Hou X, Zhang X, Zou H, Guan M, Fu C, Wang W, Zhang ZR, Geng Y, Chen Y. Differential and substrate-specific inhibition of γ-secretase by the C-terminal region of ApoE2, ApoE3, and ApoE4. Neuron. 2023 Jun 21;111(12):1898-1913.e5. Epub 2023 Apr 10 PubMed.
Recommends
Please login to recommend the paper.
Comments
K.U.Leuven and V.I.B.
This study follows up on the intriguing observation that neuronal ApoE modulates Aβ production, originally reported by Wang et al., 2018.
Here, Xianglong Hou et al. show that neuronally expressed ApoE-ε2 and ε3—but not ε4—lead to the intracellular accumulation of APP C-terminal fragments (APPCTF) while reducing secreted Aβ levels, implying the inhibition of APP processing by ApoE in an isoform-dependent manner. Their findings show that ApoE- ε2 is apparently more active than the ε3 isoform, and that the ApoE ε2 C-terminal (ApoE CT) fragment is sufficient to exert this inhibitory effect on APP processing.
Notably, the inhibitory effects triggered by the ApoE CT were more profound than those exerted by the full-length ApoE2, suggesting that the N-terminal region of ApoE modulates its inhibitory potency. Of note, the interaction between the N- and C-terminal domains in ApoE is unique across the different isoforms; it is thus possible that this "domain interaction" provides the structural basis for the observed differential effects.
Furthermore, the data point to the intracellular ApoE pool as the relevant fraction facilitating the inhibition, rather than the secreted ApoE, and provide evidence in support of a selective mechanism involving ApoE. These findings might suggest differential inhibitory effects of ApoE on the amyloidogenic versus non-amyloidogenic processing of APP, since the amyloidogenic processing occurs mostly intracellularly.
Data showing that purified ApoE2 inhibits the production of Aβ40 in a dose-dependent manner in a well-controlled γ-secretase activity assay support that a direct interaction with either the enzyme and/or the APP substrate underlies the inhibitory effect. In this regard, an interaction between ApoE and APP/Aβ could prevent γ-secretase mediated processing, explaining the apparent substrate selectivity of ApoE, and potentially its localization in amyloid plaques.
References:
Wang C, Najm R, Xu Q, Jeong DE, Walker D, Balestra ME, Yoon SY, Yuan H, Li G, Miller ZA, Miller BL, Malloy MJ, Huang Y. Gain of toxic apolipoprotein E4 effects in human iPSC-derived neurons is ameliorated by a small-molecule structure corrector. Nat Med. 2018 May;24(5):647-657. Epub 2018 Apr 9 PubMed.
University of Kansas
This new study by Hou and colleagues raises the surprising and provocative possibility that Apolipoprotein E (ApoE) interacts directly with γ-secretase and affects its proteolytic activity to selectively alter Aβ peptide production. The protective ApoE2 isoform shows the highest potency toward γ-secretase inhibition of Aβ production, followed by ApoE3, with the AD risk-associated ApoE4 showing little or no effect. The inhibitory activity appears to reside in the C-terminal region common to all three ApoE isoforms, with the variable N-terminal region differentially modulating the effect on γ-secretase cleavage of APP substrate. Cleavage of other substrates (Notch and APLP1) was not affected by the ApoE C-terminal region. The observed effects of ApoE seem to occur in a cell-autonomous manner within neurons.
The study, however, suffers from two major weaknesses. First, a majority of the experiments involved transient overexpression of ApoE, raising concerns about nonphysiological levels and subcellular locations of the protein. Only one figure shows effects of siRNA knockdown of endogenous ApoE, and these experiments mostly reveal lowering of an unusual 14 kDa C-terminal fragment (CTF) of endogenous APP, which is the result of neither α-, β-, nor η-secretase cleavage (which give 10, 12 and 30 kDa CTF APP γ-secretase substrates, respectively). Moreover, in the ApoE overexpression experiments, only α-secretase-generated APP CTF is increased, as seen by western blotting. It is perplexing that β-secretase-generated CTF is not also increased, since, by ELISA, Aβ production is shown to be decreased. This disconnect is not mentioned or explained.
The second major weakness is that nowhere are the experiments put into the context of the primary function of ApoE as a lipid carrier protein. Overexpression of ApoE is likely to lead to a substantial portion of under-lipidated protein, which is not observed physiologically. Moreover, expression of the ApoE C-terminal region likely produces a form of the protein that is not properly or fully lipidated either. Therefore, the results of these experiments may be artifacts of ApoE protein not appropriately loaded with lipids. It would be important to conduct similar experiments with primary neuronal cultures in mice deficient in endogenous ApoE, as well as with knock-in of the human ApoE isoforms, to confirm the findings of this report.
Chinese Academy of Sciences
I would like to thank Drs. Chavez-Gutierrez and Wolfe for their comments about our recent paper, and respond with these points:
1. We really appreciate Dr. Chavez-Gutierrez's comments. The Wang et al., 2018, paper from Dr. Huang's lab was indeed encouraging for us during the discovery. However, our study is not a follow-up study of it. My lab had already discovered the main phenomena about two years before the publication of Wang et al. In addition, Wang et al. only reported the phenomenon of differential Aβ production from human iPSC-induced neurons carrying different ApoE isoforms without further mechanistic explanation. In contrast, we found that ApoE2 could directly inhibit APP γ-cleavage, whereas this inhibitory activity is largely lost in ApoE4. This means ApoE4 is more a loss-of-function allele in regulating Aβ production, not a gain-of-function variant as implied by the Wang et al. paper. However, this paper did let us know that the phenomenon we observed could happen with endogenous levels of APP, γ-secretase, and ApoE in human neurons. Therefore it was very encouraging.
2. About the unusual 14 kDa C-terminal fragment (CTF) of endogenous APP mentioned by Dr. Wolfe, it is interesting that in cultured rodent neurons, this 14 kD fragment is the most robust APP CTF regulated by γ-secretase. It is so strong that it actually covered all the other CTF in western blot. However, it is also dependent on γ-secretase, therefore it is rational to use it as a biomarker to show whether the γ-cleavage of APP fragment is altered by endogenous ApoE in this experiment. In addition, we also observed that Aβ production is elevated by ApoE knockdown in these neurons, supporting the γ-cleavage of β-CTF was also enhanced. Together with the Wang et al. paper mentioned above, we think it is clear that endogenous ApoE is regulating γ-cleavage of endogenous APP in both rodent neurons and human iPSC-induced neurons.
3. About the β-CTF question Dr. Wolfe mentioned, indeed we found the β-CTF levels were relatively steady. They were not easily altered by other manipulations unless β-secretase was directly manipulated. We do not know the exact reason. I will guess this is related to the limitations of steady-state measurement of β-CTF by western blot, which is determined by both production and elimination. In comparison, Aβ was measured as a final product, which is only determined by the production in this kind of system. Therefore the measurement of Aβ is more accurate than β-CTF in this case.
4. About the suggestion of using ApoE KO and human ApoE KI mice to address the endogenous function of ApoE, the problem is that under normal conditions, mouse ApoE in WT mice and human ApoE in KI mice are only expressed in glial cells, not in neurons. That makes these models unsuitable to address the endogenous function of ApoE in neurons to regulate APP γ-cleavage (which happens in neurons almost exclusively) with cell-autonomous manners as mentioned in our paper. Indeed many labs (including our data) have shown ApoE protein and mRNA are both present in neurons at high levels in human. We also found ApoE is highly expressed in monkey cortical neurons. Therefore it is possible that ApoE has cell-autonomous functions in neurons, and rodents are just not suitable models to study this question unfortunately. Indeed we think this huge species difference in ApoE expression pattern between rodent models and humans is a big hurdle affecting ApoE research, which should be addressed in some way in the future.
Make a Comment
To make a comment you must login or register.