. The innate immunity protein IFITM3 modulates γ-secretase in Alzheimer's disease. Nature. 2020 Oct;586(7831):735-740. Epub 2020 Sep 2 PubMed.


Please login to recommend the paper.


  1. This is really nice work from Drs. Hur, Frost, and a team of collaborators led by Yue-Ming Li. Yue-Ming belongs to a small group of researchers who did pioneering work on the γ-secretase complex (or better yet, complexes, each having very different biological functions). In the current work, they are building further on their observations with various types of γ-secretase inhibitors and modulators to identify a novel, very interesting regulator of γ-secretase activity, IFITM3, which enhances APP processing (and lowers Notch processing) in cell cultures.

    It is nice to see that this protein was also identified in our own work many years ago where we tried to identify regulators and interactors of γ-secretase via a proteome approach (Wakabayashi et al., 2009). They also provide good initial correlative evidence to show human relevance of the role of IFITM3 in Aβ generation and potentially in Alzheimer’s disease.

    The paper is very convincing and springs a new surprise in the study of the γ-secretases. It actually turns the classical view that inflammation is a consequence of amyloid plaque accumulation upside-down, providing a mechanistic support for the hypothesis that inflammation causes increased Aβ generation. Yue-Ming and colleagues discuss the different implications from their observations very nicely in their manuscript, so no need to repeat them here.

    Overall, the story links together, in a very interesting way, two major trains of thought in the Alzheimer’s field—Aβ generation and inflammation. The γ-secretases are pivotal in a vicious cycle that encompasses both and provides different entry points for environmental factors and aging to initiate the Alzheimer’s disease-causing cascade.    


    . Analysis of the gamma-secretase interactome and validation of its association with tetraspanin-enriched microdomains. Nat Cell Biol. 2009 Nov;11(11):1340-6. PubMed.

    View all comments by Bart De Strooper
  2. This work on a γ-secretase-modulating protein IFITM3 published by Yue-Ming Li's group is a sequel to his Nature article on presenilin-labeling active site γ-secretase inhibitors published in 2000. The γ-secretase complex and its modulators have been extensively studied in the past 20 years (Xia, 2019). A number of γ-secretase-interacting proteins have been identified, including CD147, p23/TMP21, γ-secretase-activating protein (GSAP), as well as Hif-1α by the Li group.

    IFITM3, a member of the interferon-induced transmembrane protein family reported in the current Nature article, is different from previously identified γ-secretase interacting proteins. While efforts have been made to explore GSAP and other γ-secretase-interacting proteins as therapeutic targets for Alzheimer's disease (AD) in the past two decades, IFITM3 will likely outperform those previous targets for a number of reasons, mainly anti-inflammation/Aβ dual efficacies, existing potent compounds as candidate IFITM3 inhibitors, and anti-aging potentials.

    First, agents targeting IFITM3 probably carry similar properties of a previous γ-secretase-modulating compound that entered clinical trial, R-flurbiprofen. This nonsteroidal anti-inflammatory drug (NSAID) inhibits interferon γ (IFNγ) and selectively reduces the 42-residue of amyloid-β protein (Aβ42). However, in clinical trials, R-flurbiprofen did not achieve statistical significance on either of its primary endpoints—cognition or activities of daily living. R-flurbiprofen is a weak γ-secretase modulator (GSM) with an IC50 for Aβ reduction of approximately 300 μM. Due to its poor brain penetration, it was unlikely to have lowered brain Aβ42 levels in the clinical studies. In the future, brain-permeable potent IFITM3 inhibitors, once identified, should carry desirable efficacy against neuroinflammation akin to NSAIDs, and modulate γ-secretase activity to reduce Aβ42 production in the central nervous system.

    Second, we have a battery of potent γ-secretase modulators as candidate IFITM3 inhibitors. IFITM3 was identified through its binding to a well-characterized γ-secretase modulator E2012. E2012 is a non-NSAID-derived compound that inhibits both Aβ40 and Aβ42 production, consistent with Li's Nature report that suppressing IFITM3 exhibits a similar reduction of Aβ40 and Aβ42. Another non-NSAID derivative GSM is the former NeuroGenetics’ Compound 4, which directly interacts with presenilin and PEN-2 and inhibits all Aβ peptide production (Aβ40 and 42) in animals under chronic treatment. Other GSMs with IC50 at sub-μM include GSM1, GSM-10h, EVP-A, EVP-B, JNJ-40418677, and more potent BMS-932481 and BMS-986133 (with IC50s for reducing Aβ42 of 6.6 and 3.5 nM, respectively ), as well as BPN-15606 (IC50 of 7 nM and 17 nM to reduce Aβ42 and Aβ40, respectively). Efforts by medicinal chemists could be devoted to modifying existing GSMs, exploring their inhibitory effects on IFITM3, and establishing their structure-activity relationships.

    Third, inhibiting IFITM3 will likely suppress Aβ and other aging-related pathological processes. Since aging upregulates IFITM3 and is the primary risk factor for AD, targeting IFITM3 clearly presents a unique opportunity to potentially modulate cellular processes, especially those related to neuroinflammation and cognitive function. While we continue to explore the mechanisms of action of anti-IFITM3 agents, prevention of memory loss during aging is a key efficacy readout for any clinical trial of AD therapeutics.


    . γ-Secretase and its modulators: Twenty years and beyond. Neurosci Lett. 2019 May 14;701:162-169. Epub 2019 Feb 11 PubMed.

    View all comments by Weiming Xia
  3. In the more than 20 years of study of γ-secretase and its role in AD, many high-profile reports have touted the discovery of modulatory proteins with potential as therapeutic targets, only leading to a disappointing inability to validate and extend those findings. Despite this frustrating history, this new report by Hur et al., that interferon-induced transmembrane protein 3 (IFITM3) is a modulator of γ-secretase activity, is exciting, with important implications for AD pathogenesis, diagnostics, and treatment. Independent validation and extension are still essential, but the thorough and rigorous nature of this study and the large multi-institutional team of investigators bodes well for the future.

    In human cell lines and primary neurons, IFITM3 associated with the γ-secretase complex to increase its proteolytic processing of amyloid-β precursor protein to Aβ40 and Aβ42: Knockdown or knockout of IFITM3 reduced this activity, while overexpression increased it. In contrast, IFITM3 deficiency increased γ-secretase cleavage of Notch1, revealing promising substrate-selective modulatory effects. Photoreactive small molecule probes, derived from a γ-secretase Aβ42-lowering modulator (E2012) as well as an active site-directed inhibitor (L-685,458) covalently labeled the Presenilin-1 (PSEN1) N-terminal fragment (NTF) subunit of γ-secretase as well as IFITM3, suggesting these two proteins are proximal. This was confirmed with an L-685,458 analog containing two photoreactive groups, which was capable of crosslinking PSEN1 NTF to either the PSEN1 CTF subunit or to IFITM3, demonstrating that IFITM3 is in proximity to the protease active site.

    IFITM3 expression increased with age in mice, as did its association with γ-secretase. Knockout of IFITM3 did not affect levels of γ-secretase components but did lower Aβ40 and Aβ42 production by the protease and substantially decreased amyloid plaque deposition in the 5XFAD mouse model. Provocatively, IFITM3 expression was increased in late-onset AD human brain, and γ-secretase production of Aβ40 and Aβ42 was significantly increased in a subset of brain samples with the highest IFITM3 expression. Moreover, treatment with interferons could stimulate IFITM3 expression and increase γ-secretase-dependent production of Aβ40 and Aβ42 in cultured neurons and astrocytes, and IFITM3 expression in human brains positively correlated with expression of a variety of cytokines and with expression of two human viruses (a herpes virus and a hepatitis C virus).

    Taken together, this broad and deep study suggests that IFITM3 may be a safe target for the treatment of AD, lowering Aβ production without inhibiting Notch signaling, the latter a well-known liability of γ-secretase inhibitors. IFITM3 also has potential as a diagnostic marker for late-onset AD, with high levels of expression possibly a strong risk factor. Perhaps most exciting is the potential mechanistic link between AD and neuroinflammation, the underappreciated third pathological feature of AD after amyloid plaques and tau tangles. Inflammatory states that increase cytokine levels in the brain may also increase γ-secretase-mediated production of Aβ.

    Finally, the connection between viral expression and IFITM3 provides further evidence that increased Aβ production may be a response to microbial infection. While Aβ may help in the fight against microbes, the long-term consequences could be increased risk of AD.

    View all comments by Michael Wolfe
  4. The involvement of the immune system in Alzheimer’s disease (AD) is well-established. Immune activity is thought to play several roles in the progression of AD pathology, which may be both beneficial and detrimental, notably: neuroinflammation and microglial clearance of amyloid-β plaques (Blasko and Grubeck-Loebenstein, 2003Shi and Holtzman, 2018). 

    Furthermore, recent work has indicated that the Aβ peptide, which is considered a primary cause of AD, functions as an antimicrobial peptide (AMP), and therefore, plays a role in the innate immune system. In both human and animal models, infection leads to increased amyloid-β production in the brain and Aβ oligomers can form fibrils that entrap pathogens and damage cell membranes (Kumar et al., 2016; Eimer et al., 2018). This suggests that Aβ may be produced as a defense against invading pathogens. How Aβ production is coordinated in such an immune response remains an outstanding question in the field. In addition, the consequence of pathogen-stimulated Aβ production on risk of developing AD has yet to be investigated.

    In this Nature paper, Hur et al. identify a novel mechanism by which an immune response can stimulate Aβ production, providing a link between defense against infection and AD pathology. The authors used the γ-secretase modulator (GSM) E2012-BPyne, which interacts with presenilin 1 (PS1), to search for novel GSM binding proteins, and identified interferon-induced transmembrane protein 3 (IFITM3). IFITM3 is an interferon-stimulated protein known to restrict various viral infections (Bailey et al., 2014). The authors determined that IFITM3 directly binds to γ-secretase near the active site and reduces the production of Aβ40 and Aβ42.

    Notably, when IFITM3-/-mice were crossed to 5xFAD transgenic mice, there was a striking reduction in the number of Aβ plaques in the cortex and hippocampus. Furthermore, IFITM3 expression increases with age, the biggest risk factor for AD. Analyzing postmortem tissue, Hur et al. identified a subpopulation of human patients with late-onset Alzheimer’s disease (LOAD) who had elevated IFITM3 levels. Tissues with high IFITM3 also had significantly increased γ-secretase activity for Aβ40 and Aβ42 production, further highlighting that IFITM3 functions as a γ-secretase modulatory protein.

    This work directly links Aβ production with innate immunity and neuroinflammation and provides a novel mechanism by which Aβ secretion is stimulated in response to an invading pathogen. This not only provides insight into Aβ’s function as an AMP, but also establishes IFITM3 as a potential therapeutic target to reduce Aβ production. The identification of a subpopulation of LOAD patients in whom IFITM3 expression strongly correlates with γ-secretase activity suggests that IFITM3 may be used as a biomarker to stratify AD patients. As LOAD is a multifactorial disease, identification of biomarkers for subpopulations of AD is invaluable to studying underlying mechanisms and developing targeted therapeutics.


    . Role of the immune system in the pathogenesis, prevention and treatment of Alzheimer's disease. Drugs Aging. 2003;20(2):101-13. PubMed.

    . Interplay between innate immunity and Alzheimer disease: APOE and TREM2 in the spotlight. Nat Rev Immunol. 2018 Dec;18(12):759-772. PubMed.

    . Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer's disease. Sci Transl Med. 2016 May 25;8(340):340ra72. PubMed.

    . Alzheimer's Disease-Associated β-Amyloid Is Rapidly Seeded by Herpesviridae to Protect against Brain Infection. Neuron. 2018 Jul 11;99(1):56-63.e3. PubMed.

    . IFITM-Family Proteins: The Cell's First Line of Antiviral Defense. Annu Rev Virol. 2014 Nov 1;1:261-283. PubMed.

    View all comments by Huaxi Xu
  5. This study provides new information about a molecular pathogenic mechanism that linked innate immunity activation to increased Aβ production by an interferon-induced transmembrane protein that upregulates γ-secretase activity. Seeing the genetic evidence for the involvement of the innate immunity in the etiology of late-onset AD, these findings raise the question whether interferons, as interferon-γ, are related to the genetic background of late-onset AD.

    Twin studies have shown that the cytokine production capacity is under strong genetic control (Craen et al., 2005). We have studied the cytokine production capacity in ex vivo stimulated full blood samples from middle-aged offspring with and without a parental history of late-onset AD (van Exel et al., 2009). We found that the production capacity of some proinflammatory proteins, including interferon-γ, was significantly higher in offspring with a parental history of AD upon stimulation of whole blood with LPS. These findings were independent of ApoE4 genotype. This study provides evidence for a proinflammatory genotype for late-onset AD that is characterized by higher production capacity for proinflammatory cytokines including interferon-γ.

    It seems from a neuroinflammatory perspective that AD has two faces (Eikelenboom and van Gool, 2004). In one type such as the familial autosomal-dominant form of AD, the neuroinflammatory response follows the increased production and deposition of Aβ. In a second type, the late-onset and most common form of AD, a neuroinflammatory response precedes the process of increased of Aβ generation.


    . Heritability estimates of innate immunity: an extended twin study. Genes Immun. 2005 Mar;6(2):167-70. PubMed.

    . Vascular factors and markers of inflammation in offspring with a parental history of late-onset Alzheimer disease. Arch Gen Psychiatry. 2009 Nov;66(11):1263-70. PubMed.

    . Neuroinflammatory perspectives on the two faces of Alzheimer's disease. J Neural Transm. 2004 Mar;111(3):281-94. PubMed.

    View all comments by Eric van Exel
  6. We read the study by Hur et al. from Dr. Yue-ming Li’s group with great interest and are impressed with their insightful discoveries on IFITM3.

    Previously, we uncovered a type I interferon response in brains harboring amyloid plaques from both mouse models and human AD (Roy et al., 2020). We showed that certain amyloid plaques are capable of stimulating microglia to mount an intrinsic antiviral immune response, which upregulates a panel of interferon-stimulated genes (ISGs) including IFITM3. Moreover, interferon activates complement cascade and induces synapse loss. Given the ability of IFITM3 to further ramp up Aβ production and plaque formation, AD is seemingly afflicted by a pathogenic feed-forward loop involving APP processing, innate immune activation, and action of distinct ISGs.

    In Aβ models, early and persistent interferon and antiviral immune response was recently confirmed by an analysis of microglial proteomes (Monasor et al., 2020). We have detected IFITM3 as an ISG expressed by plaque-associated microglia, which display features of microglial neurodegenerative phenotype (Krasemann et al., 2017). Not only was IFITM3 upregulated in activated mouse microglia, IFITM3 mRNA levels were also elevated in the brain tissues of dementia patients archived in Mount Sinai Brain Bank (Roy et al., 2020). Further, we detected IFITM3 protein expression in microglia enveloping Aβ plaques in human AD brain.

    Despite being recognized as a protein involved in antiviral protection, the precise role played by microglial IFITM3 in the context of AD is not known at this time. Given the fresh interest in therapeutically targeting IFITM3, it would be pertinent to evaluate the drug candidate’s effect on microglia and other cell types in conjunction with that on γ-secretase from neurons and astrocytes.


    . The TREM2-APOE Pathway Drives the Transcriptional Phenotype of Dysfunctional Microglia in Neurodegenerative Diseases. Immunity. 2017 Sep 19;47(3):566-581.e9. PubMed.

    . Fibrillar Aβ triggers microglial proteome alterations and dysfunction in Alzheimer mouse models. Elife. 2020 Jun 8;9 PubMed.

    . Type I interferon response drives neuroinflammation and synapse loss in Alzheimer disease. J Clin Invest. 2020 Apr 1;130(4):1912-1930. PubMed.

    View all comments by Hui Zheng

Make a Comment

To make a comment you must login or register.

This paper appears in the following:


  1. IFITM3 Forges Link Between Neuroinflammation and Aβ Production