Almost 30 years after ApoE was pegged as an Alzheimer’s risk gene, researchers are still investigating fundamental questions about how the apolipoprotein sways neurodegenerative disease processes. For one, which cell type is responsible for producing forms of the protein that beckon neurodegeneration? On April 7 in Neuron, researchers pinned some of the blame on astrocytes, which release the lion’s share of ApoE in the brain. Led by David Holtzman and Jason Ulrich at Washington University in St. Louis, the study found that in human ApoE knock-in mice, silencing astrocyte ApoE4, but not ApoE3, assuaged neurodegeneration instigated by tau pathology.
- Removal of ApoE4, but not E3, from astrocytes stemmed neurodegeneration in tangle-bearing mice.
- It also lessened tau pathology, synaptic pruning by microglia.
- Biggest beneficiaries: a subset of cortical neurons that amass RNA-binding proteins.
Turning off the astrocyte ApoE4 spigot just as tau started to accumulate not only stemmed subsequent brain shrinkage, it also dampened tau accumulation and profoundly influenced gene expression in multiple cell types, including neurons, oligodendrocytes, and microglia. Without made-in-astrocytes ApoE4, microglia were less prone to assume a hostile, neurodegenerative stance, and feasted less on fragile synapses. The findings suggest—at least in this mouse model of tauopathy—that astrocytic ApoE4 exacerbates most neurodegenerative phenotypes and likely does so via cross-talk with other cells.
“The paper highlights how the apolipoprotein, when produced in astrocytes, can exert effects in multiple cell types known to participate in pathogenic pathways of AD,” said Guojun Bu of the Mayo Clinic in Jacksonville, Florida. It also reiterates the potential gain of toxic function conferred on ApoE4 relative to ApoE3, he said. Bu, Holtzman, and other investigators agreed that the study supports the idea of dampening ApoE4 expression as a therapeutic strategy (Feb 2021 news).
While early research focused on how ApoE influences Aβ accumulation, more recently attention has shifted to tau. In a dramatic example of the strong tie between ApoE and tau, a woman carried an autosomal-dominant AD mutation and also happened to carry a rare protective variant in ApoE3 called the Christchurch mutation. She had a brain full of amyloid but hardly any tau tangles (Nov 2019 news).
Alas, if toxic forms of tau are around, ApoE makes things worse. Using human ApoE knock-in mice crossed to the P301S tau model of tauopathy, Holtzman and colleagues previously reported that expression of ApoE—in particular, ApoE4— exacerbated neurodegeneration (Sep 2017 news). Mice devoid of all ApoE fared best, with hardly any signs of disease. Subsequently, the scientists reported that microglia were required to dole out ApoE4’s damaging effects in tauopathy (Oct 2019 news). Microglia only express ApoE under pathological conditions, while astrocytes churn it out constitutively (Sep 2017 news). Hence, a key question remained unanswered: Did the ApoE4-microglia axis of neurodegeneration stem from microglial expression of ApoE4, or could ApoE4 made by astrocytes instigate the noxious microglial behavior?
To get at this question, first author Chao Wang and colleagues began what became a three-year saga of mouse husbandry. The researchers requisitioned Taconic to create custom knock-in mice that carry floxed, Cre-removable human ApoE3 or ApoE4 genes instead of their endogenous mouse ApoE gene. They crossed them to Aldh1l1-Cre/ERT2 BAC transgenic mice. These can be induced to express the Cre recombinase driven by the promoter for astrocyte-specific aldehyde dehydrogenase 1 family member L1. This shuts off ApoE expression in astrocytes (Srinivasan et al., 2016). Tamoxifen added to the drinking water flipped the Cre switch. The scientists then bred these animals with control or tau P301S mice, which develop tauopathy. Ultimately, this generated P301S Tau/Aldh1l1-Cre/ApoE3flox/flox and Tau/Aldh1l1-Cre/ApoE4flox/flox—TAFE3 mice and TAFE4 mice, for short. In them, scientists can turn off human ApoE selectively in astrocytes at any time.
So what happened? The researchers waited until the TAFE3 and 4 mice reached 5.5 months of age—just as tau pathology started to appear—and then switched off their astrocytic ApoE. This reduced the total amount of ApoE in the cortex, an area vulnerable to tau accumulation, by 50 to 80 percent. In the cerebellum, which is unaffected by tau pathology, ApoE plummeted by more than 90 percent. These numbers made sense given that under physiological conditions, astrocytes are the primary source of ApoE. An oil placebo dropped into the drinking water did not affect ApoE expression.
Sparing Shrinkage. In TAFE3 female mice, turning off ApoE3 expression in astrocytes with tamoxifen (TAM) did not slow brain atrophy; turning off ApoE4 in TAFE4 mice did lessen atrophy (far right). This effect was weaker in males. [Courtesy of Wang et al., 2021.]
By 9.5 months of age, all the mice, regardless of which ApoE isoform they expressed or whether it was turned on or off, had brain shrinkage by multiple measures. This atrophy was greatest in ApoE4 mice. Shutting off astrocytic ApoE expression had no effect on atrophy in E3-expressing mice, but reduced atrophy by about 25 percent in E4-expressing mice (see image above). These effects were more pronounced in females. Unlike control mice, those whose astrocytic ApoE4 was off kept busy building nests, an innate behavior. Taken together, this suggested that astrocytic ApoE4 exacerbates neurodegeneration in response to tau accumulation.
By that age, removing ApoE from astrocytes also influenced the extent of tau pathology. In females, depleting astrocytic ApoE3 or ApoE4 reduced levels of phospho-tau-181 detected in cortical extracts. For males, there was a trend toward reduction, but only upon removal of ApoE4. Using immunohistochemistry, the researchers detected a drop in phospho-tau, as gauged by reactivity with the AT8 antibody, in cortical slices of female, but not male, mice sans astrocytic ApoE4. Together, the findings suggested that astrocytic ApoE4 exacerbated tau deposition, particularly in females. That said, removing either form of ApoE from astrocytes alone was not sufficient to completely stop tau pathology in either sex.
Tamp Down Tau. Phospho-tau levels in TAFE4 mice (left) were reduced (right) by abolishing ApoE4 expression. [Courtesy of Wang et al., Neuron, 2021.]
How did cells across the tau-tangled brain react to ApoE’s ouster? To address this, the researchers turned to single-nucleus RNA sequencing. They used hippocampal tissue from 9.5-month-old TAFE3 or TAFE4 or control mice that had ingested oil or tamoxifen at 5.5 months of age. Based on the transcriptomes of more than 71,000 nuclei isolated in toto, the researchers identified 16 clusters of cells. These included 11 subsets of neurons, and one each of oligodendrocytes, oligodendrocyte progenitor cells, choroid plexus epithelial cells, astrocytes, and microglia.
Among the neuronal subsets, the researchers spotted two clusters—1 and 3—that plummeted in response to tau pathology. Interestingly, removal of ApoE from astrocytes protected against this tau-mediated cell loss, with ApoE4 removal having the strongest effect.
In contrast, another transcriptional clique of neurons—cluster 6—dramatically expanded in mice with tau pathology. Knocking down astrocytic ApoE3 or E4 curtailed this expansion. These neurons uniquely expressed a cadre of RNA-binding proteins—Arrp21 and R3hdm1 among them. They also expressed Rorb, a protein recently reported to mark subsets of neurons vulnerable to tau pathology (Jan 2021 news). The researchers believe this transcriptome might represent the neurons’ response to ongoing tau pathology.
Indeed, RNA-binding proteins have been shown to mingle with tau and coax its aggregation. In particular, Arpp21 has been spotted in stress granules in neurons, which have also been linked to the stabilization of toxic tau oligomers and accelerated neurodegeneration (Nov 2017 news; Rehfeld et al., 2018). Investigating a potential connection between the two proteins with immunohistochemistry, the researchers spotted Arrp21 and phospho-tau co-mingling in P301S mice expressing either isoform of ApoE. Astrocytic ApoE4, but not ApoE3, depletion dramatically reduced this association in neurons.
How astrocytic ApoE4 might influence tau’s association with RNA-binding proteins in neurons remains unclear, Holtzman said. Altered neuronal signaling, perhaps in response to worse damage in ApoE4 mice, could disrupt many neuronal functions.
Astrocytes themselves also responded to tau pathology, and their response in turn was modulated by extinguishing their ApoE. The scientists defined three subclusters based on transcriptional profiling. One expressed homeostatic genes, one was reactive, and a third appeared senescent. Removing ApoE—especially E4—enlarged the pool of homeostatic astrocytes, while shrinking the reactive one. Nicholas Seyfried of Emory University in Atlanta commented that these results align with proteomic studies on thousands of human brains. In those, astrocytes expressing a similar subset of genes as those seen in the mouse reactive subtype correlated with neuropathology (May 2020 news).
Not to be outdone, oligodendrocytes also shifted their transcriptomes depending on tau and ApoE. The myelin-producing glia ratcheted up expression of the complement protein C4b in the presence of tangles, and draining astrocytes of either isoform of ApoE dampened this pro-inflammatory effect. The findings mesh with recent studies that identified C4b-expressing oligodendrocytes in models of amyloidosis (Nugent et al., 2020; Zhou et al., 2020).
Finally, the researchers examined how astrocytic ApoE molded microglial responses to tau pathology. Within the larger transcriptomic cluster of microglia, they found three subclusters. One was enriched for genes expressed under conditions of homeostasis, a small subset appeared to be peripheral macrophages, and the third had a disease-associated microglia (DAM)/neurodegenerative (MGnD) signature (Jun 2017 news; Sep 2017 news).
As expected, this third, pathological subset was dramatically expanded in the presence of tau pathology. Nixing astrocytic ApoE did not change the relative proportions of the different subsets of microglia in tau-ridden mice. That said, using immunohistochemistry of hippocampal sections, the researchers did see that when ApoE4 was gone from astrocytes, more microglia expressed the homeostatic marker P2ry12 and downshifted expression of select DAM/MGnD genes, including MHC II, Spp1, and Clec7a. Curiously, astrocytic ApoE did not significantly influence microglial expression of ApoE, a DAM gene.
Yes, this is complex, but hang in there, dear reader. Next, the scientists asked if these transcriptomic shifts influenced how microglia behave—in particular, their appetite for synapses. Compared to control mice, tangle-ridden P301Ss had about 20 percent fewer synapses in their entorhinal and piriform cortices by 9.5 months of age, and microglia were spotted having stuffed their lysosomes with synaptic material. Crucially, removing ApoE4 from astrocytes at month 5.5 meant the number of synapses was near-normal four months later, and markedly reduced the amount of synaptic material in the bellies of the microglial beasts.
Appetite for Destruction. In TAFE4 mice (left), microglia expressed more lysosomal CD68 (green), and engulfed more synapses (PSD95, red) than did microglia in TAFE4 mice without astrocytic ApoE4 (right). [Courtesy of Wang et al., Neuron, 2021.]
In all, the findings suggest that ApoE4 from astrocytes somehow riles up damaging microglial responses to tau pathology, and this despite microglia expressing their own ApoE under these conditions. What is so special about ApoE from astrocytes? Holtzman pointed out that it carries more lipids than ApoE from microglia. That could influence the way the apolipoprotein interacts with microglial cell-surface receptors. He speculated that under conditions of neuronal stress, such as tauopathy, ApoE4, along with other local factors such as complement proteins, might somehow bait microglia to eat neuronal synapses.
The findings argue against a key role for microglial ApoE during pathology in this model, commented Jesse Hanson of Genentech in South San Francisco. “Therefore, the role of activated microglia in mediating tau-dependent pathology is likely not so much via production of ApoE, but rather as a mediator of neuronal damage in response to the harmful inflammatory state of the brain that is caused by tau pathology, and exacerbated by [astrocytic] ApoE4,” he wrote.
Ulrich and Holtzman don’t think their data entirely rule out a role for microglial ApoE4 in tau-mediated neurodegeneration. After all, shutting off ApoE4 in astrocytes had a subtler benefit than complete knockout of ApoE, which, as the researchers had previously reported, erased nearly all traces of disease in P301S mice. This implies that ApoE derived from non-astrocytic cell types still influences neurodegeneration.
“This study re-emphasizes, and expands on, previous work by the Holtzman lab and others. It makes very clear that astrocytic APOE4 represents a toxic gain of function,” wrote Priyanka Narayan of the National Institutes of Health in Bethesda, Maryland. “Given APOE’s role in lipid metabolism, and the involvement of lipids in nearly every cellular process, it is possible that lipid state is a key mediator of astrocytic APOE’s inter- and intracellular effects,” she added.
Overall, the paper illustrates the striking degree of cross-talk between different cell types that steers the course of neurodegeneration. “If you modulate ApoE in astrocytes, it affects neurons, oligodendrocytes, and microglia in different ways,” Seyfried told Alzforum. In other words, when astrocytic ApoE binds to surface receptors on different cell types, different downstream signaling events may be triggered in those cell types, Seyfried noted.
On the lipidation idea, he added that astrocytic ApoE may have a unique structure and lipid content compared to ApoE derived from other cell types. Holtzman noted another possibility. Perhaps a heightened reactivity of ApoE4-producing astrocytes influences microglia via inflammation-stoking proteins, not via ApoE itself.
Bu hopes that future studies will investigate what’s behind the sex differences reported in the paper. They add to recent work suggesting that microglia in males and females react differently to all manner of insults, and that deficiency of the microglial cell surface receptor TREM2 differentially alters male and female microglial responses to tau (Jun 2018 news; Jul 2019 conference news). ApoE4 is a stronger driver of AD risk in women than in men (Sep 2017 news).—Jessica Shugart
- Would ApoE Make a Better Therapeutic Target Than Aβ?
- Can an ApoE Mutation Halt Alzheimer’s Disease?
- ApoE4 Makes All Things Tau Worse, From Beginning to End
- In Tauopathy, ApoE Destroys Neurons Via Microglia
- ApoE and Trem2 Flip a Microglial Switch in Neurodegenerative Disease
- Selective Vulnerability News: RORB Neurons Are First Victims of Tangles
- Stress Granule Protein Stabilizes Tau Oligomers, Hastens Neurodegeneration
- Massive Proteomics Studies Peg Glial Metabolism, Myelination, to AD
- Hot DAM: Specific Microglia Engulf Plaques
- Girl Power? In Mice, Female Microglia Protect Against Ischemic Injury
- Down to Sex? Boy and Girl Microglia Respond Differently
- New Look at Sex and ApoE4 Puts Women at Risk Earlier than Men
Research Models Citations
- Srinivasan R, Lu TY, Chai H, Xu J, Huang BS, Golshani P, Coppola G, Khakh BS. New Transgenic Mouse Lines for Selectively Targeting Astrocytes and Studying Calcium Signals in Astrocyte Processes In Situ and In Vivo. Neuron. 2016 Dec 21;92(6):1181-1195. Epub 2016 Dec 8 PubMed.
- Rehfeld F, Maticzka D, Grosser S, Knauff P, Eravci M, Vida I, Backofen R, Wulczyn FG. The RNA-binding protein ARPP21 controls dendritic branching by functionally opposing the miRNA it hosts. Nat Commun. 2018 Mar 26;9(1):1235. PubMed.
- Nugent AA, Lin K, van Lengerich B, Lianoglou S, Przybyla L, Davis SS, Llapashtica C, Wang J, Kim DJ, Xia D, Lucas A, Baskaran S, Haddick PC, Lenser M, Earr TK, Shi J, Dugas JC, Andreone BJ, Logan T, Solanoy HO, Chen H, Srivastava A, Poda SB, Sanchez PE, Watts RJ, Sandmann T, Astarita G, Lewcock JW, Monroe KM, Di Paolo G. TREM2 Regulates Microglial Cholesterol Metabolism upon Chronic Phagocytic Challenge. Neuron. 2020 Mar 4;105(5):837-854.e9. Epub 2020 Jan 2 PubMed.
- Zhou Y, Song WM, Andhey PS, Swain A, Levy T, Miller KR, Poliani PL, Cominelli M, Grover S, Gilfillan S, Cella M, Ulland TK, Zaitsev K, Miyashita A, Ikeuchi T, Sainouchi M, Kakita A, Bennett DA, Schneider JA, Nichols MR, Beausoleil SA, Ulrich JD, Holtzman DM, Artyomov MN, Colonna M. Human and mouse single-nucleus transcriptomics reveal TREM2-dependent and TREM2-independent cellular responses in Alzheimer's disease. Nat Med. 2020 Jan;26(1):131-142. Epub 2020 Jan 13 PubMed. Correction.
- Wang C, Xiong M, Gratuze M, Bao X, Shi Y, Andhey PS, Manis M, Schroeder C, Yin Z, Madore C, Butovsky O, Artyomov M, Ulrich JD, Holtzman DM. Selective removal of astrocytic APOE4 strongly protects against tau-mediated neurodegeneration and decreases synaptic phagocytosis by microglia. Neuron. 2021 May 19;109(10):1657-1674.e7. Epub 2021 Apr 7 PubMed.