PKR modulates sterile systemic inflammation-triggered neuroinflammation and brain glucose metabolism disturbances

Abstract

Sterile systemic inflammation may contribute to neuroinflammation and accelerate the progression of neurodegenerative diseases. The double-stranded RNA-dependent protein kinase (PKR) is a key signaling molecule that regulates immune responses by regulating macrophage activation, various inflammatory pathways, and inflammasome formation. This study aims to study the role of PKR in regulating sterile systemic inflammation-triggered neuroinflammation and cognitive dysfunctions. Here, the laparotomy mouse model was used to study neuroimmune responses triggered by sterile systemic inflammation. Our study revealed that genetic deletion of PKR in mice potently attenuated the laparotomy-induced peripheral and neural inflammation and cognitive deficits. Furthermore, intracerebroventricular injection of rAAV-DIO-PKR-K296R to inhibit PKR in cholinergic neurons of ChAT-IRES-Cre-eGFP mice rescued the laparotomy-induced changes in key metabolites of brain glucose metabolism, particularly the changes in phosphoenolpyruvate and succinate levels, and cognitive impairment in short-term and spatial working memory. Our results demonstrated the critical role of PKR in regulating neuroinflammation, brain glucose metabolism and cognitive dysfunctions in a peripheral inflammation model. PKR could be a novel pharmacological target for treating systemic inflammation-induced neuroinflammation and cognitive dysfunctions.

Keywords: laparotomy; microglia; neuroimmune responses; peripheral inflammation; postoperative cognitive dysfunction; protein kinase R; targeted metabolomics.

Figure 1

The experimental workflow. For the first part, both male wild-type C57BL/6J and C57BL/6-Tg(CD68-EGFP)1Drg/J mice were assigned into 2 groups randomly: laparotomy under sevoflurane anesthesia and control group under sevoflurane anesthesia. In the second part, PKR^-/- mice were exposed to laparotomy with sevoflurane anesthesia or sevoflurane anesthesia to examine the role of PKR in regulating systemic inflammation-triggered neuroinflammation. For the third part, intracerebroventricular injection of rAAV-DIO-PKR-K296R into the right lateral ventricle of ChAT-IRES-Cre-eGFP mice was performed to inhibit PKR activation in cholinergic neurons. Mice were then subjected to laparotomy with sevoflurane or sevoflurane anesthesia respectively. The effects of blocking PKR in cholinergic neurons on modulating glucose metabolism and cognitive functions were examined in the laparotomy model.

Figure 2

Laparotomy induced systemic inflammation and neuroinflammation. Relative cytokine mRNA expression levels were measured in the liver, frontal cortex, and hippocampus of C57BL/6J mice at (A) 4 h and (B) 1 day after laparotomy. Representative flow cytometry plots illustrating the increase in CD68-eGFP⁺ cells in the (C) frontal cortex and (D) hippocampus of the CD68-eGFP mice on postoperative day 14. Morphological gating strategy refers to dot plot SSC-H versus FSC-H to eliminate debris and aggregates in the (C) frontal cortex and (D) hippocampus of the control and laparotomy groups. Doublets are excluded by plotting FSC-H versus width and SSC-H versus width. Proportion of CD68-eGFP⁺ cells (labeled with eGFP) and cell viability (labeled with DAPI) were determined by flow cytometry with FITC and pacific blue channel respectively. The data generated are plotted in two-dimensional dot plots in which FITC-A versus Pacific-Blue-A. A four-quadrant gate was established: Q1) live, CD68-eGFP⁺ cells (FITC positive/Pacific blue negative); Q2) dead, CD68-eGFP⁺ cells (FITC positive/Pacific blue positive); Q3) live cells without CD68 expression (FITC negative/Pacific blue negative); and Q4, dead cells without CD68 expression (FITC negative/Pacific blue positive). Percentage of CD68-eGFP⁺ cells in the (E) frontal cortex and (F) hippocampus of the control and laparotomy groups. n = 6 per group. Data are expressed as mean ± S.E.M. Differences were assessed by unpaired two-tailed Student’s t-test as follows: ^* p < 0.05 and ^** p < 0.01 compared with the control group.

Figure 3

Laparotomy induced activation of microglia in the frontal cortex and hippocampus after laparotomy. Representative confocal images of the immunohistochemical staining of DAPI, and Iba1 in the (A) frontal cortex, (C) cornu ammonis (CA) 1, (E) CA3, and (G) dentate gyrus (DG) of the hippocampus sections from CD68-eGFP mice of the control and laparotomy groups on postoperative day 14. Number of Iba1 positive cells, relative fluorescence intensity of Iba1 positive cells, total number of endpoints per Iba1 positive cell, and summed process length per Iba1 positive cell in the (B) frontal cortex, (D) CA 1, (F) CA3, and (H) DG of the hippocampus sections were quantified. n = 5 per group. Data are expressed as mean ± S.E.M. Differences were assessed by unpaired two-tailed Student’s t-test denoted as follows: ^* p < 0.05, ^** p < 0.01 and ^*** p < 0.001 compared with the control group.

Figure 4

Cognitive impairment of C57BL/6J mice following laparotomy. (A) Percentage of the body weight change from baseline of the laparotomy and control groups throughout the postoperative period. (B) Total distance travelled and (C) central duration of mice in the open-field test. (D) Discrimination index, which is the ratio of exploration time of one object to two objects, was measured on the novel object recognition test day 1 (with two identical objects: A+A) and (E) day 2 (with a novel object replaced the familiar one: A+B). (F) Total number of arm entries and (G) percentage of alternation were measured in the spontaneous alternation Y-maze test. In the puzzle box test, the latency time for the mice entering the goal zone in different tasks was measured. All the tasks include (H) task for habituation (T1), (I) tasks for testing problem-solving skills (T2, T5, T8, and T11), and (J) tasks for evaluating short-term (T3, T6, T9, and T12) and (K) long-term memories (T4, T7, T10, and T13) in the puzzle box test. n = 9–11per group. Data are expressed as mean ± S.E.M. Differences were assessed by unpaired two-tailed Student’s t-test denoted as follows: ^* p < 0.05 and ^** p < 0.01 compared with the control group.

Figure 5

Knockout of PKR in mice ameliorated laparotomy-induced peripheral and neural inflammation after surgery. Relative cytokine mRNA expression levels were measured in the liver, frontal cortex, and hippocampus of the C57BL/6J mice and PKR-KO mice at (A) 4 h and (B) 1 day following laparotomy. n = 6 per group. Data are expressed as mean ± S.E.M. Differences were assessed by two-way ANOVA. Bonferroni’s post hoc tests were performed for the following comparisons: between control and laparotomy treatments within each strain, and between WT and PKR-KO mice under the same treatment condition. ^* p < 0.05 and ^** p < 0.01 indicate significant differences between indicated groups.

Figure 6

Knockout of PKR in mice ameliorated activation of microglia in the frontal cortex and hippocampus following laparotomy. Representative confocal images of the immunohistochemical staining of DAPI and Iba1 in the (A) frontal cortex, (C) cornu ammonis (CA) 1, (E) CA3, and (G) dentate gyrus (DG) of the hippocampus sections from PKR-KO mice of the control and laparotomy groups on postoperative day 14. Number of Iba1 positive cells, relative fluorescence intensity of Iba1 positive cells, total number of endpoints per Iba1 positive cell, and summed process length per Iba1 positive cell in the (B) frontal cortex, (D) CA1, (F) CA3, and (H) DG of the hippocampus sections were quantified. n = 5 per group. Data are expressed as mean ± S.E.M. Differences were assessed by unpaired two-tailed Student’s t-test.

Figure 7

Genetic deletion of PKR in mice alleviated laparotomy-induced cognitive impairment. (A) Percentage of the body weight change from baseline of the C57BL/6J mice and PKR-KO mice throughout the postoperative period. (B) Total distance travelled and (C) central duration of mice in the open-field test. (D) Discrimination index, which is the ratio of exploration time of one object to two objects, was measured on the novel object recognition test day 1 (with two identical objects: A+A) and (E) day 2 (with a novel object replaced the familiar one: A+B). (F) Total number of arm entries and (G) percentage of alternation were measured in the spontaneous alternation Y-maze test. In the puzzle box test, the latency time for the mice entering the goal zone in different tasks was measured. All the tasks include (H) task for habituation (T1), and (I) tasks for testing problem-solving skills (T2, T5, T8, and T11), (J) tasks for evaluating short-term (T3, T6, T9, and T12) and (K) long-term memories (T4, T7, T10, and T13). n = 8–11 per group. Data are expressed as mean ± S.E.M. Differences were assessed by two-way ANOVA. Bonferroni’s post hoc tests were performed for the following comparisons: between control and laparotomy treatments within each strain, and between WT and PKR-KO mice under the same treatment condition. ^* p < 0.05 and ^** p < 0.01 indicate significant differences between indicated groups.

Figure 8

Inhibition of PKR in cholinergic neurons ameliorated laparotomy-induced cognitive impairment. (A) Percentage of the body weight change from baseline of the ChAT-IRES-Cre-eGFP mice of the control group injected with saline (SAL) and the treatment group injected with the rAAV-DIO-PKR-K296R (AAV) throughout the postoperative period. (B) Total distance travelled and (C) central duration of mice in the open-field test. (D) Total number of arm entries and (E) percentage of alternation were measured in the spontaneous alternation Y-maze test. In the puzzle box test, the latency time for the mice entering the goal zone in different tasks was measured. All the tasks include (F) task for habituation (T1), (G) tasks for testing problem-solving skills (T2, T5, T8, and T11), (H) tasks for evaluating short-term (T3, T6, T9, and T12) and (I) long-term memories (T4, T7, T10, and T13). n = 6–8 per group. Data are expressed as mean ± S.E.M. Differences were assessed by two-way ANOVA. Bonferroni’s post hoc tests were performed for the following comparisons: between control and laparotomy treatments within each condition (SAL or AAV), and between SAL and AAV groups under the same treatment condition (control or laparotomy). *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001 indicate significant differences between indicated groups.

Figure 9

Inhibition of PKR in cholinergic neurons ameliorated laparotomy-induced changes in glucose metabolism in the frontal cortex. Two-dimensional score plots using the (A) principal component analysis (PCA) and (B) partial least squares discriminant analysis (PLS-DA). (C) Heatmap of the metabolites involved in central carbon metabolism in the frontal cortex of the ChAT-IRES-Cre-eGFP mice injected with saline (SAL) or rAAV-DIO-PKR-K296R (AAV). Both groups of mice were then randomly assigned into the laparotomy and control groups. The effects of AAV treatment and laparotomy on glucose metabolism were analyzed using gas chromatography-tandem mass spectrometry (GC-MS/MS). The column represents the samples, and the row displays the glucose metabolites. The heatmap scale ranges from −3 to 3. The brightness of each color corresponded to the magnitude of the log2 fold change of the measured metabolites. Concentration of (D) glucose, (E) phosphoenolpyruvate, (F) succinate, and (G) 6-phosphogluconate were quantified in nmol/g. n = 4 per group. Data are expressed as mean ± S.E.M. Differences were assessed by two-way ANOVA. Bonferroni’s post hoc tests were performed for the following comparisons: between control and laparotomy treatments within each condition (SAL or AAV), and between SAL and AAV groups under the same treatment condition (control or laparotomy). ^* p < 0.05 and ^** p < 0.01 indicate significant differences between indicated groups.