Receptors and Broken Lysosomes Explain Risk from APOE Variants

Of the 20,000+ human genes, none has greater influence on late-onset Alzheimer’s disease than APOE. But why do certain variants of this lipid transporter protect the brain, while others endanger it? A suite of experiments published in the November 7 Cell suggests that the answer lies in lysosomes. The researchers, led by Jing Guo, now at NICO Therapeutics, and Gilbert Di Paolo of Denali Therapeutics, both in South San Francisco, found that lysosomes can’t handle ApoE4 when it comes bearing certain lipids. These ApoE4 particles cause lysosomes to accumulate lipofuscin, a collection of oxidized lipids and other molecules. The low-risk ApoE2 and Christchurch variants, on the other hand, don’t ferry these fats to lysosomes because they bind the low-density lipoprotein receptor too weakly.

  • ApoE3 and ApoE4 ferry lipids into cells via lipoprotein receptor.
  • Weak receptor binding prevents ApoE2 and Christchurch variants from doing the same.
  • Peroxidation-prone lipids borne by ApoE3/4 cause lipofuscin buildup in lysosomes.
  • At low pH, lipidated ApoE4 is more prone to aggregate than is ApoE3. 

Others praised the study. “This is a very important paper and, in my view, constitutes a real advance in our understanding of how the ApoE alleles/isoforms so profoundly alter one’s risk for developing Alzheimer’s disease,” wrote Steven Paul, Washington University, St. Louis, who was not involved in the study.

“This very exciting story […] brings ApoE-receptor binding back into the limelight of Alzheimer’s disease research,” wrote Rik van der Kant of Vrije Universiteit Amsterdam to Alzforum. “While there is much more ApoE biology to uncover, this is a landmark paper that very convincingly couples major biochemical hallmarks of ApoE variants (i.e., receptor binding) to lipid dysregulation and lysosomal dysfunction.” (Comments below.)

LDLR: An ApoE Gateway
Researchers have long known that ApoE2 binds weakly to low-density lipoprotein receptor. But many have assumed that if that were important for AD pathology, then the high-risk variant ApoE4 would bind LDLR more strongly than does ApoE3. Since E3 and E4 isoforms bind LDLR with similar affinities, the receptor has been largely dismissed in AD research, said first author Guo.

Di Paolo and colleagues set out to investigate the ApoE/LDLR relationship more thoroughly. They confirmed that ApoE3 and ApoE4 had similar affinities for LDLR, while ApoE2 binding was about 50 times weaker. This binding is so pathetic, in fact, that it probably does not happen in the brain, said Di Paolo. The rare, protective Christchurch mutation also weakened ApoE3 and ApoE4 affinity for LDLR.

Next, using fluorescent dyes attached to ApoE or to its lipid cargoes, the scientists found that lipidated ApoE3 and ApoE4 readily infiltrated a variety cells, including H4 neuroglioma cells and APOE-null iPSC-derived astrocytes. In contrast, very little lipidated ApoE2 made it into those cells. Likewise, the Christchurch mutation suppressed uptake of lipidated ApoE3 and ApoE4. Furthermore, no ApoE variant accessed cells when they had LDLR knocked out, or when the scientists added the LDLR extracellular domain as a decoy, showing that uptake depended on this receptor (image below).

“While it has been known for some time that the ApoE isoforms differentially bind to LDLR, tying this to the internalization of pathogenic lipid cargoes is a real advance,” noted Paul.

Internalized Lipids. In ApoE-negative astrocytes, exogenous ApoE3 and ApoE4 carry cholesteryl esters (blue) into the cells, whereas ApoE2 does not (top panels). When LDLR-binding is blocked using LDLR extracellular domains (bottom panels), no lipidated ApoE enters the astrocytes. [Courtesy of Guo et al., 2024.]

Dysfunctional Lysosomes
What happens to cells that take up lipidated ApoE? For one, they appear more reactive. When exposed to lipopolysaccharide, an immune activator, iPSC-derived microglia that had imbibed ApoE4 produced more cytokines, including interleukin-9, IL-13, and IL-18, than did cells carrying ApoE3, which in turn produced more ILs than cells carrying ApoE2, echoing the AD risk profile.

Secondly, Guo found that the ApoE lipoproteins can poison lysosomes. She stumbled on this when experimenting with molecular conglomerates of ApoE4 bound to cholesteryl esters made with the 20-carbon polyunsaturated fatty acid, arachidonic acid. When she added those ApoE particles to human iPSC-derived neurons, she noticed a mysterious fluorescence from their lysosomes.

“If it's lysosomal, it has to be lipofuscin,” Di Paolo recalls saying at the time.

In multiple cell types that Guo and Di Paolo tested, lipoproteins containing ApoE and arachidonic acid cholesteryl esters CE(20:4) induced lipofuscin in a pattern that matched their risk for AD; that is, lots of lipofuscin with ApoE4, intermediate amounts with ApoE3, and hardly any with ApoE2. The ApoE3Ch and ApoE4Ch variants also evoked very little. The antioxidant vitamin E curtailed lipofuscin formation. The findings suggest that lipids carried into cells by ApoE, and then oxidized, are the source of lipofuscin (image below). 

“We think we may have identified a plausible molecular mechanism for lipofuscin in the brain 180 years after its discovery,” said Di Paolo. “If you think about the fact that lipofuscin is lysosomal and involves lipid peroxidation, it makes sense that a protein such as ApoE that carries lipids into the lysosomes might be a culprit.”

Lipids to Lipofuscin. In H4 neuroglioma cells, ApoE lipoprotein particles containing cholesteryl esters cause lipofuscin (blue) to form. ApoE4 is most potent, ApoE3 less so, ApoE2 is inert. LDLR extracellular domains prevent lipofuscin formation, as does the K146E mutation in ApoE, which diminishes binding to LDLR. [Courtesy of Guo et al., 2024.]

The lipofuscin discoveries confirm long-held suspicions that ApoE affects lysosomes, noted Roberto Zoncu of the University of California, Berkeley. “I think that makes a strong case for why ApoE2 is protective, because it prevents this enhanced uptake and accumulation and downstream toxic cascade,” he said.

These observations were not limited to cells in culture. The scientists also found lipofuscin in the brains of mice engineered to express human ApoE. Mice carrying ApoE4 genes had more lipofuscin in their neurons than did ApoE3 mice, according to experiments conducted in collaboration with Alexandra Litvinchuk and David Holtzman at Washington University School of Medicine in St. Louis. Furthermore, wild-type mice developed lipofuscin when the scientists injected ApoE4-CE(20:4) particles into their brains.

Curiously, mice expressing ApoE4 developed even more lipofuscin in their neurons when they carried the human tau gene with the P301S tauopathy mutation. When the scientists added tau seeds to iPSC-derived neurons exposed to a variety of ApoE lipoprotein particles, lipofuscin formation and tau fibril internalization went hand-in-hand, with cells accumulating more fibrils under conditions that produced more lipofuscin.

But what about ApoE3?
Together, the lipofuscin and LDLR-binding experiments suggested a new model to explain the risk posed by different variants of ApoE. But that model did not explain how ApoE4 could be a stronger risk factor for AD than ApoE3. Why did it cause cells to release more cytokines and accumulate more lipofuscin? LDLR binding can’t explain either, because E4 and E3 bind with the same strength and transport the same quantities of lipids into cells.

An answer emerged when the scientists tested how each ApoE variant aggregates. Prior work has shown that unlipidated ApoE aggregates and seeds amyloid, but lipidated ApoE has not previously been found to aggregate, said Di Paolo (Hatters et al., 2006; Liu et al., 2021). The scientists found that ApoE4 lipoprotein particles do indeed aggregate when they contain CE(20:4) and are exposed to acidic conditions similar to those inside lysosomes. The same type of particles made with ApoE3 aggregated much less, which could make them easier for lysosomes to process.

“It's really a first instance where we can demonstrate that lipidated ApoE can also aggregate. But the catch, and this is why we think it's so interesting, is that it only aggregates when the pH is acidic, 4.5,” said Di Paolo.

Surprisingly, ApoE2 also formed aggregates at low pH. It’s possible that if E2 carried unstable lipids to the lysosomes, it would be even worse than E4, said Di Paolo. Fortunately, the weaker LDLR-binding keeps E2 outside.

Lipidated ApoE Aggregates. At low pH, ApoE4 and ApoE2 lipoprotein particles containing cholesteryl esters aggregate more than particles made with ApoE3. [Courtesy of Guo et al., 2024.]

Scientists appreciated the study’s meticulous approach. One strength, noted Julia TCW of Boston University, was the use of cell lines engineered to lack ApoE genes, which allowed the scientists to control the ApoE particles used in the experiments and track their movements.

Of course, this is less biologically relevant than systems in which cells produce their own ApoE. Guo and Di Paolo acknowledge that more work is needed to confirm their model using systems closer to Alzheimer’s disease, including human brain.

And even if the new model is correct, it may not be the whole story. Guojun Bu of Hong Kong University of Science and Technology noted that there could be other reasons why ApoE2 protects besides its weak binding to LDLR. Scientists also said studies will need to clarify where the ApoE is produced and how ApoE and lipofuscin relate to other aspects of AD such as amyloid plaques.

Still, these new insights could help guide drug development. Strategies to increase ApoE2 levels are in trials, but the new study casts doubt on that approach, said Sarah Cohen of the University of North Carolina Chapel Hill (Dec 2022 news). “If the problem with ApoE4 is that it's causing this increased lipid uptake, then overexpressing ApoE2 may not help,” she said.

On the other hand, lowering ApoE4 levels could work, Cohen added. Paul had other suggestions, including reducing synthesis of toxic lipids, increasing their transport out of cells, keeping them from entering via LDLRs, and protecting lysosomes from oxidative stress.

“If we know why [ApoE2] is protective, then we can come up with novel therapeutic strategies to mimic that property,” said Guo.—Nala Rogers

Nala Rogers is a freelance science writer based in Silver Spring, Maryland.

Comments

  1. First off, this is a very important paper and, in my view, constitutes a real advance in our understanding of how the ApoE alleles/isoforms so profoundly alter one’s risk for developing Alzheimer’s disease. The data suggest that this lies in the differential binding of the three common ApoE isoforms. The markedly reduced binding/affinity for, and internalization of, ApoE2 by LDLR may account for its protective effect, while the avid binding and internalization of ApoE3/4 carrying a cargo of “bad/toxic” lipids, such as polyunsaturated fatty acids (PUFA) and cholesterol esters (CEs), contributes to tau-dependent neuroinflammation/neurodegeneration and thus may account for the heightened risk of AD. Because ApoE2 doesn’t bind well to LDLR, and thus doesn't get internalized, it doesn't transport these potentially toxic lipids into cells, such as microglia and neurons.

    Moreover, that the “protective” Christchurch variant of ApoE3 also binds poorly to LDLR is very interesting and likely very relevant. Tying these well-accepted and well-replicated genetic risk factors (particularly the protective alleles ApoE2 and Christchurch) to a credible pathogenic mechanism is very cool. 

    It seems ApoE2 may represent a “loss-of-function” mutation and ApoE4 may represent a “gain-of-function” mutation with respect to AD pathogenesis and risk. The field has been debating this for quite some time! 

    While it has been known for some time that the ApoE isoforms differentially bind to the LDLR (E2<<E3/E4), tying this to the internalization of pathogenic lipid cargoes is a real advance. The authors also show that lipidated ApoE3/4 carries PUFU-CE species into cells via the LDLR, resulting in lysosomal deposition of these toxic lipids as well as greater lipofuscin formation and deposition. The greater endolysosomal delivery of PUFU-CE species results in greater lipid peroxidation and lipofuscinosis, leading to increased tau pathology and presumably cell dysfunction or death. Again, tying these together into a credible pathogenic cascade is a very important contribution.

    Importantly, over time ApoE4 apparently causes greater lipid peroxidation and lipofuscinosis than does ApoE3, and presumably more tau pathology as well. The latter potentially accounts for ApoE4’s “gain-of-function” for AD pathogenesis and risk relative to ApoE3. However, since ApoE4’s relative binding affinity for LDLR is very similar to ApoE3’s, this alone cannot account for the three- to-12-fold greater risk of AD among ApoE4 carriers than ApoE3s. As the authors point out, the higher aggregation propensity of lipidated ApoE4 compared to ApoE3 may also contribute to the differences in pathogenicity, especially for amyloid pathology, and AD risk between these two isoforms. Thus ApoE4’s “gain of function” may also be due to additional pathogenic mechanisms (i.e., not just the LDLR-mediated delivery of toxic lipid species to endolysosomes). 

    The study also suggests novel therapeutic targets, approaches, and strategies that might reduce the intracellular levels of ApoE’s pathogenic lipid cargoes, such as PUFU-CEs, for example, mechanisms that might reduce their synthesis, increase their export out of cells, or even reduce their internalization into cells via the LDLR. 

    All in all, this is an important paper and series of findings.

    View all comments by Steven Paul

  2. This very exciting story from Guo, Di Paolo, and collaborators brings ApoE-receptor binding back into the limelight of Alzheimer’s disease research. In a convincing manner, the team shows how ApoE2 and ApoE Christchurch have reduced LDLR binding, thereby causing differences in lipoprotein-particle uptake. What really takes this story further from previous findings is the array of functional experiments that show the consequence of this altered receptor binding. The link to lipofuscin, lipid peroxidation, and lysosomal dysfunction is very interesting, and it is clear that the behavior of ApoE2 and ApoE Christchurch is markedly different than that of ApoE3 and ApoE4. From a drug-development perspective this story provides further rationale to try to block the uptake of ApoE-lipoprotein particles, although cell-type-specific effects of such blockage will need to be explored more.

    How ApoE and polyunsaturated fatty acids-cholesterol esters (PUFA-CE) synergistically affect lipofuscin, while PUFA-CE themselves seem to have little effect, and how PUFA-CE affect tau accumulation (in lysosomes or also outside?) needs further research, but these changes hint at some yet-to-be-discovered biology that might include interplay with other lipids and/or lysosomal processes.

    I also found it striking that only a few assays detect differences between ApoE3 and ApoE4 lipid particles. While ApoE4 leads to more lipofuscin, cholesterol homeostatic processes don't seem differentially impacted by feeding of exogenous ApoE4 versus ApoE3 particles (whereas ApoE2 has vastly different effects). One thing this indicates is that the protective effect of ApoE2/2 and ApoE Christchurch is different than the damaging effect of ApoE4, which in itself is an important conclusion. A possible caveat in this, as also noted by the authors, is that the ApoE-lipid particles generated here are likely different from those in the brain. As ApoE2, ApoE3, and ApoE4 lipid particles in the brain differ in the amounts and types of lipids they carry, this likely adds another point of biological divergence. 

    So, while there is much more ApoE biology to uncover, this is a landmark paper that definitively couples major biochemical hallmarks of ApoE variants (i.e., receptor binding) to lipid dysregulation and lysosomal dysfunction, which is highly relevant for our understanding of AD. 

    View all comments by Rik van der Kant
    • User Profile Image Guojun Bu
      Hong Kong University of Science & Technology
    • Posted:

    This is a very interesting paper exploring the protective mechanism of APOE2 in AD using mostly in vitro cell-based assays. Research on the biology and pathobiology of ApoE, the strongest genetic risk factor for AD, has become increasingly interesting, and we have strong evidence that a major pathway by which APOE4 increases AD risk is by driving early and more abundant amyloid, while additional toxic mechanisms include vascular impairment and altered immune responses. However, how APOE2 protects has proven largely elusive, with some evidence suggesting involvement of lipid metabolism. The higher level of ApoE2 in the brain and its greater lipidation support this hypothesis.

    In this paper, the authors nicely recapitulate what we know from studies decades ago, namely that ApoE2/lipoprotein has vastly reduced affinity for the LDL receptor (LDLR) compared to ApoE3 and ApoE4. This translates to reduced ApoE/lipoprotein uptake by the LDLR, and minimal downregulation of cell surface LDLR compared to ApoE3 and ApoE4. Here, the specificity of binding to, and endocytosis of, LDLR was addressed by LDLR knockout and by competition with recombinant LDLR extracellular domain. The authors further explored the effects of ApoE isoforms on the uptake and metabolism of specific lipids including cholesteryl esters (CEs) and polyunsaturated fatty acids (PUFAs), both linked to AD pathogenic events. They found that lipid peroxidation-related lipofuscinosis is greater in ApoE4 cells and less in ApoE2 cells, compared to ApoE3 cells. PUFA-CE-lipApoE4 injection into the hippocampus induced lipofuscinosis in vivo, which was also linked to increased tau accumulation in lysosomes in APOE4 mice. Finally, the authors showed evidence that ApoE3-Christchurch might also protect in this LDLR-related manner. Collectively, the authors provided strong evidence that ApoE2 limits brain lipid-related toxicity by reducing the uptake and accumulation of toxic lipids through a LDLR-mediated pathway.

    The physiological function of ApoE in the brain is to transport lipids in homeostasis and during injury repair. ApoE is essential for synaptogenesis, likely by transporting lipids from astrocytes to neurons via ApoE receptors. With reduced binding of ApoE2 to the LDLR, it is difficult to comprehend that LDLR is an important receptor compared to LRP1, the other major ApoE metabolic receptor in the brain, during physiological lipid transport. However, under pathological conditions such as AD, the transported lipid species can be vastly different, as is the cell type-specificity of transport. Under such conditions, including those described in this paper, reduced lipid uptake might be beneficial. One other aspect that was not explored in this study is the role of ApoE in lipid efflux, in which ApoE4 is deficient through either a loss of function or a gain of negative function. A pathway as critically important as lipid efflux is essential for brain lipid metabolism and repair, including that following microglia-mediated phagocytosis of amyloid, myelin debris, and degenerated synapses and neurons. Does ApoE2 have an advantage over ApoE3 and other ApoE protective variants in lipid efflux, including of toxic lipids? Another critical pathway that is gaining increasing interest is the detoxification of peroxidated lipids formed in neurons by astrocytes, a process that involves lipid efflux to ApoE through ABCA1 in neurons, and lipid uptake by ApoE receptors in astrocytes. It will be important to assess how ApoE2 functions in such events compared to ApoE3 and E4.

    We are still at the early stage of understanding the protective mechanisms of ApoE2. While the LDLR-related pathway, as shown in this paper, might be one of them, there are likely multiple pathways where ApoE2 can outperform ApoE3 and E4 to convey resiliency, or where it is more silenced when ApoE exhibits toxicity, including pathology-related pathways where ApoE4 has clear gain of toxic functions. To build new hypotheses, the field will need to gain further insights from human studies, including analyses of human biospecimens (CSF, plasma, and postmortem brains). Model systems, including conditional mouse models and iPSC-derived cell and organoid models, can provide versatility in modulating ApoE isoform levels and testing the effects in the presence or absence of pathology and injury. The potential liability of ApoE2 in eye diseases and CAA-related microbleeds will also need to be considered, particularly when the protective function of ApoE2 is championed to overcome the detrimental effects of ApoE4.

    View all comments by Guojun Bu

  4. This comprehensive study highlights protective roles of APOE2 and APOECh alleles in altered LDLR recycling and reduced lipofuscinosis, a lysosomal pathology induced with age and neurodegeneration.

    Briefly, these protective alleles lead to reduced endosomal uptake of LDLR, leading to less lipid burden in lysosomes. In particular, reduction in lysosomes of polyunsaturated fatty acid cholesterol esters, which can be peroxidated, is protective against lipofuscin buildup.

    This paper is another of several recent studies converging on lysosomal stress in aging and neurodegeneration. The endosomal recycling pathway is also clearly implicated; now it is time to find effective therapeutics to target this pathway. Reducing cargo aberrantly stuck in these organelles by promoting efficient clearance would be a promising strategy for both neurodegeneration and brain aging.

    View all comments by Jessica Young

  5. The authors propose an inverse relationship between lipidated ApoE-LDLR interactions and the risk of late-onset Alzheimer’s disease. This model is supported by observations that increased binding between lipidated ApoE and LDLR results in lysosomal impairments due to the delivery of polyunsaturated fatty acid-cholesterol esters (PUFA-CEs). Specifically, the increased delivery of PUFA-CEs correlates with the accumulation of lipofuscin in lysosomes, which parallels the enhanced LDLR binding by ApoE3 and ApoE4 compared to ApoE2. The protective ApoE Christchurch variant also exhibited decreased LDLR binding and reduced ability to induce neuronal lipofuscinosis. While both ApoE3 and ApoE4 facilitate similar uptake of PUFA-CEs, which might appear inconsistent with ApoE4’s greater disease risk, ApoE4’s higher propensity to aggregate at the acidic pH within lysosomes distinguishes its effects from those of ApoE3.

    Lipofuscin, a poorly understood fluorescent storage material whose formation is associated with lipid peroxidation, frequently signals lysosomal dysfunction. In neurons, lipofuscin is commonly observed in lysosomal storage diseases caused by mutations in genes encoding proteins essential for lysosomal degradation and recycling of cellular macromolecules. The high levels of lipofuscin in lysosomes of neurons exposed to ApoE4 loaded with PUFA-CEs strongly suggests that this route for PUFA-CE delivery has adverse effects on neuronal lysosomes.

    In mouse models, lipofuscin accumulation downstream of ApoE was exacerbated by the expression of the P301S mutant tau, suggesting a synergistic effect between tau and ApoE in perturbing lysosomal function. Furthermore, the ApoE-dependent accumulation of lipofuscin was accompanied by increased levels of tau preformed fibrils, potentially reflecting impaired lysosomal degradation of these disease-associated aggregates. Supporting this, lysosomal protease (cathepsin) activity was found to decrease in parallel with lipofuscin accumulation, further implicating lipofuscin in lysosomal dysfunction. Although the mechanisms underlying the synergy between tau and ApoE remain unclear, their convergence on lysosomal disruption underscores the potential disease relevance of these findings.

    Collectively, these results highlight how specific differences in endocytic uptake between protective and high-risk ApoE variants, along with how lysosomes process these lipoproteins, may contribute to AD pathogenesis. These findings underscore the importance and vulnerability of neuronal lysosomes in AD. Understanding the mechanisms driving lysosomal dysfunction downstream of ApoE, PUFA-CEs, and tau represents a critical area for future research. In particular, the consequences of increased ApoE4 aggregation at acidic pH requires more investigation. Additionally, the data suggest that targeting the LDLR-ApoE interaction could offer a promising therapeutic strategy for protecting against AD.

    View all comments by Shawn Ferguson

  6. In this paper, Guo and colleagues offer important mechanistic insights into the role of AD-associated ApoE variants in lipid transfer and intracellular accumulation. This is achieved using a reductionist approach, expressing various disease-associated ApoE isoforms in HEK cells, isolating and lipidating ApoE with various lipid species, and treating recipient cells with this ApoE. Key findings include: 1. ApoE with the Christchurch variant leads to significantly less lipid uptake, while the ApoE4 variant leads to increased lipid uptake; 2. ApoE-mediated lipid uptake is dependent on the LDL receptor (LDLR) in several iPSC models of CNS cell types; and 3. Uptake of lipidated ApoE induces dramatic lipofuscin accumulation in iPSC-derived neurons, and this is reduced in the ApoE Christchurch variant. Furthermore, in tau mouse models, significant neuronal lipofuscin accumulation occurs in an ApoE-dependent manner. This work also confirms prior findings that lipid accumulation in microglia induces a proinflammatory state associated with lysosomal lipid accumulation, and that cytosolic lipid droplet accumulation in microglia is induced with LPS.

    The authors present excellent mechanistic work that suggests protective and risk-associated ApoE variants alter lipofuscin, an age-related lysosomal lipid accumulation. They convincingly show that this effect is pronounced in neurons in vitro and in vivo and is mediated by LDLR. One of the strengths of this paper is the use of a well-defined reductionist system to compare receptor binding and lipid uptake rates across multiple disease-associated ApoE variants. However, while this is a strength, it also highlights a limitation: the simplification of complex CNS cell interactions. For example, in the diseased CNS, glia producing ApoE most likely lipidated it with lipids not represented in this study, have other ApoE post-translational modifications (PTMs), or transfer lipids in modes or at concentrations not reproduced by the HEK cell expression of these constructs. While the approach used here provides valuable insights, it would be interesting to contrast the effects of lipidated ApoE on neurons observed in this study with the effects of lipidated ApoE derived from astrocytes or microglia.

    The effect of lipidated ApoE isoforms on neuronal lipofuscin levels is striking, and future studies could explore whether this alters neuronal function or survival. One might assume that lysosomal lipid peroxidation, as reflected by lipofuscin, is inherently harmful. However, determining how this relates specifically to AD-associated neuronal pathology would help establish a clearer causal link between ApoE, neuronal lipofuscin, and AD pathogenesis. The study hints at a connection between lipofuscin and tau accumulation, but it would be interesting to see how this plays out in future experiments in vivo; for instance whether the authors’ in vivo cholesterol fatty acid ester-phosphatidylcholine (CE(20:4)/POPC)-lipidated ApoE4 injection paradigm induces neuronal degeneration or glial reactivity over time. Would these injections influence tau spread or neuronal atrophy in tauopathy models? Such experiments would greatly strengthen the argument for a causal relationship between neuronal lipofuscin and AD pathology.

    More broadly, questions remain regarding the causal link between lipofuscin and AD. Lipofuscin is associated with brain aging but is also ubiquitous in healthy aged individuals without AD, in other neurological diseases, and in cognitive decline. Is the association between AD and lipofuscin due to the number of neurons containing lipofuscin, the amount of lipofuscin per cell, or specific lipid species involved? Alternatively, does lipofuscin sensitize neurons to secondary insults, such as tau aggregation?

    In summary, this excellent mechanistic study provides critical insights into how AD-associated ApoE variants influence neuronal lipofuscin. It lays the groundwork for further research on the role of ApoE-induced neuronal lipofuscin in AD and highlights potential avenues for novel therapeutic interventions.

    View all comments by Michael Haney

  7. With the expansion of the number of APOE variants that affect the risk of AD, there is a great opportunity to define functional effects of the ApoE proteins that track with risk. Guo et al. took on a Herculean task of testing many aspects of neuronal lipid uptake and metabolism that are dependent on ApoE. Lipid metabolism has always seemed like the logical place to look for how APOE genotype affects risk of AD, given ApoE’s established role as the major protein in brain lipoproteins. The authors focused on the LDL receptor, which is the main receptor that modulates ApoE metabolism in the brain. Between the effects of ApoE variation on LDL receptor binding and on LDL receptor recycling, they develop a model whereby ApoE variation promotes pathogenesis through the delivery and aberrant accumulation of specific types of lipids in neurons.

    But these manipulations are acute, and mostly in cell culture; how do they affect the chronic processes seen in neurodegeneration? Here, the authors demonstrate that ApoE-lipoprotein uptake can lead to the formation of lipid aggregates and dysfunctional lysosomes inside neurons. They find that ApoE-lipoproteins can deliver polyunsaturated fatty acids (in the form of cholesterol esters), and, in the presence of oxidative processes, these lipids accumulate to generate vesicles that seem like lipofuscin. The formation of lipofuscin with aging is known to require lipids and oxidation, and it contributes to lysosomal dysfunction, fitting well with their model. It is a concise way to bring together ApoE isoforms with problems of neuronal viability and protein degradation, necessary aspects of AD pathogenesis.

    This model does not answer everything and requires more in vivo studies and testing in a larger number of human tissue samples. For example, while the work indicates other processes that may contribute to AD pathogenesis, work needs to be done to define the connection between ApoE-lipoproteins and the induction of inflammation and to determine whether ApoE accumulates inside lipofuscin, as the authors suggest. But the model integrates many old and new findings in the field: receptor binding constants of ApoE isoforms, their effects on receptor recycling, on lysosomal dysfunction, and the effects of APOE genotypes on lipid droplets and cholesterol esters. It even speculates on the observed effects of ApoE2 on macular degeneration. I am sure that I will find myself using the graphical abstract when I lecture on how APOE genotype contributes so strongly to brain dysfunction.

    View all comments by G. William Rebeck

  8. Many neurodegenerative disease-associated risk factors, including ApoE, exert key functions in endo-lysosomal and lipid homeostasis. ApoE is central to the pathology of Alzheimer´s disease since heterozygous, and even more so homozygous, ApoE4 carriers have a strongly increased disease risk, while ApoE2 carriers are protected. Despite this, mechanisms driving these isoform-dependent differences are still not understood.

    This group, led by Gilbert Di Paolo at Denali Therapeutics, set out to study the role of different ApoE isoforms in lipid homeostasis. Their work focuses on a comparison of the known alleles of ApoE, showing an AD risk in the order ApoE2<ApoE3<ApoE4. In addition to the protective effect of ApoE2, they also assessed the intriguing ApoE Christchurch variant R136S, which was shown to be protective in kindred of PSEN1 E280A mutation carriers. Using state-of-the-art technology, the authors convincingly demonstrated reduced low-density lipoprotein receptor (LDLR) binding of the lipidated protective (ApoE2 and Christchurch) variants compared to binding measured by lipidated ApoE3 and ApoE4. However, no major difference in LDLR binding was observed between ApoE3 and ApoE4 variants, suggesting that other mechanisms, including different aggregation propensities, contribute to differences in their AD-related risk potential. An interesting finding, from the perspective of the LDLR biology, is that the reduced ApoE2 binding to LDLR may lead to overall higher levels of the LDLR at the cell surface, as most clearly seen in human iPSC-derived astrocytes. In contrast to ApoE2, ApoE3 and ApoE4 alter the LDLR signalling by shifting the localization of this key lipid receptor from the cell surface and recycling endosomes toward endo-lysosomes, which is in line with their increased internalization of ApoE and its lipid cargos.

    Notably, the authors show increased levels of cholesteryl esters (CE) and in particular polyunsaturated fatty acids (PUFA)-containing CE(20:4) in CSF samples of AD patients. As PUFAs are prone to peroxidation, the authors turned toward investigating the possible effect of ApoE isoforms in lipofuscinosis, which is fuelled by peroxidised lipids. Lipofuscinosis is commonly known as a hallmark of lysosomal storage diseases, and lipofuscin accumulates in neurons during aging. However, very little is known about its contribution to neurodegeneration. Guo et al. found increasing effects in the order ApoE2<ApoE3<ApoE4, with a protective effect of ApoE2 on triggering lipofuscinosis in iPSC-derived neurons, which was also observed when the Christchurch mutation was introduced in ApoE3 (ApoE3ch) or ApoE4 (ApoE4ch). Notably, the effect detected with ApoE4ch preparations did not reach the same protective effect on lipofuscinosis seen with ApoE3ch.

    One limitation of this study, also acknowledged by the authors, is that the in vivo complexity of brain ApoE lipoproteins may not be fully recapitulated by artificially lipidated ApoE used in the current study. It is appreciated that the authors demonstrated that when using different lipid composition of CE (18:4) no lipofuscin could be observed, matching their intrinsic differences in peroxidation potential. It would be very interesting to analyze lipofuscin accumulation and composition in AD brains in relation to the ApoE genotype. The authors hypothesized that ApoE4 and pathological tau may enhance lipofuscin pathology in vivo. This is an exciting possibility to explore. The evidence presented of lipofuscinosis in vivo upon injection of CE(20:4)-ApoE4 (compared to PBS) would benefit from including a control injection of CE(20:4)-ApoE2 or CE(18:1)-ApoE4.

     A cell culture-based model of lipofuscinosis generated in this study is a valuable tool for future mechanistic studies. Of note, it would be desirable to achieve more pronounced defects in lysosomal function over time as a direct effect of lipofuscin accumulation.

    The very interesting question of the cellular source of ApoE was not investigated in this study. Upregulation of ApoE is observed in reactive microglia around Aβ plaques and co-aggregated ApoE is found within plaques and tangles of AD patients’ brains. Recently published work supports a pathological role of ApoE in microglia where it co-aggregates with Aβ peptides in lysosomes, contributing to initial Aβ plaque formation (Kaji et al., 2024). Together with another recent study that provides strong evidence for astrocytic origin of ApoE and its contribution to Aβ pathology (Preman et al., 2024), it is clear that mechanisms described here need to be considered in the context of different cellular sources of ApoE in AD. Furthermore, LDLR-mediated cellular uptake of ApoE may also differ among different brain cells influenced by LDLR expression levels, as, for example, shown in iMicroglia, where LPS stimulation upregulated LDLR.

    Indeed, it remains a challenge to sort out physiological differences between the ApoE isoforms and to fully understand the pathological role of ApoE4. Although this insightful study provides evidence of how ApoE4 could be rendered less harmful, for a therapeutic design strategy, better separation of the physiological and pathological roles of ApoE genotypes is needed. We should also keep in mind that in addition to the ApoE4 risk allele genotype, ApoE4 haplotype may also contribute to glial-cell mediated non cell-autonomous AD risk mechanisms, as nicely demonstrated by the group of Alison Goate (TCW et al., 2022).

    References:

    Kaji S, Berghoff SA, Spieth L, Schlaphoff L, Sasmita AO, Vitale S, Büschgens L, Kedia S, Zirngibl M, Nazarenko T, Damkou A, Hosang L, Depp C, Kamp F, Scholz P, Ewers D, Giera M, Ischebeck T, Wurst W, Wefers B, Schifferer M, Willem M, Nave KA, Haass C, Arzberger T, Jäkel S, Wirths O, Saher G, Simons M. Apolipoprotein E aggregation in microglia initiates Alzheimer's disease pathology by seeding β-amyloidosis. Immunity. 2024 Oct 9; PubMed.

    Preman P, Moechars D, Fertan E, Wolfs L, Serneels L, Shah D, Lamote J, Poovathingal S, Snellinx A, Mancuso R, Balusu S, Klenerman D, Arranz AM, Fiers M, De Strooper B. APOE from astrocytes restores Alzheimer's Aβ-pathology and DAM-like responses in APOE deficient microglia. EMBO Mol Med. 2024 Nov 11; Epub 2024 Nov 11 PubMed.

    Tcw J, Qian L, Pipalia NH, Chao MJ, Liang SA, Shi Y, Jain BR, Bertelsen SE, Kapoor M, Marcora E, Sikora E, Andrews EJ, Martini AC, Karch CM, Head E, Holtzman DM, Zhang B, Wang M, Maxfield FR, Poon WW, Goate AM. Cholesterol and matrisome pathways dysregulated in astrocytes and microglia. Cell. 2022 Jun 23;185(13):2213-2233.e25. PubMed. BioRxiv.

    View all comments by Michael Willem View all comments by Sabina Tahirovic

Make a Comment

To make a comment you must login or register.

References

Mutations Citations

  1. APOE R176C (ApoE2)
  2. APOE C130R (ApoE4)
  3. APOE R154S (Christchurch)
  4. MAPT P301S

News Citations

  1. In Small Trial, Gene Therapy Spurs ApoE2 Production

Paper Citations

  1. Hatters DM, Zhong N, Rutenber E, Weisgraber KH. Amino-terminal domain stability mediates apolipoprotein E aggregation into neurotoxic fibrils. J Mol Biol. 2006 Sep 1;361(5):932-44. PubMed.
  2. Liu CC, Murray ME, Li X, Zhao N, Wang N, Heckman MG, Shue F, Martens Y, Li Y, Raulin AC, Rosenberg CL, Doss SV, Zhao J, Wren MC, Jia L, Ren Y, Ikezu TC, Lu W, Fu Y, Caulfield T, Trottier ZA, Knight J, Chen Y, Linares C, Wang X, Kurti A, Asmann YW, Wszolek ZK, Smith GE, Vemuri P, Kantarci K, Knopman DS, Lowe VJ, Jack CR Jr, Parisi JE, Ferman TJ, Boeve BF, Graff-Radford NR, Petersen RC, Younkin SG, Fryer JD, Wang H, Han X, Frieden C, Dickson DW, Ross OA, Bu G. APOE3-Jacksonville (V236E) variant reduces self-aggregation and risk of dementia. Sci Transl Med. 2021 Sep 29;13(613):eabc9375. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. Guo JL, Braun D, Fitzgerald GA, Hsieh YT, Rougé L, Litvinchuk A, Steffek M, Propson NE, Heffner CM, Discenza C, Han SJ, Rana A, Skuja LL, Lin BQ, Sun EW, Davis SS, Balasundar S, Becerra I, Dugas JC, Ha C, Hsiao-Nakamoto J, Huang F, Jain S, Kung JE, Liau NP, Mahon CS, Nguyen HN, Nguyen N, Samaddar M, Shi Y, Tatarakis D, Tian Y, Zhu Y, Suh JH, Sandmann T, Calvert ME, Arguello A, Kane LA, Lewcock JW, Holtzman DM, Koth CM, Di Paolo G. Decreased lipidated ApoE-receptor interactions confer protection against pathogenicity of ApoE and its lipid cargoes in lysosomes. Cell. 2024 Nov 11; Epub 2024 Nov 11 PubMed.

Annotate

To make an annotation you must Login or Register.