Early lysosomal maturation deficits in microglia triggers enhanced lysosomal activity in other brain cells of progranulin knockout mice

Abstract

Background: Heterozygous loss-of-function mutations in the progranulin gene (GRN) lead to frontotemporal lobar degeneration (FTLD) while the complete loss of progranulin (PGRN) function results in neuronal ceroid lipofuscinosis (NCL), a lysosomal storage disease. Thus the growth factor-like protein PGRN may play an important role in lysosomal degradation. In line with a potential lysosomal function, PGRN is partially localized and processed in lysosomes. In the central nervous system (CNS), PGRN is like other lysosomal proteins highly expressed in microglia, further supporting an important role in protein degradation. We have previously reported that cathepsin (Cat) D is elevated in GRN-associated FTLD patients and Grn knockout mice. However, the primary mechanism that causes impaired protein degradation and elevated CatD levels upon PGRN deficiency in NCL and FTLD remains unclear.

Methods: mRNA expression analysis of selected lysosomal hydrolases, lysosomal membrane proteins and autophagy-related genes was performed by NanoString nCounter panel. Protein expression, maturation and in vitro activity of Cat D, B and L in mouse embryonic fibroblasts (MEF) and brains of Grn knockout mice were investigated. To selectively characterize microglial and non-microglial brain cells, an acutely isolated microglia fraction using MACS microbeads (Miltenyi Biotec) conjugated with CD11b antibody and a microglia-depleted fraction were analyzed for protein expression and maturation of selected cathepsins.

Results: We demonstrate that loss of PGRN results in enhanced expression, maturation and in vitro activity of Cat D, B and L in mouse embryonic fibroblasts and brain extracts of aged Grn knockout mice. Consistent with an overall enhanced expression and activity of lysosomal proteases in brain of Grn knockout mice, we observed an age-dependent transcriptional upregulation of certain lysosomal proteases. Thus, lysosomal dysfunction is not reflected by transcriptional downregulation of lysosomal proteases but rather by the upregulation of certain lysosomal proteases in an age-dependent manner. Surprisingly, cell specific analyses identified early lysosomal deficits in microglia before enhanced cathepsin levels could be detected in other brain cells, suggesting different functional consequences on lysosomal homeostasis in microglia and other brain cells upon lack of PGRN.

Conclusions: The present study uncovers early and selective lysosomal dysfunctions in Grn knockout microglia/macrophages. Dysregulated lysosomal homeostasis in microglia might trigger compensatory lysosomal changes in other brain cells.

Keywords: Cathepsin; Frontotemporal lobar degeneration; Lysosome; Microglia; Neurodegeneration; Progranulin.

Fig. 1

Minor alterations in expression of lysosomal and autophagy-related genes in brain of Grn^−/− mice. a mRNA expression for 45 selected genes associated with the lysosome-autophagy degradation pathway [45, 46] in brain of Grn^+/+ (wt) and Grn^−/− (ko) mice at 6 and 12 months of age (for original data see Additional file 2). Genes are grouped by their function within these pathways or by their FTLD-association. Previously identified TFEB targets are labeled by red boxes [46, 47]. N = 3 mice per group, notice the low expression differences between individual mice and between 6- and 12-month-old mice. Trem2 and Atp6v0d2 were below the detection limit. b Fold change of gene expression which show at least one significand change either at 6 or 12 months of age. Data were normalized to the corresponding mean value of Grn^+/+ (wt) mice and are shown as mean ± SD. For statistical analysis the unpaired, two-tailed student’s t-test was used (n = 3) (*, p < 0.05; **, p < 0.01; ***, p < 0.001)

Fig. 2

Altered protein expression, maturation and activity of cathepsins in brain of Grn^−/− mice. a Schematic presentation of cathepsin maturation [53]. Cathepsins (Cat) are synthesized as an inactive pre-pro-form (pp), translocated into the endoplasmic reticulum (ER) by signal peptide (SP). After SP removal the pro-form (p) becomes co-transitionally modified and is transported to lysosomes predominantly via the manose-6-phosphate pathway. With increasing acidification, the pro-peptide is removed, either autocatalytically or by other enzymes, leading to an active single chain variant (sc, green). For most cathepsins this sc-variant can be further proteolytically processed to a heavy (hc) and a light chain (lc), as long as hc and lc are linked by disulfide bridges or hydrophobic interaction the double chain variant remains active (green) but will be inactivated by separation of hc and lc. Representative blots of brain lysate from 3-month-old and 20-month-old Grn^+/+ (wt) and Grn^−/− (ko) mice probed for cathepsin D (CatD) (b, d) cathepsin B (CatB) (f, h) and cathepsin L (CatL) (j, l). The molecular weight standards in kilo Daltons (kDa) are indicated on the left side of the blots. Quantification of blots for total cathepsin or maturation variants normalized to Grn^+/+ are shown as mean ± SD. N = 5 mice per genotype (b, d, f, h, j, l). Statistical significance was set at *, p < 0.05; **, p < 0.01; and ***, p < 0.001; ns, not significant using an unpaired, two-tailed student’s t-test. In vitro enzyme activity of CatD, CatB and CatL in lysates of mouse brains used for immunoblot analysis. Equal amounts of enzyme optimized brain lysates from Grn^+/+ and Grn^−/− (n = 3–5) mice were incubated with quenched fluorogenic substrate (c, e, g, i, k, m). The increase of fluorescence signal was continuously measured and for a linear turnover time period normalized to Grn^+/+ set as 100% activity, mean ± SD. Statistical significance was set at ***, p < 0.001 and ns, not significant using unpaired, two-tailed student’s t-test

Fig. 3

PGRN loss alters maturation and elevates activity of cathepsins in MEF. a CatD expression and maturation in MEF_pool Grn^+/+ (wt) and Grn^−/− (ko) shown in representative immunoblots. The pro-form CatD_p, single chain form CatD_sc, and heavy chain form CatD_hc are indicated. Bar graphs show the quantification of blots for total CatD or maturation variants normalized to Grn^+/+. b CatD activity measured as cleavage of a quenched fluorogenic substrate. The increase of fluorescence signal was continuously measured and during a linear turnover time period normalized to Grn^+/+. Note that extracts of CatD deficient MEF (Ctsd_ko) show no CatD activity and therefore confirm the specificity of the assay. c CatB expression and maturation in MEF_pool Grn^+/+ (wt) and Grn^−/− (ko) shown in representative immunoblots. The pro-form CatB_p, single chain form CatB_sc are indicated. Bar graphs show the quantification of blots for total CatB or maturation variants normalized to Grn^+/+. d CatB activity normalized to Grn^+/+. e CatL expression and maturation in MEF_pool Grn^+/+ (wt) and Grn^−/− (ko) shown in representative immunoblots. The pro-form CatL_p, single chain form CatL_sc, heavy chain form CatL_hc are indicated. Bar graphs show the quantification of blots for total CatL or maturation variants normalized to Grn^+/+. f CatL activity normalized to Grn^+/+. g PGRN deficient MEF were stably transfected with mouse PGRN (mGrn) and low PGRN expressing single cell clones (#5, #8) were analyzed for CatD, CatB and CatL in vitro activity. Notice that very low expression of PGRN (#8) lowers cathepsin activities and thereby partially rescues the phenotype of the Grn^−/− MEF, while the higher expressing clone (#5) allows a full rescue for CatB and CatL. The molecular weight standards in kilo Daltons (kDa) are indicated on the left side of the blots. All bar graphs are shown as mean ± SD. Statistical significance was set at *, p < 0.05; **, p < 0.01; ***, p < 0.001; and ****, p < 0.0001 with ns as not significant using a-f unpaired, two-tailed student’s t-test (n = 3), g one-way ANOVA with Dunnett’s post hoc test (n = 3–6)

Fig. 4

Elevated lysosomal activity results in enhanced fast protein degradation in Grn^−/−MEF. a Turnover of ³⁵S-methionine radiolabeled proteins. MEF at 70–80% confluency were metabolically pulse-labeled with ³⁵S-methionine/cysteine for 1 h, followed by indicated chase periods. Radioactivity of ³⁵S- labeled proteins at chase time point 0 h was set to 100% and remaining radioactive-labeled proteins at later chase points were normalized to the initial radioactivity at time point 0 h. For statistical analysis the unpaired, two-tailed student’s t-test was used to compare Grn^−/− to Grn^+/+ MEF (n = 5), (*, p < 0.05). b APP holoprotein (APP_holo) and C-terminal fragments (APP_CTF) detected by immunoblotting of MEF_pool Grn^+/+ (wt) and Grn^−/− (ko) lysates. The molecular weight standards in kilo Daltons (kDa) are indicated on the left side of the blots. Bar graphs show the quantification of the blots for APP_holo and APP_CTF normalized to Grn ^+/+ (wt) as mean ± SD. c Quantification of App mRNA of MEF_pool Grn^−/− (ko) normalized to Grn^+/+ (wt) as mean ± SD. d Ubiquitin (Ub), p62, LC3-I and LC3-II detected by immunoblotting of MEF_pool Grn^+/+ (wt) and Grn^−/− (ko) lysates. Bar graphs show the quantification of the blots normalized to wt as mean ± SD. b-d For statistical analysis the unpaired, two-tailed student’s t-test was used to compare ko to wt cells (n = 3) (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ns, not significant)

Fig. 5

Cathepsin maturation is selectively impaired in Grn^−/− microglia. a Schematic representation of the brain cell isolation using MACS Technology (Miltenyi Biotec) b PGRN expression in acutely isolated microglia, astrocytes and neurons enriched fractions of 4- month-old wt mice detected by immunoblotting. The identity of neural cell types was verified by detection of Iba1 for microglia, GFAP for astrocytes and Tuj1 for neurons. c-i Cathepsin expression and maturation in the CD11b-positive, microglia enriched, fraction and the CD11b-negative, microglia depleted cellular fraction isolated form cortices of brain from 3-month-old Grn^+/+ (wt) and Grn^−/− (ko) mice. Representative immunoblots for the cathepsin expression of CatD (c, d), CatB (e, f), CatL (g, h) and for CatS (i) (only microglia enriched fraction). The molecular weight standards in kilo Daltons (kDa) are indicated on the left side of all immunoblots. A dotted line in the blot indicates that samples of heterozygous mice were cut out, but all samples were loaded on one gel. Quantification of immunoblots for total cathepsin or maturation variants pro-form (p), single chain (sc) and heavy chain (hc) normalized to wt are shown as mean ± SD. For statistical analysis the unpaired, two-tailed student’s t-test was used to compare ko to wt mice (n = 3–5) (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ns, not significant) (c-i)

Fig. 6

Cathepsins are elevated in the microglia depleted cell fraction of aged Grn^−/− mice. a-d Cathepsin expression and maturation in the CD11b-positive, microglia enriched, fraction and the CD11b-negative, microglia depleted cellular fraction isolated form cortices of brains from 12-month-old Grn^+/+ (wt) and Grn^−/− (ko) mice. Representative immunoblots for the expression of CatD (a, b), CatB (c, d). The molecular weight standards in kilo Daltons (kDa) are indicated on the left side of all immunoblots. A dotted line in the blot indicates that samples of heterozygous mice were cut out, but all samples were loaded on one gel. Quantification of immunoblots for total cathepsin or pro-form (p), single chain (sc) and heavy chain (hc) normalized to wt are shown as mean ± SD. For statistical analysis the unpaired, two-tailed student’s t-test was used to compare ko to wt mice (n = 3–5) (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ns, not significant) (a-d)

Fig. 7

Enhanced accumulation of lysosomal proteins in microglia of 3-month-old Grn^−/− mice. Representative blots of LAMP1 (a, b) and saposin D (SapD) (c, d) in the CD11b-positive, microglia enriched fraction and the CD11b-negative, microglia depleted fraction isolated form brain cortices of 3-month-old Grn^+/+ (wt) and Grn^−/− (ko) mice. A dotted line in the blot indicates that samples of heterozygous mice were cut out, but all samples were loaded on one gel. Data were normalized to the corresponding mean value of wt mice and are shown as mean ± SD. For statistical analysis the unpaired, two-tailed student’s t-test was used to compare ko to wt mice (n = 4–5) (*, p < 0.05; **, p < 0.01; ***, p < 0.001; ns, not significant)

Fig. 8

Schematic summary of differential effects of PGRN deficiency on brain cell types. Lysosomal impairment in microglia observed in young Grn^−/− mice result in enhanced lysosomal activity in non-microglial brain cells like neurons in aged Grn^−/− mice. Basal lysosomal functions of microglia and neurons are indicated in orange, changes to lower activity (yellow) in microglia or higher activity (red) in neurons or other non-microglial brain cells are caused by PGRN deficiency and aging