Anti-Excitotoxic Effects of N-Butylidenephthalide Revealed by Chemically Insulted Purkinje Progenitor Cells Derived from SCA3 iPSCs

Abstract

Spinocerebellar ataxia type 3 (SCA3) is characterized by the over-repetitive CAG codon in the ataxin-3 gene (ATXN3), which encodes the mutant ATXN3 protein. The pathological defects of SCA3 such as the impaired aggresomes, autophagy, and the proteasome have been reported previously. To date, no effective treatment is available for SCA3 disease. This study aimed to study anti-excitotoxic effects of n-butylidenephthalide by chemically insulted Purkinje progenitor cells derived from SCA3 iPSCs. We successfully generated Purkinje progenitor cells (PPs) from SCA3 patient-derived iPSCs. The PPs, expressing both neural and Purkinje progenitor's markers, were acquired after 35 days of differentiation. In comparison with the PPs derived from control iPSCs, SCA3 iPSCs-derived PPs were more sensitive to the excitotoxicity induced by quinolinic acid (QA). The observations of QA-treated SCA3 PPs showing neural degeneration including neurite shrinkage and cell number decrease could be used to quickly and efficiently identify drug candidates. Given that the QA-induced neural cell death of SCA3 PPs was established, the activity of calpain in SCA3 PPs was revealed. Furthermore, the expression of cleaved poly (ADP-ribose) polymerase 1 (PARP1), a marker of apoptotic pathway, and the accumulation of ATXN3 proteolytic fragments were observed. When SCA3 PPs were treated with n-butylidenephthalide (n-BP), upregulated expression of calpain 2 and concurrent decreased level of calpastatin could be reversed, and the overall calpain activity was accordingly suppressed. Such findings reveal that n-BP could not only inhibit the cleavage of ATXN3 but also protect the QA-induced excitotoxicity from the Purkinje progenitor loss.

Keywords: ATXN3; Purkinje progenitor; SCA3; iPSCs; n-butylidenephthalide; quinolinic acid.

Figure 1

The differentiation of control and SCA3 iPSCs into Purkinje progenitor cells (PPs). The control and SCA3 iPSCs were differentiated with gfCDM as described in Materials and Methods. At 35 days of differentiation, both control and SCA3 iPSCs exhibited similar characteristics of PPs when compared with a previous study [29]; (A) morphological patterns of neural markers β-tubulin III and MAP2, and Purkinje lineage markers KIRREL2, as reflected by ICC staining. Scale bar = 100 μm; (B) the flow cytometry examination of neural marker β-tubulin III and Purkinje lineage marker KIRREL2 to determine the differentiation yield.

Figure 2

The SCA3 PPs exhibited pathological characteristics after the treatment of QA. The control and SCA3 PPs were treated with 1 μM of QA for 12 h. (A) The representative morphological observation and the quantification of neurite length analysis of control- and SCA3 PPs from three independent images. Scale bar = 100 μm. Data are presented as the means ± standard deviation; t(28) = 6.945, p < 0.01, for Control PPs vs. Control PPs with QA; t(28) = 14.16, p < 0.01, for Control PPs vs. SCA3 PPs with QA; (B) protein analysis of PARP1 in cell lysates of control and SCA3 PPs by immunoblots under the treatment of QA in the presence of variable concentrations of n-BP. Statistics refer to lane 1 to lane 3 of Figure 3D. Data are presented as the means ± standard deviation. t(4) = 5.876, p < 0.01, for SCA3 PPs vs. SCA3 PPs with QA; (C) the representative image of ATXN3 nucleus co-localization assessed by ICC assay. The co-localization is indicated using white arrowheads and the quantification data from three independent images are presented as the means ± standard deviation. t(4) = 13.74, p < 0.01, for SCA3 PPs vs. SCA3 PPs with QA; Scale bar = 100 μm; (D) protein analysis of wild-type, mutant ATXN3 and its proteolytic fragment in cell lysates of control and SCA3 PPs by immunoblots. Sol, soluble protein; Insol, insoluble protein form. **, p < 0.01.

Figure 3

The treatment of n-BP rescued the progression of QA-induced excitotoxicity in SCA3 PPs. The control and/or SCA3 PPs were treated with or without QA (1 μM) in the presence of n-BP (0, 10, 20, and 40 μg/mL) for 12 h. (A) Images of QA-treated SCA3 PPs in the presence of BP (40 μg/mL). (B) Confocal images of poly Q localization in BP (40 μg/mL)-treated SCA3 PPs in the presence of QA (1 um). When compared with Figure 2C, 40 μg/mL of n-BP prevented the amounts of colocalized polyQ in the cell nucleus of PPs. The distribution of polyQ in the nucleus is indicated by white arrowheads. Scale bar = 100 μm; (C) protein analysis of wild-type, mutant ATXN3 and its proteolytic fragment in cell lysates of control and SCA3 PPs, as assessed by immunoblots. Soluble, soluble protein; Insoluble, insoluble protein; (D,E) representative immunoblot of control and cleaved PARP1 in cell lysates of control and SCA3 PPs, as assessed by immunoblots. The HDAC2 were used as loading controls. The quantification from three independent images is presented as the means ± standard deviation. t(4) = 2.824, p < 0.05, for lane 1 vs. lane 2; t(4) = 5.613, p < 0.01, for lane 1 vs. lane 3; t(4) = 3.522, p < 0.05, for lane 1 vs. lane 4; (F) the quantification result of calcium concentration in SCA3 PPs, as assessed by Fura-2 indicator. Cells were treated with QA (1 μM) in the presence of n-BP (0, 10, 20, and 40 μg/mL). Data are presented as the means ± standard deviation. t(4) = 2.95, p < 0.05; (G) the calpain activity in control and SCA3 PPs cell lysates, as assessed by ELISA assay. The quantification results from three independent replicates are presented as the means ± standard deviation. t(4) = 3.27, p < 0.05, for lane 3 vs. lane 4; t(4) = 17.55, p < 0.01, for lane 3 vs. lane 5; t(4) = 9.019, p < 0.01, for lane 3 vs. lane 6. (H) Protein analysis of calpain 1, calpain 2 and calpastatin in cell lysates of SCA3 PPs, as assessed by immunoblots. Cells were treated with or without QA (1 μM) in the presence of n-BP (0, 10, 20, and 40 μg/mL). *, p < 0.05; **, p < 0.01.