The diphenylpyrazole compound anle138b blocks Aβ channels and rescues disease phenotypes in a mouse model for amyloid pathology

Abstract

Alzheimer's disease is a devastating neurodegenerative disease eventually leading to dementia. An effective treatment does not yet exist. Here we show that oral application of the compound anle138b restores hippocampal synaptic and transcriptional plasticity as well as spatial memory in a mouse model for Alzheimer's disease, when given orally before or after the onset of pathology. At the mechanistic level, we provide evidence that anle138b blocks the activity of conducting Aβ pores without changing the membrane embedded Aβ-oligomer structure. In conclusion, our data suggest that anle138b is a novel and promising compound to treat AD-related pathology that should be investigated further.

Keywords: Alzheimer's disease; Aβ channels; amyloid pathology; gene expression; membrane pores.

Conflict of interest statement

AG and CG are co‐founders of MODAG. AL is partly employed by MODAG.

© 2017 The Authors. Published under the terms of the CC BY 4.0 license.

Figures

Figure EV1. Experimental design

Overview of the experimental approach employed for the pre‐plaque and the post‐plaque group. Red arrows underneath the time line indicate onset of pathology in APPPS1△9 mice.

Figure 1. Anle138b rescues hippocampal LTP deficits spatial memory in the pre‐plaque group

1. In wild‐type mice of the pre‐plaque group (treated from 2 to 6 months of age with anle138b), robust LTP that lasts for at least 3 h is elicited upon a strong tetanization (STET) (3 trains of 100 pulses at 100 Hz given 10 min apart, arrows) at the Schaffer collateral CA3‐CA1 synapse (t‐test, P = 0.00005; n = 16 comparing before vs. after STET).

2. LTP is not maintained in APPPS1Δ9 mice treated with placebo. Here, the potentiation declined to baseline after 3 h (t‐test, P = 0.08; n = 20 comparing before vs. after STET).

3. APPPS1Δ9 treated with anle138b show a rescue of the LTP impairment (t‐test, P = 0.0001; n = 23 comparing before vs. after STET).

4. Escape latency in the Morris water maze test is impaired in placebo‐ but not in anle138b‐treated APP mice (one‐way ANOVA F = 16.01, **P = 0.0008; n = 15/group).

5. Probe test performed 24 h after the last training session. The lower panel shows representative swimming path during the probe test. T = target quadrant compared vs. other quadrants (t‐test, ***P = 0.00002; n = 15/group).

6. Average swim speed during water maze training (n = 15/group).

7. Explorative behavior in the open field test. Upper panel: Representative motion tracks during the test session. Lower panels show the total distance travelled during the 5‐min test session (n = 15/group).

8. Bar graph showing the time spent in the center vs. the corner of the open field (n = 15/group).

Data information: Error bars indicate SEM.

Figure 2. Anle138b rescues hippocampal LTP deficits and spatial memory in the post‐plaque group

1. Wild‐type mice of the post‐plaque group (treated from 6 to 10 months of age) display robust LTP upon STET (arrows) that was maintained throughout the recording session (t‐test, P = 0.00006; n = 30 comparing before vs. after STET).

2. Lasting LTP induced by STET was not observed in 10‐month‐old APPPS1Δ9 placebo‐treated mice. The potentiation decayed to baseline after 3 h (t‐test, P = 0.16; n = 19 comparing before vs. after STET).

3. Treatment with anle138b starting at 6 months of age rescues LTP deficit in 10‐month‐old APPPS1Δ9 mice (t‐test, P = 0.0003; n = 20 comparing before vs. after STET).

4. Escape latency in the Morris water maze test is impaired in placebo‐treated APP mice and partially restored to WT levels in anle138b‐treated APP mice (one‐way ANOVA, F = 35.94, P = 0.00006; *P = 0.0309 for APP + anle138b vs. APP + placebo and P = 0.4 for APP + placebo vs. WT + anle138b; n = 7/group).

5. Probe test performed 24 h after the last training session. The lower panel shows representative swimming path during the probe test. T = target quadrant compared vs. other quadrants (t‐test, **P = 0.0007, *P = 0.0031; n = 7/group).

6. Average swim speed during water maze training (n = 7/group).

7. Explorative behavior in the open field test. Upper panel: Representative motion tracks during the test session. Lower panels show the total distance travelled during the 5‐min test session (n = 7/group).

8. Bar graph showing the time spent in the center vs. the corner of the open field (n = 7/group).

Data information: Error bars indicate SEM. In panels (A–C), triplets of arrows represent STET applied for inducing L‐LTP. Insets in each graph represent typical fEPSP traces recorded 15 min before (dotted line), 30 min after (broken line), and 3 h after (full line) STET.

Figure 3. Hippocampal transcriptome analysis in anle138b‐treated mice

1. Upper panel: Heat map showing differentially expressed genes in placebo‐treated WT (n = 4), placebo‐treated APPPS1ΔE9 mice (n = 4), and anle138b‐treated APPPS1ΔE9 mice (n = 4) of the pre‐plaque group. Note that a WT‐like gene expression profile is to a large extent reinstated in APPPS1ΔE9 mice upon treatment with anle138b. Lower panel: Bar graph showing the number of up and down‐regulated genes.

2. Pathways down‐regulated (blue) or up‐regulated (red) in 6‐month‐old APPPS1ΔE9 mice when compared to age‐matched WT controls (pre‐plaque group; see Fig EV1).

3. Expression of APP and PS1 genes in WT and APPPS1ΔE9 transgenic mice, placebo‐ or anle138b‐treated in pre‐ and post‐plaque groups. Note that anle138b treatment does not affect the expression level of APP or PS1. Post hoc analysis revealed a significantly increased expression of APP and PS1 in transgenic mice when compared to wild‐type mice (*P < 0.05). Error bars indicate SEM.

4. Upper panel: Heat map showing differentially expressed genes in placebo‐treated WT (n = 4), placebo‐treated APPPS1ΔE9 mice (n = 5), and anle138b‐treated APPPS1ΔE9 mice (n = 5) of the post‐plaque group. Lower panel: Bar graph showing the number of up‐ and down‐regulated genes.

5. Venn diagram showing that 10‐month‐old mice treated with placebo or anle138b show very similar changes in hippocampal gene expression (post‐plaque group; see Fig EV1).

6. Pathway analysis based on the 103 genes commonly increased in placebo‐ and anle138b‐treated APPPS1ΔE9 mice at 10 months of age. Note that the pathways are exclusively linked to neuroinflammation.

Figure 4. Amyloid plaque pathology is ameliorated by anle138b treatment in the pre‐plaque (A) and post‐plaque (B) groups

1. Reduced amyloid pathology as indicated by reduced number of plaques (left panel) and reduced area covered by plaques (right panel) in the pre‐plaque group (t‐test ***P = 0.00003; n = 5/group).

2. Reduced amyloid pathology as indicated by reduced number of plaques (left panel) and reduced area covered by plaques in the post‐plaque group (t‐test ***P = 0.00007, **P = 0.005; n = 5/group).

3. Representative images showing thioflavin S staining in the hippocampus and cortex of mice of the pre‐ and post‐plaque groups. Scale bar = 200 μm.

Data information: Error bars indicate SEM.

Figure 5. Anle138b ameliorates Aβ1‐40 and Aβ1‐42‐induced membrane integrity

1. Aβ1‐42 induces pore‐like step ionic current increases across lipid bilayer membranes and grows with inverted sign after voltage inversion to amplifier saturation current. In the presence of anle138b, the current remains at low amplitude and does not increase beyond 30 pA.

2. Expanded trace of anle138b‐treated membrane presented in (A). Discreet conductance levels are highlighted suggesting multiple opening and closing events for three individual pores.

3. Hippocampal neurons (DIV 10) were treated with anle138b (1 μM) or vehicle before Aβ1‐40 oligomers or monomers were added (10 μM, n = 4/group). After 48 h, membrane integrity was measured as fluorescence intensity using a CyQUANT assay (Thermo Fisher). In the vehicle group, membrane integrity was significantly impaired when treated with Aβ oligomers comparing with control neurons or Aβ monomer‐treated neurons, anle138b‐treated neurons did not exhibit a difference between addition of Aβ monomers or Aβ oligomers (t‐test, **P = 0.007).

4. Same experimental setting as in (C) but cell viability was measured using the MTT assay (t‐test, *P = 0.002 vs. control). No difference in cell viability was observed for Aβ monomer or oligomer treatment in the absence or presence of anle138b.

5. Schematic of potential mechanisms of activity inhibition for anle138b. In the absence of anle138b, Aβ monomers and/or oligomers insert in the membrane and form conducting pores. Treatment with anle138b renders these conductive channels inactive most probably due to reduced life time of the open state.

Data information: Error bars indicate SEM.

Figure EV2. Observation of Aβ pores by AFM in anle138b‐doped membranes

AFM tapping mode images showing similar pore structures in membranes with and without anle138b.

1. A

   POPE/DOPS (1:1) membranes without compound and not exposed to Aβ1‐42 showing lateral phase separation. Scale bar is as indicated and z scale is 5 nm.

2. B

   Membrane with the same lipid composition as in (A) without compound reconstituted with Aβ1‐42 at a 1:60 peptide/lipid molar ratio. White circles represent the position of pore‐like structures. Scale bar is 100 nm and z scale is 5 nm.

3. C

   Membrane with the same lipid composition as above with anle138b (10 mM with respect to lipid volume) and reconstituted with Aβ1‐42 (1:60 peptide/lipid molar ratio). White circles represent the position of pore‐like structures. Scale bar is 100 nm and z scale is 5 nm.

4. D, E

   Four representative pores extracted from (B) and (C), respectively. Image size is 23.4 nm × 23.4 nm.

5. F

   Table summarizing that Aβ pores show the same dimension in the presence or absence of anle138b. Values are represented as average ±SD.