INPP5D/SHIP1 regulates inflammasome activation in human microglia

Microglia and neuroinflammation are implicated in the development and progression of Alzheimer’s disease (AD). To better understand microglia-mediated processes in AD, we studied the function of INPP5D/SHIP1, a gene linked to AD through GWAS. Immunostaining and single nucleus RNA sequencing confirmed that INPP5D expression in the adult human brain is enriched in microglia. Examination of prefrontal cortex across a large cohort revealed reduced full-length soluble INPP5D protein levels in AD patient brains compared to cognitively normal controls. However, elevated INPP5D immunostaining in microglia in the AD brain was observed, which was driven by elevations in plaque-associated microglia suggesting that these two methods quantify different pools of INPP5D. Examination of INPP5D overexpression and bi-allelic loss-of-function in induced pluripotent stem cell derived microglia (iMGs) revealed that INPP5D LOF most closely resemble microglia in the AD brain with respect to immune signaling alterations, supporting the hypothesis that INPP5D function is reduced in AD microglia. The functional consequences of reduced INPP5D activity were evaluated in human iMGs, using both pharmacological inhibition of the phosphatase activity of INPP5D and genetic reduction in copy number. Unbiased transcriptional and proteomic profiling of these iMGs suggested an upregulation of innate immune signaling pathways, lower levels of scavenger receptors, reduced lysosomal proteins and altered inflammasome signaling with INPP5D reduction. INPP5D inhibition induced the secretion of IL-1ß and IL-18, further implicating inflammasome activation. Inflammasome activation was confirmed through visualization of inflammasome formation through ASC immunostaining in INPP5D-inhibited iMGs, increased cleaved caspase-1 and through rescue of elevated IL-1ß and IL-18 with caspase-1 and NLRP3 inhibitors. In accord, lower INPP5D is associated with higher IL-18 levels and an elevation in microglia with ASC specks in the AD brain. This work implicates INPP5D as a regulator of inflammasome signaling in human microglia. For INPP5D acute inhibition, samples include iMGs from 2 genetic backgrounds (BR01, BR33), treated with 3AC (1.25 uM 3AC, 6 hours) or vehicle. INPP5D HETs were in BR33 background. The INPP5D biallelic loss-of-function monoclones were in the BR24 iPSC line and the INPP5D overexpression samples were also collected from the BR24 iPSC line. For all samples 500ngs of total RNA input was pA selected and sequenced through the Genewiz polyA selection, HiSeq 2x150 single index sequencing platform. RNAseq reads were quality tested using fastqc, quality trimmed then quantified using the Kallisto pseudoalignment quantification program (v0.43.1) running 50 bootstraps against a Kallisto index generated from GRCh38. Kallisto quantified samples were analyzed using the “Sleuth” package (v0.30.0) in R Studio (v3.6.1 of R; v1.2.5019 of R Studio). Expression values were exported from the Sleuth object as normalized TPM values. The final RNAseq master expression matrix has 16 samples and quantifies expression of 43,122 genes. To identify differentially expressed genes the above matrix was filtered to remove low expressers (greater than 5 TPM in at least 2 samples), leaving 13,532 quantified genes. Conditions were compared by linear modeling with empirical Bayesian analysis using the “limma” package.