Cell autonomous microglia defects in a stem cell model of frontotemporal dementia

Abstract

Neuronal dysfunction has been extensively studied as a central feature of neurodegenerative tauopathies. However, across neurodegenerative diseases, there is strong evidence for active involvement of immune cells like microglia in driving disease pathophysiology. Here, we demonstrate that tau mRNA and protein are expressed in microglia in human brains and in human induced pluripotent stem cell (iPSC)-derived microglia like cells (iMGLs). Using iMGLs harboring the MAPT IVS10+16 mutation and isogenic controls, we demonstrate that a tau mutation is sufficient to alter microglial transcriptional states. We discovered that MAPT IVS10+16 microglia exhibit cytoskeletal abnormalities, stalled phagocytosis, disrupted TREM2/TYROBP networks, and altered metabolism. Additionally, we found that secretory factors from MAPT IVS10+16 iMGLs impact neuronal health, reducing synaptic density in neurons. Key features observed in vitro were recapitulated in human brain tissue and cerebrospinal fluid from MAPT mutations carriers. Together, our findings that MAPT IVS10+16 drives cell-intrinsic dysfunction in microglia that impacts neuronal health has major implications for development of therapeutic strategies.

Figure 1.. iPSC-derived microglia expressing MAPT IVS10+16 reveal shifts in microglia states compared with isogenic controls.

A-C. Microglia isolated from human brains and analyzed by bulk RNA sequencing were analyzed for MAPT expression (n=10) as previously reported in Olah et al. A. Workflow for microglia isolation. B. MAPT expression (TPM). Each dot represents microglia sampled from an individual brain. C. MAPT expression, indicated by the red dashed line, relative to the 19,182 protein coding genes in microglia isolated from human brains. D. Schematic representing how iPSC-derived human microglia-like cells (iMGLs) were generated from hematopoietic progenitor cells (HPCs) differentiated from two independent donors of MAPT IVS10+16 carriers: FTD1 and FTD2 and their isogenic controls (4 lines total). E. Dotplots of CD45 and CD11b expression reveals that >95% of the iMGLs are double positive for CD45 and CD11b across donors and genotypes. F. Immunocytochemistry of iMGLs for TMEM119. Scale bar, 12μm. G-H. Whole cell lysates were isolated by detergent extraction and analyzed by immunoblotting for total tau (Tau5) and GAPDH (G) and by immunoprecipitation (Tau1/HJ8.5/HJ8.7 antibodies) and mass spectrometry (H). Peptides monitored by mass spectrometry are plotted relative to the mid-domain peptide at amino acid 181–190 (TPSL). Tau protein domains are annotated above the graph N term (N-terminus), Mid domain, MTBR (microtubule binding region) and C term (C-terminus). Open circles, MAPT IVS10+16. Closed circles, MAPT WT (isogenic). I. RNAseq and differential gene expression were performed on iMGLs from MAPT IVS10+16 and isogenic controls. Volcano plot. Red dots, FDR-BH ≤ 0.05. Blue dots, p-value ≤ 0.05. Gray dots, not statistically significant. J. Percentage of genes defining specific microglia functions/states: chemokines (n=12 genes), phagocytosis (n= 14 genes), DAM (n= 17 genes), LAM (n=12 genes) and homeostatic (n= 10 genes) were explored in the differential gene expression. Red, significantly upregulated. Blue, significantly down regulated. White, not statistically significant. FDR-BH ≤ 0.05. DAM, disease associated microglia. LAM, lipid-associated myeloid cells.

Figure 2.. MAPT IVS10+16 alters the cytoskeleton in iMGLs.

A. Pathway analysis of genes differentially expressed between MAPT IVS10+16 and isogenic controls shows an enrichment of genes regulating the cellular cytoskeletal network. Red bars, significantly upregulated genes. Blue bars, significantly downregulated genes. Significance defined as FDR-BH≤0.05. Three biological replicates were included for each donor. B. Immunocytochemistry for -tubulin. Representative images from a Nikon spatial array confocal images acquired at 60x. Scale bar, 12μm. C. Quantification of mean fluorescence intensity (MFI) of -tubulin staining across cellular area using the straight-line tool in ImageJ. MFI was plotted relative to distance across the cell (normalized and expressed as a percentage). D. Immunocytochemistry for filamentous actin (F-actin; phalloidin; green) and nucleus (NucBlue). Scale bar, 12μm. E. Quantification of mean fluorescence intensity of phalloidin within each cell. ****, p<0.0001; *, p<0.05. F. Total intracellular levels of F-actin were quantified by flow cytometry using Phalloidin-488 represented as dot plots reveal the percentage of F-actin positive cells within CD45+CD11b+ iMGLs. Representative dot plots from FTD1 donor pair. G. Quantification of phalloidin (MFI). MFI normalized to WT. Results represent three independent experiments. **, p<0.005; ***, p<0.0005; p-values generated by one-way ANOVA with Sidak’s multiple comparisons test in (E) and (G).

Figure 3.. MAPT IVS10+16 iMGL exhibit reduced phagocytosis of myelin and tau fibrils.

A. Pathway analysis of genes differentially expressed between MAPT IVS10+16 and isogenic controls shows an enrichment of genes regulating endocytosis and phagocytosis. Red bars, significantly upregulated genes. Blue bars, significantly downregulated genes. Significance defined as FDR-BH≤0.05. Three biological replicates were included for each donor. B. Phagocytic uptake of human myelin (20μg/ml, C-E) or recombinant human tau preformed fibrils, Tau-PFF (500nM, F-H) analyzed by IncuCyte live cell imaging over 24 hours. C. Representative images 12 hours post myelin treatment. Scale bar, 100μm. D. Representative quantification of myelin area over time. E. Quantification of myelin area at 24 hours post-treatment. ****, p<0.0001. F. Representative images 12 hours post Tau-PFF treatment. Scale bar, 100μm. G. Representative quantification of Tau-PFF area over time. H. Quantification of Tau-PFF area at 12 hours post-treatment. *, p<0.05; **, p<0.005. Data in C-H is representative of 8 replicates per donor line per treatment group. I-J. Tau-PFF uptake was measured by flow cytometry. I. Histogram of Tau-PFF mean fluorescence intensity (MFI) within CD45+CD11b+ cells 24 hours post-treatment. J. Quantification of Tau-PFF MFI. Data is representative of three biological replicates per line. ***, p<0.005; ****, p<0.0001. p-values by two-ANOVA with Sidak’s multiple comparisons test in (E), (H) and (J).

Figure 4.. MAPT IVS10+16 disrupts TREM2 signaling in iMGLs.

Pathway analysis of genes differentially expressed between MAPT IVS10+16 and isogenic controls shows an enrichment of genes involved in TREM2 signaling. Red bars, significantly upregulated genes. Blue bars, significantly downregulated genes. Significance defined as FDR-BH≤0.05. Three biological replicates were included for each donor. B. Immunocytochemistry for TREM2 (cyan) and nucleus (NucBlue) Scale bar represents 12μm. C-D. TREM2 on the cell membrane (mTREM2) was measured by flow cytometry. C. Histogram overlay of mTREM2 MFI. D. mTREM2 MFI quantification in both donor pairs. *, p<0.05. E. Soluble TREM2 (sTREM2) levels measured by ELISA in media from both donor pairs. ***, p<0.05; p-values were calculated using an unpaired t-test with Welch’s correction in (D) and (E). F. TREM2 mediates pathways in microglia that regulate actin/cytoskeletal organization, proliferation and survival, cytokines, autophagy, and metabolism. Those genes involved in these pathways that were significantly differentially expressed FDR-BH<0.05 are annotated in red (upregulated in mutant iMGL) or blue (downregulated in mutant iMGL).

Figure 5.. MAPT IVS10+16 iMGLs produce secretory factors that alter synapse density.

A. Genes shared between MAPT IVS10+16 iMGLs and brains of patients with FTLD-tau with MAPT mutations (n=100 genes) were enriched in pathways involved in synaptic function and immune response. Genes defined by BH-FDR ≤0.05 in each dataset. B-C. Secretory proteins were monitored using OLINK. B. Volcano plot showing secretory factors altered in comparisons of media from MAPT IVS10+16 iMGLs and controls. Three technical replicates were included for each donor pair. Red dots, significantly upregulated (p<0.05). Blue dots, significantly downregulated (p≤0.05). Gray dots, not significant. C. Pathway analysis of differentially expressed secretory factors (p<0.05). Red, upregulated genes. Blue, downregulated genes. D. Proteins in cerebrospinal fluid (CSF) were measured by SomaScan proteomics platform in MAPT mutation carriers compared with normal controls. Network analyses revealed several modules that were significantly enriched in CSF from MAPT mutation carriers compared with controls. Sx, symptomatic. PreSx, pre-symptomatic. E. Diagram of paradigm. F. Media from control neurons, WT iMGL, or MAPT IVS10+16 iMGL was added to iPSC-derived neurons for 7 days. Immunocytochemistry was performed for dendritic marker (MAP2; green), pan-synaptic marker (Synapsin; red) and nucleus (NucBlue). Scale bar represents 50μm. G. Quantification of synaptic density/mm². Data was normalized to neurons cultured in neuronal media. **, p<0.05; ****, p-value<0.0001; one-way ANOVA with Sidak’s multiple comparisons test. H. Immunocytochemistry was performed for dendritic marker (MAP2; green) and nucleus (NucBlue). Scale bar represents 50μm. I. Quantification of dendritic length (μm). Data was normalized to neurons cultured in neuronal media. *, p<0.05; ****, p-value<0.0001; one-way ANOVA with Sidak’s multiple comparisons test.