A somatic mutation in erythro-myeloid progenitors causes neurodegenerative disease

Abstract

The pathophysiology of neurodegenerative diseases is poorly understood and there are few therapeutic options. Neurodegenerative diseases are characterized by progressive neuronal dysfunction and loss, and chronic glial activation. Whether microglial activation, which is generally viewed as a secondary process, is harmful or protective in neurodegeneration remains unclear. Late-onset neurodegenerative disease observed in patients with histiocytoses, which are clonal myeloid diseases associated with somatic mutations in the RAS-MEK-ERK pathway such as BRAF(V600E), suggests a possible role of somatic mutations in myeloid cells in neurodegeneration. Yet the expression of BRAF(V600E) in the haematopoietic stem cell lineage causes leukaemic and tumoural diseases but not neurodegenerative disease. Microglia belong to a lineage of adult tissue-resident myeloid cells that develop during organogenesis from yolk-sac erythro-myeloid progenitors (EMPs) distinct from haematopoietic stem cells. We therefore hypothesized that a somatic BRAF(V600E) mutation in the EMP lineage may cause neurodegeneration. Here we show that mosaic expression of BRAF(V600E) in mouse EMPs results in clonal expansion of tissue-resident macrophages and a severe late-onset neurodegenerative disorder. This is associated with accumulation of ERK-activated amoeboid microglia in mice, and is also observed in human patients with histiocytoses. In the mouse model, neurobehavioural signs, astrogliosis, deposition of amyloid precursor protein, synaptic loss and neuronal death were driven by ERK-activated microglia and were preventable by BRAF inhibition. These results identify the fetal precursors of tissue-resident macrophages as a potential cell-of-origin for histiocytoses and demonstrate that a somatic mutation in the EMP lineage in mice can drive late-onset neurodegeneration. Moreover, these data identify activation of the MAP kinase pathway in microglia as a cause of neurodegeneration and this offers opportunities for therapeutic intervention aimed at the prevention of neuronal death in neurodegenerative diseases.

Extended Data Figure 1. Analysis of one-month old Csf1r^MeriCreMer; BRAF^LSL-V600E; Rosa26^LSL-YFP mice

(a) % of mice born from the cross depicted in Figure 1a according to their genotype (n=42), but no injection of hydroxy-tamoxifen (4-OHT) to test for adverse effects of 4-OHT administration. (b) Flow cytometry analysis of YFP expression on blood leukocytes. Representative for n=8 per genotype. (c) Flow cytometry analysis of YFP⁺ cells in the liver. YFP⁺ cells, present only in Csf1r^MeriCreMer+ (Cre⁺) mice (upper panels), fall into the F4/80⁺CD11b⁺ Kupffer cell gate (lower panels). Representative for n=8 per genotype. (d) YFP expression by immunofluorescence in the liver of BRAF^VE mice and BRAF^WT. YFP⁺ cells are F4/80⁺ Kupffer cells. Representative of n=6 mice per genotype. Scale bars=200 µm (5 µm for insets). (e) Total tissue-resident macrophages cell numbers per gram (g) of tissue were analyzed by flow cytometry in BRAF^VE mice (n=4) and BRAF^WT (n=6). Circles represent individual mice. Unpaired two-tailed t-test. (f) In situ analysis of phospho-Histone H3 (pHis3) staining in YFP⁺ cells from brains of BRAF^VE and BRAF^WT. Circles represent individual mice (n=3). Unpaired two-tailed t-test. (g) RNA-seq analysis, heatmap representation of MAPK target genes in YFP⁺ microglia from BRAF^VE (n=3) and BRAF^WT (n=2) mice, values are displayed as z-score. (h) Histological analysis of liver, lung, kidney and spleen in BRAF^VE mice and BRAF^WT. Representative of n=4 mice per genotype. Scale bars=200µm (10µm for insets).

Extended Data Figure 2. Effect of constitutive BRAF^V600E expression in Csf1r-expressing cells

(a) Breeding scheme. (b) Embryonic lethality of Csf1r^iCre+; BRAF^LSL-V600E; Rosa26^LSL-YFP mice, bars represent the % of mice born from the cross depicted in (a) according to their genotype (n=39). (c) Brightlight (upper panel) and epifluorescence microscopy (lower panel) of Csf1r^iCre+ BRAF^VE and Csf1r^iCre+; BRAF^WT embryos showing haemorrhagic foci in the liver (arrow) and accumulation of YFP⁺ cells in fetal liver. † indicates dead embryo. Pictures are representative of n=3 per genotype. (d) Mouse embryos found alive during different developmental stages. Csf1r^iCre+; BRAF^LSL-V600E; Rosa26^LSL-YFP mice are associated with 100 % lethality beyond E14.5. (e) Liver weight of E12.5 embryos. Circles represent individual mice. n=8 for WT; Cre⁻, n=14 for VE; Cre⁻, n=16 for VE; Cre⁻, n=12 for VE; Cre⁺. One-way ANOVA. (f) Flow cytometry analysis of Lin⁻ Kit⁺ blast, erythroid cell (Ter119) and hematopoietic stem cell numbers (LSK CD150⁺CD48⁻ and CD150⁻CD48⁻) in the E12.5 fetal liver and of E12.5 tissue-resident macrophages in the limbs, head and liver. Circles represent individual mice. n=4 for BRAF^WT and n=6 for BRAF^VE. Unpaired two-tailed t-test.

Extended Data Figure 3. Analysis of CD11c^Cre; BRAF^V600E mouse model

(a) Kaplan-Meier survival curve of BRAF^VE (n=16) and BRAF^WT (n=66) controls. Log-rank (Mantel-Cox) test. (b) Representative photographs of lung and spleen from BRAF^VE mice at time of death with representative BRAF^WT control organs. (c–d) Hematoxylin and eosin (HE) stain of lung tissue from BRAF^VE and littermate controls. (e) CD68 immunohistochemistry (IHC) of BRAF^VE lung tissue. (f) HE stain of liver tissue from BRAF^VE and littermate controls with magnified view of granuloma in the BRAF^VE liver. (g) HE stain of bone marrow (BM) from BRAF^VE and littermate controls with CD68 IHC of BRAF^VE mouse. All pictures for (b-g) are representative of n=5 per genotype.

Extended Data Figure 4. Longitudinal study and PLX treatment of the Csf1r^MeriCreMer; BRAF^V600E; Rosa26^LSL-YFP mice

(a–b) Latency to fall in the rotarod assay and footprint assay quantification for BRAF^VE mice (n=7) and BRAF^WT littermates (n=8). (a) Rotarod assay at 1 to 4 months of age. Values are mean ± s.d. (b) Rotarod and footprint assay at 4 months of age displaying single values. Mice that are score 1 are labeled in red. (c) Footprint assay quantification of BRAF^VE mice at score 1 and littermate controls. Circles represent individual mice. n=10 for BRAF^WT and n=11 for BRAF^VE. (d) Representative weight curve of BRAF^WT and BRAF^VE mice on control or PLX4720 diet. (e) PLX4720 concentration in serum (ng/ml), liver and brain (ng/g) of 7–9 month old BRAF^WT (n=9) and BRAF^VE mice placed on the diet at 1 (n=8) or at 3 months (n=3) of age. Circles represent individual mice. (f) Footprint assay quantification from BRAF^VE mice on PLX4720 diet at 1 month (n=8) or at 3 months (n=6) and control (Ctrl) diet (n=13) and BRAF^WT (n=32, black). Mice reaching paralysis were excluded from further analysis. See also (g) where † indicates when BRAF^VE animals were euthanized. Values are mean ± s.d. 2way ANOVA comparing treated and not treated BRAF^VE mice. * p<0.05, **p<0.01, ***p<0.001. (g) Disease progression for BRAF^VE mice on control or PLX4720 diet. † indicates animal death due to paralysis.

Extended Data Figure 5. Microglia activation in the brain starts at early, preclinical stages

(a) Histological analysis by hematoxylin and eosin (HE) and Luxol-fast-blue-PAS (LFB-PAS) and immunohistochemistry analysis of T-cells (CD3), B-cells (B220) and astrocyte activation (GFAP) in one-month old BRAF^VE mice and BRAF^WT littermates. Representative of n=5 per for BRAF^WT and n=4 for BRAF^VE. (b) Immunohistochemistry analysis and quantification of Iba1⁺ cell density, cortical neurons (NeuN) and expression of amyloid precursor protein (APP), a positive signal for neurodegeneration in one-month old BRAF^VE mice and BRAF^WT. Representative of n=5 per for BRAF^WT and n=4 for BRAF^VE. Circles represent individual mice. Scale bars=100µm (10µm for insets). Unpaired two-tailed t-test.

Extended Data Figure 6. Neurodegenerative process in BRAF^VE mice

(a) Iba1 and GFAP immunohistochemistry of brain and spinal cord from 6-month-old BRAF^VE and BRAF^WT. Anatomical regions of insets are indicated. Representative for n=5 BRAF^WT and n=4 BRAF^VE. Scale bars=500µm for spinal cords and 1mm for brains (50µm for insets). (b) Immunohistochemistry and immunofluorescence as used for quantification in Figure 2h of brain stem for NeuN (neurons), APP (amyloid precursor protein) and GFAP (astrocytes), Iba1⁺/Lamp2⁺ cells (phagocytosis), Synaptophysin (Syn) and Homer1 (synapse density) and staining with Luxol-fast-blue (LFB)-PAS. Scale bars=100µm (10µm for insets). Iba1/Lamp2 scale bar=25µm, Syn/Homer1 scale bar=10µm. Representative for 6–9 month old BRAF^WT (n=5), BRAF^VE (n=4), and for BRAF^VE on PLX diet (n=4–6). (c) LFB staining of spinal cord samples from (a). Scale bar=100µm. (d) Immunohistochemistry of brain stem for B220 (B-cells) from BRAF^VE on control and PLX diet. Representative for n=4 per genotype. Scale bars=10 µm (10µm for insets).

Extended Data Figure 7. Microglia and T-cell phenotype in BRAF^VE mice

(a) Representative pERK staining in Iba1⁺ microglia as used for the quantification in Figure 3b in brain stem of 5–9 month old BRAF^WT and BRAF^VE mice on control or PLX diet. Scale bar=50 µm. (b) Representative tSNE analysis of flow cytometry staining of CD45⁺ cells from the brain of paralyzed BRAF^VE mice and littermate controls. Arrow indicates expansion of F4/80⁺ YFP⁺ cells. Representative for n=3 per genotype. (c) FSC profile of YFP⁺ and YFP⁻ microglia from (b) from BRAF^VE and BRAF^WT mice indicates an increase of YFP⁺ microglia cell size. Representative for n=3 per genotype. (d) Proportion of YFP⁺ F4/80⁺ cells in indicated organs analyzed by flow cytometry. The proportion of YFP⁺ among F4/80⁺ cells from Cre⁺ BRAF^WT on control diet was normalized and set to 1. Analysis was performed on 5–8 month-old BRAF^VE mice (n=5–6) and BRAF^WT mice (n=6) on control diet, and 7–9 month-old BRAF^VE mice (n=6) and BRAF^WT mice (n=4) on PLX diet. Circles represent values for individual mice. One-way ANOVA. * p<0.05, **p<0.01, ***p<0.001. (e) CD3 immunohistochemistry of brain and spinal cord from 6-month-old BRAF^VE and BRAF^WT. Anatomical regions of insets are indicated. Representative for n=5 BRAF^WT and n=4 BRAF^VE. Scale bars=500µm for spinal cords and 1mm for brains (50µm for insets). (f–g) Analysis of CD8⁺, CD4⁺ and Foxp3⁺ T-cell numbers (e) and proliferation (f) in brain/spinal cord by flow cytometry in 5–8 month-old BRAF^VE (n=4) and BRAF^WT (n=6) on control diet, and 7–9 month-old BRAF^VE (n=6) and BRAF^WT (n=5) on PLX diet. Circles represent values for individual mice. One-way ANOVA.

Extended Data Figure 8. BRAF^VE mice analysis outside the central nervous system

(a) Proportion of YFP⁺ F4/80⁺ cells in indicated organs from analyzed by flow cytometry. The proportion of YFP⁺ among F4/80⁺ cells from Cre⁺ BRAF^WT (n=6) was normalized and set to 1. Circles represent values for individual BRAF^VE mice (n=7). Unpaired two-tailed t-test. (b) Analysis of liver Kupffer cells as in (a) was performed on 5–8 month-old BRAF^VE (n=5) and BRAF^WT (n=4) on control diet, and 7–9 month-old BRAF^VE (n=6) and BRAF^WT mice (n=4) on PLX diet. Circles represent values for individual mice. One-way ANOVA. * p<0.05, **p<0.01, ***p<0.001. (c) Immunofluorescent analysis of pERK in F4/80⁺ Kupffer cells from 5–8 month-old BRAF^VE. Results are representative for n=3. (d) Serum analysis of BRAF^VE mice (score 1, n=6) and their littermates controls (n=6). ALB: albumin, ALP: alkaline phosphatase, ALT: alanine aminotransferase, AST: aspartate aminotransferase. (e) Gross liver, lung, kidney and spleen structure (HE, Trichrome) of paralyzed BRAF^VE and BRAF^WT. Representative of n=7 per genotype. Scale bars=200µm (10µm for insets). (f) Liver and spleen gross organs from paralyzed BRAF^VE and BRAF^WT. Representative of n=5 per genotype.

Extended Data Figure 9. ECD patients

(a) Table summarizing observed pathological and molecular findings in brain tissue of three ECD patients with neurologic presentations. BRAF status was determined by immunohistochemical analysis and by sequencing. Neuronal loss and demyelination was determined by immunohistochemistry of neurofilament and myelin basic protein (MBP). RF: Rosenthal fiber. n/a: not applicable/no tissue available for further analysis. (b) Immunohistochemistry and immmunofluorescent analysis of ECD brain tissue for CD163, pERK and BRAF^V600E (anti-BRAFVE1 antibody). Upper panel scale bar=200µm. Lower panel scale bar=5µm. (c) Immunohistochemistry analysis of ECD brain tissue for neurofilament and MBP shows areas of myelin deficits with preserved axons in the same region. Scale bar=200µm.

Figure 1. Targeting BRAF^V600E in tissue-resident macrophages

(a, b) Breeding scheme for experimental mice and genotype distribution (n=342). (c, d) YFP expression on BM LSK, blood leukocytes and microglia from 1-month-old mice, representative of n=5 per group. (e) Proportion of YFP⁺ F4/80⁺ cells in tissues from 1-month-old mice. Circles represent individual mice. Unpaired two-tailed t-test. (f) A>T transversion encoding BRAF^V600E in YFP⁺ Kupffer cells at the Braf locus. Red and blue bars indicate forward and reverse strands. (g) Ki67 and cleaved Caspase 3 (Casp3) expression in YFP⁺ microglia from 1-month-old brains. n=5 per group. Unpaired two-tailed t-test. (h) GSEA of differentially expressed genes in YFP⁺ microglia from BRAF^VE (n=3) and littermates (n=2) mice. q-value < 0.01. (i) Heatmap representation of selected genes from (h), values are displayed as z-score. See also Extended Data Fig. 1.

Figure 2. Neurodegenerative disease in BRAF^VE mice

(a) Footprint assays. n=20 mice per group. (b) Limb-clasping reflexes in 6–8-month old mice. n=10 per group. (c) Cumulative incidence of behavioral abnormalities in BRAF^VE mice and controls. Log-rank (Mantel-Cox) test. (d) Overlap distance, and stride length of mice on control or PLX4270 diet from 1 month or 3 months of age. Average values ±s.d. for each group, 2way ANOVA, *p<0.05, **p<0.01, ***p<0.001. (e) Disease progression in mice from (d), average score excluding mice euthanized due to paralysis (†). (f) Cumulative incidence of behavioral abnormalities in mice from (d). Log-rank (Mantel-Cox) test (g) Scheme depicting microglia (Iba1) and neuronal (NeuN) densities in brain regions from BRAF^VE and BRAF^WT (n=4 per group). (h) Iba1 staining and quantitative analysis of microglial accumulation (Iba1⁺), phagocytosis (Iba1⁺/Lamp2⁺), astrogliosis (GFAP⁺), relative synapse density (determined by Synaptophysin and Homer1), neuronal loss (NeuN) and amyloid precursor protein (APP) in brainstem from 5–9 month-old BRAF^VE mice on control diet (n=4), BRAF^VE mice on PLX diet (n=4–6), and BRAF^WT (n=4). One-way ANOVA. See also Extended data Fig. 4, 6.

Figure 3. ERK activation in BRAF^V600E microglia

(a) CD68, YFP and pERK staining in spinal cord from 7-month old mice. Scale bars=500µm, 10µm for insets. n=4 per group. (b) pERK⁺ microglia in brainstem. Circles represent individual mice. One-way ANOVA. (c) ERK phosphorylation in spinal cords and brains from 6–9 month-old mice. Top: representative western blot, bottom: pERK/ERK ratio, n=5 per group. One-way ANOVA. (d) pERK expression in YFP⁺ microglia from BRAF^VE mice. n=5 per group. Scale bars=5µm. (e) Numbers of microglia from 5–9 month-old mice Circles represent individual mice. One-way ANOVA. (f) Heatmap representation of cell frequency among CD45⁺ cells in the brain. n=3 per group. (g) Ki67⁺ and cleaved Caspase 3⁺ (Casp3) expression in YFP⁺ microglia from 5–9 month-old BRAF^VE mice, n=6 per group. Unpaired two-tailed t-test. See also Extended data Fig. 7.

Figure 4. Molecular features of ERK-activated microglia, and their presence in histiocytoses patients

(a) GSEA of differentially expressed genes in YFP⁺ microglia from 6–7 month-old littermates. q-value <0.05. EMT: epithelial-mesenchymal transition. (b) Heatmap representation of selected DEG, values are displayed as z-score. (c) Validation of gene expression in YFP⁺ microglia. n=5 per group. FMO: fluorescence minus one. (d) Il1b expression in spinal cords from 6–7 month-old mice. n=3 per group. Scale bars=10µm. (e) Bioplex analysis of Il1b and Il17a in spinal cords from 6–9 month-old mice. Circles represent individual mice. Unpaired two-tailed t-test. (f) Collagen IV and collagen VI expression in spinal cord from (d). n=3 per group. Scale bars=10µm. (g) CD163, pERK and BRAF^V600E expression in ECD brain tissue. Upper panels: 400× objective, lower panels: scale bar is 10µm. (h) Quantification of pERK microglia in control (n=6) and ECD (n=3) brains. Mann-Whitney test. (i) Heatmap representation of selected genes from RNA-seq analysis of brain tissue from 5 control brains and 2 histiocytoses patients (JXG and LCH), values are displayed as z-score. q-value <0.01. See also Extended Data Fig. 9.