4-Phenylbutyrate ameliorates apoptotic neural cell death in Down syndrome by reducing protein aggregates

Abstract

Individuals with Down syndrome (DS) commonly show unique pathological phenotypes throughout their life span. Besides the specific effects of dosage-sensitive genes on chromosome 21, recent studies have demonstrated that the gain of a chromosome exerts an adverse impact on cell physiology, regardless of the karyotype. Although dysregulated transcription and perturbed protein homeostasis are observed in common in human fibroblasts with trisomy 21, 18, and 13, whether and how this aneuploidy-associated stress acts on other cell lineages and affects the pathophysiology are unknown. Here, we investigated cellular stress responses in human trisomy 21 and 13 neurons differentiated from patient-derived induced pluripotent stem cells. Neurons of both trisomies showed increased vulnerability to apoptotic cell death, accompanied by dysregulated protein homeostasis and upregulation of the endoplasmic reticulum stress pathway. In addition, misfolded protein aggregates, comprising various types of neurodegenerative disease-related proteins, were abnormally accumulated in trisomic neurons. Intriguingly, treatment with sodium 4-phenylbutyrate, a chemical chaperone, successfully decreased the formation of protein aggregates and prevented the progression of cell apoptosis in trisomic neurons. These results suggest that aneuploidy-associated stress might be a therapeutic target for the neurodegenerative phenotypes in DS.

Figure 1

Establishment of trisomy-specific iPSCs and early enhanced gliogenesis in trisomy 21 NPCs. (a) Schematic diagram showing the reprogramming processes of diploid control, trisomy 21, and trisomy 13 iPSCs. (b) iPSCs generated from diploid, trisomy 21, and trisomy 13 patients showed ESC-like morphology (phase image). NPCs differentiated from iPSCs expressed the NPC markers SOX2 and PAX6. Neurons differentiated from NPCs after 3 months expressed the pan-neural marker MAP2 and cortical neural marker TBR1. Scale bar = 200 μm. (c) Representative image of immunocytochemistry for a neural marker (TUJ1), glial marker (GFAP), and NPC marker (SOX1) after 3 weeks of differentiation from NPCs. Scale bar = 200 μm. (d) Quantification of the ratio of each marker-positive cell/Hoechst 33342-positive cell after 3 weeks of differentiation from NPCs. Data are presented as mean ± SEM. n = 3 per clone.

Figure 2

Perturbed protein homeostasis in human iPSC-derived NPCs in trisomy syndromes. (a) Relative protein synthesis per cell in NPCs. Data are presented as mean ± SEM normalized to the average of Dip #1 clone. *P < 0.05, n = 4–6 per clone. (b) Relative aggregate intensity per cell in NPCs. Data are presented as mean ± SEM normalized to the average of Dip #1 clone. *P < 0.05, n = 4–6 per clone. (c) Representative images showing immunocytochemical staining for aggregates (red); p62 (green); and merged images of aggregates, p62, and Hoechst 33342 in NPCs. Scale bar = 20 μm. (d) Western blot of ubiquitinylated (Ub) proteins and p62 in NPCs, using anti-Ub-protein and anti-p62 antibodies. Full immunoblot images of them are shown in Figure S2. (e, f) Quantification of the western blot bands of the Ub proteins (e) and p62 (f). β-Actin, loading control. Data are presented as mean ± SEM normalized to the average of Dip #1 clone. *P < 0.05, n = 4 per clone.

Figure 3

Neurogenesis with Ngn2-inducible iPSCs and decreased cell survival in trisomic neurons. (a) NGN2 under the control of the tetracycline operator was introduced into iPSCs using the piggyBac vector. Neurons were generated from iPSCs within 14 days. Representative images of neurons from Dip-NGN2-iPSCs and the construct of the NGN2 inducible piggyBac vector are shown. Scale bar = 200 μm. (b) Representative images showing immunocytochemical staining of TUJ1 and NeuN on days 7 and 14 in diploid control and trisomy neurons. Scale bar = 200 μm. (c) Quantitative data of the NeuN-positive cells on days 7 and 14. Data are presented as mean ± SEM normalized to the average of diploid clones on day 7. *P < 0.05, n = 4–6 per clone. (d) Representative images showing immunocytochemical staining for cleaved caspase 3 (red), TUJ1 (green), and Hoechst 33342 (blue) in NGN2 neurons on day 14. Scale bar = 100 μm. (e) Quantification of relative cleaved caspase 3 intensity in NGN2 neurons on day 14. Data are presented as mean ± SEM normalized to the average of diploid clones. *P < 0.05, **P < 0.001, n = 4 per clone.

Figure 4

Aggregation of neurodegenerative disorder-associated proteins in trisomic neurons. (a) Representative images showing immunocytochemical staining for aggregates (red), p62 (green), and MAP2 (white) in NGN2-neurons on day 14. Scale bar = 20 μm. (b) Relative aggregate intensity per cell in NGN2-neurons (D14). *P < 0.05, **P < 0.001, n = 5–13 per clone. (c) PLA to detect the interaction between p62 and α-Syn, Parkin, or Htt in NGN2-neurons. PLA signals are seen as red dots. Hoechst 33342 counterstaining appears in blue. Scale bar = 20 μm. (d) Quantification of PLA signals (red dots) per cell. Data are presented as mean ± SEM. *P < 0.05, n = 3–6 per clone.

Figure 5

Elevated ER stress in trisomic neurons, which was ameliorated by 4-phenylbutyric acid. (a) Representative images showing immunocytochemical staining for GRP78 (BIP) (red) and TUJ1 (green) in diploid and trisomic NGN2-neurons treated with vehicle or 4-PBA. Hoechst 33342 (blue), counterstaining. Scale bar = 200 μm. (b) Quantification of the relative intensity of GRP78 (BIP) in NGN2-neurons treated with vehicle or 4-PBA. Data are presented as mean ± SEM normalized to the average of diploid clones. *P < 0.05, n = 4 per clone. (c) Representative images showing immunocytochemical staining for GAD154 (CHOP) (red) and TUJ1 (green) in diploid and trisomic NGN2-neurons treated with vehicle or 4-PBA. Hoechst 33342 (blue), counterstaining. Scale bar = 200 μm. (d) Quantification of the relative intensity of GAD153 (CHOP) in NGN2-neurons treated with vehicle or 4-PBA. Data are presented as mean ± SEM normalized to the average of diploid clones. *P < 0.05, n = 4 per clone.

Figure 6

4-PBA ameliorated aggregation and neuronal loss in trisomic neurons. (a) Representative images showing immunocytochemical staining for aggregates (red) in diploid and trisomic NGN2-neurons treated with vehicle or 4-PBA. CellMask (green) and Hoechst 33342 (blue), counterstaining. Scale bar = 50 μm. (b) Quantification of relative aggregate intensity per cell in NGN2-neurons on day 14. Data are presented as mean ± SEM normalized to the average of diploid clones. *P < 0.05, n = 5–7 per clone. (c) Quantification of relative aggregate intensity per cell in diploid and trisomic NGN2-neurons treated with vehicle or HDAC inhibitors on day 14. Data are presented as the mean ± SEM normalised to the average of diploid clones treated with vehicle. *P < 0.05, n = 3–6 per clone. (d) Quantification of relative aggregate intensity per cell in diploid and trisomic NGN2-neurons treated with vehicle, or 4-PBA with or without 3-MA. Data are presented as the mean ± SEM normalised to the average of diploid clones treated with vehicle. *P < 0.05, n = 3–6 per clone. (e) Representative images showing immunocytochemical staining of TUJ1 and cleaved caspase 3 in diploid and trisomic NGN2-neurons treated with vehicle or 4-PBA. Scale bar = 200 μm. (f) Quantification of relative cleaved caspase 3 intensity in NGN2-neurons treated with vehicle or 4-PBA. Data are presented as mean ± SEM normalized to the average of diploid clones. *P < 0.05, n = 10–13 per clone. 4-PBA, sodium 4-phenylbutyrate; VPA, valproic acid; 3-MA, 3-methyladenine.