Differentiation Drives Widespread Rewiring of the Neural Stem Cell Chaperone Network

Abstract

Neural stem and progenitor cells (NSPCs) are critical for continued cellular replacement in the adult brain. Lifelong maintenance of a functional NSPC pool necessitates stringent mechanisms to preserve a pristine proteome. We find that the NSPC chaperone network robustly maintains misfolded protein solubility and stress resilience through high levels of the ATP-dependent chaperonin TRiC/CCT. Strikingly, NSPC differentiation rewires the cellular chaperone network, reducing TRiC/CCT levels and inducing those of the ATP-independent small heat shock proteins (sHSPs). This switches the proteostasis strategy in neural progeny cells to promote sequestration of misfolded proteins into protective inclusions. The chaperone network of NSPCs is more effective than that of differentiated cells, leading to improved management of proteotoxic stress and amyloidogenic proteins. However, NSPC proteostasis is impaired by brain aging. The less efficient chaperone network of differentiated neural progeny may contribute to their enhanced susceptibility to neurodegenerative diseases characterized by aberrant protein misfolding and aggregation.

Keywords: CRYAB; HSPB5; NSPC; TRiC/CCT; aging; neural stem cells; neurodegeneration; protein aggregation; protein quality control; proteostasis.

Figure 1.. Robust proteostasis ability of NSPCs vs. differentiated lineages

A, Regenerative NSPCs, isolated from the SVZ and SGZ stem cell niches of the young adult (3 month old) mouse brain, were differentiated into three main neural lineages (i.e. astrocytes, neurons, oligodendrocytes). B, Stress resilience of NSPCs and their differentiated progeny, determined upon exposure to (i) oxidative stress ([H₂O₂]) and (ii) heat stress. Viability was normalized for basal conditions. Basal unstressed conditions yield lower MTT reduction into formazan by differentiated progeny relative to NSPCs (metabolic activity of differentiated cells was attenuated by 49 ± 3% compared to NSPCs (Figure 5E)). n.s., non-significant, *** p < 0.001, **** p < 0.0001. C, (i) Proteostat fluorescence and nuclear DNA (Hoechst). Scale bars: 10 μm; traces: cell outlines. Insets: Magnification of Proteostat⁺ deposits. (ii) Quantified Proteostat fluorescence (mean ± SD). **** p < 0.0001. D, (i) Proteostat and Hoechst in NSPCs isolated from brains of 3, 12, and 24 month old mice and differentiated progeny cells. Insets: Magnification of Proteostat⁺ deposits. (ii) Total Proteostat fluorescence (mean ± SD; quantified by fluorimeter). Proteostat fluorescence of young NSPCs is set at 1. * p < 0.05; ** p < 0.01; n.s., non-significant. E, Relative cell viability (mean ± SD) of young and old NSPCs under basal and oxidative stress conditions. Viability for young NSPCs at basal conditions is set at 1. ** p < 0.01; *** p < 0.005; n.s. non-significant. F, (i) Schematic overview of SGZ region in the dentate gyrus (DG) of the hippocampus. GCL: granular cell layer. (ii) Proteostat⁺ deposits in the SGZ (red marked area in i) of 3, 12, and 24 month old mouse brains. Bottom row: Proteostat fluorescence signal masks. Scale bars: 50 μm. (iii) Relative Proteostat fluorescence in hippocampal brain sections (whole frame; mean ± SD). Fluorescence at 3 months is set to 1. * p < 0.05, ** p < 0.01. Insets: Magnification of Proteostat⁺ depositions. See Figure S1D for additional ages. G, H, Sox2⁺ (G) or ubiquitin⁺ (H)/Proteostat⁺ inclusions in the DG of the aged brain. Scale bar: 50 μm. Insets: Magnification of Proteostat⁺ deposits.

Figure 2.. High capacity of NSPC to maintain proteome integrity

A, (i) HttQ97-GFP, transfected into NSPCs and maintained under self-renewal or differentiation conditions, and DNA (Hoechst). Traces: cell outlines; arrows: HttQ97 aggregates. (ii) Quantification (mean ± SD) of cells with diffusely vs. aggregated HttQ97-GFP (≥ 150 cells/condition/experiment). B, HttQ97-GFP, DNA (Hoechst), and specific markers for differentiated lineages. Traces: cell outlines. Insets: Magnification of perinuclear aggregation. C, Dead (7AAD⁺) cells (mean ± SD) in NSPC and differentiated HttQ97-GFP^- (Ctrl) and HttQ97-GFP⁺ cell populations. n.s., non-significant, * p < 0.05, ** p < 0.01. D, AR-Q113, VHL^L158P, and SOD1^G37R-GFP fusion proteins and DNA (Hoechst) in NSPCs and differentiated progeny. Traces: cell outlines; arrows: aggregates. Insets: Magnification of inclusion. Quantification (mean ± SD) of diffusely localized vs. aggregated proteins (≥ 100 cells/condition/experiment). E, HttQ97-GFP in control and MMC-treated (1μg/ml, 24 h) NSPCs. Scale bars: 10 μm. Quantification (mean ± SD) of diffuse vs. aggregated HttQ97-GFP (≥ 60 cells/condition/experiment). See also Figure S2E. F, (i) Ub-R-eGFP and mCherry-empty vector control plasmid (transfection control) in NSPCs and differentiated progeny. UPS-mediated clearance was determined at basal conditions and upon proteasomal blockage (+ MG132; 15 μM, 4 h) by relative quantification of GFP⁺/mCherry⁺ cells. (ii) Graph (mean ± SD) quantifies Ub-R-eGFP as well as Ub-G76V-eGFP and their stable counterpart Ub-M-eGFP (≥ 100 cells/condition). ** p < 0.01; **** p < 0.0001; n.s., non-significant. G, HttQ97-GFP in NSPCs at vehicle control-treated (DMSO) or proteasome-inhibited (MG132; 15 μM, 4 h prior to fixation) conditions. Graph (mean ± SD) quantifies diffusely localized vs. aggregated HttQ97-GFP (≥ 60 transfected cells/condition/experiment). n.s., non-significant. H, i) Amyloid-like aggregates (Proteostat⁺) and DNA (Hoechst) in NSPCs and differentiated progeny at control (DMSO) or proteasome-inhibited conditions (MG132; 15 μM, 4 h prior to fixation). Insets: Magnification of Proteostat⁺ deposits. Quantification (mean ± SD) shows Proteostat fluorescence. ** p < 0.01; n.s. non-significant. See also Figure S2H. ii) Immunoblot for Ubiquitin in NSPCs at basal or proteasome-inhibited conditions (MG132; 15 μM, 4 h prior to fixation). Gapdh serves as loading control.

Figure 3.. NSPC differentiation rewires cellular chaperone circuits

A, (i) The robust NSPC proteostasis capacity (i.e. high stress resistance, ability to maintain proteome solubility) is drastically impaired upon differentiation. The differential transcriptomes of NSPCs and their progeny was compared by RNA-Seq. (ii) Proof of concept: marked enrichment of transcripts specifically expressed in the differentiated lineages following NSPC differentiation. B, Cellular proteostasis relies on the balance of protein synthesis, folding, and clearance, mediated by molecular chaperones and degradation pathways acting during protein maturation. Volcano plots show differentially expressed ribosomal, proteasomal, autophagy or chaperome genes in differentiated progeny compared to NSPCs. X-axis: log2 (fold change); y-axis: –log10 of q-value. Triangles indicate q-values below 10⁻³⁰. C, (i) Bulk rates of protein synthesis, determined by O-propargyl-puromycin (OPP) incorporation into newly translated proteins. (ii) Quantification (mean ± SD) of the relative protein synthesis rates in NSPCs and differentiated progeny. Value of NSPCs was arbitrary set at 1. p = 0.041. See Figure S3B for FACS plots. D, Volcano plots of differentially expressed genes of all five major cytosolic chaperone classes in differentiated cells compared to NSPCs. X-axis: log2 (fold change); y-axis: –log10 of q-value. E, (i) Immunoblots for indicated chaperones for three independent samples of NSPCs and early, intermediate, and fully differentiated cells. Gapdh serves as loading control; (ii) Heat map of immunoblot quantifications. F, Eukaryotic cells contain different chaperone classes guiding folding of non-native polypeptides or effecting quality control through protein refolding, degradation, and sequestration. We postulate a model in which high TRiC in NSPCs ensures maintenance of a soluble proteome, by TRiC binding to nascent and non-native polypeptides to promote their folding into native proteins or clearance by proteasome or autophagy pathways. NSPC differentiation remodels the cellular chaperone networks and induces sHSP expression. This may change the cellular proteostasis strategy and promote stabilization of non-native proteins by spatial sequestration into defined protective inclusions.

Figure 4.. Differential expression of TRiC and HspB5 in NSPCs and differentiated progeny

A, B, Immunoblots for all 8 Cct subunits of the TRiC chaperonin complex (A) or various sHsps (B) for three independent samples of NSPCs and progeny cells (differentiated for 7 days (A) or 2, 4, and 7 days (B). Gapdh (A) and actin (B) serve as loading controls. C, Cct1 and Sox2 and Nestin (NSPC markers) in NSPCs. Scale bar: 10 μm. D, TRiC (i; Cct1 and Cct2 subunits) and HspB5 (ii), and Hoechst in NSPCs and differentiated cells. Traces: cell outlines. E, HspB5 with Dcx (neuronal marker) in in vitro differentiated cells (single channels in Figure S4G). F, Mouse brain sections of the SGZ (i) and SVZ (ii) NSPC niches, immunostained for Sox2 (NSPC marker) and TRiC (Cct1). Scale bars: 50 μm. Insets: Magnification of Sox2⁺/Cct1⁺ cells at indicated areas. G, Mouse SGZ brain sections immunostained for Sox2 (NSPC marker) and HspB5. Scale bar: 50 μm. Insets: Magnification of Sox2⁺/HspB5^- cells at indicated area.

Figure 5.. TRiC in NSPCs ensures proteome solubility and cellular fitness

A, (i) Immunoblot analyses of TRiC (Cct1), NSPC markers (Nestin, Sox2), and neuronal marker (NeuN). Gapdh serves as loading control. (ii) Epifluorescent images of HttQ97-GFP in NSPCs or during NSPC differentiation (at day 1, 2, 4, or 7). Traces: cell outlines; arrows: aggregates. B, Cct1 immunoblot of control (siCtrl)- and siCct1-targeted NSPCs as well as differentiated cells. Gapdh serves as loading control. Expression in siCtrl-treated NSPCs is set at 1. Degree of Cct1 knockdown is representative for all experiments presented. C, Aggregation of endogenous proteins in NSPC treated with control or a pool of siRNAs targeting Cct1; assessed by confocal imaging of Proteostat and DNA (Hoechst; (i). Insets: Magnification of Proteostat⁺ depositions at indicated areas); and quantified by fluorescence (mean ± SD; ii). Proteostat fluorescence of siCtrl-targeted NSPCs is set at 1. ** p < 0.01; **** p < 0.001; n.s., non-significant. A similar increase in Proteostat⁺ aggregates was observed in NSPCs treated with single siRNAs targeting Cct1 (Figure S5C) or a siCct2 SMARTPool (Figure S5F). D, siRNA-mediated Cct1 depletion from NSPCs impairs NSPC proteome solubility. Relative Coomassie signal, determined by densitometry, is displayed for each lane. E, MTT assay on control (siCtrl) and siCct1-treated NSPCs prior to or after differentiation (represented as mean ± SD). Metabolic activity of siCtrl-targeted NSPCs is set at 1. ** p < 0.01; n.s., non-significant. F, HttQ97-GFP and siGloRED (indicator of siRNA transfection) in control (siCtrl) and siCct1-treated NSPCs and differentiated cells. Arrows: aggregates. Insets: Magnification of HttQ97-GFP inclusions. Quantification (mean ± SD) of diffusely localized vs. aggregated HttQ97-GFP in double-transfected siCtrl/siGloRED⁺- and siCct1/siGloRED⁺-treated NSPCs (≥ 40 cells/condition/experiment). Similar results were obtained upon TRiC depletion using single siRNAs targeting Cct1 (Figure S5D). G, Viability of control (siCtrl) and siCct1-treated NSPCs at basal conditions or upon exogenous expression of HttQ97-GFP (represented as mean ± SD). siCtrl-targeted NSPCs is set at 1. * p < 0.05; ** p < 0.01; n.s., non-significant. Similar results were obtained in NSPCs treated with a siCct2 SMARTPool (Figure S5G).

Figure 6.. HSPB5 promotes sequestration of misfolded proteins in protective inclusions in differentiated cells

A, Relative viability (mean ± SD) of differentiated cells, treated with control siRNAs or a siHspB5 SMARTPool, at basal and oxidative stress (16 h, 10 or 150 μM H₂O₂) conditions. Viability of untreated cells for each condition is set at 1 (i). Viability of siCtrl- and siHspB5-treated cells at basal conditions is set at 1 (ii). * p < 0.05; ** p < 0.01; n.s. non-significant. B, Aggregation of endogenous proteins quantified by Proteostat fluorescence (mean ± SD) in differentiated progeny of control (siCtrl) and siHspB5-treated NSPCs. Fluorescence of siCtrl-targeted cells is set at 1. ** p < 0.01. Similar results were obtained upon HspB5 depletion using single siRNAs (Figure S6C). C, HspB5 and HttQ97-GFP inclusions in differentiated cells (single channels in Figure S6D). Trace: cell outline; arrow: HspB5 localization to HttQ97-GFP aggregate. D, HttQ97-GFP, siGloRED, and Hoechst in differentiated progeny of siCtrl- and siHspB5-treated NSPCs. Quantification (mean ± SD) shows diffusely localized vs. aggregated HttQ97-GFP (≥ 50 siRNA-targeted NSPCs and differentiated cells/experiment). Similar results were obtained in differentiated cells treated with various single siRNAs targeting HspB5 (Figure S6E). E, Relative viability (mean ± SD) of control (siCtrl) and siHspB5-treated differentiated cells under basal and proteotoxic stress (exogenous HttQ97-GFP) conditions. Viability of siCtrl-treated cells for each condition is set at 1. * p < 0.05; ** p < 0.01; n.s. non-significant. F, NSPCs exogenously expressing HttQ97-GFP and EV or HA-HSPB5. DNA stained with Hoechst. Arrows: HttQ97-GFP inclusions. Inset: Magnification of HA-HSPB5 co-localization at HttQ97-GFP inclusion at indicated area. Quantification (mean ± SD) shows diffuse vs. aggregated HttQ97-GFP (≥ 50 NSPCs/condition/experiment). G, Fluorescent quantification (mean ± SD) of amyloid-like aggregates in control (EV) and HA-HSPB5 expressing NSPCs. p = 0.0251.

Figure 7.. Aging attenuates TRiC expression in NSPCs as well as the overall brain

A, Schematic overview of coronal brain section. Areas of images shown in B and C are indicated. B, Proteostat⁺ amyloid deposits in brain sections from young and old mice in Dentate Gyrus, Striatum, and Cortex. See Figure S7A for stitched images of full brain sections. C, (i) Quantification (mean ± SD) of Cct1 fluorescence in Sox2⁺ cells (whole frame; 6 mice/age group). Cct1 fluorescence in young mouse brain is set at 1. p = 0.0006. See Figure 7B for stitched images of Cct1/Sox2/Neun stained brain sections; (ii) Cct1 in brain sections from young and old mice in Dentate Gyrus, Striatum, and Cortex. D, TRiC (Cct1) and Hoechst in young and old isolated NSPCs. Insets: Magnification of cellular Cct1. Quantification (mean ± SD) shows Cct1 fluorescence in young and old NSPCs. Fluorescence of young NSPCs is set at 1. p = 0.0029. E, Cct1 expression (mean ± SD) in young, middle-aged, and old NSPCs. mRNA abundance in young NSPCs is set at 1. * p < 0.05; ** p < 0.01. F, Differentiation drastically remodels the robust chaperone network and proteostasis logic of NSPCs to enhance cellular fitness. NSPCs express high levels of TRiC, which is central to maintain a soluble proteome and contributes to the NSPC ability to effective manage non-native proteins and withstand proteotoxic stress. Differentiation drastically remodels the proteostasis logic. In contrast to NSPCs, the load of misfolded proteins is markedly increased in differentiated cells; most likely due to a marked attenuation of TRiC expression, protein synthesis rates, and UPS clearance activities. sHSPs are induced during NSPC differentiation. This promotes spatial sequestration of misfolded proteins into defined protective inclusions to clear the cellular milieu of potentially harmful, misfolded proteins.