Alzheimer's disease presenilin-1 exon 9 deletion and L250S mutations sensitiZe SH-SY5Y neuroblastoma cells to hyperosmotic stress-induced apoptosis

Abstract

Mutations in the presenilin-1 (PS1) and presenilin-2 (PS2) genes account for the majority of early-onset familial Alzheimer's disease cases. Recent studies suggest that presenilin gene mutations predispose cells to apoptosis by mechanisms involving altered calcium homeostasis and oxidative damage. In the present study, we determined whether PS1 mutations also sensitize cells to hyperosmotic stress-induced apoptosis. For this, we established SH-SY5Y neuroblastoma cell lines stably transfected with wild-type PS1 or either the PS1 exon 9 deletion (ΔE9) or PS1 L250S mutants. Cultured cells were exposed to an overnight (17 h) serum deprivation, followed by a 30 min treatment with either 20 mM glucose, 10 nM insulin-like growth factor-1 or 20 mM glucose +10 nM insulin-like growth factor-1. Cells were then cultured for a further 3, 6 or 24 h and stained for apoptotic condensed nuclei using propidium iodide. Confirmation that cells were undergoing an active apoptotic process was achieved by labelling of DNA strand breaks using the terminal dUTP nick end labelling (TUNEL) technique. We also determined cell viability using 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction. Propidium iodide staining revealed that all cell lines and controls showed an increased number of apoptotic cells appearing with condensed nuclei at 24 h compared with 6 h and 3 h. High glucose-induced hyperosmotic stress resulted in significantly more apoptotic cells in the PS1 ΔE9 and PS1 L250S mutation cell lines at 24 h, compared with the wild-type PS1 lines ( P<0.001, ANOVA for both comparisons). Mean values (±S.D.) for the percentage number of apoptotic cells at 24 h following high glucose treatment were 16.1±3.5%, 26.7±5.5% and 31.0±5.7% for the wild-type PS1, PS1 ΔE9 and PS1 L250S lines, respectively. The pro-apoptotic effects of high glucose treatment were reversed by 10 nM insulin-like growth factor-1, although to a lesser extent in the mutation cell lines (5.8±2.4%, 15.2±7.3% and 13.2±2.0% for the wild-type PS1, PS1 ΔE9 ( P<0.01 for comparison with wild-type PS1) and PS1 L250S ( P<0.01 for comparison with wild-type PS1) transfected lines, respectively. TUNEL labelling of cells at 24 h following treatment gave essentially the same results pattern as obtained using propidium iodide. The percentage number of apoptotic cells with DNA strand breaks (means±S.D.) following high glucose treatment was 15.4±2.6% for the wild-type PS1, 26.8±3.2% for the PS1 ΔE9 ( P<0.001 for comparison with wild-type PS1) and 29.7±6.1% for the PS1 L250S transfected lines ( P<0.001 for comparison with wild-type PS1). The PS1 ΔE9 and PS1 L250S transfected lines also showed a higher number of apoptotic cells with DNA strand breaks at 24 h following high glucose plus insulin-like growth factor-1 treatment (11.4±2.0% and 14.3±2.8%, respectively), compared with values for the wild-type PS1 lines (8.5±2.4%). These differences were significant ( P<0.01) for the comparision of wild-type PS1 and PS1 L250S, but not PS1 ΔE9 lines. The mutation-related increases in number of apoptotic cells at 24 h following high glucose treatment were not accompanied by significant differences in cell viability at this time-point. Our results indicate that PS1 mutations predispose to hyperosmotic stress-induced apoptosis and that the anti-apoptotic effects of insulin-like growth factor-1 are compromised by these mutations. Perturbations of insulin-like growth factor-1 signalling may be involved in PS1 mutation-related apoptotic neuronal cell death in Alzheimer's disease.