RIPK1 mediates a disease-associated microglial response in Alzheimer's disease - PubMed (original) (raw)

. 2017 Oct 10;114(41):E8788-E8797.

doi: 10.1073/pnas.1714175114. Epub 2017 Sep 13.

Sonia Mazzitelli 1, Yasushi Ito 1, Judy Park DeWitt 1, Lauren Mifflin 1, Chengyu Zou 1, Sudeshna Das 2 3, Xian Adiconis 4, Hongbo Chen 1, Hong Zhu 1, Michelle A Kelliher 5, Joshua Z Levin 4, Junying Yuan 6

Affiliations

RIPK1 mediates a disease-associated microglial response in Alzheimer's disease

Dimitry Ofengeim et al. Proc Natl Acad Sci U S A. 2017.

Abstract

Dysfunction of microglia is known to play an important role in Alzheimer's disease (AD). Here, we investigated the role of RIPK1 in microglia mediating the pathogenesis of AD. RIPK1 is highly expressed by microglial cells in human AD brains. Using the amyloid precursor protein (APP)/presenilin 1 (PS1) transgenic mouse model, we found that inhibition of RIPK1, using both pharmacological and genetic means, reduced amyloid burden, the levels of inflammatory cytokines, and memory deficits. Furthermore, inhibition of RIPK1 promoted microglial degradation of Aβ in vitro. We characterized the transcriptional profiles of adult microglia from APP/PS1 mice and identified a role for RIPK1 in regulating the microglial expression of CH25H and Cst7, a marker for disease-associated microglia (DAM), which encodes an endosomal/lysosomal cathepsin inhibitor named Cystatin F. We present evidence that RIPK1-mediated induction of Cst7 leads to an impairment in the lysosomal pathway. These data suggest that RIPK1 may mediate a critical checkpoint in the transition to the DAM state. Together, our study highlights a non-cell death mechanism by which the activation of RIPK1 mediates the induction of a DAM phenotype, including an inflammatory response and a reduction in phagocytic activity, and connects RIPK1-mediated transcription in microglia to the etiology of AD. Our results support that RIPK1 is an important therapeutic target for the treatment of AD.

Keywords: Alzheimer’s disease; RIP1; RIPK1; inflammation; microglia.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest statement: Denali Therapeutics has licensed the Necrostatin program (including Nec-1s) from Harvard Medical School. J.Y. is a consultant of Denali Therapeutics.

Figures

Fig. 1.

Fig. 1.

Elevated levels of active RIPK1 in cortical samples from AD patients. (A) Immunostaining using anti-RIPK1 antibody in slices from the temporal lobe from control and AD patients. (Inset, Bottom) Higher magnification images. (B) Western blot analysis of RIPK1 from control (Con) and AD patient postmortem temporal lobe-derived tissue; the soluble fraction was isolated from buffer with Triton (1%), and the insoluble fraction was subsequently solubilized using buffer with 6 M urea. (C, Top) Representative Western blot analysis of control and AD cortical tissues for RIPK1 and p-S166-RIPK1 levels in insoluble fractions (dissolved in 6 M urea) from five control and five AD patients. (Bottom) Quantifications of samples from 10 control and 10 AD cases probed with antibodies against RIPK1 and normalized to actin loading control. Data are represented as the normalized means ± SEM, n = 10 replicates per group (**P < 0.01, Student’s t test). (D) RNA was isolated from postmortem tissue from control and AD patients. Levels of RIPK1 and TNFα transcripts were examined. GAPDH was used as a housekeeping gene, and fold change was determined using the ΔΔCT method (**P < 0.01, Student’s t test). (E, Top) Coimmunostaining using anti-RIPK1 antibody and the microglial marker IBA1 in slices from the temporal lobe from a control and an AD patient. (Bottom) Higher magnification images. (F, Top) Western blot analysis from the soluble and insoluble fractions from brains of WT and APP/PS1 mice (7 mo old) probed with antibodies against APP, RIPK1, and Actin, used as a loading control. (Bottom) Graph representing the quantification of Western blot results of RIPK1 normalized to actin (data are means ± SEM, n = 6 to 8 animals per group). RNA was isolated from WT and APP/PS1 mouse brains. Levels of RIPK1 transcript were examined. GAPDH was used as a housekeeping gene, and fold change was determined using the ΔΔCT method (*P < 0.05, Student’s t test). (G, Top) Coimmunostaining using anti-RIPK1 antibody and the microglial marker IBA1 close to an Aβ plaque in slices from 7-mo-old APP-PS1 mice. (Bottom) Higher magnification images. Images are representative from experiments done on eight animals.

Fig. S1.

Fig. S1.

(A_–_C) Low magnification immunostaining using anti-RIPK1 antibody in slices from the frontal and temporal lobes from human control and AD patients. (D) Western blot analysis of RIPK1 from control and AD patient’s postmortem temporal lobe-derived tissue: the soluble fraction–Triton (1%) soluble and the urea soluble and urea insoluble fractions. Three control and three AD patient samples were used. (E) Coimmunostaining using anti-RIPK1 antibody and the astrocyte marker GFAP from the temporal AD patients. Four AD patient’s slices were used. (F, Top) Coimmunostaining using anti-RIPK1 antibody and the astrocyte marker GFAP close to an amyloid plaque in slices from 7-mo-old APP/PS1 mice. (Bottom) Higher magnification images. (G) Quantification of RIPK1-positive IBA1 and GFAP cells in human AD and mouse APP/PS1 tissue (n = 3 brain slices per group). (H, Top) Coimmunostaining using anti-RIPK1 antibody and axonal marker neurofilament close to an Aβ plaque in slices from 7-mo-old APP/PS1 mice. (Bottom) Higher magnification images.

Fig. S1.

Fig. S1.

(A_–_C) Low magnification immunostaining using anti-RIPK1 antibody in slices from the frontal and temporal lobes from human control and AD patients. (D) Western blot analysis of RIPK1 from control and AD patient’s postmortem temporal lobe-derived tissue: the soluble fraction–Triton (1%) soluble and the urea soluble and urea insoluble fractions. Three control and three AD patient samples were used. (E) Coimmunostaining using anti-RIPK1 antibody and the astrocyte marker GFAP from the temporal AD patients. Four AD patient’s slices were used. (F, Top) Coimmunostaining using anti-RIPK1 antibody and the astrocyte marker GFAP close to an amyloid plaque in slices from 7-mo-old APP/PS1 mice. (Bottom) Higher magnification images. (G) Quantification of RIPK1-positive IBA1 and GFAP cells in human AD and mouse APP/PS1 tissue (n = 3 brain slices per group). (H, Top) Coimmunostaining using anti-RIPK1 antibody and axonal marker neurofilament close to an Aβ plaque in slices from 7-mo-old APP/PS1 mice. (Bottom) Higher magnification images.

Fig. 2.

Fig. 2.

Pharmacological inhibition of RIPK1 attenuates biochemical pathology and behavioral deficits in the APP/PS1 mouse. Five-month-old WT and APP/PS1 mice were given vehicle or Nec-1s for 1 mo. The brain slices from these mice were examined by histology for ThS+ inclusions in the cortex (A and B) and hippocampus (B). (C) Nec-1s inhibited the APP/PS1-induced increase of ThS+ plaques. Furthermore, Nec-1s attenuated the increase of both soluble and insoluble (guanidine-soluble) Aβ1–42 in APP/PS1 mice (n = 6 to 8 mice per group). ns, not significant. (D) WT and APP/PS1 mice (∼5 mo old) were given vehicle or Nec-1s for 1 mo and examined in the open field test (n = 7 to 9 male mice per group) for 60 min; the total ambulatory distance the mice traveled was quantified. (E) The distance the mice traveled during 5-min bins was also assessed in the APP/PS1 mice. (F) WT and APP-PS1 mice (∼6 mo old) were assessed using the water T-maze to evaluate spatial working memory. A significant decrease was observed in both the acquisition phase (Left) and the reversal phase (Middle) in ∼6-mo-old APP/PS1 mice compared with WT mice (n = 12 to 14 male mice per group). (Right) The same mice were randomly split into two groups per genotype and given either vehicle or Nec-1s for 1 mo and then reassessed in the reversal phase of the water T-maze. *P < 0.05; **P < 0.01; one-way ANOVA followed by Bonferroni's post hoc test.

Fig. S2.

Fig. S2.

WT and APP/PS1 mice (5 mo old) were given vehicle or Nec-1s for 1 mo. Brain slices from these mice were immunostained with an Aβ1–42 antibody (Top). Plaque numbers were quantified in each brain slice inclusion in the cortex (Bottom). Data are represented as the number of plaques per slice, n = 3 to 4 slices per animal, seven to eight animals per group (*P < 0.05, Student’s t test).

Fig. 3.

Fig. 3.

Genetic inhibition of RIPK1 attenuates biochemical pathology and behavioral deficits in the APP/PS1 mouse. WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice were examined by histology for ThS+ inclusion in the cortex (A and B) and hippocampus (B). Genetic inhibition of RIPK1 kinase by D138N mutation reduced the levels of ThS+ plaques. (C) Additionally, there were reduced levels of both soluble and insoluble (guanidine-soluble) Aβ1–42 in the APP/PS1;RIPK1D138N mice compared with the APP/PS1;WT mice (n = 6 to 8 mice per group). ns, not significant. (D) Six-month-old male WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice were examined in the open field test (n = 7 to 8 male mice per group) for 60 min; the total ambulatory distance the mice traveled was quantified. (E) The distance the mice traveled during 5-min bins was also assessed. (F) The water T-maze was used to evaluate spatial working memory in APP/PS1 mice. A significant decrease was observed in both the acquisition phase (Left) and the reversal phase (Right) in 5- to 6-mo-old APP/PS1 mice (n = 7 to 9 male mice per group. *P < 0.05, one-way ANOVA followed by Bonferroni's post hoc test.

Fig. 4.

Fig. 4.

Genetic inhibition of RIPK1 reduces the microglial inflammatory response in the APP/PS1 mouse. (A) Representative images of immunostaining using anti-Aβ 6E10 antibody and the microglial marker IBA1 in slices from APP/PS1 and APP/PS1;RIPK1D138N mice (7 mo old). (B) Quantification of IBA1-positive cells that are associated with Aβ plaques (15 slices were quantified per genotype; *P < 0.05, Student’s t test). (C) The levels of TNFα and IL1β protein in the brains of WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice were examined by ELISA. The quantification is represented ±SEM; n = 6 to 9 animals per group in the graph. *P < 0.05, Student's t test. (D) Primary microglia were isolated from WT or RIPK1D138N and were treated with oligomeric Aβ1–42 in the presence of either vehicle or Nec-1s. The levels of TNFα were examined by ELISA, 24 h after treatment. *P < 0.05 [Ab vehicle vs. Ab+Nec-1s]; #P < 0.05 [WT Ab vehicle vs. RIPK1D138B Ab vehicle]. (E) WT primary microglia were treated with increasing concentrations of oligomeric Aβ1–42 for either 6 h or 24 h in the presence of either vehicle or Nec-1s. Western blot analysis was used to examine RIPK1 activation by immunoblotting for p-S166 RIPK1. (F_–_H) WT primary microglia were treated with oligomeric Aβ1–42. Both extracellular (F) and intracellular (G) levels of Aβ1–42 were examined by both dot-blot and by ELISA. *P < 0.05, Student's t test. (H) The time course of the intracellular levels of Aβ1–42 was examined by Western blot analysis in the presence of either vehicle or Nec-1s. Images are representative of two to three experiments.

Fig. S3.

Fig. S3.

(A) Primary mouse cortical neurons (DIV 10) were treated with oligomeric Aβ1–42 in the presence of either vehicle, Nec-1s, or zVAD.fmk. The cell viability was assessed by using the Cell Titer-Glo assay. (B) Primary microglia were isolated from WT mice (two to three replicate experiments). The cells were treated with oligomeric Aβ1–42 for 24 h in the presence of either vehicle or Nec-1s. RNA levels of TNFα and IL6 were assessed by quantitative real time PCR (Top). Primary microglia were isolated from WT and RIPK1D138N mice and treated with oligomeric Aβ1–42 for 24 h. The RNA levels of TNFα and IL6 were assessed by quantitative real time PCR (Bottom). (C) Primary microglia were isolated from WT mice and treated with fluorescently labeled oligomeric Aβ1–42 for 8 h in the presence of either vehicle or Nec-1s. High content imaging and analysis were used to quantify the amount of Aβ1–42 associated with the microglia in each condition (two replicate experiments). (D) Primary microglia were isolated from WT and RIPK1D138N mice. The cells were treated with fluorescently labeled oligomeric Aβ1–42 for 8 h (four replicate experiments). *P < 0.05, **P < 0.01.

Fig. S4.

Fig. S4.

(A, Left) Representative FACS plot of the population of CD45-positive CD11b-positive brain-derived microglia that were used for subsequent RNA-seq experiments. (Right) Distribution of the number of genes identified in the RNA-seq experiment (x axis), relative to the read counts for that gene (y axis), based on all 13 samples used for this experiment. (B) Microglia were isolated from WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice (5 to 6 mo old), and the RNA levels of CCL3, CD200R1, CAP1, and ARSB were assessed by quantitative real time PCR.

Fig. 5.

Fig. 5.

Defects in protein turnover promote the activation of RIPK1 in APP/PS1 microglia. (A) Pathways and processes that are enriched in APP/PS1-derived microglia transcriptome compared with WT mice. Analysis was done using gene set enrichment analysis (MSigDB; Broad). (B) WT primary microglia were treated with either vehicle chloroquine (50 μM), MG132 (10 μM), or LPS (10 ng/mL) for the indicated amount of time (min). Western blot analysis was used to examine RIPK1 activation by immunoblotting for p-S166 RIPK1. Images are representative of two to three experiments. (C) WT primary microglia were treated with chloroquine (50 μM) in the presence of either vehicle or Nec-1s. Western blot analysis was used to examine RIPK1 activation by immunoblotting for p-S166 RIPK1. Images are representative of two experiments. (D) Microglia were isolated from adult WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice (5 to 6 mo old) and analyzed by RNA-seq. Our coexpression analysis identified a module with ∼149 genes that were up-regulated in APP/PS1 microglia and suppressed in APP/PS1;RIPK1D138N microglia. (E) Pathways and processes that are enriched in the genes that are up-regulated in the APP/PS1-derived microglia vs. WT-derived microglia and the increase is attenuated in the APP/PS1;RIPK1D138N-derived microglia. Analysis was done using gene set enrichment analysis (MSigDB; Broad). (F) The transcription factors that can regulate the expression of these 149 genes were also examined. Analysis was done using gene set enrichment analysis (MSigDB; Broad). Data are represented as an FDR q-value (×10−6).

Fig. 6.

Fig. 6.

RIPK1-mediated transcriptional program impairs lysosomal function. (A) A multistudy comparison of the transcription levels of several genes up-regulated in a RIPK1-dependent manner in APP/PS1 microglia vs. APP/PS1;RIPK1D138N microglia with the published datasets. Column 2, APP/PS1 vs. WT microglia; column 3, APP/PS1;RIPK1D138N vs. WT microglia; column 4, APP/PS1 vs. WT microglia (64); column 5, 5xFAD vs. WT microglia (65); column 6, LPS-stimulated vs. control microglia (66); column 7 microglia isolated from 4-mo-old mice vs. microglia isolated from 24-mo-old mice (67); column 8, microglia of SOD1G93A transgenic mice vs. WT control microglia (39). (B) A comparison of the transcript levels of CH25H in our microglia-derived RNA-seq dataset, whole brain tissue, and also a qPCR from a new cohort of brain-derived microglia in WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice. (C) BV2 cells were treated with IFNγ in the presence or absence of Nec-1s. CH25H transcript levels were assessed by qPCR. (D) A comparison of the Cst7 transcript levels in our brain-derived microglia RNA-seq dataset, whole brain tissue, as well as qPCR from a new cohort of brain-derived microglia from WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice. (E) BV2 cells were treated with IFNγ in the presence or absence of Nec-1s. CST7 transcript levels were assessed by qPCR. *P < 0.05, **P < 0.01, ***P < 0.001, Student’s t test.

Fig. 7.

Fig. 7.

CST7 is increased in AD. (A) Immunostaining using anti-CST7 antibody in slices from the WT and APP/PS1 mice in both the cortex and hippocampus (Hipp). (Bottom) Higher magnification of APP/PS1 slices in the cortex. Three WT and three APP/PS1 animals were used to validate this finding. (B) Immunostaining using anti-CST7 antibody in slices from APP/PS1 mice in the cortex followed by ThS+ staining. Two magnifications are shown. Three WT and three APP/PS1 animals were used to validate this finding. (C) Coimmunostaining using an anti-CST7 antibody and the microglial marker IBA1 in slices from WT, APP/PS1, and APP/PS1;RIPK1D138N. DAPI was used to stain nuclei. Three WT and three APP/PS1 animals were used to validate this finding. (D) RNA was isolated from postmortem tissue from control and AD patients. Levels of CST7 were examined by q-PCR. GAPDH was used as a housekeeping gene, and fold change was determined using the ΔΔCT method (**P < 0.01, Student’s t test; five control and five AD patients). (E) Immunostaining using anti-CST7 antibody in slices from the temporal lobe from control and AD patients. (Bottom) Higher magnification images from AD tissue. (F) Brain cells derived from WT or APP/PS1 animals (5 to 6 mo old) treated with either Nec-1s or vehicle for 1 mo were then FACS sorted based as CD11b+ CD45low/intermediate vs. CD11b+ CD45High cells. RNA levels of CST7 and Ch25H in both populations of cells (n = 3 to 4 animals per group, **P < 0.01, #P < 0.05 [APP/PS1 CD45 low/inter vs. APP/PS1 CD45 high], Student’s t test). (G) BV2 cells were transfected with either empty vector or a CST7 expressing plasmid overnight. The cells were then treated with either vehicle or chloroquine for 4 h. Western blot analysis was performed to examine RIPK1, p62, LC3, p-cJun and cJun levels. Beta actin was used as a loading control. Western blot images are representative of two experiments. (H) Vector or CST7 was transfected in HEK293T or BV2 cells, and cathepsin C activity was assessed by a fluorometric assay.

Fig. S5.

Fig. S5.

A Comparison of RIPK1-regulated transcriptional changes with those reported to be associated with DAM. A comparison of the transcript levels of DAM-associated genes (12) in the microglia derived RNA-seq dataset in WT, RIPK1D138N, APP/PS1;WT, and APP/PS1;RIPK1D138N mice that were attenuated with inhibition of RIPK1 kinase activity by RIPK1D138N mutation (A) and those that were not (B).

Fig. S6.

Fig. S6.

(A) Representative FACS plot of CD45 and CD11b immunostained that were sorted as CD45low/intermediate vs. CD45High cells, which were used for qRT-PCR. (B, Top) Representative FACS plot of TREM2 surface expression in CD11b+/CD45low/intermediate and CD11b+/CD45High cell populations. (Bottom) Mean fluorescent intensity was quantified (n = 3 to 4 animals per group, **P < 0.01, Student’s t test). (C) FACS sorted microglia/macrophages from WT or APP/PS1 animals (5 to 6 mo old) treated with either Nec-1s or vehicle for 1 mo and the RNA levels of TMEM119, RIPK1 were assessed by real time qPCR in both CD11b+ CD45low/intermediate and CD11b+ CD45High cells (n = 3 to 4 animals per group).

Fig. S7.

Fig. S7.

BV2 cells were transfected with either empty vector (HA), GFP, or an HA-CST7–expressing plasmid overnight. To confirm the expression of Cst7, anti-HA conjugated beads were used to immunoprecipitate HA-Cst7, treated with endoH to remove glycosylation and then analyzed by immunoblotting using anti-CST7 (two replicates experiments were done).

Comment in

References

    1. Mandrekar-Colucci S, Landreth GE. Microglia and inflammation in Alzheimer’s disease. CNS Neurol Disord Drug Targets. 2010;9:156–167. - PMC - PubMed
    1. Karch CM, Goate AM. Alzheimer’s disease risk genes and mechanisms of disease pathogenesis. Biol Psychiatry. 2015;77:43–51. - PMC - PubMed
    1. Abduljaleel Z, et al. Evidence of trem2 variant associated with triple risk of Alzheimer’s disease. PLoS One. 2014;9:e92648. - PMC - PubMed
    1. Hollingworth P, et al. Alzheimer’s Disease Neuroimaging Initiative. CHARGE consortium. EADI1 consortium Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with Alzheimer’s disease. Nat Genet. 2011;43:429–435. - PMC - PubMed
    1. Jonsson T, et al. Variant of TREM2 associated with the risk of Alzheimer’s disease. N Engl J Med. 2013;368:107–116. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources