Figure 1.
Mass spectra of Aβ isoform patterns in all cell models investigated.
SH-SY5Y APP695wt cells treated with DMSO (Panels a and c), 5 µM β-secretase inhibitor IV (Panel b) or 10 µM AZ-20 (Panel d). SH-SY5Y APP695swe cells treated with DMSO (Panel e) or 10 µM AZ-20 (Panel f). 7PA2 APP751 V717F cells treated with DMSO (Panel g) or 10 µM AZ-20 (Panel h). HeLa APPswe cells treated with DMSO (Panel i) or 10 µM β-secretase inhibitor IV (Panel j). HeLa APPswe scrambled siRNA transfected control cells (Panel k) and cells transfected with single oligo siRNAs against BACE1 (Panel l). The mass-to-charge ratio (m/z) of the [M+H]+ ion of Aβ5-38 is very close to that of Aβ1-33, causing the peaks to partially overlap and making quantification difficult, wherefore both isoforms were excluded from quantitative analysis. Those peptides are instead presented in these mass spectra as expanded inserts (except for panels g-h where they are clearly visible).
Figure 2.
SH-SY5Y APP695wt cells treated with β-secretase inhibitor IV.
Peak intensities of all Aβ isoforms detected (Panel a). Normalized (against the sum for all the Aβ peaks in the spectrum) levels in relation to controls (DMSO) (Panel b). Immunoassay measurements of Aβ1-40 and AβX-40 (Panels c–e), Aβ1-42 and AβX-42 (Panels f–g), sAPP-α and sAPP-β (Panels h–i). Note that the X-40/42 groups may consist of a large variety of species not detected with the MALDI-TOF-MS method utilized in this work, including such spanning the β-secretase site. N = 3 for each concentration. Data are means and error bars are SD in panels a–b.
Figure 3.
Cell medium Aβ isoform patterns.
SH-SY5Y APP695wt cells treated with AZ-20 (Panel a). SHSY-5Y APP695swe cells treated with AZ-20 (Panel b). 7PA2 APP751 V717F cells treated with AZ-20 (Panel c). HeLa-APPswe cells treated with β-secretase inhibitor IV (Panel d). HeLa-APPswe mock and scrambled siRNA-transfected control cells, and cells transfected with single oligo siRNA or pooled siRNA against BACE1 (Panels e). N = 1 for each concentration and treatment.
Figure 4.
Dogs treated with vehicle (N = 15), NB-C8 (N = 3 at 3 hours, N = 3 at 16 hours) and NB-B4 (N = 5 at 6 hours). Mass spectra of the Aβ isoform pattern from dogs treated with placebo (Panel a, upper panel) or NB-C8 16 hours post treatment (Panel a, lower panel). The expanded sections show Aβ1-33, Aβ1-34, Aβ5-38 and Aβ5-40. Aβ1-33 and Aβ5-38 were excluded from quantitative analysis since their peaks partially overlap, making quantification difficult. Absolute (Panels b–c) and normalized (Panels d–e) mass spectral peak intensities of all detected Aβ isoforms. Statistical significances were tested for normalized peak intensities comparing different treatments. For NB-B4, significant differences were seen for Aβ5-40 (P = 0.001), Aβ1-34 (P = 0.001) and Aβ11-40 (P = 0.002). For NB-C8, significant differences were seen between vehicle and treatment at 16 h for Aβ5-40 (P = 0.01) and Aβ1-34 (P = 0.05). Data are means; error bars are SD.
Figure 5.
Multivariate analysis and CSF Aβ5-40/Aβ1-34.
Multivariate discriminant analysis of the CSF Aβ pattern for treatment compared with placebo in dogs treated with vehicle (N = 15), NB-C8 (N = 3 at 3 h, N = 3 at 16 h) and NB-B4 (N = 5 at 6 h) (Panels a–b). Score vector results including medians (Panel a). Samples were taken 3 h (open triangles), 6 h (filled squares) or 16 h (filled triangles) after treatment. The groups denoted by “T” were used for constructing the multivariate model while “P” represents the prediction set used for testing the stability of the model. In the construction of the model, all dogs treated with NB-C8 were regarded equal, independent of time after drug administration. Relative contributions of different isoforms to group separations, with increased or decreased relative levels in the treated groups, are shown by white or grey bars, respectively (Panel b). Change of the CSF Aβ5-40/Aβ1-34 ratio in relation to change of CSF Aβ1-40 (Panel c) and CSF Aβ1-42 (Panel d). The Aβ5-40/Aβ1-34 ratio was a more sensitive biomarker than Aβ1-40 and Aβ1-42 in terms of change from baseline (the dotted lines indicate predicted correlations for biomarkers affected equally by treatment). Dogs treated with vehicle (N = 4) and BACE1 inhibitor S obtained from Janssen (N = 4 for each dosage) (Panel e). The CSF Aβ5-40/Aβ1-34 ratio completely separated dogs on active treatment versus placebo (P = 0.005 using the Mann-Whitney U test for comparison of all animals on active treatment versus placebo). Time-dependent dynamics of the ratio in high dose treatment (Panel f).
Figure 6.
Summary of processing pathways.
The main pathways of Aβ peptide release and how these are affected by BACE1 inhibition (arrows indicate absolute and/or relative changes). The APP box shows major APP cleaving secretases with selected cleavage sites that depend on them.