Role of Endoplasmic Reticulum Stress in Learning and Memory Impairment and Alzheimer's Disease-Like Neuropathology in the PS19 and APPSwe Mouse Models of Tauopathy and Amyloidosis

Abstract

Emerging evidence suggests that endoplasmic reticulum (ER) stress may be involved in the pathogenesis of Alzheimer's disease (AD). Recently, pharmacological modulation of the eukaryotic translation initiation factor-2 (eIF2α) pathway was achieved using an integrated stress response inhibitor (ISRIB). While members of this signaling cascade have been suggested as potential therapeutic targets for neurodegeneration, the biological significance of this pathway has not been comprehensively assessed in animal models of AD. The present study investigated the ER stress pathway and its long-term modulation utilizing in vitro and in vivo experimental models of tauopathy (MAPT P301S)PS19 and amyloidosis (APP^Swe). We report that thapsigargin induces activating transcription factor-4 (ATF4) in primary cortical neurons (PCNs) derived from rat and APP^Swe nontransgenic (nTg) and transgenic (Tg) mice. ISRIB mitigated the induction of ATF4 in PCNs generated from wild-type (WT) but not APP^Swe mice despite partially restoring thapsigargin-induced translational repression in nTg PCNs. In vivo, C57BL/6J and PS19 mice received prolonged, once-daily administration of ISRIB. While the compound was well tolerated by PS19 and C57BL/6J mice, APP^Swe mice treated per this schedule displayed significant mortality. Thus, the dose was reduced and administered only on behavioral test days. ISRIB did not improve learning and memory function in APP^Swe Tg mice. While ISRIB did not reduce tau-related neuropathology in PS19 Tg mice, no evidence of ER stress-related dysfunction was observed in either of these Tg models. Taken together, the significance of ER stress and the relevance of these models to the etiology of AD require further investigation.

Keywords: ER stress; amyloidosis; cognition; integrated stress response; neurodegeneration; tauopathy.

Figure 1.

Timeline of behavioral studies conducted in PS19 and APP^Swe mice. FC, fear conditioning; AC, activity chamber.

Figure 2.

Thapsigargin induced ER stress and target engagement in vitro. A–D, ER stress-induced ATF4 translation, but not CHOP activation or cytotoxicity, is reduced by ISRIB in rat PCNs. A, Representative immunoblots of primary cortical cell lysates derived from E17 Sprague Dawley rats probed using antibodies directed against ATF4, CHOP, and tubulin. B, Quantification of ATF4 levels normalized to tubulin. ATF4 is increased in cells treated with 1 μM thapsigargin or 1 μM thapsigargin + 200 nM ISRIB compared to vehicle control. Cells treated with 1 μM thapsigargin have more ATF4 compared to cells treated with 1 μM thapsigargin + 200 nM ISRIB. Vehicle, n = 6; 1 μM thapsigargin, n = 5; 1 μM thapsigargin + 200 nM ISRIB, n = 4. C, Quantification of CHOP levels normalized to tubulin. CHOP is increased in cells treated with 1 μM thapsigargin compared to vehicle control. Vehicle, n = 4; 1 μM thapsigargin, n = 3; 1 μM thapsigargin + 200 nM ISRIB, n = 3. D, Quantification of cytotoxicity by LDH. Cells treated with 10 μM thapsigargin, 10 μM thapsigargin + 200 nM ISRIB, or 10 μM thapsigargin + 1 μM ISRIB have higher percentages of cytotoxicity compared to vehicle control. Vehicle, n = 6; no vehicle, n = 6; 10 μM thapsigargin, n = 6; 10 μM thapsigargin + 200 nM ISRIB, n = 6; 10 μM thapsigargin + 1 μM ISRIB, n = 6. Error bars indicate SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 3.

ER stress-related dysfunction is not observed in the APP^Swe model in vitro despite evidence of ISRIB target engagement. A, Thapsigargin-induced ER stress is not mitigated by ISRIB in APP^Swe mouse PCNs. Cells from nTg and Tg mice cultured for 13 DIV treated with 1 μM thapsigargin or 1 μM thapsigargin + 200 nM ISRIB have increased levels of ATF4 compared to vehicle control. nTg vehicle, n = 3; nTg + 1 μM thapsigargin, n = 3; nTg + 1 μM thapsigargin + 200 nM ISRIB, n = 3; Tg vehicle, n = 3; Tg + 1 μM thapsigargin, n = 3; Tg + 1 μM thapsigargin + 200 nM ISRIB, n = 3. B, Thapsigargin (1 μM) attenuates protein synthesis in nTg PCNs cultured for 7 DIV. ISRIB (200 nM) provides partial restoration of protein synthesis in PCNs challenged with thapsigargin. nTg vehicle + 10 μg/ml puromycin, n = 4; nTg + 1 μM thapsigargin + 10 μg/ml puromycin, n = 5; nTg + 1 μM thapsigargin + 200 nM ISRIB, n = 5. C, Thapsigargin-induced ER stress is not mitigated by ISRIB in nTg mouse PCNs. Cells from nTg mice cultured for 7 DIV treated with 1 μM thapsigargin or 1 μM thapsigargin + 200 nM ISRIB have increased levels of ATF4 compared to vehicle control. nTg vehicle, n = 6; nTg + 1 μM thapsigargin, n = 6; nTg + 1 μM thapsigargin + 200 nM ISRIB, n = 6. Error bars indicate SEM; **p < 0.01, *** p = 0.001, ****p < 0.0001.

Figure 4.

Peripherally administered ISRIB crosses the BBB and prolonged administration is well tolerated by C57BL/6J mice. A, ISRIB concentration over time in brain and plasma of five-month-old C57BL/6J mice after a single intraperitoneal injection (5 mg/kg). Plasma collected at 0.5 h (n = 4), 2 h (n = 4), 4 h (n = 3), and 8 h (n = 4) after injection. Brain tissue collected at 0.5 h (n = 4), 2 h (n = 4), 4 h (n = 4), and 8 h (n = 4) after injection. B, Body weights of mice receiving daily injections of vehicle or ISRIB for nine weeks. Vehicle, n = 6; ISRIB, n = 6. C, Survival rates of vehicle or ISRIB-treated mice. Error bars indicate SEM.

Figure 5.

Effects of genotype and treatment on body weight and mortality in PS19 mice. A, Tg mice have lower body weights compared to nTg mice. Daily administration of ISRIB reduced body weight in nTg mice over the course of nine weeks. B, Tg mice have reduced survival rates compared to nTg mice. ISRIB improved survival in nTg mice. nTg vehicle, n = 27; nTg ISRIB, n = 25; Tg vehicle, n = 15; Tg ISRIB, n = 25. Asterisks indicate comparisons to nTg vehicle. Error bars indicate SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 6.

Despite severe behavioral impairments, ISRIB provides modest restoration of spatial acquisition deficits in PS19 Tg mice. A–C, Tg mice exhibit a characteristically hyperactive phenotype. A, Quantification of ambulatory distance. Tg mice treated with ISRIB ambulate a greater distance on AC test 3 compared to nTg mice treated with ISRIB. B, Quantification of ambulatory duration. Tg mice treated with ISRIB spend more time moving on AC tests 2 and 3 compared to nTg mice treated with ISRIB. C, Quantification of rear frequency. Tg mice treated with vehicle rear more frequently on AC test 2 compared to nTg mice treated with vehicle. Tg mice treated with ISRIB rear more frequently on AC tests 3 and 4 compared to nTg mice treated with ISRIB. nTg vehicle, n = 27; nTg ISRIB, n = 25; Tg vehicle, n = 25; Tg ISRIB, n = 25. D, Tg mice exhibit diminished anxiety-like behavior. Tg mice treated with vehicle or ISRIB spend more time in the open arms of the EPM compared to nTg mice treated with vehicle. nTg vehicle, n = 26; nTg ISRIB, n = 25; Tg vehicle, n = 24; Tg ISRIB, n = 22. E–G, Tg mice display impaired spatial learning and memory with modest restoration of acquisition by ISRIB. E, Quantification of escape latency during the acquisition/training phase of the MWM. Tg mice treated with vehicle took longer to locate the platform on each day compared to nTg mice treated with vehicle. Tg mice treated with ISRIB took longer to locate the platform on days 1 and 2 compared to nTg mice treated with vehicle. Tg mice treated with ISRIB took longer to locate the platform on day 2 compared to nTg mice treated with ISRIB. All groups located the platform significantly faster on day 5 compared to day 1. F, G, Quantification of quadrant duration during the MWM probe test conducted 24 and 72 h after training. nTg mice treated with vehicle or ISRIB spent significantly more time in the target quadrant compared to the nontarget quadrant during the 24 h (F) and 72 h (G) probe tests. nTg vehicle, n = 23; nTg ISRIB, n = 24; Tg vehicle, n = 15; Tg ISRIB, n = 12. H, Tg mice do not exhibit fear-based associative learning and memory deficits. Quantification of latency to cross during the habituation, training, and testing phases of the passive PA test are shown. All groups took significantly longer to cross during testing compared with the training phase. nTg vehicle, n = 24; nTg ISRIB, n = 23; Tg vehicle, n = 17; Tg ISRIB, n = 17. Asterisks indicate comparisons to nTg vehicle; $ indicate comparisons to nTg ISRIB; # indicate within group comparisons to trial 1. Error bars indicate SEM; $, *p < 0.05; $$, **p < 0.01, ***p < 0.001; #, ****p < 0.0001.

Figure 7.

ER stress does not appear to be implicated in PS19 Tg neuropathology. A, Representative immunoblot of cortical tissue homogenates from nTg and Tg mice probed using antibodies directed against ATF4 and CHOP. CHOP was detected only in PCN lysate derived from E17 Sprague Dawley rats treated with 1 μM thapsigargin, which served as a positive control (+ control). B, Quantification of ATF4 normalized to tubulin. No significant differences in levels of ATF4 were found. nTg vehicle, n = 12; nTg ISRIB, n = 15; Tg vehicle, n = 6; Tg ISRIB, n = 8; + control (n = 2). C, Representative immunoblot of hippocampal tissue homogenate from nTg and Tg mice probed using antibodies directed against p-tau (AT8) and total tau (Tau5). D, Quantification of AT8 normalized to Tau5 and Tau5 normalized to GAPDH. Tg mice treated with ISRIB have significantly more AT8 compared to Tg mice treated with vehicle. Tg + vehicle, n = 6; Tg + ISRIB, n = 8. Error bars indicate SEM; *p < 0.05.

Figure 8.

PS19 Tg mice have increased p-tau and evidence of hippocampal degeneration. A, Representative photomicrographs (5×) of the hippocampus in nTg and Tg mice treated with vehicle or ISRIB. Sections stained using DAPI, fluoro-Nissl, and antibody directed at p-tau (AT8). Boxes indicate the ROIs. B, Representative photomicrographs (40×) of the endal limb of the DG in nTg and Tg mice treated with vehicle or ISRIB. Sections stained using DAPI, fluoro-Nissl, and antibody directed at p-tau (AT8). C, Quantification of AT8 in the DG. Tg mice treated with ISRIB have an increased number of p-tau-immunoreactive inclusions in the DG compared to nTg mice treated with vehicle. Error bars indicate SEM; *p < 0.05. D, Quantification of AT8 in the hippocampus. Tg mice have an increased number of p-tau-immunoreactive inclusions throughout CA1, CA3, and the DG compared to nTg mice treated with vehicle. Error bars indicate SEM; **p < 0.01. E, Representative photomicrographs (2.5×) of brain sections from nTg and Tg mice treated with vehicle or ISRIB. Arrows indicate the pyramidal cell layer of CA1. Sections stained using fluoro-Nissl. F, Quantification of CA1 layer thickness. Tg mice have significant reductions in CA1 compared to nTg mice treated with vehicle. nTg vehicle, n = 7; nTg ISRIB, n = 7; Tg vehicle, n = 7; Tg ISRIB, n = 7. Error bars indicate SD; **p < 0.01.

Figure 9.

Effects of genotype and treatment on body weight and mortality in APP^Swe mice. A, Tg mice have reductions in body weight compared with nTg mice. B, ISRIB reduced survival in nTg and Tg mice by the eighth day of treatment. nTg vehicle, n = 14; nTg ISRIB, n = 15; Tg vehicle, n = 15; Tg ISRIB, n = 14. Asterisks indicate comparisons to nTg vehicle; # indicate comparisons to Tg vehicle. Error bars indicate SEM; *p < 0.05; #, **p < 0.01.

Figure 10.

APP^Swe Tg mice exhibit learning and memory deficits and behavioral impairments. A, B, Tg mice display locomotor hyperactivity. A, Quantification of ambulatory distance. Tg mice treated with ISRIB ambulate a greater distance compared to nTg mice treated with ISRIB. B, Quantification of ambulatory duration in center. Tg mice spend more time in the center of the AC compared to nTg mice treated with vehicle. nTg vehicle, n = 14; nTg ISRIB, n = 13 (n = 15 for center duration); Tg vehicle, n = 14; Tg ISRIB, n = 14. C, D, Tg mice exhibit deficits in fear-based learning and memory. C, Quantification of freezing behavior during the acquisition phase of the fear conditioning paradigm. Tg mice treated with ISRIB freeze less during tone 2 compared to nTg mice treated with ISRIB. During ITI2, Tg mice treated with vehicle or ISRIB freeze less compared to nTg mice treated with vehicle, and Tg mice treated with ISRIB freeze less compared to nTg mice treated with ISRIB. D, Quantification of freezing behavior during context- or cue-based retrieval. Tg mice freeze significantly less compared to nTg mice treated with vehicle during context-based retrieval. No significant differences were found in time spent freezing to tone. nTg vehicle, n = 14; nTg ISRIB, n = 15; Tg vehicle, n = 15; Tg ISRIB, n = 14. Asterisks indicate comparisons to nTg vehicle, $ indicate comparisons to nTg ISRIB. E, F, Tg mice do not display deficits in spatial learning and memory. E, Quantification of escape latency during the second MWM acquisition phase and reversal trial (inset). nTg mice treated with vehicle located the platform significantly faster on day 5 compared with day 1. Tg mice treated with vehicle located the platform significantly faster on days 3–5 compared with day 1. nTg mice treated with vehicle and Tg mice treated with vehicle or ISRIB took longer to locate the platform on the reversal trial (inset) compared with day 5. Additional between group comparisons were made for each training day and no significant differences were found; # indicates comparisons made within group to trial 1. F, Quantification of escape latency during the second MWM probe trial. All groups spent significantly more time in the target quadrant compared with the nontarget quadrant. nTg vehicle, n = 10; nTg ISRIB, n = 4; Tg vehicle, n = 11; Tg ISRIB, n = 6. G, H, Tg mice exhibit phenotypic deficits in spatial working memory and short-term recognition memory. G, Quantification of percentage alternation during the Y-maze. nTg mice treated with vehicle had a significantly greater percentage alternation compared with 50% chance alternation. No significant differences were found between groups in total number of entries or percentage of correct entries in the Y-maze (data not shown). nTg vehicle, n = 14; nTg ISRIB, n = 11; Tg vehicle, n = 15; Tg ISRIB, n = 10. H, Quantification of percentage of time spent investigating the novel or familiar object during the NOR test. nTg mice treated with vehicle spent more time investigating the novel object compared with the familiar object. nTg vehicle, n = 12; nTg ISRIB, n = 10; Tg vehicle, n = 15; Tg ISRIB, n = 10. Error bars indicate SEM; #, *p < 0.05; ##, $$, **p < 0.01, ***p < 0.001; $$$$, ****p < 0.0001.