Intranasal insulin rescues repeated anesthesia-induced deficits in synaptic plasticity and memory and prevents apoptosis in neonatal mice via mTORC1

Abstract

Long-lasting cognitive impairment in juveniles undergoing repeated general anesthesia has been observed in numerous preclinical and clinical studies, yet, the underlying mechanisms remain unknown and no preventive treatment is available. We found that daily intranasal insulin administration to juvenile mice for 7 days prior to repeated isoflurane anesthesia rescues deficits in hippocampus-dependent memory and synaptic plasticity in adulthood. Moreover, intranasal insulin prevented anesthesia-induced apoptosis of hippocampal cells, which is thought to underlie cognitive impairment. Inhibition of the mechanistic target of rapamycin complex 1 (mTORC1), a major intracellular effector of insulin receptor, blocked the beneficial effects of intranasal insulin on anesthesia-induced apoptosis. Consistent with this finding, mice lacking mTORC1 downstream translational repressor 4E-BP2 showed no induction of repeated anesthesia-induced apoptosis. Our study demonstrates that intranasal insulin prevents general anesthesia-induced apoptosis of hippocampal cells, and deficits in synaptic plasticity and memory, and suggests that the rescue effect is mediated via mTORC1/4E-BP2 signaling.

Figure 1

Pre-treatment with intranasal insulin prevents repeated general anesthesia-induced impairment in memory formation in male and female mice. (a) Diagram depicting experimental design. Mice were pre-treated daily (postnatal days 8–14) with intranasal insulin or saline before repeated general anesthesia (1.5% isoflurane, 2 h/day) at postnatal days 15–17. (b) Schematic representation of a novel object location test. Male (c) and female (d) adult mice, subjected to repeated general anesthesia postnatally, exhibit decreased discrimination index (time exploring moved object minus time exploring unmoved object divided by total time exploring both objects) and the decrease is prevented by pre-treatment with intranasal insulin (Home cage: n = 9 males and 6 females; Anesthesia + saline: n = 9 males and 7 females; Anesthesia + insulin: n = 6 males and 7 females; Insulin: n = 6 males and 6 females, one-way ANOVA followed by Tukey’s multiple comparisons post hoc test). (e) Comparison between males and females shows no sex-specific effects. Total exploration time (both sexes combined) (f) and exploration of moved and unmoved objects (g) are shown. (h) Schematic representation of a contextual fear conditioning test. Long-term fear memory (24 h after training, recorded during 5 min period) is impaired in anesthesia-exposed mice in males (i) and females (j), and this deficit is prevented by pre-treatment with intranasal insulin (Home cage: n = 9 males and 8 females; Anesthesia + saline: n = 8 males and 10 females; Anesthesia + insulin: n = 6 males and 9 females; Insulin: n = 8 males and 6 females, one-way ANOVA followed by Tukey’s multiple comparisons post hoc test). (k) comparison between males and females. Data are presented as mean ± SEM. n.s. (not significant), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

Insulin applied intranasally rapidly penetrates the brain and strongly activates mTORC1 activity. (a) Postnatal day 8 mice were treated with intranasal FITC-insulin or saline and their brains were fixed 30 min post-administration and immunostained with phospho-S6-specific antibody. Imaging of FITC fluorescence reveals widespread FITC distribution in the brain, indicating the presence of the FITC-insulin in the brain (left). (b) FITC-insulin activates mTORC1 signaling as evidenced by increased phosphorylation of mTORC1 downstream effector, S6. Bottom images represent magnification of the grey rectangle-marked area in the middle images. Scale bar (middle) 4 mm, (bottom) 700 µm.

Figure 3

Intranasal insulin prevents repeated anesthesia-induced impairment in L-LTP. (a) Acute hippocampal slices were prepared from adult mice and field excitatory postsynaptic potentials (fEPSPs) were recorded in the CA1 hippocampal area following theta-burst stimulation (TBS) of the Schaffer collateral pathway. Representative traces at 3-h time-point are shown for each condition (baseline in black and 3 h post-TBS in grey). Column bars on the right show quantification of fEPSP slope (% of baseline) during the last 10 min of recording (average responses during 170–180 min post-TBS). Whereas the late-phase of long-term potentiation (L-LTP, 3 h after stimulation) is impaired in anesthesia-exposed mice, this impairment is rescued in mice pretreated with intranasal insulin (F_{2, 23} = 7.204; control (n = 8 slices from 8 mice) versus Anesthesia + saline (n = 10 slices from 10 mice), **p < 0.01; Anesthesia + saline versus Anesthesia + insulin (n = 8 slices from 8 mice), *p < 0.05; one-way ANOVA followed by Tukey’s multiple comparisons post hoc test). (b) No differences were found between the three experimental groups in input–output curves over a wide range of stimulus intensities. (c) Paired-pulse facilitation is not different between groups. For statistical analysis in (b) and (c), two-way ANOVA followed by Tukey’s multiple comparisons post hoc test was used. Data are presented as mean ± SEM.

Figure 4

Intranasal insulin prevents the induction of apoptosis following repeated general anesthesia in juvenile mice via activation of mTORC1. (a) Repeated general anesthesia (1.5% isoflurane, 2 h/day) was performed on postnatal days 15–17 mice, and the brains were fixed one hour after the last anesthesia session at postnatal day 17 and dorsal hippocampus sections were immunostained against cleaved caspase 3 (CC3). Representative images of hippocampus showing CC3 at low (left) and high (middle) magnification. Yellow circles indicate the location of CC3-positive cells. (Right) Quantification of CC3-positve cells. The numbers of CC3-positive cells in three sequential dorsal hippocampus sections (separated by 200 µm, averaged for both hippocampi) was calculated per animal. Anesthesia increases the number of CC3-positive cells in the hippocampus (****p < 0.0001, a student’s t-test, control n = 6, anesthesia n = 7). (b) Repeated general anesthesia (1.5% isoflurane, 2 h/day, postnatal days 15–17) was performed on mice pre-treated daily during postnatal days 8–14 with saline, insulin, insulin and temsirolimus (applied ICV 20 min before intranasal insulin), or temsirolimus alone. The brains were fixed one hour after the last anesthesia session at postnatal day 17 and hippocampal sections were immunostained against cleaved caspase 3 (CC3). Representative images of CA1 hippocampal area showing CC3 (white) and DAPI (blue). (c) Quantification of CC3-positve cells. The increase in the number of CC3-positive cells following repeated general anesthesia is prevented by intranasal insulin (F_{5, 15} = 76.09; Control (n = 4) versus Anesthesia + vehicle (n = 4), ***p < 0.001; Anesthesia + vehicle versus Anesthesia + insulin (n = 4), ****p < 0.0001, one-way ANOVA followed by Tukey’s multiple comparisons post hoc test). Temsirolimus (2 µl at 50 µg/µl injected icv 20 min before intranasal insulin administration) prevents the rescue effect of intranasal insulin (Anesthesia + insulin (n = 4) versus Anesthesia + insulin + temsirolimus (n = 3), ****p < 0.0001). Data points represent individual mice. Data are presented as mean ± SEM. (d) Diagram showing the activation of mTORC1 by insulin and its inhibition by temsirolimus. 4E-BPs and S6K are two major downstream effectors of mTORC1.

Figure 5

4E-BP2 knockout juvenile mice do not show apoptosis induction following repeated general anesthesia. (a) Western blot analysis shows the lack of 4E-BP2 protein expression in the hippocampus of 4E-BP2 knockout (KO) mice. See Supplementary Fig. 1 for full-length blots. (b) Representative images showing immunostaining against CC3 (white) and DAPI (blue) in hippocampal sections from wild type (WT, left) and 4E-BP2 KO (right) control and repeated anesthesia-exposed mice (1.5% isoflurane, 2 h a day at postnatal days 15–17, fixed one hour after the last anesthesia session). (c) Quantification of CC3-positive cells. Repeated anesthesia induces apoptosis in juvenile wild type but not 4E-BP2 knockout mice (F_{3, 9} = 10.13; WT control (n = 3) versus WT anesthesia (n = 3), *p < 0.05; 4E-BP2 control (n = 3) versus 4E-BP2 anesthesia (n = 4), ns (not significant), p > 0.05. Data points represent individual mice. Data are presented as mean ± SEM.