Enhanced neprilysin-mediated degradation of hippocampal Aβ42 with a somatostatin peptide that enters the brain

Abstract

Background: Aggregation of the amyloid-beta (Aβ) peptide is one of the main neuropathological events in Alzheimer's disease (AD). Neprilysin is the major enzyme degrading Aβ, with its activity enhanced by the neuropeptide somatostatin (SST). SST levels are decreased in the brains of AD patients. The poor delivery of SST over the blood-brain barrier (BBB) and its extremely short half-life of only 3 min limit its therapeutic significance. Methods: We recombinantly fused SST to a BBB transporter binding to the transferrin receptor. Using primary neuronal cultures and neuroblastoma cell lines, the ability of the formed fusion protein to activate neprilysin was studied. SST-scFv8D3 was administered to mice overexpressing the Aβ-precursor protein (AβPP) with the Swedish mutation (APPswe) as a single injection or as a course of three injections over a 72 h period. Levels of neprilysin and Aβ were quantified using an Enzyme-linked immunosorbent assay (ELISA). Distribution of SST-scFv8D3 in the brain, blood and peripheral organs was studied by radiolabeling with iodine-125. Results: The construct, SST-scFv8D3, exhibited 120 times longer half-life than SST alone, reached the brain in high amounts when injected intravenously and significantly increased the brain concentration of neprilysin in APPswe mice. A significant decrease in the levels of membrane-bound Aβ42 was detected in the hippocampus and the adjacent cortical area after only three injections. Conclusion: With intravenous injections of our BBB permeable SST peptide, we were able to significantly increase the levels neprilysin, an effect that was followed by a significant and selective degradation of membrane-bound Aβ42 in the hippocampus. Being that membrane-bound Aβ triggers neuronal toxicity and the hippocampus is the central brain area in the progression of AD, the study has illuminated a new potential treatment paradigm with a promising safety profile targeting only the disease affected areas.

Keywords: Amyloid-β (Aβ) / blood-brain barrier (BBB) / neprilysin / somatostatin (SST) / transferrin receptor (TfR).

Figure 1

Design of SST-scFv8D3 and characterisation of its purity and binding properties to transferrin receptor TfR. (A). Schematic picture of SST-scFv8D3, where scFv of 8D3 is added to the N-terminal end of SST. A poly-6 His-tag is added to the N-terminal end of scFv8D3 for purification purposes. (B). SDS-PAGE of the purified SST-scFv8D3 showing a single band at the expected size, approximately 31 kDa. (C). ELISA displaying similar binding to TfR between SST-scFv8D3 and RmAb158-scFv8D3, using SST (without scFv8D3) as a negative control.

Figure 2

Enhanced cell-surface activity of neprilysin by SST-scFv8D3 in primary neurons and SHSY5Y cells. (A). Primary murine neurons incubated for 24 h with 1 μM of SST, SST-scFv8D3 and scFv8D3, using PBS as a negative control. The recombinant protein SST-scFv8D3 significantly enhances the activity of neprilysin when compared to scFv8D3 and PBS. (B). Human SHSY5Y cells incubated for 24 h with 10 μM of SST and SST-scFv8D3, with or without cyclo-SST, using PBS as a negative control. The recombinant protein SST-scFv8D3 significantly enhances the activity of neprilysin when compared to PBS. Co-administration of cyclo-SST abolished neprilysin activation. Results are presented as mean ±SD. One-way ANOVA was used followed by Dunnett's post hoc test. A significant p-value is defined as: >0.5 (ns), <0.05 (*), <0.01 (**), <0.005 (***).

Figure 3

Pharmacokinetics and distribution of iodine-125 labelled SST-scFv8D3 when given intravenously to 7-8 month old C57Bl/6 mice (n=3 per group). (A). Plasma concentrations, expressed as percentage of injected dose (%ID) per gram blood of [¹²⁵I]SST-scFv8D3. The plasma-half-life was estimated to 6 h, which is 120 times longer than the blood half-life of SST peptide alone. (B). Brain concentration, expressed as %ID per gram brain tissue, at 2, 6 and 24 h after intravenous injection of [¹²⁵I]SST-scFv8D3. High brain uptakes observed at 2 and 6 h post injection which then declined at 24 h. (C). Brain-to-blood concentration ratio of SST-scFv8D3 at 2, 6 and 24 h post intravenous injection showing that the construct was retained in the brain longer than in the blood. (D). Concentration in the peripheral organs, expressed %ID per gram tissue. Spleen and kidney displayed high uptake 2 h post intravenous injection of [¹²⁵I]SST-scFv8D3. (E). Regional distribution of SST-scFv8D3 in the brain 24 h post injection, detected with anti-His tag immunostaining (scale bar 1 mM). A high signal was mainly detected in the hippocampal area, cerebral cortex and the cerebellum. Boxed areas show the distribution of the drug in two regions of the cerebral cortex and the hippocampus (scale bar 100 μM). (F). Confocal images of neprilysin immunostaining in coronal sections prepared from the hippocampus of C57Bl/6 and neprilysin knock-out mice (scale bar 500 μM).

Figure 4

Single intravenous injection of SST-scFv8D3 increases brain levels of neprilysin but has no detectable effect on Aβ levels. (A). Comparison of neprilysin expression in the brain of wild-type mice (n=4), eight-months old APPswe (n=7) and fourteen-months old APPswe mice (n=4). APPswe mice exhibit a significant reduction in neprilysin levels compared to wild-type. A significant age-dependent decline in neprilysin protein levels is detected between the APPswe groups. (B). Schematic representation of the first therapeutic study with SST-scFv8D3. Fourteen-month old transgenic mice harbouring the Swedish mutation in APP (APPswe) injected intravenously with 2 mg/kg of SST-scFv8D3 (n=4) or PBS (n=4) as a negative control. Mice euthanized 24 h post injection by perfusion with physiological saline. (C). Significant increase in the levels of neprilysin in the membrane fraction (TBS-Triton soluble) of the brain of the treatment group. No differences detected in the levels of Aβ40 and Aβ42, neither in the soluble pool (D), nor in the membrane fraction (E) between SST-scFv8D3 treated group and the controls. Results are presented as mean ±SD. One-way ANOVA followed by Tukey's post hoc test (A) and unpaired t-test (C-E) were applied to measure the presence of statistically significant differences in the results. A significant p-value is defined as: >0.5 (ns), <0.05 (*), <0.01 (**), <0.005 (***).

Figure 5

Three intravenous injections of 1 mg/kg SST-scFv8D3 increase neprilysin levels and decrease Aβ42 in the hippocampal area. (A) Schematic representation of the therapeutic study with repeated injections of SST-scFv8D3. Eight-month old APPswe mice were injected intravenously with 1 mg/kg of SST-scFv8D3 (n=7), an equimolar dose of scFv8D3 (n=7) or PBS (n=7) every 36 h (in total three injections). The mice were euthanized 24 h after the last injection and the brains dissected into three parts: Hippocampal area, rest of cerebrum and cerebellum. (B) A significant increase could be detected in neprilysin concentrations in the membrane fraction of hippocampus of the SST-scFv8D3 treated group compared to the controls measured with sandwich ELISA. No significant differences were detected in the rest of the cerebrum or the cerebellum. (C) A significant increase could be detected in neprilysin activity in the membrane fraction of the SST-scFv8D3 treated group compared to the controls. The differences were specific to the hippocampal area. No differences were detected in Aβ40 levels in the soluble pool (D) and membrane fraction (E), or in Aβ42 levels in the soluble pool (F) between SST-scFv8D3 treated group and the two control groups. (G) A significant decrease in Aβ42 levels detected in the membrane fraction of hippocampus in the treatment group. (H). A significant inverse correlation between the levels of neprilysin and Aβ42 in the membrane fraction of the hippocampus was obtained with Pearson's correlation. (I) A significant positive correlation between the concentration of Aβ40 and Aβ42 in the membrane fraction of the hippocampus of the SST-scFv8D3 treated mice only, obtained with Pearson's correlation. (J) A significant increase could be detected in neprilysin concentration in spleen and kidneys of the SST-scFv8D3 treated group (n=4) compared to scFv8D3 (n=7) and PBS (n=4) groups. No differences detected in liver and heart. Results are generated from two repetitive experiments, where four mice were used in first round and three mice in the second round. Results are presented as mean ±SD. One-way ANOVA was used followed by Tukey's post hoc test. A significant p-value is defined as: >0.5 (ns), <0.05 (*), <0.01 (**), <0.005 (***).