Early and lifelong effects of APOE4 on neuronal gene expression networks relevant to Alzheimer’s disease

  1. Brian P. Grone
  2. Kelly A. Zalocusky
  3. Yanxia Hao
  4. Seo Yeon Yoon
  5. Patrick Arriola
  6. Yadong Huang

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This manuscript is of broad interest to readers in the field of Alzheimer's disease, neurodegeneration, and single-cell omics. The identification of shared pathways across different cell types and ages is an important contribution to our understanding of APOE4 gene regulation in a cell type-specific manner. A combination of snRNAseq in APOE mouse models and human iPSC cells supports the key claims in the paper.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Apolipoprotein E4 ( APOE4 ) genotype and aging are critical risk factors for Alzheimer’s disease (AD). Aged APOE4 knock-in (APOE4-KI) mice have phenotypes reflecting features of AD. We conducted a large-scale single nucleus RNA-sequencing study to identify cell-type-specific effects of APOE4 on hippocampal gene expression during aging. APOE4-KI mice showed prominent alterations, relative to APOE3-KI mice, in neuronal transcriptome related to synaptic function, calcium signaling, and MAPK/Rap1/Pld signal transduction, starting by 5 months and persisting during aging. Mice with the APOE4 gene removed specifically from neurons failed to show most of these neuronal transcriptomic changes, suggesting a specific effect of neuron-derived APOE4 on the transcriptome. APOE4 affects similar cellular pathways in induced pluripotent stem cell-derived human neurons transplanted into APOE4-KI mouse hippocampus and in cortical neurons from aged human brains. Thus, neuronal APOE4 has early and persistent effects on neuronal transcriptomes, suggesting the requirement of early interventions for successfully treating APOE4 -related AD.

Article activity feed

  1. eLife

    Evaluation Summary:

    This manuscript is of broad interest to readers in the field of Alzheimer's disease, neurodegeneration, and single-cell omics. The identification of shared pathways across different cell types and ages is an important contribution to our understanding of APOE4 gene regulation in a cell type-specific manner. A combination of snRNAseq in APOE mouse models and human iPSC cells supports the key claims in the paper.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  2. eLife

    Reviewer #1 (Public Review):

    The authors generated valuable snRNAseq data sets from the hippocampus region in APO E4 and E3 mouse models. Through bioinformatics analysis, they identified a list of differentially expressed genes between E4 and E3at 5, 10, 15, and 20 months. In addition, changes in cell type distribution were observed across different time points, and the number of differentially gene expressions varied across multiple cell types. Through pathway enrichment analysis, the authors identified shared pathways such as calcium signaling and MAPK/Raps1/Pld pathways. To determine the relevance of these observations with respect to human Alzheimer's disease, they verified that genes/pathways identified in their mouse models are largely conserved in APOE4-Knockin and human APE4 iPSC-derived Neurons.

    A major strength of this study comes from the combination of mouse and human models using snRNAseq analysis. In addition, the authors also used comprehensive bioinformatics tools to dissect the shared genes/pathways during disease progression. While a major weakness of the study is the lack of experimental validation of the specific pathways and their impact on disease. The observational gene expression analysis cannot provide any casual information. It is unclear whether the genes and pathways identified are primary events of disease etiology or secondary events due to disease progression.

    Finally, I want to congratulate the authors on creating and sharing such a comprehensive set of snRNAseq data of the APO e4 allele. This set of omics data will become a reference point for the Alzheimer's research community. Their initial analysis of this rich dataset has yielded many interesting findings that may be validated by other groups.

  3. eLife

    Reviewer #2 (Public Review):

    The goal of this paper is to understand how neuronal APOE4 and aging interact to elicit Alzheimer's disease phenotypes. The authors utilize a previously published snRNA-seq dataset of APOE3 and APOE4 KI mice at different ages to explore this question. Interestingly, they find that neurons are the only cell type to show substantive transcriptomic changes in all time points analyzed. These transcriptomic changes center around synaptic and calcium signaling pathways. Using new and previously published datasets from mice with neuron-specific deletion of APOE4, the authors find that neuronal transcriptomic changes are primarily dependent upon neuronal APOE4, not APOE4 from other cell types. Finally, these findings are supported by additional previously published snRNA-seq datasets of iPSC-derived APOE4 neurons transplanted into APOE4KI mice and human prefrontal cortex. The authors posit that these early and persistent APOE4-dependent changes they observe in neurons necessitate early intervention for APOE4-related Alzheimer's disease.

    Strengths:
    The authors take advantage of multiple APOE3/4 snRNA-seq datasets to identify a strikingly consistent number of pathways dysregulated in excitatory neurons that are dependent upon neuronal APOE4. The number of human and mouse datasets utilized for this study makes the identification of these transcriptional changes convincing.

    The use of APOE4 KI mice with neuron-specific deletion of APOE4 is a very clever way to demonstrate the dependence of the dysregulated pathways on neuronal APOE4.

    The paper is clearly written and the figures are well-formatted and legible.

    Weaknesses:
    While the authors have access to multiple snRNA-seq datasets, their analyses are rather incomplete. The APOE4-fKI/Syn1-Cre snRNA-seq dataset is used to identify neuronal ApoE4-induced genes. However, the authors only use it as a filter to identify neuronal DE genes dependent upon neuronal APOE4. There is no discussion of the neuronal DE genes that aren't dependent upon neuronal APOE4 (and thus must be dependent upon APOE4 in other cell types, which could generate some interesting hypotheses regarding the effect of cell-type specific APOE4 on neuron-glia signaling). Conversely, the potential changes in non-neuronal DE genes (genes that may rely on APOE4 expression in neurons) are also not discussed. Because the authors are interested in the neuronal effects of APOE4, it seems that this type of analysis and discussion would strengthen their case.

    The authors emphasize 16 pathways that are persistently dysregulated at all ages, which are categorized into three main biological modules (synaptic function, calcium signaling, and signal transduction). While they reference previous literature demonstrating aberrant neuronal activity in young adult APOE4 KI mice, follow-up experiments confirming the present findings are absent.

    Only the 5-month and 10-month APOE4-fKI/Syn1-Cre snRNA-seq was generated for this paper. The rest of the data presented is publicly available. Furthermore, no attempt is made to validate any of the identified DEGs or pathways via independent methods such as IHC, qPCR, etc.

  4. eLife

    Reviewer #3 (Public Review):

    The study by Grone and colleagues proposes to understand how APOE4 contributes to Alzheimer's disease risk by understanding how different cell types within the brain are affected at the level of the transcriptome across the lifespan. There are several strengths of the study, including the concept of profiling different cell types across the lifespan using advanced sequencing methods and the use of a model incorporating neuron-specific deletion of APOE to understand how distinct pools of APOE affect the networks identified according to the form of APOE allele being expressed. There are a number of pathways identified that may inform the field in terms of the elusive role of neuronal APOE in shaping brain function. There are a number of issues in this work that limit many of the conclusions made. For example, the ages chosen to study how APOE alleles affect gene expression in different cell types are limiting and do not unfortunately include earlier ages representing developmental or young adult ages or very advanced age, two ends of lifespan where many functional changes occur in the brain that may be regulated by APOE. Additionally, sex is not studied as a biological variable in the study, leaving the results in question as to whether the findings are limited to one sex. There are a number of other methodological issues, including a lack of clarity on how variance from different sequencing datasets generated at different times for ages within the same comparisons has been handled. In terms of the impact of the study, there is a missing functional validation of key networks that have been identified. We do not know if any of the gene expression differences identified here translate to changes in brain function, limiting our ability to know whether neuronal APOE regulates the brain and may play a role in AD as claimed. Finally, constitutive deletion of APOE within neurons may result in changes in gene expression that are shaped by developmental changes mediated by APOE. Overall, this is an interesting resource that may be useful for scientists seeking to understand the non-canonical roles of APOE in shaping gene expression in the hippocampus.

  5. Version published to 10.1101/2022.06.16.496371v1 on bioRxiv