Microglia in the aged brain develop a hypoactive molecular phenotype after surgery

Abstract

Background: Microglia, the resident immune cells of the brain, play a crucial role in maintaining homeostasis in the central nervous system (CNS). However, they can also contribute to neurodegeneration through their pro-inflammatory properties and phagocytic functions. Acute post-operative cognitive deficits have been associated with inflammation, and microglia have been implicated primarily based on morphological changes. We investigated the impact of surgery on the microglial transcriptome to test the hypothesis that surgery produces an age-dependent pro-inflammatory phenotype in these cells.

Methods: Three-to-five and 20-to-22-month-old C57BL/6 mice were anesthetized with isoflurane for an abdominal laparotomy, followed by sacrifice either 6 or 48 h post-surgery. Age-matched controls were exposed to carrier gas. Cytokine concentrations in plasma and brain tissue were evaluated using enzyme-linked immunosorbent assays (ELISA). Iba1⁺ cell density and morphology were determined by immunohistochemistry. Microglia from both surgically treated mice and age-matched controls were isolated by a well-established fluorescence-activated cell sorting (FACS) protocol. The microglial transcriptome was then analyzed using quantitative polymerase chain reaction (qPCR) and RNA sequencing (RNAseq).

Results: Surgery induced an elevation in plasma cytokines in both age groups. Notably, increased CCL2 was observed in the brain post-surgery, with a greater change in old compared to young mice. Age, rather than the surgical procedure, increased Iba1 immunoreactivity and the number of Iba1⁺ cells in the hippocampus. Both qPCR and RNAseq analysis demonstrated suppression of neuroinflammation at 6 h after surgery in microglia isolated from aged mice. A comparative analysis of differentially expressed genes (DEGs) with previously published neurodegenerative microglia phenotype (MGnD), also referred to disease-associated microglia (DAM), revealed that surgery upregulates genes typically downregulated in the context of neurodegenerative diseases. These surgery-induced changes resolved by 48 h post-surgery and only a few DEGs were detected at that time point, indicating that the hypoactive phenotype of microglia is transient.

Conclusions: While anesthesia and surgery induce pro-inflammatory changes in the plasma and brain of mice, microglia adopt a homeostatic molecular phenotype following surgery. This effect seems to be more pronounced in aged mice and is transient. These results challenge the prevailing assumption that surgery activates microglia in the aged brain.

Keywords: Aging; Microglia; Neuroinflammation; Surgery.

Fig. 1

Schematic of experimental design. Young (3–5 months old) or old (20–22 months old) male C57BL/6 mice received a 2-h exposure to 1.2% isoflurane and a 15-min abdominal laparotomy; age-matched controls received 2 h of carrier gas alone. Perfusions were performed at 6- or 48-h post-induction of anesthesia. Plasma was collected for cytokine ELISA. Brain tissue was harvested for thin sectioning and evaluation of microglial phenotype by IHC, and for isolation of microglia by FACS and subsequent analysis by qPCR or RNAseq. RNAseq datasets were compared to previously published data on MGnD and DAM transcriptomic phenotypes. Created with Biorender.com

Fig. 2

Surgery induces systemic and CNS inflammation. a, b IL-6 and CCL2 were quantified by ELISA in the plasma, frontal cortex, and hippocampus of young and old mice 6 h (a) and 48 h (b) after surgery. c and d. 10 × microscopic images of Iba1⁺ cells in the hippocampus (CA1) of young and old mice 6 h (c) or 48 h (d) after surgery. e, f Quantification of Iba1 immunohistochemistry (% area stained; mean gray value, and # of cells) in the CA1, CA3, and DG of the hippocampus 6 h (e) and 48 h (f) after surgery. Scale bars: 100 μm. Data are mean ± SD for 5–7 animals per group. Main effects calculated by two-way ANOVA, Šídák’s multiple comparisons test. *P < 0.05; ***P < 0.001; **** P < 0.0001

Fig. 3

Suppressed neuroinflammation after surgery in aging microglia. a Tgfbr1 expression and b Clec7a expression in the hippocampus (HP) or hemisphere (HM) at 6 and 48 h, quantified by qPCR. Data are mean ± SD for 8–10 animals per group. Two-way ANOVA, Šídák’s multiple comparisons test, *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. c Number of DEGs in surgery versus control mice for each age group, region, and timepoint, as identified by DESeq2 analysis of SmartSeq2-generated gene counts (FDR < 0.2, |log2FC|> 0.58). d, e Heatmaps showing count data, expressed as Z score, for DEGs in the hemisphere and hippocampus of young (d) and old (e) mice, with genes of interest labeled. IPA analysis of top-affected canonical by surgery in the microglia isolated from the hemisphere (f) and hippocampus (g) of old mice

Fig. 4

Overlap of DEGs with profile of DAM. a The number of DEGs affected by surgery among old mice and young mice, in the hemisphere and hippocampus, that overlap with genes associated with DAM. b Comparison of fold changes for DEGs that overlap with DAM in the hemisphere and hippocampus of old mice. c Comparison of fold changes for DEGs that overlap with DAM in the hemisphere and hippocampus of young mice