Although researchers know that most of the genetic risk for late-onset Alzheimer’s disease resides in microglia, exactly how these genes work their mischief remains something of a black box. In the November 29 Nature Communications, researchers led by Tracy Young-Pearse at Brigham and Women’s Hospital, Boston, cracked open the box to shed light on one such microglial gene, inositol polyphosphate-5-phosphatase. INPP5D makes the protein SHIP1, a membrane-associated phosphatase that dampens cell-signaling pathways.

  • Inhibiting microglial AD gene INPP5D activates the inflammasome.
  • In AD brain, INPP5D expression is up, but active protein is down.
  • This triggers the inflammasome, worsening inflammation.

In human microglia generated from iPSCs, dunking SHIP1 drove up microglial activation. In particular, it triggered the NLRP3 inflammasome and spurred the release of pro-inflammatory cytokines.

In postmortem AD brain, the situation was complicated. Though more SHIP1 was made, less of it was functional. This meant that overall, SHIP1 activity was down, and again, more inflammasomes formed. Together, the findings suggest that INPP5D normally taps the brakes on inflammation, keeping microglia in check.

Michael Heneka at the University of Luxembourg, Germany, praised the paper. “Elucidating the regulatory role of INPPD5 on NLRP3 inflammasome activation and its possible mechanism of action is quite intriguing, opening new avenues for novel therapeutic approaches for AD and other pathologies in which NLRP3 inflammasome activation is involved,” Heneka and postdoc Bora Tastan wrote to Alzforum (full comment below).

More? Or Less? In AD brain, microglia (green) contacting amyloid plaques (pink) express INPP5D (red); alas, much of that protein is inactive. Nuclei are blue. [Courtesy of Chou et al., Nature Communications.]

INPP5D was first identified as an AD gene a decade ago, but few studies have focused on it (Jul 2013 news). Researchers led by Adrian Oblak and Kwangsik Nho at Indiana University School of Medicine, Indianapolis, showed that INPP5D expression was up in AD brain and amyloidosis mice, particularly around plaques (Tsai et al., 2021). The same group found that knocking out one copy of INPP5D in mice reduced plaques and preserved memory, implying that suppressing it might be therapeutic (Lin et al., 2023). 

However, in the hands of Michelle Ehrlich, Icahn School of Medicine at Mount Sinai, New York, the opposite happened. INPP5D knockdown in a different amyloidosis model worsened plaques (Castranio et al., 2022). 

What’s going on? To investigate, first author Vicky Chou generated microglia from human iPSCs, dubbed iMGs, using a protocol from Mathew Blurton-Jones’ group (Abud et al., 2017; McQuade et al., 2018). When she inactivated SHIP1 in these microglia with a drug, more inflammasomes formed, and the cells secreted more of the pro-inflammatory cytokines IL-1β and IL-18.

What about AD brain? The authors immunostained temporal cortex samples from eight AD and eight control brains for SHIP1. Confirming previous work, they found higher levels of SHIP1 in AD than control. SHIP1 was concentrated in microglia around plaques. However, when they isolated SHIP1 and analyzed it by mass spectrometry, they detected less full-length protein in AD than control brain. Instead, much of it was truncated, missing the phosphatase domain. This creates an inactive version of the protein (Rohrschneider et al., 2000). In keeping with this, inflammasomes and IL-18 were up in AD brain, matching findings from cell culture.

A previous study had also reported the presence of truncated SHIP1 in AD brain (Zajac et al., 2023). How does this shortened version get made? AD risk SNPs occur in introns of INPP5D and affect splicing. Potentially, they could promote formation of the truncated protein, though this remains to be proven, Young-Pearse told Alzforum.

It is also possible that other AD risk genes influence INPP5D. Suppressing SHIP1 in iMGs affected expression of several of these, including CD33, TMEM106B, SORL1, and GRN. The functional consequences are unknown.

Meanwhile, although TREM2 levels stayed the same, SHIP1 is known to dampen TREM2 signaling, suggesting that its loss could amplify TREM2 activity. TREM2 promotes microglial survival and phagocytosis, hence suppressing SHIP1 could promote plaque clearance. “Further research efforts should concentrate on investigating the link between INPP5D and other GWAS hits,” Heneka and Tastan suggested.

SHIP1’s opposing effects on inflammation and phagocytosis might help explain previous conflicting findings, where knocking down INPP5D worsened plaque in some mouse models and improved it in others. “Our data suggest there’s a delicate balance of SHIP1 activity needed to keep microglia clearing Aβ without making them dysfunctional,” Young-Pearse told Alzforum.

For her part, Young-Pearse is burrowing deeper into the effects of genetic background. She has generated iMG lines from 100 ROSMAP participants of different racial and ethnic heritages. Young-Pearse will look for factors that could predict who would benefit most from modulating SHIP1.

Both agonists and antagonists of the protein have been developed within the MODEL-AD and TREAT-AD programs.—Madolyn Bowman Rogers

Comments

  1. This study provides new insights into the role of INPP5D in AD pathology and its potential mechanism of action within the axis of microglial NLRP3 inflammasome activation. The study used multiomics approaches to profile transcriptomic and proteomic landscapes in which acute and chronic INPPD5 inhibition has been achieved using pharmacological and genome-editing tools. Although the association of INPP5D with AD has been presented to the literature by genome-wide association studies (Deming et al., 2017) and original research articles (Sierksma et al., 2020; Tsai et al., 2021), how INNPP5D functions in AD brain and impacts molecular pathways contributing to AD pathology has been largely unclear.

    Chou and colleagues have tackled this question and addressed the controversies surrounding INPP5D's role in the AD brain. Their findings demonstrated heterogeneity in INNDP5 expression as phosphatase domain-lacking truncated isoforms in AD brain overtook the functional INPP5D protein, causing alteration in innate immune processes, including NLRP3 inflammasome activation. The implication of microglial NLRP3 inflammasome activation has been shown in AD. Microglial NLRP3 inflammasome activation enhances Aβ pathology and subsequent AD progression (Heneka et al., 2013) and drives tau pathology (Ising et al., 2019). The deficiency of NLRP3 protein or other components of NLRP3 inflammasome complex, namely CASP1 or ASC, protects against these pathologies. Thus, elucidating the regulatory role of INPPD5 on NLRP3 inflammasome activation and its possible mechanism of action is quite intriguing, opening new avenues for novel therapeutic approaches for AD and other pathologies in which NLRP3 inflammasome activation is involved. Eventually, this study by Chou and colleagues will contribute to an extensive array of existing literature on the regulatory network of NLRP3 inflammasome activation.

    Last but not least, this study once again reminds us of the necessity of investigating risk factors identified by GWAS. We might know the risk factors enriched in disease pathology, but we might not have a complete comprehension of the implications of those risk factors in underlying pathologies. In the present study, downregulation of INPPD5 expression resulted in altered protein expression enriched in lysosomal proteins, including TMEM106B, GLA, LAMP1, and altered autophagic capacity possessing a regulatory role in NLRP3 inflammasome activation (Harris et al., 2011). 

    Additionally, not only in AD (Gao et al., 2018) but also in other neurodegenerative diseases (Wainberg et al., 2023), risk factors are enriched in endo-lysosomal networks. Further research efforts should concentrate on investigating the link between INPP5D and other GWAS hits whose expressions are altered with functional INPP5D deficiency and the autonomous/non-autonomous effects of these factors on microglia, neurons, and astrocytes. 

    References:

    . Genome-wide association study identifies four novel loci associated with Alzheimer's endophenotypes and disease modifiers. Acta Neuropathol. 2017 May;133(5):839-856. Epub 2017 Feb 28 PubMed.

    . Genetic variation within endolysosomal system is associated with late-onset Alzheimer's disease. Brain. 2018 Sep 1;141(9):2711-2720. PubMed.

    . Autophagy controls IL-1beta secretion by targeting pro-IL-1beta for degradation. J Biol Chem. 2011 Mar 18;286(11):9587-97. Epub 2011 Jan 12 PubMed.

    . NLRP3 is activated in Alzheimer's disease and contributes to pathology in APP/PS1 mice. Nature. 2013 Jan 31;493(7434):674-8. Epub 2012 Dec 19 PubMed.

    . NLRP3 inflammasome activation drives tau pathology. Nature. 2019 Nov;575(7784):669-673. Epub 2019 Nov 20 PubMed.

    . Novel Alzheimer risk genes determine the microglia response to amyloid-β but not to TAU pathology. EMBO Mol Med. 2020 Mar 6;12(3):e10606. Epub 2020 Jan 17 PubMed.

    . INPP5D expression is associated with risk for Alzheimer's disease and induced by plaque-associated microglia. Neurobiol Dis. 2021 Jun;153:105303. Epub 2021 Feb 22 PubMed.

    . Shared genetic risk loci between Alzheimer's disease and related dementias, Parkinson's disease, and amyotrophic lateral sclerosis. Alzheimers Res Ther. 2023 Jun 16;15(1):113. PubMed.

  2. This study identifies a new player in the regulation of neuroinflammation in AD. In a GWAS, mutations of the myeloid-specific gene INPP5D have been found to be associated with the risk of developing late-onset AD. INPP5D encodes the phosphatase SHIP1, which has been shown previously to be involved in phagocytosis and in TREM2 signaling (reviewed by Terzioglu and Young-Pearse, 2023).

    In postmortem brain tissue of AD patients, Chou et al. describe expression of full-length SHIP1 to be decreased. To mimic the situation in the AD brain, they downregulated INPP5D in iPSC-derived microglia by treatment with a selective inhibitor of INPP5D or by introduction of a heterozygous loss-of-function mutation. Both interventions were, without any additional stimulus, sufficient to increase the release of IL-1β and IL-18. These two cytokines are processed by the NLRP3 inflammasome. Correspondingly, NLRP3 protein, cleaved Caspase-1, and ASC-stained inflammasomes were also increased in iPSC microglia with reduced INPP5D activity. From a translational viewpoint, reduced INPP5D levels correlated with a higher percentage of microglia with ASC foci in AD brain tissue. Thus, this exciting study indicates that INPP5D can influence inflammasome action and can thus negatively regulate IL-1β and IL-18 processing and production of the mature cytokines.

    These data raise several questions. One relates to the molecular mechanisms relevant for the observed effects. The authors propose that diminished INPP5D levels may mediate alterations of CLEC7A and/or PLA2G7 signaling, which could then affect the inflammasome. Interestingly, the authors found that reduced INPP5D activity lowered the autophagic flux.

    We and others were able to show that NLRP3 (Houtman et al., 2019) and the inflammasome itself (Shi et al., 2012) are targets for autophagic degradation. It would thus be interesting to assess the effect of INPP5D expression levels and activity on microglial autophagy in general and on NLRP3/inflammation degradation specifically.

    Another interesting aspect is the potential interventional-translational value of these data. Lack of the NLRP3, or of the inflammasome, has been shown to reduce neuroinflammation and to ameliorate AD disease pathology in mouse models (Heneka et al., 2013; Ising et al., 2019). Various NLRP3 inhibitors are already being assessed in clinical trials for the treatment of inflammation in different disease contexts (recently reviewed by Li et al., 2023). Thus, despite the fact that the efficacy of anti-inflammatory treatments in clinical AD trials is still being assessed (i.e., not yet proven to function), an inflammasome-targeted approach could be a promising additional tool.

    As is the case for many other interventional AD approaches, the right time point of administration will be a matter of debate. Most likely, treatment should take place prior to cognitive alterations, which, in turn, stresses the need for early biomarkers in AD. Along that line, it would be required and important to track changes of INPP5D and the NLRP3-inflammasome axis in the course of the disease.

    References:

    . NLRP3 is activated in Alzheimer's disease and contributes to pathology in APP/PS1 mice. Nature. 2013 Jan 31;493(7434):674-8. Epub 2012 Dec 19 PubMed.

    . Beclin1-driven autophagy modulates the inflammatory response of microglia via NLRP3. EMBO J. 2019 Feb 15;38(4) Epub 2019 Jan 7 PubMed.

    . NLRP3 inflammasome activation drives tau pathology. Nature. 2019 Nov;575(7784):669-673. Epub 2019 Nov 20 PubMed.

    . Recent Progress and Prospects of Small Molecules for NLRP3 Inflammasome Inhibition. J Med Chem. 2023 Nov 9;66(21):14447-14473. Epub 2023 Oct 25 PubMed.

    . Activation of autophagy by inflammatory signals limits IL-1β production by targeting ubiquitinated inflammasomes for destruction. Nat Immunol. 2012 Jan 29;13(3):255-63. PubMed.

    . Microglial function, INPP5D/SHIP1 signaling, and NLRP3 inflammasome activation: implications for Alzheimer's disease. Mol Neurodegener. 2023 Nov 29;18(1):89. PubMed.

  3. Chou and colleagues have added important insights into the role of INPP5D in Alzheimer's disease. Their study shows that in AD brains, despite increased INPP5D expression, the predominant protein is in a truncated, potentially dysfunctional form. This finding, together with the observation that reduced INPP5D activity in human iPSC-derived microglia activates neurotoxic inflammasomes, provides a new perspective on how INPP5D may exacerbate neuroinflammation in AD.

    However, this contrasts with the results of several studies using Inpp5d knockout mice: These studies, including ours (Iguchi et al., 2023), demonstrated neuroprotective effects of Inpp5d deficiency in amyloid mouse models (Lin et al., 2023; Samuels et al., 2023; Yin et al., 2023), with one exception showing increased amyloid pathology (Castranio et al., 2022). 

    This discrepancy raises two important questions: First, what explains the differences between the neuroprotective effects observed in the mouse studies and the adverse effects reported by Chou et al.? This could be due to different research foci—Chou et al. investigated cellular phenotypes by INPP5D alterations, whereas the knockout studies, like ours, assessed broader neurological outcomes. Species differences may also play a role.

    Second, what are the nature and functional significances of the truncated INPP5D isoform? While increased INPP5D expression from an alternative internal transcription start site has been proposed (Zajac et al., 2023), similar isoforms have been identified by different groups with no consensus on their origin (Antignano et al., 2010). It is also noteworthy that in addition to its role as a lipid phosphatase, INPP5D exerts a complex biological function via scaffolding activity (Pauls and Marshall, 2017). Given the multifunctionality of INPP5D, understanding the consequences of its upregulation is crucial to determine its precise role in AD pathogenesis.

    In conclusion, the study by Chou et al., while illuminating, underscores the need to integrate these diverse findings to fully understand the complex role of INPP5D in AD pathology.

    References:

    . INPP5D modulates TREM2 loss-of-function phenotypes in a β-amyloidosis mouse model. iScience. 2023 Apr 21;26(4):106375. Epub 2023 Mar 13 PubMed.

    . INPP5D deficiency attenuates amyloid pathology in a mouse model of Alzheimer's disease. Alzheimers Dement. 2023 Jun;19(6):2528-2537. Epub 2022 Dec 16 PubMed.

    . The Alzheimer's disease risk factor INPP5D restricts neuroprotective microglial responses in amyloid beta-mediated pathology. Alzheimers Dement. 2023 Apr 15; PubMed.

    . APOE4 impairs the microglial response in Alzheimer's disease by inducing TGFβ-mediated checkpoints. Nat Immunol. 2023 Nov;24(11):1839-1853. Epub 2023 Sep 25 PubMed.

    . Microglial INPP5D limits plaque formation and glial reactivity in the PSAPP mouse model of Alzheimer's disease. Alzheimers Dement. 2022 Nov 30; PubMed.

    . Expression of INPP5D Isoforms in Human Brain: Impact of Alzheimer's Disease Neuropathology and Genetics. Genes (Basel). 2023 Mar 21;14(3) PubMed.

    . Chapter 134 - The Src Homology 2 Containing Inositol 5' Phosphatases. Handbook of Cell Signaling, 2010 Handbook of Cell Signaling (Second Edition)

    . Regulation of immune cell signaling by SHIP1: A phosphatase, scaffold protein, and potential therapeutic target. Eur J Immunol. 2017 Jun;47(6):932-945. Epub 2017 May 26 PubMed.

  4. This publication carefully identifies and validates downstream effects of knockdown or pharmacologic inhibition of INPP5D in microglia induced from human iPSCs (iMGs). It follows on the heels of multiple studies examining effects in constitutive and conditional Inpp5d knockdowns crossed with various mouse models of genetic Alzheimer’s disease, e.g., 5xFAD and APP/PS1. Though constitutive knockout was ultimately lethal, comparisons between constitutive heterozygotes and conditional Inpp5d deletions gave markedly different outcomes, with worsening of some aspects of the phenotype and improvement in others.

    In this iMG study, major findings included activation of the NLRP3 inflammasome, upregulation of autophagy, and increased secretion of IL-1β and IL-18. Notably, the “genetic” downregulation in the iMGs is “constitutive.” Comparing data from mouse models in vivo to corresponding results from iMGs in vitro, there were some opposite effects on specific cytokine production in individual models, and neither autophagy pathways nor NLRP3 inflammasome pathways were highlighted as enriched pathways or relevant genes in vivo.

    These comparisons highlight the importance of studies in human cells, the results of which are not recapitulated even when FAD mutant forms of hAPP are expressed in mice. However, the apparent species differences may also be caused by the comparison between iMGs and the whole brain, with its associated cell-cell interactions, cerebrovascular milieu, and many other factors not present in either iMGs or in organoids with both iMGs and neurons. Despite not being able to distinguish between these possibilities based on current studies, the Young-Pearse study encourages the specific analysis of the NLRP3 inflammasome and other identified pathways in the mouse models. In addition, studies should be continued to analyze the effects of the actual INPP5D SNPs identified by GWAS.

    Finally, a very important facet of this study is that it further highlights the complexity of the interactions between genes and systems known to impact the AD phenotype, both within microglia and in a non-cell-autonomous fashion. The NLRP3 inflammasome has previously been identified as impacting both amyloid and tau pathology. One is left to wonder whether upregulation of INPP5D activity would inhibit the inflammasome. INPP5D has also been shown to downregulate TLR4 and TREM2/TYROBP(DAP12) signaling, either of which potentially could have detrimental effects on the AD phenotype. 

    Studies of both mouse models and iMGs are required to predict not only what possible effects of INPP5D deficiency and/or mutations might occur in the brains of humans but also the timing, disease-stage specificity, and/or cell-type specificity of therapeutic interventions that may be helpful.

  5. We are excited to see the work of Chou et al. suggesting a link between INPP5D and inflammasomes (Chou et al., 2023). Their findings dovetail nicely with our recent paper, which found the INPP5D in human brain has multiple transcription start sites and two variably spliced exons, which overall generate multiple SHIP1 isoforms (Zajac et al., 2023). These isoforms encode full-length SHIP1 (SH2 domain/phosphatase domain/COOH domain), as well as isoforms encoding only the SH2 domain, only the phosphatase and COOH domains, and only the COOH domain.

    As indicated by Chou et al., we found that brains with high AD neuropathology had increased expression of INPP5D exons encoding the SH2 domain and the COOH domain. We did not detect a directional SNP effect on INPP5D isoform expression. However, we did find that the AD risk factor rs35349669 was associated with INPP5D expression by comparing allelic expression in people heterozygous for an exon 2 SNP as a function of rs35349669 status. 

    We couldn't ascribe a particular rs35349669 allele as increasing or decreasing INPP5D expression because rs35349669 is too remote from the reporter SNP for us to determine which rs35349669 allele was co-inherited with which allele of the exon 2 SNP. 

    Lastly, we noticed that the Western blots in Chou et al. used antibodies against epitopes encoded by exon 25 or by exon 26. The antibody against the exon 25-encoded epitope labeled an extra protein that was not labeled by the antibody against the exon-26 encoded epitope.  Within our paper, we found that exon 26 was subject to alternative splicing, which could be the cause of this interesting Western blot pattern.

    Overall, our study and that of Chou et al. suggest that differential INPP5D isoform expression may impact functional SHIP1 protein. We look forward to further research examining this relationship and how the presence/absence of functional domains impacts inflammasome formation and other molecular pathways.

    References:

    . INPP5D regulates inflammasome activation in human microglia. Nat Commun. 2023 Nov 29;14(1):7552. PubMed.

    . Expression of INPP5D Isoforms in Human Brain: Impact of Alzheimer's Disease Neuropathology and Genetics. Genes (Basel). 2023 Mar 21;14(3) PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Pooled GWAS Reveals New Alzheimer’s Genes and Pathways

Paper Citations

  1. . INPP5D expression is associated with risk for Alzheimer's disease and induced by plaque-associated microglia. Neurobiol Dis. 2021 Jun;153:105303. Epub 2021 Feb 22 PubMed.
  2. . INPP5D deficiency attenuates amyloid pathology in a mouse model of Alzheimer's disease. Alzheimers Dement. 2023 Jun;19(6):2528-2537. Epub 2022 Dec 16 PubMed.
  3. . Microglial INPP5D limits plaque formation and glial reactivity in the PSAPP mouse model of Alzheimer's disease. Alzheimers Dement. 2022 Nov 30; PubMed.
  4. . iPSC-Derived Human Microglia-like Cells to Study Neurological Diseases. Neuron. 2017 Apr 19;94(2):278-293.e9. PubMed.
  5. . Development and validation of a simplified method to generate human microglia from pluripotent stem cells. Mol Neurodegener. 2018 Dec 22;13(1):67. PubMed.
  6. . Structure, function, and biology of SHIP proteins. Genes Dev. 2000 Mar 1;14(5):505-20. PubMed.
  7. . Expression of INPP5D Isoforms in Human Brain: Impact of Alzheimer's Disease Neuropathology and Genetics. Genes (Basel). 2023 Mar 21;14(3) PubMed.

External Citations

  1. MODEL-AD
  2. TREAT-AD

Further Reading

Primary Papers

  1. . INPP5D regulates inflammasome activation in human microglia. Nat Commun. 2023 Nov 29;14(1):7552. PubMed.