The amyloid β protein (Aβ) is derived by sequential β- and λ-secretase cleavage of amyloid precursor protein and is widely regarded as the principal toxic factor of Alzheimer’s disease (AD). Newly developed inhibitors of β- and λ-secretase prevent Aβ production and represent a potential therapeutic strategy in AD (Sinha et al. 1999; Selkoe, 2001). We have recently proposed a physiological role for Aβ (Ramsden et al. 2001), and so chose to investigate the impact of secretase inhibitors on the viability of cultured cells.
Primary cultures of rat cortical or cerebellar granule neurones (CGN) were prepared from humanely killed animals as previously described (Ramsden et al. 2001). Immortalised cell lines (SH-SY5Y, DDT1FM2, HEK293) and primary cultures of a teratorhabdoid tumour were grown using the methods of Shukla et al. (2001). Cell viability was assessed using the MTT assay (Shukla et al. 2001). All values are given as means ± S.E.M.
Following 24 h treatment with 2-naphthoyl-VF-CHO (λ-IV; 10 µM), an inhibitor of λ-secretase activity, the viability of cortical neurones was reduced by 57 ± 5 % (n = 9, P < 0.01, Student’s unpaired t test). This neurotoxic effect was also seen with the β secretase inhibitor H-KTEEISEVN-stat-VAEF-OH (βSI; 100 nM, 54 ± 4 % cell death, n = 9) and could be reproduced using cerebellar granule neurones and the human neuroblastoma cell line SH-SY5Y (62 ± 6 and 43 ± 10 % cell death, respectively). In contrast, cells of a non-neuronal origin were unaffected by inhibition of amyloidogenic pathways. In HEK293 cells 24 h incubation with λ-IV resulted in a small, non-significant increase in cell viability of 6.4 ± 1.9 % (n = 8) and in DDT1FM2 cells and a teratorhabdoid carcinoma similar increases (2.1 ± 3.0 and 2.0 ± 2.5 %, n = 8 in each case) were seen, respectively. Crucially, the cell death observed in neurones could be prevented by co-application of Aβ1-40 at physiologically relevant concentrations. At a concentration of 1 nM, Aβ reduced cell death in cortical neurones caused by λ-IV to 4 ± 7 % (n = 9). This protective effect was not seen with the pathological 1-42 or synthetic 25-35 size forms.
These data suggest that the production of Aβ is a requirement for neuronal viability and support a physiological role for this peptide. Importantly, these data suggest that interfering with amyloidogenesis as a therapeutic approach in AD may result in unforeseen neurotoxic effects.
This work was supported by the MRC and The Wellcome Trust.
All procedures accord with current UK legislation.