to include innate immune genes in myeloid cells (Brady et al., 2018), which may contribute to neurotoxic activities. Assessing the role of this subpopulation of microglia would require identification of transcription factors responsible for upregulation of lysosomal genes, likely within the TFEB family. 

    Overall, this study raises several critical issues that pave the way for exciting new research and will ultimately lead to a better understanding of mechanisms controlling the beneficial versus deleterious role of microglia in health and disease. It also suggests that therapeutic strategies involving Trem2 should take into consideration the fact that murine ApoE may not fully replicate the physiology and pathophysiology of human ApoE, and ApoE4 in particular. 

    References:

    . TREM2 Haplodeficiency in Mice and Humans Impairs the Microglia Barrier Function Leading to Decreased Amyloid Compaction and Severe Axonal Dystrophy. Neuron. 2016 May 18;90(4):724-39. PubMed.

    . TREM2-mediated early microglial response limits diffusion and toxicity of amyloid plaques. J Exp Med. 2016 May 2;213(5):667-75. Epub 2016 Apr 18 PubMed.

    . Disease Progression-Dependent Effects of TREM2 Deficiency in a Mouse Model of Alzheimer's Disease. J Neurosci. 2017 Jan 18;37(3):637-647. PubMed.

    . Loss of TREM2 function increases amyloid seeding but reduces plaque-associated ApoE. Nat Neurosci. 2019 Feb;22(2):191-204. Epub 2019 Jan 7 PubMed.

    . Trem2 Deletion Reduces Late-Stage Amyloid Plaque Accumulation, Elevates the Aβ42:Aβ40 Ratio, and Exacerbates Axonal Dystrophy and Dendritic Spine Loss in the PS2APP Alzheimer's Mouse Model. J Neurosci. 2020 Feb 26;40(9):1956-1974. Epub 2020 Jan 24 PubMed.

    . Impact of TREM2R47H variant on tau pathology-induced gliosis and neurodegeneration. J Clin Invest. 2020 Sep 1;130(9):4954-4968. PubMed.

    . TREM2 deficiency attenuates neuroinflammation and protects against neurodegeneration in a mouse model of tauopathy. Proc Natl Acad Sci U S A. 2017 Oct 24;114(43):11524-11529. Epub 2017 Oct 9 PubMed.

    . Differential effects of partial and complete loss of TREM2 on microglial injury response and tauopathy. Proc Natl Acad Sci U S A. 2018 Oct 2;115(40):10172-10177. Epub 2018 Sep 19 PubMed.

    . Interplay between innate immunity and Alzheimer disease: APOE and TREM2 in the spotlight. Nat Rev Immunol. 2018 Dec;18(12):759-772. PubMed.

    . Microglia drive APOE-dependent neurodegeneration in a tauopathy mouse model. J Exp Med. 2019 Nov 4;216(11):2546-2561. Epub 2019 Oct 10 PubMed.

    . ApoE4 markedly exacerbates tau-mediated neurodegeneration in a mouse model of tauopathy. Nature. 2017 Sep 28;549(7673):523-527. Epub 2017 Sep 20 PubMed.

    . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

    . TREM2 Regulates Microglial Cholesterol Metabolism upon Chronic Phagocytic Challenge. Neuron. 2020 Mar 4;105(5):837-854.e9. Epub 2020 Jan 2 PubMed.

    . Propagation of Protein Aggregation in Neurodegenerative Diseases. Annu Rev Biochem. 2019 Jun 20;88:785-810. Epub 2019 Mar 27 PubMed.

    . Loss of TREM2 rescues hyperactivation of microglia, but not lysosomal deficits and neurotoxicity in models of progranulin deficiency. EMBO J. 2022 Feb 15;41(4):e109108. Epub 2022 Jan 12 PubMed.

    . Rescue of a lysosomal storage disorder caused by Grn loss of function with a brain penetrant progranulin biologic. Cell. 2021 Sep 2;184(18):4651-4668.e25. Epub 2021 Aug 26 PubMed.

    . Alzheimer's-associated PLCγ2 is a signaling node required for both TREM2 function and the inflammatory response in human microglia. Nat Neurosci. 2020 Aug;23(8):927-938. Epub 2020 Jun 8 PubMed.

    . Defective cholesterol clearance limits remyelination in the aged central nervous system. Science. 2018 Feb 9;359(6376):684-688. Epub 2018 Jan 4 PubMed.

    . Changes in the Synaptic Proteome in Tauopathy and Rescue of Tau-Induced Synapse Loss by C1q Antibodies. Neuron. 2018 Dec 19;100(6):1322-1336.e7. Epub 2018 Nov 1 PubMed.

    . Complement C3 Is Activated in Human AD Brain and Is Required for Neurodegeneration in Mouse Models of Amyloidosis and Tauopathy. Cell Rep. 2019 Aug 20;28(8):2111-2123.e6. PubMed.

    . Progranulin Deficiency Promotes Circuit-Specific Synaptic Pruning by Microglia via Complement Activation. Cell. 2016 May 5;165(4):921-35. Epub 2016 Apr 21 PubMed.

    . Neurotoxic microglia promote TDP-43 proteinopathy in progranulin deficiency. Nature. 2020 Dec;588(7838):459-465. Epub 2020 Aug 31 PubMed.

    . Type I interferon response drives neuroinflammation and synapse loss in Alzheimer disease. J Clin Invest. 2020 Apr 1;130(4):1912-1930. PubMed.

    . Oxidized phosphatidylcholines found in multiple sclerosis lesions mediate neurodegeneration and are neutralized by microglia. Nat Neurosci. 2021 Apr;24(4):489-503. Epub 2021 Feb 18 PubMed.

    . Astrocyte-immune cell interactions in physiology and pathology. Immunity. 2021 Feb 9;54(2):211-224. PubMed.

    . Emerging roles for TFEB in the immune response and inflammation. Autophagy. 2018;14(2):181-189. Epub 2017 Jul 24 PubMed.

Make a Comment

To make a comment you must login or register.

References

Research Models Citations

  1. Tau P301S (Line PS19)

News Citations

  1. ApoE4 Makes All Things Tau Worse, From Beginning to End
  2. Changing With the Times: Disease Stage Alters TREM2 Effect on Tau

Mutation Interactive Images Citations

  1. TREM2

Antibody Citations

  1. Tau (AT8)
  2. Tau (AT180)
  3. Tau (MC1)

Mutations Citations

  1. TREM2 R47H
  2. TREM2 R62H

Paper Citations

  1. . A landscape of the genetic and cellular heterogeneity in Alzheimer disease. medRxiv 2021.11.30. medRxiv

Further Reading

Primary Papers

  1. . TREM2-independent microgliosis promotes tau-mediated neurodegeneration in the presence of ApoE4. Neuron. 2022 Nov 4; PubMed.