NHE6 depletion corrects ApoE4-mediated synaptic impairments and reduces amyloid plaque load

  1. Theresa Pohlkamp
  2. Xunde Xian
  3. Connie H Wong
  4. Murat S Durakoglugil
  5. Gordon Chandler Werthmann
  6. Takaomi C Saido
  7. Bret M Evers
  8. Charles L White
  9. Jade Connor
  10. Robert E Hammer
  11. Joachim Herz

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This paper is of interest to a broad range of neuroscientists, particularly those interested in ApoE biology and Alzheimer's disease (AD), as it reveals a novel mechanism that counteracts AD-linked amyloid plaque burden and synapse dysfunction in mice. Overall, the methodology is sound, sophisticated, and employs animal models that more closely mimic human diseases, and the results are interesting and compelling. Whilst the mechanistic hypothesis proposed by the authors is consistent with the data, plausible alternative explanations remain.

    (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. Reviewer #2 agreed to share their name with the authors.)

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Apolipoprotein E4 (ApoE4) is the most important and prevalent risk factor for late-onset Alzheimer’s disease (AD). The isoelectric point of ApoE4 matches the pH of the early endosome (EE), causing its delayed dissociation from ApoE receptors and hence impaired endolysosomal trafficking, disruption of synaptic homeostasis, and reduced amyloid clearance. We have shown that enhancing endosomal acidification by inhibiting the EE-specific sodium-hydrogen exchanger 6 (NHE6) restores vesicular trafficking and normalizes synaptic homeostasis. Remarkably and unexpectedly, loss of NHE6 (encoded by the gene Slc9a6 ) in mice effectively suppressed amyloid deposition even in the absence of ApoE4, suggesting that accelerated acidification of EEs caused by the absence of NHE6 occludes the effect of ApoE on amyloid plaque formation. NHE6 suppression or inhibition may thus be a universal, ApoE-independent approach to prevent amyloid buildup in the brain. These findings suggest a novel therapeutic approach for the prevention of AD by which partial NHE6 inhibition reverses the ApoE4-induced endolysosomal trafficking defect and reduces plaque load.

Article activity feed

  1. eLife

    Author Response

    Reviewer #1 (Public Review):

    Weaknesses of this paper include:

    1. The authors conclude that reducing NHE6 clears plaques by activating resident microglia, shifting them from a dormant state to a damage-associated activated state that phagocytoses Abeta plaques. However, there is no data presented to demonstrate this. In a supplemental figure, the authors show there are more Iba1-expressing microglia and GFAP-expressing astrocytes in APP mice and in APP/ApoE4KI mice in which NHE6 has been ablated, but this does not prove that this is the mechanism by which plaques are cleared.

    We apologize for the overstatement. We agree, we have not evaluated whether NHE6 depletion causes a signature of damage-associated microglia. Thus, we have removed this comment from the manuscript (abstract).

    1. The mechanisms underlying the increase in Iba1 and GFAP are not clear. The authors cite a previous paper from another group that demonstrated in their own NHE6 KO mice, there was an increase in GFAP and in activated microglia expressing CD68, which may relate to the cell loss in hippocampus and other brain regions documented in those mice. However, in the current study, the authors indicate that in their NHE6 KO lines, there is no overt cell loss. It is therefore unclear how reductions in NHE6 expression led to microglial/astrocyte activation. This is an important point to work out, since the authors conclude that it is microglial activation that is responsible for the reduction in Abeta plaques.

    We agree that identifying the mechanism how NHE6 depletion causes glial activation is crucial. We and others show that germline NHE6 depletion causes glial activation. Moreover, our current data suggest that genetic deletion of NHE6 in both germline and from adulthood on causes glial activation. Neuronal cell loss is a potential explanation for glial activation. We found cerebellar Purkinje cell loss in both of our NHE6 mutant lines. As stated in the old version of our manuscript we find “Normal Gross Anatomical Brain Structure in Both NHE6-KO and NHE6cKO Mice” (Supplementary Figure S3). To address whether neuronal loss occurs in the hippocampus or cortex, as described for NHE6-KO mice in Xu et al., 2017, we measured neuronal loss in both NHE6 mutant lines. Comparable to Xu et al., in the NHE6-KO line we detect a reduction in total brain area, HC area, cortical thickness and CA1 thickness. By contrast, in our NHE6-cKO;APP-KI;ApoE4-KI mice we do not observe any neuronal loss when compared to NHE6-floxed,APPNL-F,ApoE4-KI littermate controls; however, we detect similar glial activation in the NHE6-cKO;APPNL-F;ApoE4-KI mice as compared to the germline NHE6KO,APPNL-F mice. This suggests that the neuronal loss in the germline NHE6-KO model does not mediate glial activation. Lastly, we have removed the statement that the microglial activation is the reason why we detect Aβ reduction and included a discussion of our new findings.

    1. What might be some of the underlying explanations be for the differences between the published NHE6-KO mice, which has fairly widespread cell loss, and the current KO mice generated in this paper, which did not exhibit noticeable cell loss in brain regions other than the cerebellum?

    Our previous manuscript stated that there are no gross anatomical abnormalities in the NHE6KO mice. However, we appreciate the reviewer’s concerns as it prompted us to analyze neuronal loss in NHE6-KO versus NHE6-cKO mice. Besides Purkinje cell loss in both lines, and as stated above, we do detect cell loss in the hippocampus and cortex in our germline NHE6KO mouse model, but not in the tamoxifen induced NHE6-cKO mice.

    1. There are a number of mechanistic links that have not been worked out, as indicated above. Until these links are identified and characterized, a number of the conclusions drawn by the authors are not yet supported.

    We thank the reviewer for the constructive feedback. We have removed these conclusions.

    Reviewer #3 (Public Review):

    1. The leading hypothesis of this work is that APOE4 impairs synapse function through prolonged association with endosomes, thereby making brain cells vulnerable to AD-related pathological changes. However, the positive effects of NHE6 in a mouse model of Aβ accumulation occurs regardless of APOE4. This suggests that NHE6 may contribute to pathology by mechanisms other than APOE4-mediated retention of endosomal trafficking.

    We agree with the reviewer that NHE6 depletion plays a protective role in AD both by protecting against synaptic impairments in ApoE4-KI mice and Aβ toxicity in an Aβ overproducing mouse model. This may reflect a beneficial effect of endosomal compartment acidification through NHE6 depletion. Our current work and studies by others (Fagan, A.M., et al., Neurobio. of Dis. 2002) show that human Aβ-overproducing ApoE4-KI mice generate plaques at a much later age than mice with wildtype, mouse ApoE, but the mechanism is unknown. Since both of our mouse models, NHE6-KO and NHE6-cKO;ApoE4-KI show a comparable reduction in plaque load, this might be the result of a maximally accelerated early endosomal maturation and cargo transport in the absence of NHE6. We elaborated on this in topic in the discussion of our manuscript.

    1. With the current data, it is not possible to exclude possible nonspecific effects resulting from NHE6 genetic deletion. Additional experiments to measure the endosomal pH would add support to the hypothesis.

    We agree with the reviewer’s concern and addressed this in the discussion accordingly.

    1. The authors attribute reduced amyloid plaque load in NHE6-deficient APP KI mice to increased glial responses, which would promote plaque clearance. This is a very interesting hypothesis, but it is not supported by the experimental data reported in Supplemental Figure 6. Additional experimentation is needed to more thoroughly characterize astrocytic and microglial phenotypes caused by NHE6 genetic depletion in APP KI mice. Functional assays, including cytokine release, nitric oxide production (Griess reaction), and Aβ uptake experiments would be desired to strengthen these conclusions.

    We thank the reviewer for this valuable feedback. In our revised manuscript, we evaluated whether there is a change of microglial Ab content in the NHE6 depletion mouse model. We also quantified the immunoreactivity of Iba1 and GFAP in plaque areas. We found no change between NHE6-KO or control littermate APPNL-F controls when we co-stained with Aβ and Iba1 (microglia) or GFAP (astrocyte). However, when considering a massively reduced Aβ signal in NHE6-KO brains overall, yet the proportion of microglia containing Aβ is comparable to control, this indirectly indicates that NHE6 deficient microglia are more efficient in Aβ uptake and degradation. We agree with the reviewers that future studies will be required to evaluate Aβ uptake in primary microglia derived from NHE6-KO mice to properly conclude that the reduction of Aβ is mediated by enhance glial activation. Thus, we have adjusted our conclusions in the manuscript accordingly.

    1. The authors demonstrate that global or conditional NHE6 deletion causes severe Purkinje cell loss in the mouse cerebellum (Figure 2). Although the authors included representative images of H&E staining indicating no gross histological abnormalities (Supplemental Figure 3), a more detailed investigation is required to assess neuronal survival in the hippocampus and cortex upon NHE6 suppression, given the relevance of these regions to AD pathology. Indeed, previous evidence (Xu et al., eNeuro, 2017) showed that NHE6 depletion leads to significant cortical and hippocampal atrophy, in addition to the cerebellum. Could the reductions in plaque load in NHE6 depleted mice (Figure 5, 6; Supplemental Figure 5) be somehow a reflection of neuronal loss? It is important that the authors discuss this issue in the manuscript.

    We thank the reviewer for this suggestion. We have now measured brain area, hippocampal area, cortical and CA1 thickness. Comparable to Xu et al., we detect a reduction in total brain area, HC area, cortical thickness and CA1 thickness. Contrary, in our NHE6-cKO;APPNFL;ApoE4-KI mice we do not see any neuronal loss; however, we detect similar plaque reduction and glial activation in the NHE6-KO;APPNL-F mice. These findings suggest that the neuronal loss does not mediate the reduction in plaque load or glial activation. We discussed our findings in our manuscript accordingly.

  2. eLife

    Evaluation Summary:

    This paper is of interest to a broad range of neuroscientists, particularly those interested in ApoE biology and Alzheimer's disease (AD), as it reveals a novel mechanism that counteracts AD-linked amyloid plaque burden and synapse dysfunction in mice. Overall, the methodology is sound, sophisticated, and employs animal models that more closely mimic human diseases, and the results are interesting and compelling. Whilst the mechanistic hypothesis proposed by the authors is consistent with the data, plausible alternative explanations remain.

    (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. Reviewer #2 agreed to share their name with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    In this study, the authors confirm and extend their previous work demonstrating that ApoE4, a major risk for for Alzheimer's disease, impairs endocytic recycling of membrane receptors, leading to synaptic dysfunction. Previously, the authors demonstrated in vitro that upon binding to ApoER2 at the plasma membrane and internalization, ApoE4 along with ApoER2 and glutamate receptors become trapped in the early endosome due to the similarity between the isoelectric point of ApoE4 and the pH of early endosomes. Enhancing acidification by inhibiting NHE6, a proton leak channel in the early endosome, restored vesicle recycling and improved synaptic plasticity in AD extract-treated hippocampal slices from ApoE4-KI mice. In the current study, the authors create and use novel NHE6 germline knockout and conditional knockout mouse lines to reduce NHE6 expression and enhance acidification of early endosomes. They confirm their previous findings and also extend their work by crossing NHE6 KO or cKO mice to a knockin, humanized APP mouse that expresses mutant human amyloid precursor protein under control of the endogenous APP promoter, alone or crossed with ApoE4-KI mice. In both cases, reduction of NHE6 resulted in increased Iba1-expressing microglia and GFAP-expressing astrocytes as well as a reduction in plaques. Together these findings highlight the importance of ApoE4's detrimental effect on endosomal recycling in vivo, with consequences for accumulation of AD-related pathology. While the studies presented are well-done and robust, some mechanistic links are missing, which make it difficult to fully support the conclusions drawn.

    Major strengths of this paper include:
    1. The creation of novel germline KO and conditional KO NHE6 mice allow for a number of in vivo investigations that would be difficult to complete otherwise. These mice represent a valuable resource to the field.

    2. Use of the NHE6-KO and cKO mice to confirm the previous findings (that used pharmacological inhibition of NHE6) that enhancing endosomal acidification ameliorates ApoE4-induced deficits in vesicle cycling and synaptic plasticity in vivo. In addition, the finding that NHE6 ablation in APP and APP/ApoE4KI mice robustly reduced plaque accumulation is striking.

    3. The demonstration that BACE-mediated production of APP CTFs is unaltered by NHE6 ablation supports the conclusion that the reduction in plaques is unlikely due to reduction in Abeta generation, but more likely due to clearance of Abeta.

    Weaknesses of this paper include:
    1. The authors conclude that reducing NHE6 clears plaques by activating resident microglia, shifting them from a dormant state to a damage-associated activated state that phagocytoses Abeta plaques. However, there is no data presented to demonstrate this. In a supplemental figure, the authors show there are more Iba1-expressing microglia and GFAP-expressing astroctyes in APP mice and in APP/ApoE4KI mice in which NHE6 has been ablated, but this does not prove that this is the mechanism by which plaques are cleared.

    2. The mechanisms underlying the increase in Iba1 and GFAP are not clear. The authors cite a previous paper from another group that demonstrated in their own NHE6 KO mice, there was an increase in GFAP and in activated microglia expressing CD68, which may relate to the cell loss in hippocampus and other brain regions documented in those mice. However, in the current study, the authors indicate that in their NHE6 KO lines, there is no overt cell loss. It is therefore unclear how reductions in NHE6 expression lead to microglial/astrocyte activation. This is an important point to work out, since the authors conclude that it is microglial activation that is responsible for the reduction in Abeta plaques.

    3. What might be some of the underlying explanations be for the differences between the published NHE6-KO mice, which has fairly widespread cell loss, and the current KO mice generated in this paper, which did not exhibit noticeable cell loss in brain regions other than the cerebellum?

    4. There are a number of mechanistic links that have not been worked out, as indicated above. Until these links are identified and characterized, a number of the conclusions drawn by the authors are not yet supported.

  4. eLife

    Reviewer #2 (Public Review):

    This is a strong and interesting manuscript which examines innovative new hypotheses that have broad relevance to Alzheimer's disease pathophysiology as well as potential new therapeutics. Pohlkamp, Herz and colleagues study the role NHE6 in several orthogonal studies related to production and deposition of amyloid plaques. The use of various different experimental approaches as well as the use of advance mouse genetics is a strength.

    The authors demonstrate several important findings that are robustly supported by the data including: late loss of NHE6 leads to Purkinje cell degeneration; recycling defects in surface receptors relevant to AD and APOE4, namely APOER2 and Glu receptors is improved by deletion of NHE6; NHE6 KO restores Reelin enhancement of LTP inhibited by APOE4; and profound decrease in plaque deposition due to NHE6 mutation.

    The data are well presented in general and compelling. There are many strengths. The PC findings are important in the field of Christianson Syndrome. The reductions in plaque load in the NHE6 null brains are VERY interesting. The mouse genetics, including the conditional mutation -- presentation of a new cKO NHE6 mouse, the humanized Abeta and APOE4 alleles, are truly elegant. Some of the experiments are uniquely supported by the prior findings from the lab relating to Reelin effects on endocytosis and trafficking and effects on LTP, and this is a very important strength. I do not see major weaknesses with the experiments as presented.

    I believe that this work will have broad interest and this work and prior work of the Herz lab in the area of NHE6 as it may relate to therapeutics in AD is developing into a unique niche with potential strong impact in AD therapeutics.

  5. eLife

    Reviewer #3 (Public Review):

    In this manuscript, Pohlkamp, Xian, Wang et al. investigated the role of the sodium-hydrogen exchanger NHE6 in synaptic plasticity and Aβ plaque load in a mouse model of Alzheimer's disease (AD) in the presence or absence of Apolipoprotein E4 (APOE4), a major genetic risk factor for sporadic AD. They initially report that NHE6 deletion causes cerebellar neurodegeneration. They find that genetic deletion of NHE6 alleviates impairments in reelin-induced synaptic plasticity in mice expressing human APOE4. The main novelty of this study is that NHE6 suppression significantly reduced amyloid plaque load in a mouse model of AD expressing humanized Aβ, either in the presence or absence of ApoE4. This is interesting, as it potentially opens new roads to understand and control amyloid pathology in the AD brain. Although the data are intriguing and relevant for the community, some issues need to be addressed so that conclusions are justified by data:

    1. The leading hypothesis of this work is that APOE4 impairs synapse function through prolonged association with endosomes, thereby making brain cells vulnerable to AD-related pathological changes. However, the positive effects of NHE6 in a mouse model of Aβ accumulation occurs regardless of APOE4. This suggests that NHE6 may contribute to pathology by mechanisms other than APOE4-mediated retention of endosomal trafficking.

    2. With the current data, it is not possible to exclude possible nonspecific effects resulting from NHE6 genetic deletion. Additional experiments to measure the endosomal pH would add support to the hypothesis.

    3. The authors attribute reduced amyloid plaque load in NHE6-deficient APP KI mice to increased glial responses, which would promote plaque clearance. This is a very interesting hypothesis, but it is not supported by the experimental data reported in Supplemental Figure 6. Additional experimentation is needed to more thoroughly characterize astrocytic and microglial phenotypes caused by NHE6 genetic depletion in APP KI mice. Functional assays, including cytokine release, nitric oxide production (Griess reaction), and Aβ uptake experiments would be desired to strengthen these conclusions.

    4. The authors demonstrate that global or conditional NHE6 deletion causes severe Purkinje cell loss in the mouse cerebellum (Figure 2). Although the authors included representative images of H&E staining indicating no gross histological abnormalities (Supplemental Figure 3), a more detailed investigation is required to assess neuronal survival in the hippocampus and cortex upon NHE6 suppression, given the relevance of these regions to AD pathology. Indeed, previous evidence (Xu et al., eNeuro, 2017) showed that NHE6 depletion leads to significant cortical and hippocampal atrophy, in addition to the cerebellum. Could the reductions in plaque load in NHE6 depleted mice (Figure 5, 6; Supplemental Figure 5) be somehow a reflection of neuronal loss? It is important that the authors discuss this issue in the manuscript.

  6. Version published to 10.7554/elife.72034 on eLife
  7. Version published to 10.1101/2021.03.22.436385v2 on bioRxiv
  8. Version published to 10.1101/2021.03.22.436385v1 on bioRxiv