Microglia normally gobble up and break down Aβ. However, in Alzheimer’s disease, an altered inflammatory state causes them to stop clearing the aggregated peptide. How does this happen, and can it be stopped? Junying Yuan of Harvard Medical School, Boston, blames the microglial enzyme RIPK1, and believes that blocking it may help return microglia to their normal state. According to a paper Yuan and colleagues published September 13 in the Proceedings of the National Academy of Sciences, the kinase appears to set off transcriptional changes that cripple the microglial lysosome system. The cells start producing new gene products, some characteristic of the recently identified disease-associated microglia (DAM) surrounding plaques in AD model mice (Jun 2017 news). Genetically deleting or pharmacologically inhibiting RIPK1 both sped up Aβ clearance and improved memory in an AD mouse model. The findings lay the groundwork for a new treatment for AD, and, since RIPK1 has been implicated in amyotrophic lateral sclerosis (ALS) and multiple sclerosis (MS), for those diseases as well, the authors believe.

  • Microglia produce RIPK1 kinase in response to inflammatory signals.
  • RIPK1 inhibits endosomes, lysosomes, Aβ degradation.
  • Blocking RIPK1 reduced Aβ, improved memory in AD mouse model.

“It’s the first paper that shows blocking RIPK1 alleviates the inflammatory response, reduces plaque, and improves behavior in AD mice,” Yuan said. “It points out directly the beneficial effects of inhibiting RIPK1 for the treatment of multiple diseases characterized by inflammation and cell death.”

“A crucial role of RIPK1 in the switch between phagocytosis and inflammatory capacity in microglia is novel and exciting,” said Marco Colonna, Washington University in St. Louis. “The authors convincingly show that inhibition of RIPK1 increases the turnover of Aβ1–42 and reduces production of proinflammatory cytokines by microglia.”

“Microglial lysosome biology is poised to become the next hot topic in Alzheimer’s research,” said Terrence Town, University of Southern California, Los Angeles, who was not involved in the study. “A lot of recent data are pointing to failure of the lysosome in microglia and other innate immune cells as the problem in AD, and rebalancing that as the way forward.”

You’re Surrounded! Microglia (IBA1/green) surrounding Aβ plaques (Thioflavin T/blue) from seven-month-old APP/PS1 mice express RIPK1 (red). Nuclei stained with DAPI (blue). [Image courtesy of Ofengeim, PNAS]

RIPK1, short for Receptor-interacting protein 1 kinase, gets induced in response to the inflammatory signals tumor necrosis factor (TNFa) and ligands of the toll-like receptor (TLR) family. It causes an inflammatory response, controls inflammation-induced cell death (necroptosis), and leads to some forms of apoptosis (for a review, see Ofengeim and Yuan, 2013). The kinase has already been shown to mediate the necrotic cell death and inflammation observed in amyotrophic lateral sclerosis and multiple sclerosis (Ito et al., 2016Ofengeim et al., 2015). Since inflammation also plays a big role in Alzheimer’s disease, Yuan wondered if RIPK1 was involved there, as well. They also wondered how it might affect disease pathology.

To find out, co-first authors Dimitry Ofengeim and Sonia Mazzitelli first peered into postmortem human brains and found more phosphorylated RIPK1 in slices from AD patients than controls. This implied that the kinase was activated in the disease. That RIPK1 co-localized with microglial markers suggested that it was expressed primarily in these cells. The same proved true in the brains of six-month-old APP/PS1 mice, where RIPK1 appeared in microglia, particularly those surrounding plaques, but not in astrocytes or neurons (see image above).

What did the kinase do? The authors tested this in APP/PS1 mice by adding to their drinking water a RIPK1 inhibitor the group had previously developed called necrostatin-1 (Nec-1s). After a month, the treated mice had fewer plaques and less soluble and insoluble Aβ in the brain. What’s more, whereas five-month-old APP/PS1 mice scurried around an open field in a hyperactive state, a month of Nec-1s treatment calmed them down. The researchers also examined spatial memory with a T-shaped water maze, where mice are trained to find a hidden platform at the end of one arm, then retrained to find it in another. At five months, APP/PS1 mice had trouble learning a new platform location, but a month of Nec-1s restored their performances to match those of wild-type mice.

The kinase activity of RIPK1 likely lay at the root of these effects, as APP/PS1 mice that expressed the protein with a mutated, inactive kinase region reaped the same physiological and behavioral benefits of receiving Nec-1s. Together the data suggested that thwarting RIPK1 action lightened amyloid pathology and subsequent symptoms.

How does a microglial kinase do this? The researchers added Aβ1-42 to microglia isolated from wild-type and kinase-dead knock-in mice. Wild-type cells bumped up production of the inflammatory cytokines TNFa and IL6, mutant cells less so. Wild-type microglia pretreated with Nec-1s also produced less TNFα and IL6.

Intriguingly, microglia lacking RIPK1 action better digested synthetic Aβ1-42 oligomers. What whetted their appetites? Analyzing the microglial transcriptomes, Ofengeim and Mazzitelli found that one of the proteins upregulated by RIPK1 was cystatin F. Encoded by the Cst7 gene, cystatin F inhibits the endosomal/lysosomal system. Microglia that expressed Cst7 surrounded plaques in APP/PS1 mouse and AD postmortem tissue.

RIPK1-regulated genes were also altered in microglia from the 5XFAD model of AD, the SOD1G93A model of ALS, and aging microglia, implying that RIPK1 went into overdrive there, as well. One of those genes was CH25h, an enzyme on the cell surface responsible for cholesterol and lipid metabolism. CH25h is an AD GWAS hit (Wollmer, 2010). 

Cst7 recently drew notice as a marker of DAM cells that surround plaques in 5xFAD AD mouse models. The results suggest that active RIPK1 triggers microglia to adopt a DAM state, which slows the ability to break down Aβ and raises interest in RIPK1 inhibition as a new treatment approach for AD, the authors write.

Researchers at Denali Therapeutics in South San Francisco picked up this idea and are testing a RIPK1 inhibitor in a Phase 1 trial of healthy volunteers in Europe. Yuan actively collaborates with the group. If the compound proves safe, trials in ALS and AD patients are next.

The current paper provides “indirect evidence that releasing the brakes on cerebral innate immunity is a valid therapeutic approach,” Town told Alzforum. Once controversial, multiple lines of evidence are now converging on this notion, he said. “What used to be a fringe theory is coming into the mainstream in a big way.” He cautioned that before concluding that the lysosomal effect occurs through Cst7, the authors need to show that blocking Cst7 has the same effect as the kinase mutant.

Salvatore Oddo, Arizona State University, Tempe, said the results jibe with those his group reported earlier this year (Jul 2017 news). He suggested an acceleration of RIPK1-induced necroptosis in AD. While the current paper focuses instead on resulting inflammation, both studies converge on the idea that RIPK1 could be a therapeutic target for AD, he said. “It is likely that in AD, RIPK1 influences multiple pathways, having to do with both cell death and non-cell death mechanisms.”—Gwyneth Dickey Zakaib

Comments

  1. RIPK is an interesting molecule. It is involved in multiple biological responses that can be broadly divided into three major pathways. Surprisingly, only one of those three depends on its kinase activity. One of the new ideas in protein biology is that kinases can serve multiple functions, and only some may depend on their kinase activity.

    We became interested in RIPK because it was a hit in a screen we did to discover genes that modify progranulin expression levels (Mason et al., 2017). Progranulin deficiency leads to neurodegenerative disease. Haploinsufficiency causes frontotemporal dementia and rare patients with no progranulin develop neuronal ceroid lipofucinosis. There are genetic variants of progranulin that some believe may affect the risk of developing AD, and there are preclinical results suggesting that the level of progranulin in mice affects amyloid formation in mutant APP mouse models. Consistent with the main findings of this study, the conclusion of that work was that progranulin levels were working primarily by regulating microglial function to affect amyloid clearance.

    It’s interesting to consider the possibility that the effects described in this paper of RIPK1 inhibition on amyloid might be mediated by microglia. However, one wrinkle is that we found that the kinase activity of RIPK1 was not required for its ability to regulate progranulin levels. We did not find any effect of the necrostatin RIPK1 inhibitor on progranulin levels. So, at the moment, it’s unclear whether and how the effects of RIPK1 on progranulin and progranulin on amyloid might be related to the observations described here.

    That said, there is potentially another interesting connection. The authors here believe that the effects might be mediated through an effect of Cst7 and detrimental effects on the lysosome. That is interesting because we and others have found that progranulin haploinsufficiency impairs lysosomal function, which can cause hyper activation of microglia and poor neuronal proteostasis. That could contribute to neurodegeneration by making neurons more vulnerable and by enhancing the toxicity of the environment.

    References:

    . The Receptor-interacting Serine/Threonine Protein Kinase 1 (RIPK1) Regulates Progranulin Levels. J Biol Chem. 2017 Feb 24;292(8):3262-3272. Epub 2017 Jan 9 PubMed.

  2. The finding that genetic and pharmacological inhibition of RIPK1, a kinase overexpressed in plaque-associated microglia recently defined as disease-associated microglia (DAM), has therapeutic benefits in APP/PS1 mice adds little to the debate as to whether DAM are a dead-end product, as I believe, or a protective reaction optimized by evolution in the context of “neuroinflammation,” as the creators of the notion of DAM support. This study strictly shows that RIPK1 inhibition eliminates DAM, suggesting that RIPK1 participates in their generation. But this is hardly a breakthrough: when inhibited in mouse models, numerous microglia pathways result in the disappearance of DAM, along with reduction of plaque burden and memory deficits. Translation of these therapies to the clinic is still work in progress.

    What, then, might be unique about RIPK1?

    First, RIPK1 could be a superior target because it appears to be a signaling hub. An asset of the study is the network analysis of adult isolated microglia from mice with genetic inactivation of RIPK1. The analysis has led to the identification of several transcription factors as downstream targets of the kinase. This suggests that RIPK1 is a pleiotropic kinase controlling microglia phenotype through multiple transcriptional programs. It follows that therapeutical inhibition of RIPK1 may globally hamper maladaptive stress reactions of microglia.

    Second, the in-depth characterization of RIP1K-dependent pathways may provide insight into which microglia functions are altered in AD. Based upon a series of in vitro studies, the authors favor the notion that aberrant RIP1K activation impairs lysosomal function in microglia by overexpression of the protease inhibitor cystatin F, thereby reducing Aβ clearance and contributing to plaque formation. However, their unbiased network analysis in RIPK1-deficient microglia shows dysregulation of up to 20 pathways different from “lysosomal function.” It would be interesting to clarify the roles of these pathways in healthy and diseased microglia, and their relationship with pathways regulated by microglia genes identified as genetic risk factors in AD by GWAS.

    Even if the authors are right about RIP1K causing impairment of Aβ clearance in DAM, other possibilities cannot be excluded. For example, aberrant RIPK1 activation in microglia might impair their crosstalk with neurons and astrocytes, which may lead, in turn, to increased levels of soluble Aβ due to enhanced production by neurons (Cirrito et al., 2005) and reduced clearance by astrocytes (Iliff et al., 2012). All in all, we do not know how pharmacologically or genetically restored microglia reverse pathology. Any conclusion, like the very reasonable one that microglia take up Aβ, is based upon indirect evidence. When it comes to microglia, the early stages of AD are a black box and we need to refine our experimental tools and questions to cast a light on it.

  3. These findings are particularly interesting, as they provide novel insights into the microglial phenotypic changes implicated in AD pathology.

    Whether Cst7 can be considered a biomarker of the DAM phenotype should be investigated further. Cst7 has been shown to be upregulated by the DAM (Keren-Shaul et al., 2017), but it still remains undetermined whether all DAM do express Cystatin F, and whether Cst7 is exclusive to this microglial subpopulation.

    Determining whether inhibiting the RIPK1 pathway could be used to prevent the DAM phenotype, and perhaps to generate animal models devoid of DAM, is a promising research area to pursue. It will be important in future investigations to assess the impact of DAM microglia on synaptic loss and cognition across various disease contexts.

    References:

    . A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Hot DAM: Specific Microglia Engulf Plaques
  2. Necroptosis Rampant in the Alzheimer’s Brain?

Research Models Citations

  1. APPswe/PSEN1dE9

Paper Citations

  1. . Regulation of RIP1 kinase signalling at the crossroads of inflammation and cell death. Nat Rev Mol Cell Biol. 2013 Nov;14(11):727-36. Epub 2013 Oct 16 PubMed.
  2. . RIPK1 mediates axonal degeneration by promoting inflammation and necroptosis in ALS. Science. 2016 Aug 5;353(6299):603-8. PubMed.
  3. . Activation of necroptosis in multiple sclerosis. Cell Rep. 2015 Mar 24;10(11):1836-49. PubMed.
  4. . Cholesterol-related genes in Alzheimer's disease. Biochim Biophys Acta. 2010 Aug;1801(8):762-73. PubMed.

External Citations

  1. Phase 1 trial

Further Reading

Papers

  1. . Microglia emerge as central players in brain disease. Nat Med. 2017 Sep 8;23(9):1018-1027. PubMed.
  2. . Microglia in Alzheimer's disease. J Clin Invest. 2017 Sep 1;127(9):3240-3249. PubMed.

Primary Papers

  1. . RIPK1 mediates a disease-associated microglial response in Alzheimer's disease. Proc Natl Acad Sci U S A. 2017 Oct 10;114(41):E8788-E8797. Epub 2017 Sep 13 PubMed.