Surgery-induced neuroinflammatory transcriptional programs in medial prefrontal cortex of mice during early phase of perioperative neurocognitive disorders

Animals and experimental design

A total of 61 male C57BL/6J mice (10–14 week-old; 25–30 g) were purchased from Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China) and acclimated for 2 weeks. All mice were housed in ventilated cages (485 × 200 × 200 mm; 4–5 mice per cage) in standard conditions (12-h light/12-h dark cycle, lights on at 8 am; temperature, 22–24 °C; humidity, 55 ± 10%) with ad libitum access to water and food. Mice were tagged and randomly allocated to each group before any procedure. For open-field test, 10 mice were used in control and surgery groups; for fear conditioning test, 11 mice were used in control and surgery groups; for RNA sequencing (RNA-seq) and quantitative real-time polymerase chain reaction (qRT–PCR), three mice were used in control and surgery groups; for enzyme-linked immunosorbent assay (ELISA), 6–7 mice were used in control and surgery groups. Since whole blood and mPFC tissue collection were required for this study, mice were euthanized at the conclusion of the experimental period. Euthanasia was performed using 1% pentobarbital sodium (50 mg/kg, intraperitoneal injection). All animal experiments were approved by the Animal Care Committee of the First Affiliated Hospital at Zhejiang University School of Medicine (approval number: 2022-1435) and were conducted in compliance with the Animal Research Reporting In Vivo Experiments (ARRIVE) guidelines and the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Surgery of tibial fracture

We used a clinically relevant mouse model of tibial fracture. The postoperative changes in these mice mimic certain cognitive impairments commonly observed in humans after routine orthopedic surgery, while exhibiting no apparent impact on sensory function and motor activities during behavioral tests (Xiong et al., 2018; Xiang et al., 2022). Mice underwent aseptic open tibial fracture surgery with intramedullary fixation under general anesthesia, as previously described (Xiang et al., 2022). Briefly, general anesthesia consisted of induction with 3.0% isoflurane followed by maintenance with 2.0% isoflurane in 30% FiO₂. To expose the tibia, a median incision was made on the disinfected left paw. A 0.38-mm needle was inserted into the tibial medullary cavity, followed by osteotomy of the tibia at the junction of the middle and distal thirds. The fixation needle was left in place, trimmed flush with the tibial cortex, and the skin was sutured. The entire procedure from induction of anesthesia to the end of surgery was completed within 15 min. Mice were allowed to recover spontaneously from anesthesia. Throughout the surgery and recovery, body temperature was maintained at 37 °C using a heating pad. Analgesia (30 mg/kg tramadol) was subcutaneously administered after induction of anesthesia and prior to skin incision. Mice in the control group received identical anesthesia and analgesia but underwent no surgical procedures.

Open-field test

An open-field test was conducted to evaluate the spontaneous locomotor activity and anxiety levels in mice 24 h post-surgery. The apparatus consisted of a 50 cm × 50 cm open arena with 50 cm high walls and a floor divided into 25 equal squares. Briefly, mice were placed individually in the center of the apparatus and allowed to habituate and explore freely for 10 min. The total distance traveled and time spent in the center were automatically recorded using ANY-maze behavior tracking software (Stoelting Co., Wood Dale, IL, USA). The arena was thoroughly cleaned with 75% ethanol between trials.

Fear conditioning test

The contextual and cued fear conditioning test is a behavioral paradigm used to assess learning and memory. Freezing behavior, defined as complete immobility except for breathing, is a typical natural response to fearful situations. The extent of fear learning was recorded in a fear-conditioning chamber (Shanghai Xinruan Information Technology Co. Ltd., Shanghai, China). A decrease in freezing time indicated memory impairment.

RNA-seq analysis

Total RNA was extracted from the mPFC of mice of the Control and Surgery groups (three mice in each group) using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA), and digested with RNase-Free DNase to remove residual DNAs. RNA quality was assayed on an Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and all samples sequenced exhibited an RIN > 8.0. Then, the poly (A) mRNA was enriched by oligo (dT)-attached magnetic beads and followed by fragmented into small pieces using fragmentation reagent. The RNA fragments were copied into first-strand cDNA using reverse transcriptase and random primers. Then second-strand cDNA was synthesized using the prepared second-strand synthesis reaction system. These cDNA fragments then had the addition of a single “A” base and subsequent ligation of the adapter. Library construction and sequencing were performed using a BGISEQ500 platform (BGI-Shenzhen, China). All raw data were deposited in the National Center for Biotechnology Information Sequence Read Archive database (accession ID: PRJNA1077729). The sequencing data were filtered using SOAPnuke (v1.5.2), and clean reads were mapped to the mouse reference genome (mm10) using HISAT2 (v2.0.4). Gene quantification was performed using RSEM (v1.3.1), and expression was quantified as fragments per kilobase of transcript per million mapped reads (FPKM). DESeq2 (v1.4.5) was used to identify differentially expressed genes (DEGs) between the groups. The adjusted p-value, termed Q value, was calculated by controlling false discovery rate. Fold changes (FC) of DEGs were determined based on the FPKM values. DEGs were identified using the criteria of |log₂FC| ≥ 1 and Q value ≤ 0.05. To gain insight into phenotypic changes, Gene Ontology (GO, http://www.geneontology.org/) and Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.kegg.jp/) enrichment analyses of the annotated DEGs were performed. GO and KEGG terms were ranked by Q value, and a Q value ≤ 0.05 served as the cutoff criterion. The statistical power of this experimental design, calculated in RNASeqPower (An online implementation of RNASeqPower is available at https://rodrigo-arcoverde.shinyapps.io/rnaseq_power_calc/) is 0.83. There are six biological and technical replicates used to achieve the claimed statistical power.

qRT–PCR

RNA was reverse-transcribed into cDNA using the PrimeScriptTMRT Reagent Kit (Takara, Shiga, Japan), in accordance with the manufacturer’s instructions. qRT–PCR was performed using a QuantStudio5 thermocycler (Thermo Fisher Scientific, Waltham, MA, USA) and the SYBR Premix EX TaqTM II kit (Takara, Shiga, Japan) following the manufacturer’s instructions. Amplification was performed starting with an initial denaturation step at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 31 s, in 10 μL reaction volume. All samples were amplified in triplicate, and the mean value was used for subsequent calculations. The fold changes in mRNA expression were calculated using the 2^−ΔΔCt method, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. The sequences of primer pairs for the target genes are listed in Table S1. All PCR experiments and analyses adhered to the MIQE guidelines.

Measurement of cytokine levels

After mice were deeply anesthetized, whole blood was collected via left ventricular puncture. The blood samples were centrifuged at 3,000×g for 15 min at 4 °C, and serum was separated and immediately stored at –80 °C until analysis. Subsequently, mice were transcardially perfused with ice cold saline, and the mPFC tissues were dissected on ice under microscopic observation as previously described (Xu et al., 2016), immediately frozen in liquid nitrogen, and stored at –80 °C for further processing. Specifically, 1-mm-thick coronal brain sections were prepared by using a mouse brain mold. The mPFC tissue was then identified and collected with a biopsy punch based on the mouse brain atlas (Paxinos & Franklin, 2019), targeting the region approximately 2.3–1.3 mm anterior to the bregma, encompassing the cingulate, prelimbic and infralimbic subregions. The mPFC tissues were homogenized in NP-40 lysis buffer containing protease and phosphatase inhibitors (Beyotime Biotech, Shanghai, China), and were centrifuged at 12,000×g for 10 min at 4 °C. The supernatant was collected for analysis. Total protein content in the supernatant was quantified using a BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA). A commercially available ELISA kit for detecting Interleukin (IL)-6 level in mice (R&D Systems, Minneapolis, MN, USA) was used according to the manufacturer’s instructions.

Statistical analysis

Statistical analyses were performed using GraphPad Prism software (v7.0; Prism Software Inc., San Diego, CA, USA). Data are expressed as mean ± standard error of the mean (SEM). Comparisons between two groups were performed using an unpaired two-tailed student’s t-test, assuming equal variance. No data points were excluded from the statistical analysis. P < 0.05 was considered statistically significant.

Cognitive decline after surgery

An aseptic model of tibial fracture surgery was established to investigate the underlying mechanisms of PND. Cognitive performance was assessed using a fear conditioning paradigm (Fig. 1A). Compared to that of the anesthesia-matched control group, the surgery group exhibited a significant decrease in freezing time in contextual and cued fear memory tests (Figs. 1B and 1C), suggesting that fear memory performance was impaired after surgery. In a separate cohort, in the open-field test (Figs. 1D and 1E), there was no significant difference in the total distance traveled between the control and surgery group (Fig. 1F), indicating spontaneous locomotor activity was not impaired after surgery. Additionally, time spent in the center zone of the open field was comparable between the two groups (Fig. 1G), suggesting no substantial difference in anxiety-like behaviors.

Transcriptome profile in the mPFC

To reveal the surgery-induced changes in the mPFC at the transcriptional level, RNA-seq analysis and DEG screening were performed on six samples collected 6 h after surgery (three mice in each group, Fig. 2A). The sequencing generated, on an average, 66.90 million clean reads with a clean read ratio of 93.02%. From the mouse reference genome, 18,029 mRNAs were identified, corresponding to 202,923 genes.

In total, 178 genes were differentially expressed between control and surgery mice (Table S2), and 105 were upregulated (log₂FC ≥ 1, Q value ≤ 0.05), while 73 were downregulated (log₂FC ≤ –1, Q value ≤ 0.05) (Fig. 2B). These findings indicated that significant reprogramming of gene expression in the mPFC was acutely induced after surgery.

GO enrichment analysis

To gain insight into the functions of these DEGs, GO analysis was performed, and the DEGs were categorized into three distinct ontologies such as biological process (BP), cellular component (CC), and molecular function (MF).

In the BP category (Fig. 3A), among the 20 top enriched terms, the majority were related with immune processes, including leukocyte chemotaxis, lymphocyte chemotaxis, monocyte chemotaxis, chemokine-mediated signaling pathway, neutrophil chemotaxis, cellular response to IL-1, positive regulation of neutrophil differentiation, cellular response to tumor necrosis factor (TNF), response to bacterium, positive regulation of mononuclear cell migration, astrocyte activation, inflammatory response and cellular response to interferon-gamma (IFN-γ), which suggest a significant involvement of immune activity. In the CC category (Fig. 3B), the enriched terms included NADPH oxidase complex, protein complex involved in cell adhesion, T cell receptor complex and immunological synapse. Regarding the MF class (Fig. 3C), several enriched GO terms pertained to inflammation-related binding processes, including chemokine activity, CCR7 chemokine receptor binding, CCR chemokine receptor binding, chemokine receptor binding, cytokine activity and IL-6 receptor binding. These findings suggest an active inflammatory response in the mPFC region.

KEGG pathway analysis

To further elucidate the biological functions of the identified DEGs, KEGG pathway enrichment analysis was performed. Surprisingly, the three top significantly enriched pathway terms were cytokine–cytokine receptor interaction, nuclear factor kappa beta signaling pathway, and chemokine signaling pathway (Fig. 3D). Moreover, several pathways associated with neurogenic diseases (neuroactive ligand–receptor interaction, cell adhesion molecules, and tight junction) were among the 20 top enriched terms. Of note, several inflammation-related pathways involved in signal transduction were also enriched, including cAMP, Jak-STAT, TNF and IL-17 signaling pathways, Th1 and Th2 cell differentiation, complement and coagulation cascades, and Toll-like receptor signaling pathway. Together with GO analysis, these results suggest the presence of neuroinflammation and significant neuroinflammatory transcriptional reprogramming in the mPFC during the early phase of PND.

qRT–PCR validation of RNA-seq results

To validate the RNA-seq data, 15 DEGs of interest were selected for qRT–PCR analysis. The DEGs were both statistically robust and biologically relevant to neuroinflammatory pathways or cognitive function, aligning with our study’s focus (Table S3). Their expression analyzed by qRT–PCR exhibited changes similar to the RNA-seq data (Figs. 4B and 4C), with 10 genes (Vipr2, Lcn2, Cxcl5, Ccl21c,Ttr, Cybb, Ano2, Scgn, Lrg1, and Cd28) upregulated and five (Nts, Gh, Stat4, Htr1d, and Agt) downregulated. Among them, Vipr2, Lcn2, Ttr, Ano2, Scgn, Gh, Nts, and Agt have been implicated in neurological dysfunction and disease, and Lcn2, Cxcl5, Ccl21c, Cybb, and Cd28 play roles in immune responses. Other genes selected for qRT–PCR analysis included Lrg1, Stat4, and Htr1d, which are involved in signal transduction and neurotransmission. Importantly, all of these genes exhibited substantial regulation in both RNA-seq and qRT–PCR analyses.

Levels of inflammatory cytokines

To verify the immune response mediators, levels of representative proinflammatory cytokine were evaluated by ELISA and qRT–PCR (Fig. 4A). Notably, ELISA indicated that IL-6 level was significantly elevated in serum and mPFC 6 h post-surgery (Figs. 4D and 4E). However, no significant alternations were noticed in the mRNA levels of Il-6, Tnf-α and Il-1β in the mPFC after surgery (Fig. 4F). These results suggested that a distinct neuroinflammatory response was induced during early phase of PND.