Introduction
SAE is a common and severe neurological complication of sepsis that mainly manifests as long-term cognitive impairment and mental illness and is closely related to increases in morbidity and mortality [
1,
2]. Approximately 70% of patients with severe infection develop cognitive impairment after rehabilitation [
3]. Despite the tremendous efforts of researchers, the pathological mechanism of SAE is still not fully understood. Therefore, studying the pathogenesis of SAE and finding potential targets to improve the quality of life and survival rate of SAE patients have scientific importance and clinical value [
4,
5].
Many studies have reported that excessive neuroinflammation induced by sepsis-induced microglial activation in the brain is an important factor in cognitive impairment. Microglias are central innate immune cells and the main source of inflammatory mediators in the brain. Therefore, these cells are the focus of studies on neuroinflammation. Microglia is rapidly activated in response to various stimuli, including infectious and pathological stimuli or Aβ peptides. Activated microglias secrete many proinflammatory cytokines, such as TNF-α, IL-6 and IL-1β. Significantly increasing in neuroinflammation exacerbates neuronal damage and death, which leads to the behavioural and psychological symptoms of SAE [
6,
7].
Activation of the NLRP3 inflammasome and the occurrence of pyroptosis in microglia are related to the pathogenesis of SAE [
8,
9]. The NLRP3 inflammasome is the most well-defined inflammasome and is a multiprotein complex. The NLRP3 inflammasome is composed of Nod-like receptor (NLR), apoptosis-related spot-like protein containing caspase recruitment domain (ASC) and the caspase-1 precursor. NLR recognizes multiple stimuli, which forms complexes and cleaves pre-caspase-1 into activated caspase-1. Activated caspase-1 then cleaves the pore-forming protein gasdermin D (Gsdmd) before cleaving pre-IL-1 β and pre-IL-18, resulting in pyroptosis and IL-1 β and IL-18 secretion [
10,
11]. Pyroptosis is closely related to SAE. Erbin protects against sepsis-associated encephalopathy by attenuating microglial pyroptosis via the IRE1α/Xbp1s-Ca
2+ axis [
12], and NLRP3/Caspase-1 pathway-induced pyroptosis mediates cognitive deficits in a mouse model of sepsis-associated encephalopathy [
13]. However, the mechanism of NLRP3-mediated pyroptosis in SAE is still unclear.
Previous studies have shown that the motif (TRIM) protein family characterized by a ring finger, B-box zinc finger and spiral coil domain with conventional ubiquitin E3 ligase activity plays an indispensable role in regulating the inflammatory response, innate immunity, cell proliferation and apoptosis [
14]. As a member of the TRIM family, TRIM45 stabilizes p53 through ubiquitin linked to K63 and is a brain tumour suppressor through its E3 ligase activity [
15]. TRIM45 activates the NF-κB pathway through ubiquitin-induced TAB2 during cerebral ischaemia and reperfusion injury, which exacerbates microglia-mediated neuroinflammation and causes neuronal injury [
16]. TRIM45 is highly expressed in human adult and embryonic brains [
17]. However, whether and how TRIM45 plays a role in SAE remain unknown.
In this study, we found that the expression of TRIM45 was increased in septic encephalopathy and confirmed the colocalization of microglia and TRIM45 and that TRIM45 could regulate microglial pyroptosis and neuroinflammation in the brain. In summary, these findings showed that silencing TRIM45 could protect brain function from SAE-related damage and revealed the underlying mechanisms of TRIM45 in SAE.
Materials and methods
Animals
Male C57BL/6 mice (6–8 weeks old) were obtained from Beijing Vital River Laboratory Animal Technology (China). The mice were placed in controlled environments (12-h light/dark cycle; 22 °C; 50–60% humidity) and had free access to bacteria-free water and food. All animal housing and experiments were conducted in accordance with the ethical guidelines formulated by the Animal Experimental Committee of the First Affiliated Hospital of Wenzhou Medical University.
Sepsis model
The sepsis mouse model was established by intraperitoneal injections of LPS (10 mg/kg; from Escherichia coli 055:B5, L2880, Sigma-Aldrich), and mice in the Sham group received intraperitoneal injections of an equal volume of PBS. Twenty-four hours later, the mice were anaesthetized and perfused with normal saline until the lungs were whitened. The hippocampus were collected for histological analysis or frozen in liquid nitrogen for rapid cryopreservation.
Behavioural tests
Morris water maze (MWM) test
Ten days after the intraperitoneal injection of LPS, the MWM test was performed to evaluate the spatial learning and memory abilities of the mice. First, the mice were trained for 4 consecutive days, and then the experiment was carried out on the fifth day. The MWM consisted of a round steel pool (1.2 m in diameter and 0.6 m in height) with a hidden platform (0.1 m in diameter). The water in the pool was maintained at 23 °C, and the hidden platform was located in the southwest quadrant of the pool, approximately 1 cm under the surface. Propylene dye was used to make the water opaque. During the training period, each mouse was randomly placed in a different quadrant each day. The mice were allowed to find the platform within 60 s, and the time to reach the platform was recorded. If the platform was not found within 60 s, to the animal was placed on the platform to rest for 10 s. On the fifth day, the platform was removed, and the mouse was placed in the water in the quadrant opposite the platform. Each mouse swam freely for 60 s, and the number of passes and time spent in the target quadrant were recorded.
Nissl staining
Nissl staining was performed to evaluate neuronal damage and loss. After paraffin embedding and sectioning (4 μm), brain tissues were stained with a 1% toluidine blue solution (Beyotime, C0117, China).
Immunocytochemistry
BV2 cells were treated, soaked in 4% paraformaldehyde for 30 min and then permeabilized for 5 min with 0.2% Triton X 100. After being sealed with 5% BSA for 1 h, the cells were incubated with the indicated antibodies at 4 ℃ overnight. After being washed with PBS three times, the cells were incubated with Daylight 488-coupled secondary antibodies (1:500, CL488-10,594, Proteintech) or 594-coupled secondary antibodies (1:500, CL594-10594, Proteintech) for 1 h, and the nucleus was stained with DAPI. Images were recorded with a Leica confocal microscope.
Frozen embedded brain tissue sections were prepared. After the tissue sections were blocked with 5% bovine serum albumin for 1 h, 4-μm-thick brain slices were prepared and treated with rabbit anti-Gsdmd (1:100, A10164, ABclonal) or mouse anti-NLRP3 (1:100, AG-20B-0042, AdipoGen) overnight. Then, the sections were incubated with Daylight 488 -coupled secondary antibodies (1:500, CL488-10594, Proteintech) or 594-coupled secondary antibodies (1:500, CL594-10594, Proteintech) at room temperature for 1 h. After the sections were washed with PBS 3 times, the nuclei were stained with DAPI. Images were recorded with a Leica confocal microscope.
Cell culture and treatments
The mouse microglia BV2 line, mouse astrocytes MA line and mouse hippocampal neuronal HT22 cell line were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and cultured in DMEM (Gibco, USA) supplemented with 10% FBS and 1% penicillin/streptomycin at 37 °C in a humidified atmosphere containing 5% CO2. Lipopolysaccharide (LPS, L2880, Sigma‒Aldrich) and adenosine triphosphate (ATP, A3377, Sigma‒Aldrich) were used to activate the NLRP3 inflammasome in the cell model. For NLRP3 inflammasome activation, BV2 cells were treated with LPS (1 µg/mL) for 6 h and then treated with ATP (5 mM) for 30 min. DMSO was used as a vehicle control for the treatment conditions.
Plasmids and siRNA transfection
Small interfering RNAs against TRIM45 were purchased from Hanbio Biotechnology (Shanghai, China). Small interfering RNAs against Atg5 (sc-41446, Santa Cruz) were purchased from Santa Cruz Biotechnology. Flag-Atg5, HA-TRIM45, HA-Ub, HA-Ub-K48 and HA-Ub-K63 were purchased from Limibio Biotechnology. These agents were transfected into BV2 cells using Lipofectamine RNAiMAX transfection reagent (13778075, Thermo Fisher Scientific) or Lipofectamine 2000 transfection reagent (11668500, Thermo Fisher Scientific). After 48 h, the cells were used for further experiments. The following sequences were used: siTRIM45 sense: 5′-GGTGGAGTGAAGGCTTTAACG-3′ and negative control siRNA (siNC) sense: 5′- UUCUCCGAACGUGUCACGUTT-3′.
Determination of ROS by flow cytometry
ROS were examined by an ROS assay kit (Beyotime, S0033S, China). A total of 1 × 106 BV2 cells were plated in 6-well plates and treated with LPS (1 μg/mL) for 6 h and ATP (5 mM) for 30 min. Then, the cell culture fluid was discarded, and the cells were carefully washed three times with DMEM. The cells were incubated with DMEM containing 10 µM dichlorodihydrofluorescein diacetate (DCFH-DA) for 30 min at 37 °C. Then, the cells were washed with DMEM three times to eliminate the excess DCFH-DA and collected in centrifuge tubes. Finally, the DCF fluorescence intensity was measured by flow cytometry at wavelengths of 485 nm and 535 nm. ROS levels were analysed by FlowJo Software.
Detection of apoptosis in HT22 and BV2 cells by flow cytometry
Apoptosis in HT22 and BV2 cells were tested by Annexin V-FITC/PI Apoptosis Kit (AP101, MULTI SCIENCES). Adherent cells were collected by 0.25% EDTA digestion and centrifugation (4 °C, 1000 g, for 3–5 min), and 105–106 cells were collected and centrifuged (4 °C, 1000 g, for 3–5 min). Cell pellets were resuspended with 0.8–1 mL of cell staining buffer. Then, 5 μL of Annexin V-FITC staining solution was added, and 10 μL of PI staining solution was added. The mixture was mixed well and incubated at 37 °C for 5 min. Red fluorescence and blue fluorescence were detected by flow cytometry.
JC-1 staining
The mitochondrial membrane potential in BV2 cells was determined via a JC-1 fluorescent probe (Beyotime, C2003S, China), and these cells were incubated with JC-1 working solution for 20 min at 37 °C. After being treated with LPS (1 μg/mL) for 6 h, the cells were treated with ATP (5 mM) for 30 min. Then, JC-1 buffer solution was used to wash the cells at least three times. The results were determined by calculating the rate of the green/red fluorescence intensity, which represented the level of mitochondrial disruption.
PBMC collection, cytokine detection, antibodies, cell staining, and flow cytometry
Mouse eyeball blood (PBMC) was collected after isoflurane anaesthesia. Leukocytes were isolated from whole blood with red blood cell lysis buffer, and the remaining cells were stained. The antibodies and reagents used for flow cytometry are listed in Resources Table
1. Surface staining was performed in PBS containing 2% BSA or FBS (w/v). To detect cytokine production (IL-6 and TNF-α), lymphocytes were stimulated for 5 h in the presence of Cell Stimulation Cocktail (plus protein transport) (00-4975-93, Thermo Fisher). Intracellular cytokine staining (ICS) of IL-6 and TNF-α was performed with the Cytofix/Cytoperm Fixation/Permeabilization Kit (554714, BD Biosciences). Flow cytometry data were acquired with a BD Fortessa (BD Biosciences) and analysed using FlowJo (Tree Star).
Table 1
Flow antibody catalogue
Fixable viability stain 700 | BD | 564997 |
Anti-Mouse CD45 APC-cy7 | BD | 557659 |
Anti-Mouse CD11b AF488 | BD | 557672 |
Anti-Mouse CD86 PE-cy7 | BD | 560582 |
Anti-Mouse F4/80 APC | Biolegend | 123116 |
Anti-Mouse TNF-α PE | Biolegend | 506306 |
Anti-Mouse IL-6 PE | BD | 5204807 |
TRIM45 adeno-associated virus infection
To downregulate the expression of TRIM45 in the mouse brain, we used an AAV9 vector carrying shRNA targeting mouse TRIM45 mRNA (Shanghai Genechem Co., Ltd.) and the core sequence of AAV-shTRIM45 was 5'-GGTGGAGTGAAGGCTTTAACG-3'. Stereotactic surgery to transfer the AAV vector was performed on male mice aged 11–12 weeks (25–30 g) that were anaesthetized with 350 mg/kg 4% chloral hydrate (i.p.). The mice were fixed on a stereotactic instrument, the skull was pierced by a drill, and the microsyringe was driven by a stepper motor. A total of 500 nL (2.5 × 1012 vg/mL) of virus solution was injected into the hippocampus. The speed was 50 nL/min. Two weeks later, the model was established in transfected mice.
Reverse transcription real-time quantitative polymerase chain reaction (qRT-PCR)
Total RNA was extracted from BV2 cells with TRIzol reagent (Invitrogen, USA). RNA was reverse-transcribed to complementary DNA using RT-qPCR Fast Master Mix (Vazyme, China). Real-time fluorescence quantitative PCR was performed according to the manufacturer’s instructions. β-Actin and gapdh were used as internal controls for IL-1β, IL-18, TRIM45, Atg5, P62 and beclin1 mRNA expression analysis. Gene expression was quantified using the 2−ΔΔCt method. The gene primer sequences are listed in Table
2.
Table 2
Base sequence of each gene
Mouse |
TRIM45-F | TCAGGCAAGACTCATTGTCCT |
TRIM45-R | ACGGATGTCCACTACTGAGAAT |
Atg5-F | TGTGCTTCGAGATGTGTGGTT |
Atg5-R | GTCAAATAGCTGACTCTTGGCAA |
IL-18-F | GACTCTTGCGTCAACTTCAAGG |
IL-18-R | CAGGCTGTCTTTTGTCAACGA |
IL-1β-F | GCAACTGTTCCTGAACTCAACT |
IL-1β-R | ATCTTTTGGGGTCCGTCAACT |
P62-F | ATGTGGAACATGGAGGGAAGA |
P62-R | GGAGTTCACCTGTAGATGGGT |
beclin1-F | ATGGAGGGGTCTAAGGCGTC |
beclin1-R | TCCTCTCCTGAGTTAGCCTCT |
Huamn |
TRIM45-F | ACAAGCTCTGAGGGGTCAATA |
TRIM45-R | CCACCTGAGCATCACATACAG |
gapdh-F | GGAGCGAGATCCCTCCAAAAT |
gapdh-R | GGCTGTTGTCATACTTCTCATGG |
ELISA
The levels of IL-1β and IL-18 in medium of BV2 cells and peripheral blood of mice were determined using an ELISA kit (MEIMIAN) according to the manufacturer’s instructions. The absorbance of the samples at a wavelength of 450 nm was measured with a BioTek microplate reader.
Western blot analysis
RIPA lysis buffer was used to extract protein from hippocampal tissue or cultured cells, and a BCA protein detection kit was used to measure the concentration. Approximately 25 mg of protein was boiled for 5 min at 100 °C, separated by 10% SDS-PAGE, and transferred to PVDF membranes (1620177, Bio-Rad) The PVDF membranes were then blocked with 5% skim milk for 2 h at room temperature and incubated with primary antibodies (TRIM45 (ab169036, Abcam), NLRP3 (ab263899, Abcam), Gsdmd (ab219800, Abcam), IL-1β (ab16288, ABclonal), Atg5 (ET1611-38, HuaBio), LC3 (#4108, CST), SQSTM1/p62 (#23214, CST), beclin1 (JE59-31, Huabio), β-actin (HRP-60008, Proteintech), HA (51064-2-AP, Proteintech), Flag (80010-1-RR, Proteintech), Caspase3 (Abmart, T40044), cl-Caspase3 (Affinity, Asp175)) overnight at 4 °C. Then, the membranes were washed three times with TBST and then incubated with the appropriate horseradish peroxidase-labelled secondary antibodies. An enhanced chemiluminescence reagent was used to view the reaction. We measured the signal intensity using ImageJ AQ7. Standardization was performed using β-actin.
Coimmunoprecipitation
The supernatants were incubated with Atg5 primary antibodies overnight at 4 °C, followed by the addition of 30 µL of protein A/G PLUS-agarose. After that, the solutions were incubated at 4 °C for 6–8 h. Then, the protein A/G PLUS-agarose was washed three times, and the protein attached to the agarose and linked to keap1 was extracted. We discarded the supernatant and resuspended the pellet in 45 µL of 2 × PAGE loading buffer, and boiled it for 5 min at 100 °C.
Human sepsis specimens
To observe cell activity in the peripheral blood of normal subjects and patients with septicaemia, all septic patients hospitalized in the ICU from January 2023 to June 2023 were selected. Sepsis was determined according to the third internationally recognized definition of sepsis and septic shock, and the patients were minors, pregnant or had type 1 diabetes, aplastic diseases or immunosuppressive diseases or patients receiving immunosuppressive therapy were outside the scope of our study. Ethical review of human studies (KU2022-126) was performed by the Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University, and the study was performed in accordance with the Helsinki Declaration and federal policy to protect human subjects. Each participating patient provided informed consent. For further study, peripheral blood mononuclear cells were obtained from blood by density gradient centrifugation.
Statistical analysis
GraphPad Prism 8.3.0 was used to analyse the data and construct the graphs. The data are expressed as the mean ± SEM. Experiments with only 2 groups were analysed by the unpaired two-tailed Student’s t test. Single-factor experiments with > 2 groups were analysed with one-way analysis of variance (ANOVA) with Dunnett’s post hoc test. P < 0.05 was considered statistically significant.
Discussion
TRIM45 functions as an E3 ubiquitin ligase and participates in a variety of cellular signalling pathways. TRIM45 plays an inhibitory role in brain tumours, and overexpression of TRIM45 can inhibit the proliferation and tumorigenicity of glioblastoma cells in vivo and in vitro. Further analysis showed that TRIM45 induced p53 polyubiquitination with a K63 linkage [
15]. Other studies have shown that TRIM45 plays a role in ischaemic stroke. TRIM45 also uses K63 ubiquitin and TAB2 to control microglial NF-kB signalling. Microglia-specific knockdown of TRIM45 in mice can significantly reduce infarct size and neurological deficit scores [
16]. Although TRIM45 is highly expressed in human and embryonic tissues, its role in the central nervous system of mice with septic encephalopathy has not been explored. In this study, we showed that TRIM45 regulated the TRIM45–Atg5–NLRP3 axis to regulate microglial pyroptosis in the central nervous system of mice with SAE.
SAE is a common and serious complication of sepsis with high morbidity and mortality. Some studies have shown that there is a link between brain damage and cognitive impairment in SAE [
38,
39]. Consistent with previous studies, intraperitoneal injection of LPS resulted in severe learning and memory impairment mice, as shown by the water maze test, and a lack of TRIM45 alleviated cognitive impairment in septic mice. More importantly, we found that TRIM45 was upregulated in the hippocampus of septic mice compared to that of untreated mice. We hypothesized that this improvement was due to SAE. The hippocampus is the structure associated with cognitive impairment in sepsis, and impairment of the CA1, CA3, and DG regions is used in many studies to explore cognitive function [
40,
41]. In the Transwell coculture system in vitro, the downregulation of TRIM45 expression inhibited microglial activation induced by LPS + ATP and reduced neuronal apoptosis. We observed severe neuronal damage in the CA1 and DG regions of mice in the LPS group, and AAV-shTRIM45 mice had reduced damage caused by neuronal inflammation in the brain. This study was the first to show the role of TRIM45 in SAE and that it is important in alleviating cognitive impairment induced by SAE.
During septic encephalopathy, neuroinflammation in the brain causes microglial activation, and the secretion of proinflammatory factors can damage surrounding brain tissue [
42]. Microglia are the main sites of abnormal activation and pyroptosis mediated by NLRP3 inflammatory bodies. Once the NLRP3 inflammatory body is activated, its downstream active proteins Gsdmd-N-terminal and cl-caspase-1 are increased, resulting in pyroptosis and IL-1β and IL-18 secretion [
43,
44]. We analysed the GSE76328 dataset and found that TRIM45-related differentially expressed genes were clustered with nod-like pathways. Then, we explored the relationship between TRIM45 and the NLRP3-mediated pyroptosis pathway and found that knockdown of TRIM45 downregulated pyroptosis proteins in BV2 cells stimulated with LPS + ATP. Overexpression of TRIM45 upregulated pyroptosis proteins. In vivo, pyroptosis-related proteins were decreased in AAV-shTRIM45-treated septic mice compared to septic mice. These results indicate that TRIM45 is related to pyroptosis and that knocking down TRIM45 can inhibit the pyroptosis pathway.
Autophagy is the main catabolic process in cells. The fate of damaged cells is determined by the rapid degradation of harmful factors and the coordination of survival and death processes [
45,
46]. Autophagy dysfunction is associated with a variety of inflammation-related diseases [
47,
48]. Studies have shown that the TRIM family can affect disease progression through autophagy. Knockout of TRIM65 can inhibit autophagy and cisplatin resistance in A549/DDP cells by regulating miR-138-5p/ATG7 [
49]. TRIM22 inhibits osteosarcoma progression by destroying NRF2 and activating ROS/AMPK/mTOR/autophagy signalling [
50]. TRIM39 deficiency can inhibit tumour progression and autophagy in colorectal cancer by inhibiting Rab7 activity [
51]. However, the relationship between TRIM45 and autophagy is still unclear. We found that in the model, knocking down or overexpressing TRIM45 affected the protein levels of p62, beclin1, and Atg5, but at the transcriptional level, knocking down TRIM45 had no effect on beclin1 or Atg5. We hypothesize that there is a protein modification process. Autophagy is regulated by ATG proteins, and Atg5 is indispensable for the formation of autophagic vesicles [
52]. Some studies have shown a relationship between Atg5 and NLRP3 inflammatory bodies [
53]. In ulcerative colitis, EZH2 reduces colonic inflammation through the Atg5-NLRP3 axis [
31]. In NLRP3 inflammasome-related diseases, USP22 inhibits NLRP3 inflammatory bodies through Atg5-dependent autophagy to degrade NLRP3 [
23]. Therefore, we further explored the relationship between TRIM45 and Atg5.
TRIM45 has E3 ubiquitin ligase activity. Previous studies have shown that TRIM45 can modify downstream proteins via ubiquitin. We hypothesize that there is also an interaction between TRIM45 and Atg5. We found that there was binding between TRIM45 and Atg5. We further explored the type of ubiquitin modification of Atg5 mediated by TRIM45. The results show that TRIM45 modifies Atg5 with K63 ubiquitin but not K48.
After verifying the relationship between TRIM45 and Atg5, we focused on how Atg5 affects NLRP3. We overexpressed or knocked down TRIM45 and examined changes in NLRP3 and downstream proteins. The results showed that TRIM45 regulated NLRP3 in an Atg5-dependent manner. Considering the correlation between autophagy, ROS and mitochondrial membrane potential [
54,
55], we also examined the changes in these indices in the different groups. The results showed that knocking down TRIM45 could improve mitochondrial membrane potential and reduce ROS production. However, these two indicators are the prerequisite for activating NLRP3. In addition, IP experiments showed that Atg5 could directly act on NLRP3. Based on these results, we believe that TRIM45 plays a role in autophagy, mitochondrial membrane potential and ROS.
However, there are some limitations in our study that cannot be ignored. We used BV2 cell line to replace primary microglia for in vitro experiments. Even though BV2 cell line has been widely used in neuroscience research, it can be used as an alternative model of primary microglia. In the future, we will continue to use primary microglia to conduct in-depth research and increase the reliability of the conclusions. Some studies have shown that TRIM50, which is a member of the TRIM family, can directly induce NLRP3 oligomerization to promote the activation of NLRP3 inflammatory bodies [
56]. This study did not examine whether TRIM45 could directly act on NLRP3 inflammatory bodies, similar to TRIM50. We did not explore how TRIM45 affects IL-1β and IL-18 mRNA levels. We only revealed the relationship between TRIM45 and NLRP3 in autophagy and mitochondrial dysfunction. We showed that there was an interaction between TRIM45 and Atg5, but we did not explore which domain of TRIM45 could bind to Atg5. There are many core autophagy proteins, and whether there are binding sites between TRIM45 and other proteins remains unclear. Although it has been reported that Atg5 can affect the activation of NLRP3, how Atg5 affects NLRP3 was not examined in this study. Therefore, although this study is helpful for understanding the role of TRIM45 in SAE, the findings need further experimental verification.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.