Inhibition of the ISR abrogates mGluR5-dependent long-term depression and spatial memory deficits in a rat model of Alzheimer's disease

Abstract

Soluble amyloid-β-protein (Aβ) oligomers, a major hallmark of AD, trigger the integrated stress response (ISR) via multiple pathologies including neuronal hyperactivation, microvascular hypoxia, and neuroinflammation. Increasing eIF2α phosphorylation, the core event of ISR, facilitates metabotropic glutamate receptor (mGluR)-dependent long-term depression (LTD), and suppressing its phosphorylation has the opposite effect. Having found the facilitation of mGluR5-LTD by Aβ in live rats, we wondered if suppressing eIF2α phosphorylation cascade would protect against the synaptic plasticity and cognitive disrupting effects of Aβ. We demonstrate here that the facilitation of mGluR5-LTD in a delayed rat model by single i.c.v. injection of synthetic Aβ_1-42. Systemic administration of the small-molecule inhibitor of the ISR called ISRIB (trans-isomer) prevents Aβ-facilitated LTD and abrogates spatial learning and memory deficits in the hippocampus in exogenous synthetic Aβ-injected rats. Moreover, ex vivo evidence indicates that ISRIB normalizes protein synthesis in the hippocampus. Targeting the ISR by suppressing the eIF2α phosphorylation cascade with the eIF2B activator ISRIB may provide protective effects against the synaptic and cognitive disruptive effects of Aβ which likely mediate the early stage of sporadic AD.

Fig. 1. mGluR5-dependent LTD facilitation in Aβ_1–42-injected rats.

A Experimental scheme: Rats received i.c.v. injection of soluble Aβ_1–42 (10 μL each side, 75 μM) under ketamine/xylazine anesthesia. Two weeks after single i.c.v. injection of soluble Aβ_1-42, the animals were re-anaesthetized with urethane and in vivo electrophysiology (EP) experiments were performed. B Application of a peri-threshold weak LFS (bar, LFS-300; 300 high-intensity pulses at 1 Hz) induced robust and stable LTD in Aβ_1–42-injected rats. One hour post systemic administration of the selective mGluR5 antagonist MTEP (hash; 3 mg/kg, i.p.) completely prevented the induction of LTD by LFS-300 in animals injected i.c.v. with soluble Aβ_1–42. In contrast, systemic injection of the NMDAR competitive antagonist CPP (hash;10 mg/kg, i.p.) did not affect the induction of LTD by LFS-300 in animals injected i.c.v. with soluble Aβ_1–42. As summarized in (C), the EPSP at 1.5 h measured 71.9 ± 3.7% in Aβ_1–42 + Veh group (n = 4, P = 0.0019 compared with Pre, paired t), 95.5 ± 5.8% in Aβ_1–42 + MTEP group (n = 5, P = 0.6185 compared with Pre and P = 0.0171 compared with Aβ_1-42 + Veh group; paired t and one-way ANOVA) and 68.6 ± 5.8% in Aβ_1–42 + CPP group (n = 4, P = 0.0092 compared with Pre, and P > 0.9999 compared with Aβ_1–42 + Veh group; paired t and one-way ANOVA). Calibration bars for EPSP traces: vertical, 2 mV; horizontal, 10 ms.

Fig. 2. The integrated stress response inhibitor ISRIB reverses facilitation of LTD in Aβ_1–42-injected rats.

A Plasma concentration (ng/ml) of ISRIB were determined by high-performance liquid chromatography (HPLC) 2, 6, and 24 h after a single intraperitoneal injection (0.25 mg/kg) in rats (n = 3). B Experimental scheme: Starting one week after single i.c.v. injection of synthetic Aβ_1–42 or reverse sequence control peptide Aβ_42–1 (10 μL each side) rats received a systemic injection of ISRIB (0.25 mg/kg, i.p.) or vehicle for 5 consecutive days. Three days after the final injection of ISRIB, in vivo electrophysiology (EP) experiments were performed under anesthesia. C Whereas the application of LFS-300 induced small LTD in control peptide Aβ_42–1-injected rats, the same conditioning stimulation trigged a robust and persistent LTD in Aβ_1–42-injected rats. Treatment of ISRIB successfully reversed LTD facilitation in Aβ_1–42-injected rats. As summarized in (D) at 90 min, the EPSP measured 84.1 ± 3.5% in Aβ_42–1 + Veh (n = 8), 67.8 ± 3.7% in Aβ_1–42 + Veh group (n = 7, P = 0.0167 compared with Aβ_42–1 + Veh group; one-way ANOVA) and 90.1 ± 3.8% in Aβ_1–42 + ISRIB (n = 8, P = 0.0011 compared with Aβ_1–42 + Veh and P = 0.7370 compared with Aβ_42–1 + Veh group; one-way ANOVA). E Plasma concentration (ng/ml) of ISRIB were determined by HPLC 2, 6, and 24 h after a single intraperitoneal injection (2.5 mg/kg) in rats (n = 3). F Experimental scheme: Rats were allowed to recover for one week after single i.c.v. injection of synthetic Aβ_1–42 or reverse sequence control peptide Aβ_42–1 (10 μL each side) and received an injection of a higher dose of ISRIB (2.5 mg/kg, i.p.) or vehicle for 3 consecutive days. Five days after the final injection of ISRIB, in vivo electrophysiology experiments were performed under anesthesia of urethane. G Application of LFS-300 induced robust and persistent LTD in Aβ_1–42-injected rats. Treatment of ISRIB successfully reversed LTD facilitation in Aβ_1–42-injected rats. As summarized in (H) at 90 min, the EPSP measured 91.6 ± 4.5% in Aβ_42–1 + Veh (n = 5), 63.8 ± 4.5% in Aβ_1–42 + Veh group (n = 6, P = 0.0069 compared with Aβ_42–1 + Veh group; one-way ANOVA) and 87.7 ± 6.8% in Aβ_1–42 + ISRIB (n = 5, P = 0.0188 compared with Aβ_1–42 + Veh and P > 0.9999 compared with Aβ_42–1 + Veh group; one-way ANOVA). Calibration bars for EPSP traces: vertical, 2 mV; horizontal, 10 ms.

Fig. 3. Promotion of learning and memory by ISRIB in Aβ_1–42-injected rats using a standard MWM protocol.

A The timeline of experimental design. Water maze training (4 trials per day for 5 days) was performed 2 weeks after i.c.v. injection of Aβ_1-42 or reverse control Aβ_42–1. Vehicle (1% DMSO in saline) or ISRIB (0.25 mg/kg, i.p.) were injected immediately after the last training trial in the MWM every day. B Escape latency in the navigation trial plotted against the training days. Two-way ANOVA followed by a post hoc Bonferroni multiple comparison test, P < 0.0001, F_5,36 = 15.38 (n = 7 rats per group). During training, compared with the Aβ_42–1 + Veh and Sham+Veh group, the Aβ_1–42 + Veh group spent more time to escape to the hidden platform from day 3 (Aβ_1–42 + Veh versus Sham + Veh: P = 0.0079 on day 3, P = 0.0101 on day 4, P = 0.0229 on day 5; Aβ_1–42 + Veh versus Aβ_42–1 + Veh: P = 0.0165 on day 3, P = 0.0118 on day 4, P = 0.0107 on day 5) but not in the first 2 days. However, a large reduction in escape latency was caused by ISRIB in rats injected with Aβ_1–42 from day 2 (Aβ_1-42 + Veh versus Aβ_1-42 + ISRIB: P = 0.0362). C In the probe trial (n = 7 rats per group), Aβ_1–42 + Veh animals appeared to cross the platform less frequently compared with the Aβ_42–1 + Veh and Sham + Veh group and ISRIB significantly improved performance (P = 0.0001, one-way ANOVA followed by a post hoc Bonferroni multiple comparison test). D In the case of the probe trial quadrant bias, ISRIB significantly enhanced target quadrant occupancy in the Aβ_1–42-injected animals (Aβ_1–42 + Veh versus Aβ_1–42 + ISRIB: P = 0.0020, one-way ANOVA followed by a post hoc Bonferroni multiple comparison test). E, F Both total swimming distance (P = 0.4351, one-way ANOVA) and swimming speed (P = 0.3626, one-way ANOVA) are comparable in all the groups. Error bars, s.e.m.

Fig. 4. Spatial learning and memory deficits in Aβ_1–42-injected rats are abrogated by ISRIB using a weak MWM training protocol.

A The timeline of experimental design. All animals were trained with 1-trial/day. Vehicle or ISRIB (0.25 mg/kg, i.p.) were injected immediately after the training session in the MWM every day. B All the rats spent less time gradually to find the hidden platform after each training trial. Aβ_1–42-injected rats spent more time to find the hidden platform from day 6 and on day 14 and day 15 after one-week break (n = 8–10 rats per group, Two-way ANOVA followed by a post hoc Bonferroni multiple comparison test, P < 0.0001, F_5,48 = 8.778. Aβ_1–42 + Veh versus Sham+Veh: P = 0.0681 on day 6, P = 0.0325 on day 14, P = 0.0415 on day 15; Aβ_1–42 + Veh versus Aβ_42–1 + Veh: P = 0.1142 on day 6, P = 0.0207 on day 14, P = 0.0410 on day 15) and ISRIB significantly improved performance (Aβ_1–42 + Veh versus Aβ_1–42 + ISRIB: P = 0.0146 on day 6, P = 0.0231 on day 14, P = 0.0158 on day 15). No difference was detected among Sham + Veh, Aβ_42–1 + Veh, Sham + ISRIB, Aβ_42–1 + ISRIB and Aβ_1–42 + ISRIB groups (Repeated measures ANOVA. P = 0.0569, F_4,41 = 2.504). C In the probe trial (n = 8–10 rats per group), Aβ_1–42-injected animals crossed the platform much less compared with control groups (Aβ_1–42 + Veh versus Sham + Veh: P = 0.0129; Aβ_1–42 + Veh versus Aβ_42–1 + Veh: P = 0.0028) and ISRIB significantly enhanced platform crossing in Aβ_1–42-injected rats (Aβ_1–42 + Veh versus Aβ_1–42 + ISRIB: P = 0.0050, One-way ANOVA followed by a post hoc Bonferroni multiple comparison test). D–F All the groups were similar in target quadrant occupancy (P = 0.9409, One-way ANOVA) (D) and total swimming distance (P = 0.7056, One-way ANOVA) (E) and swimming speed (P = 0.7876, One-way ANOVA) (F). Error bars, s.e.m.

Fig. 5. ISRIB prevented the disruption of protein synthesis induced by Aβ_1–42.

A Experimental timeline: Rats received systemic and i.c.v. injections under isoflurane anesthesia. ISRIB (2.5 mg/kg, i.p.) was administrated 1 h before Aβ injection (i.c.v., 10 μL each side). Rats were re-anesthetized 24 h after Aβ injection with isoflurane and received an i.c.v. injection of puromycin (5 μL each side, 10 μg/μL). Brain samples were collected 2 h after puromycin injection. B Protein extracts were separated by electrophoresis and analyzed by western blot with antibody to puromycin. GAPHD immunoblot is shown as a loading control (bottom). C Levels of newly synthesized proteins labeled with puromycin were significantly lower in Aβ_1–42 + Veh group compared with Aβ_42–1 + Veh group (n = 8, P = 0.0386, one-way ANOVA). In ISRIB treatment group, protein synthesis was markedly restored (n = 8, P = 0.0018, Aβ_1–42 + ISRIB compared with Aβ_1–42 + Veh; P = 0.6087, Aβ_1–42 + ISRIB compared with Aβ_42–1 + Veh group; one-way ANOVA). Error bars, s.e.m.