Dichloroacetate Prevents Sepsis Associated Encephalopathy by Inhibiting Microglia Pyroptosis through PDK4/NLRP3

Abstract

Dichloroacetate (DCA), a pyruvate dehydrogenase kinase inhibitor, is often used to treat lactic acidosis and malignant tumors. Increasing studies have shown that DCA has neuroprotective effects. Here, we explored the role and mechanism of DCA in Sepsis associated encephalopathy (SAE). Single-cell analysis was used to determine the important role of PDK4 in SAE and identify the cell type. GO and GSEA analysis were used to determine the correlation between DCA and pyroptosis. Through LPS + ATP stimulation, a microglia pyroptosis model was established to observe the expression level of intracellular pyroptosis-related proteins under DCA intervention, and further detect the changes in intracellular ROS and JC-1. Additionally, a co-culture environment of microglia and neuron was simply constructed to evaluate the effect of DCA on activated microglia-mediated neuronal apoptosis. Finally, Novel object recognition test and the Morris water maze were used to explore the effect of DCA on cognitive function in mice from different groups after intervention. Based on the above experiments, this study concludes that DCA can improve the ratio of peripheral and central M1 macrophages, inhibit NLRP3-mediated pyroptosis through ROS and mitochondrial membrane potential (MMP). DCA can reduce neuron death caused by SAE and improve cognitive function in LPS mice. In SAE, DCA may be a potential candidate drug for the treatment of microglia-mediated neuroinflammation.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s10753-024-02105-3.

Introduction

Sepsis associated encephalopathy (SAE) is one of the major public health issues that threaten public health and hinder social development and its solution is urgently needed [1]. Cognitive dysfunction is one of the main clinical manifestations of SAE [2, 3], which is associated with high morbidity and mortality [4]. Although research has confirmed that neuroinflammation [5, 6], autophagy [7] and neuronal damage [8] are factors related to the occurrence of SAE, the potential pathological mechanisms of SAE involve multiple factors. Therefore, we need to gain a deeper understanding of the physiological and pathological mechanisms of SAE and develop new drugs for practical treatment.

PDK4 is a key mitochondrial energy metabolism kinase [9]. Existing research has shown that PDK4 plays a crucial role in regulating cellular metabolism. For example, PDK4 can effectively promote the transition of cancer cells to glycolytic metabolism, thereby driving cancer cell proliferation [10, 11]. Additionally, in sepsis-related diseases, studies have indicated that PDK4 levels show an upward trend in both the peripheral blood of patients with septic cardiomyopathy and the heart tissue of septic cardiomyopathy mice [12, 13]. Our preliminary experiments have also observed increased expression of PDK4 in peripheral blood mononuclear macrophages of sepsis patients and in the hippocampus tissue of septic mice. It’s worth noting that current research on the association between PDK4 and brain diseases remains relatively scarce. Therefore, deeply exploring the connection between PDK4 and SAE has become an important research direction for this study.

DCA can inhibit four pyruvate dehydrogenase kinases [14] and is commonly used to treat lactic acidosis and malignant tumors [15, 16]. Increasing evidence suggests that DCA has neuroprotective effects [17–19], including regulating metabolism, improving oxidative stress and reducing neuroinflammation [20–22]. However, the underlying mechanism of the neuroprotective effects of the DCA/PDK4 axis on SAE remains to be elucidated.

In this study, we found that DCA can inhibit the activation of M1 macrophages, decrease neuronal death and improve the cognitive defects caused by SAE. Mechanistically, DCA may exert its neuroprotective effects by inhibiting NLRP3-mediated cell pyroptosis through PDK4 and suppressing the secretion of inflammatory factors.

MATERIALS AND METHODS

Experimental Animals

Male C57BL/6 mice (6–8 weeks old) were purchased from Beijing Vital River Laboratory Animal Technology. We would like to express our gratitude to the School of Pharmacy for providing NLRP3 KO mice. The experimental animal license number of Wenzhou Medical University is WYYY-IACUC-AEC-2024-048. The mice were housed in a controlled environment (12-hour light/dark cycle; 22 °C; 50–60% humidity) with free access to sterile water and food. All animal husbandry and experiments were conducted in accordance with the ethical guidelines established by the Animal Experimentation Committee of the First Affiliated Hospital of Wenzhou Medical University.

Sepsis Model

A sepsis mouse model was established by intraperitoneal injection of LPS (10 mg/kg; L2880, Sigma–Aldrich). Mice in the sham-operated group received an intraperitoneal injection of an equal volume of PBS. After 24 hours, the mice were anesthetized and the lungs were perfused with physiological saline until they appeared white. Subsequently, the hippocampus was collected for histological analysis or rapidly frozen in liquid nitrogen for cryopreservation.

Novel Object Recognition Test

The Novel Object Recognition Test consists of three stages. The first stage is the adaptation phase, where mice are allowed to explore the open field for 5 minutes. After 24 hours, the training phase begins, allowing mice to explore the field for 5 minutes with two identical objects placed in parallel. One hour later, the testing phase takes place. Mice are permitted to explore the field for 3 minutes, where one familiar object and one novel object are placed in parallel. Before each trial, the field is cleaned with 70% ethanol to minimize olfactory cues. Testing is conducted in a soundproof room with a temperature controlled between 22 and 25 °C. A light source (40-watt fluorescent lamp) is positioned 130 cm above the field, providing uniform illumination of approximately 55 lux. The Novel Object Discrimination Index (NODI) is calculated using the formula: the number of contacts with the novel object / the total number of contacts with both old and novel objects × 100.

Morris Water Maze (MWM) Test

Seven days after intraperitoneal injection of LPS, the MWM test was conducted to assess the spatial learning and memory abilities of the mice. First, the mice were trained for four consecutive days, followed by the experimental session on the fifth day. The MWM consisted of a circular steel pool (diameter of 1.2 m, height of 0.6 m) and a submerged platform (diameter of 0.1 m). The water temperature in the pool was maintained at 23 °C and the submerged platform was located in the northeastern quadrant of the pool, approximately 1 cm below the water surface. The water was made opaque using acrylic dye. During the training period, each mouse was randomly placed in a different quadrant each day. The mouse was allowed to search for the platform for 60 seconds and the time taken to reach the platform was recorded. If the mouse did not find the platform within 60 seconds, it was placed on the platform for 10 seconds. On the fifth day, the platform was removedand the mouse was placed into the water in the quadrant opposite to where the platform had been. Each mouse was allowed to swim freely for 60 seconds and the number of crossings over the platform location and the percentage of time spent in the target quadrant were recorded.

Construction of BV2 Cell Pyroptosis Model and Steps for Drug Intervention in BV2 Cell

Before constructing the pyroptosis model, BV2 cells were pretreated with DCA drug (10 μM; Sigma, 347,795, St. Louis, MO, USA) and/or NLRP3 agonist (1 μM; BMS-986299, MedChemExpress, United States) [23] or H₂O₂ (100 μM；Sigma, 18,304, St. Louis, MO, USA) [24]for 1 hour.

BV2 cells grow to a density of 70%–80%, Pyroptosis model is constructed by stimulating the cells with LPS (1μg/ml) for 6 hours followed by ATP (5 mM) for 30 minutes.

Cell Plasmid Transfection

To establish a cell-based small interference system, appropriate numbers of cells are seeded in medium-sized dishes. Prior to transfection, prepare two 1.5 mL centrifuge tubes, each containing 250 μl of Opti-MEM culture medium. Add 10 μl of Lip2000 to one tube and 2 μg of plasmid or 2.5 μl of siRNA (10 μM) to the other. After 5–10 minutes, gently add the contents of the plasmid tube to the tube containing Lip2000 and incubate for 30 minutes. After 30 minutes, replace the culture medium in the medium-sized dish with 1.5 mL of fresh Opti-MEM. Slowly add the mixture of plasmid and Lip2000 to the dish and drop it in the center while hovering over the dish. Gently swirl the dish in a figure-eight motion to mix the culture medium. After 6–8 hours of transfection, replace the medium with fresh culture medium without antibiotics and continue culturing. Cells are collected for subsequent experiments 48 hours after transfection. SiRNAs targeting mouse Pdk4 [Sense 5’-CUCUACUCUAUGUCAGGUU(dTdT) − 3′; Antisense 5’-AACCUGACAUAGAGUAGAG(dTdT) − 3′].

Detection of Intracellular Reactive Oxygen Species (ROS)

DCFH-DA is diluted with serum-free culture medium at a ratio of 1:1000. The extracellular fluid is removed and an appropriate volume of diluted DCFH-DA is added to cover the cells. Typically, for one well of a six-well plate, the diluted DCFH-DA added should be no less than 1 ml. The treated cells are then incubated in a 37 °C cell culture incubator for 20 minutes. After three washes with serum-free cell culture medium or PBS to remove residual extracellular DCFH-DA, the washed cells are collected into a flow tube for intracellular ROS detection using a flow cytometer.

Enhanced Mitochondrial Membrane Potential Detection Kit (JC-1)

Preparation of JC-1 staining working solution: Take an appropriate amount of JC-1 (200X) and mix it with 1 ml of JC-1 staining buffer for every 5 μl of JC-1 (200X). Use a pipette to pipette multiple times until the mixture is uniform to form the JC-1 staining working solution. For a six-well plate, the cell supernatant is removed and the cells are washed once with PBS. Then, 1 ml of cell culture medium and 1 ml of JC-1 staining working solution are added and mixed well. The stained cells are incubated in a 37 °C cell culture incubator for 20 minutes. After incubation, the supernatant is aspirated and the cells are washed twice with JC-1 staining buffer or PBS. 2 ml of cell culture medium is added and the cells are observed under a fluorescence microscope or collected into a flow tube for detection using a flow cytometer.

Detection of Apoptosis

After rinsing the cells with pre-cooled PBS, they are centrifuged and 1–10 × 10⁵ cells (including cells in the culture supernatant) are collected using a cell counter. 5× Binding Buffer is diluted with double-distilled water to a 1× working solution and 500 μl of 1× Binding Buffer is used to resuspend the cells. 5 μl of Annexin-V and 10 μl of PI/7-AAD are added to each tube. After gently vortexing to mix, the cells are incubated in the dark at room temperature for 5 minutes. Flow cytometry analysis is then performed according to the experimental protocol.

Cytokine Detection, Antibody, Cell Staining and Flow Cytometry

After anesthesia with isoflurane (RWD), mouse head twitching was observed. Once the mouse entered a state of shallow anesthesia, it was promptly removed and the eyeballs were extracted using a forceps. With the head end pointing down, blood was collected from the eyeballs into a venous blood collection tube. Red blood cells were separated from the whole blood using a red blood cell lysis buffer and the remaining cells were stained with antibodies for flow cytometry. After termination of the red blood cell lysis buffer, the cells were centrifuged and the cell pellet was surface-stained in PBS containing 2% BSA or FBS (w/v). To detect cytokine production (IL-6 and TNF-α), lymphocytes were stimulated for 5 hours in the presence of a cell stimulation mixture (plus protein transport, 00–4975-93, ThermoFisher). Intracellular cytokine staining (ICS) for IL-6 and TNF-α was performed using the Cytofix/Cytoperm Fixation/Permeabilization Kit (554,714, BD Biosciences). Flow cytometry data were collected using a BD Fortessa (BD Biosciences) and analyzed using FlowJO (Tree Star).

Western Blot Analysis

Equal amounts of protein samples (30 μg for cellular protein and 50 μg for animal protein) were loaded into SDS-PAGE gel wells. The proteins were flattened at a low voltage of 30 V and then electrophoresis was continued at 80 V. Electrophoresis was stopped when the marker separated to a suitable position. Subsequently, the proteins were transferred to a PVDF membrane (Millipore, Billerica, MA, USA) and the membrane was incubated with 5% skimmed milk at room temperature for 1 hour. After washing with TBST three times, the membrane was incubated overnight with the corresponding primary antibody at 4 °C. The primary antibodies used are listed in Supplementary Table 1. The membrane was then incubated with the corresponding secondary antibodies at room temperature for 1 hour. The immunoblot bands were detected using ECL (NCM Biotech, China). Finally, we used ImageJ software to analyze the grayscale values of the protein bands.

Real-Time PCR Analysis

Total RNA was extracted from cells using Trizol (Invitrogen). Reverse transcription and transcription systems were configured using TOROIVD® qRT Master Mix 2.0 and TOROGreen® 5G qPCR Premix 2.0 reagents, respectively. Real-time PCR was performed using the ABI Q6 real-time PCR system (Applied Biosystems, Foster City, CA, USA). β-actin expression (B661202–0001, Sangon Biotech) was used as the housekeeping gene and data were normalized using the 2-ΔΔCT method. All primers used in this study were chemically synthesized by Sangon Biotech (Shanghai). The sequences of all primers are shown in Supplementary Table 2.

Immunofluorescence Staining

Mice were anesthetized 24 hours after LPS stimulation and perfused with 0.9% saline and 4% paraformaldehyde through the heart. The entire brain was removed and fixed in 4% paraformaldehyde for 24 hours, followed by paraffin or frozen sectioning. The sections were permeabilized with 0.5% Triton X-100 and blocked with 2% bovine serum albumin, then incubated overnight at 4 °C with the primary antibodies. The primary antibodies used for cell and tissue fluorescence in Supplementary Table 1. The sections were washed with PBST and incubated with the secondary antibodies in the dark for 1 hour. Finally, the sections were stained with DAPI for 5 minutes at room temperature and images were captured using a fluorescence microscope.

Flow Cytometry Methods

Detection of Peripheral and Central Macrophages and their Subtypes

After anesthesia with isoflurane, the mouse eyeballs were clamped using a forceps and blood was collected into heparin tubes. The blood was diluted and processed through Ficoll density gradient centrifugation to obtain mouse immune cells. After euthanasia, the mouse brain tissue was obtained through dissection, digested and processed through Percoll density gradient centrifugation to obtain immune cells from the brain tissue. The immune cells were surface-stained with flow cytometry antibodies such as Fixable viability stain 700, CD45, CD11b, F4/80 and CD86 for identification. Microglia were defined as Live⁺ (Fixable viability stain 700⁻) CD45^intCD11b⁺F4/80⁺. CD86⁺ indicated the M1 subtype. The gating strategy of flow cytometry is shown in Supplementary Table 3.

Detection of Apoptosis Status of Mouse Neurons

After perfusion through the body, the brain tissues of experimental mice were extracted and the connective tissue and surface blood vessels were carefully removed. After digestion and resuspension, some cells were taken for flow cytometry staining. According to previous literature, CD24 is mainly expressed on the surface of myeloid cells, lymphocytes and neurons [25–27]. Firstly, CD45 was set up through flow cytometry staining to exclude myeloid cells and lymophocytes. In addition, apoptosis status of neuronal cells was determined through surface staining of CD24, 7-AAD and Annexin-V. Apoptotic neuron cells were defined as CD45⁻CD24⁺Annexin-V⁺7-AAD⁻.

Cytokine Measurement

Cell supernatants and mouse peripheral blood were thawed at room temperature. The concentrations of IL-1β and IL-18 in the samples were then measured using ELISA kits (4403, 3748, MEIMIAN).

FJB Staining

FJB staining is a fluorescent substance that binds to degenerated neurons, resulting in green fluorescence, while normal neurons do not bind to FJB and do not emit fluorescence. The paraffin sections of each group were dewaxed to water and FJB working solution was added. A histological pen was used to draw a circle around the tissue periphery and FJB green fluorescent probe was added at a dilution of 1:400 using 50% glacial acetic acid as the solvent. The sections were incubated overnight at 4 °C. After air-drying the sections, they were transparentized with xylene for 1 minute and a drop of neutral resin was added for sealing. Images were captured using a fluorescence microscope.

Statistical Analysis

All quantitative data in this study are presented as Mean ± SEM. GraphPad Prism 8 was used for statistical analysis. One-way analysis of variance (ANOVA) was used to assess differences among three or more groups. When comparing only two groups, an unpaired t-test was used. Statistics are deemed significant when ^*p < 0.05, ^**p < 0.01 or ^***p < 0.001.

RESULTS

PDK4 Is Associated with Microglia Activation and May Be a Potential Target for SAE

This study utilized the single-cell transcriptomics dataset (GSE224095) of sepsis patients from the GEO database, dividing them into normal group and sepsis group. Using the relevant single-cell data packages in Rstudio software, peripheral blood mononuclear cells were clustered into seven different cell types and automatically annotated and labeled using SingleR (Fig. 1a). Some studies have indicated that PDK4 can serve as a potential research protein in the serum of sepsis neonates [13, 28], while other studies have found that PDK4 is associated with sepsis-induced lung injury and sepsis-induced liver injury [29–31]. Therefore, our aim was to investigate whether PDK4 is associated with SAE. Using the above dataset to explore the changes of PDK4 gene in peripheral blood mononuclear cells of patients with sepsis, we found that the PDK4 gene was mainly expressed in monocytes. Compared with the control group, the number of PDK4-positive cells in monocytes of the sepsis group increased (Fig. 1b). Simultaneously, this study conducted an in-depth exploration of the expression of PDK1, PDK2 and PDK3 proteins in monocytes. The results showed that there were no significant differences in the expression levels of these three proteins among sepsis patients. This finding further underscores the potentially crucial role of PDK4 in sepsis (Fig. 1c). Monocytes are the precursors of macrophages, differentiating into specific macrophages after reaching tissues and differentiating into microglia in the brain [32, 33]. Therefore, based on the increased expression of PDK4 in peripheral monocytes, this study speculated whether there would be synchronized changes in microglia. Immunofluorescence showed that compared to the control group, the number of Iba-1⁺/PDK4⁺ double-positive cells in the brain increased in the LPS group and there was fluorescent colocalization between Iba-1 and PDK4 (Fig. 1d-e). Based on the above experimental results, we observed increased expression of PDK4 in microglia during SAE, which may suggest a potential correlation between PDK4 and the activation state of microglia. Research indicates that pyruvate dehydrogenase kinase plays a crucial metabolic regulatory role in the polarization of macrophages into the M1 phenotype, serving as an important metabolic checkpoint. When the PDK4 gene is knocked out, it effectively blocks the polarization process of macrophages towards a proinflammatory phenotype [26]. Western blot experiments both in vivo and in vitro demonstrated that the expression of PDK4 and CD86 increased in the LPS group compared to the control group. With DCA intervention, as the expression of PDK4 decreased, CD86 also decreased, indicating a correlation between CD86 and PDK4 at the protein level (Fig. 1f-g). PCR experiments found that compared to the model group, knocking down PDK4 in BV2 cells reduced markers related to M1-type polarization (Fig. 1h). Therefore, we speculated that there is a close relationship between PDK4 and the activation of M1 microglia and this study began to investigate the intervention effects of DCA on PDK4.

Fig. 1.

PDK4 is associated with microglia activation and may be a potential target for SAE. A Single-cell analysis of cell subset clustering and annotation. B Expression of PDK4 in clustered cells. C Expression of PDK1–4 in monocytes. D Immunofluorescence staining showed the number of PDK4⁺Iba-1⁺ double-positive cells in the hippocampus and cortex (scale bar = 20 μm); N = 3. E Immunocolocalization of PDK4 and Iba-1 in BV2 cells (scale bar = 11.6 μm). F-G Western blot images and quantitation showing protein levels and correlation analysis of CD86 and PDK4 in the hippocampus and BV2; N = 3. H Expression of M1 markers CD86, TNF-α, IL-6 and IL-1β in BV2 cells after knockdown of PDK4 was evaluated by RT-qPCR; N = 3. ^*P < 0.05, ^**p < 0.01 and ^***p < 0.001.

DCA Can Reduce the M1 Type Proportion and Inflammatory Cytokines Release of Peripheral Macrophages and Central Microglia Induced by LPS

DCA is the target of PDK4 and can play a role in different diseases by inhibiting the activity of PDK4 [34, 35]. According to literature searches, excessive doses of DCA can have neurotoxic effects [36, 37]. Many previous experimental animal studies have shown that administering DCA at doses of 50–200 mg/kg has beneficial effects [38–41]. After comprehensive consideration, we adopted doses of 50 mg/kg and 100 mg/kg to explore the relationship between DCA and peripheral and central macrophages under SAE. In the periphery, compared to the LPS group, the D50 and D100 groups reduced the proportion of M1 macrophage polarization. In the central nervous system, compared to the LPS group, the D50 and D100 groups reduced the proportion of M1 microglia polarization (Fig. 2a-b). The above results indicate that DCA can attenuate the polarization of peripheral and intracranial M1 macrophages caused by LPS. In addition, we conducted intervention experiments on mice using two doses of DCA, 50 mg/kg and 100 mg/kg, with neuronal apoptosis as an indicator to evaluate brain damage. The experimental results showed that, compared to the control group, these two doses of DCA did not cause damage to neurons in the mouse brain (Fig. S1d-e). Since DCA itself can cause adverse neurological symptoms, we chose 50 mg/kg as the therapeutic dose for studying sepsis-associated encephalopathy in mice and will continue to study with this dose in the future. M1 macrophage polarization produces a series of proinflammatory cytokines [42, 43]. We studied the changes in peripheral and central inflammatory cytokines by examining IL-6 and TNF-α in response to DCA. Compared to the LPS group, the DCA group reduced the expression levels of IL-6 and TNF-α in the periphery and the central nervous system (Fig. 2c-d). Therefore, DCA can effectively inhibit macrophage M1 polarization and the release of proinflammatory factors induced by LPS.

Fig. 2.

DCA can reduce the M1 type proportion and inflammatory cytokines release of peripheral macrophages and central microglia induced by LPS. A Flow cytometry analysis was conducted on M1 type (CD45^intCD11b⁺F4/80⁺CD86⁺) microglia in the brain tissues of mice in the LPS and LPS + DCA groups. Statistical analysis was performed on the proportion of M1 in each group; N ≥ 3. B Flow cytometry analysis was conducted on M1 type (CD45⁺CD11b⁺F4/80⁺CD86⁺) macrophages in the peripheral blood of mice in the LPS and LPS + DCA groups. Statistical analysis was performed on the proportion of M1 in each group; N ≥ 3. C-D Flow cytometry analysis was conducted on the levels of inflammatory cytokines TNF-α and IL-6 in the brain or peripheral blood of mice in the LPS and LPS + DCA groups. Statistical analysis was performed for each group; N ≥ 3. ^*P < 0.05, ^**p < 0.01 and ^***p < 0.001.

DCA Can Reduce the Pyroptosis Pathway Mediated by NLRP3 Activation In Vivo and In Vitro after SAE

This study has demonstrated that DCA treatment can alleviate the polarization of peripheral and central M1 macrophages caused by LPS. To further explore the potential mechanism between DCA and M1 macrophages, we obtained the GSE112595 dataset from the GEO database, extracting data from the M1 macrophage group induced by LPS + IFN-γ and the PDK2/4 knockout M1 macrophage group. Differential analysis was performed using the Limma software package (parameters: P value <0.05 and |log2FC| > 0.25), yielding 2304 DEGS. Among them, 1199 genes were downregulated (Green) and 1105 genes were upregulated (Red), with the visualization results presented in a volcano plot (Fig. 3a). We performed GO (FC) analysis on the obtained DEGS and the results showed that knockout of PDK2/4 primarily plays a role in immune responses, neural development and synaptic function (Fig. 3b). GSEA enrichment analysis revealed the top 10 enriched pathways (Fig. 3c). Considering the close relationship between pyroptosis and M1 macrophage polarization [44, 45] and the crucial role of microglia-mediated pyroptosis-induced neuroinflammation in SAE [46–48], we chose the NOD-like pathway as the focus of our exploration of the mechanism of DCA in SAE. NLRP3 is an important member of the NOD-like receptor family, primarily involved in intracellular inflammatory responses. Studies have demonstrated that NLRP3-mediated pyroptosis plays a crucial role in various diseases [49–51]. Therefore, this study primarily aimed to investigate the impact of DCA on the NLRP3-mediated pyroptosis pathway in SAE.

Fig. 3.

DCA can reduce the pyroptosis pathway mediated by NLRP3 activation in vivo and in vitro after SAE. A Volcano plots for the comparisons of M1-PDK2/4 KO vs. M1-WT. dotted lines represent the cutoffs for fold change and p value. Red represents up-regulated genes, while green represents down-regulated genes. B Gene ontology enrichment analysis. C Gene set enrichment analysis. D Western blot images and quantitation showing protein levels of NLRP3, GSDMD-N and IL-1β in each group; N = 3. E Immunofluorescence revealed the average fluorescence intensity levels of NLRP3, GSDMD and IL-1β in various groups of BV2 cells (scale bar = 20 μm). F Western blot images and quantitation showing protein levels of NLRP3, GSDMD-N and IL-1β in each group; N = 3. G Flow cytometry was used to detect the annexin-V and PI double-positive region in BV2 cells. H Western blot images and quantitation showing protein levels of NLRP3, GSDMD-N and Cl-caspase1 in BV2; N = 3. SN = cell supernatant. I ELISA experiment was conducted to detect IL-1β and caspase 1 in the supernatant of different BV2 cell groups; N = 4. J Western blot images and quantitation showing protein levels of NLRP3, GSDMD-N in hippocampus; N = 3. K ELISA was used to detect the secretion levels of IL-18 and IL-1β in the peripheral blood of mice in each group; N ≥ 3. L Immunofluorescence staining showed the number of GSDMD⁺Iba-1⁺ double-positive cells in the hippocampus (scale bar = 20 μm); N = 3. ^*P < 0.05, ^**p < 0.01 and ^***p < 0.001.

Western blot experiments in vivo showed that compared to the LPS group, the DCA group reduced protein expression levels of NLRP3, Gsdmd-N and IL-1β in the hippocampus tissue. Immunohistofluorescence results indicated that DCA treatment reduced the number of Iba-1⁺/Gsdmd⁺ double-positive cells stimulated by LPS. ELISA experiments demonstrated that compared to the LPS group, the DCA group reduced the secretion levels of IL-1β and IL-18 in the peripheral blood of septic mice (Fig. 3j-l). Subsequently, we further validated this conclusion in vitro. We used LPS + ATP to activate the pyroptosis model in BV2 cells. Based on literature evidence, the effective concentration of DCA on microglia is 30 μM. However, due to differences in models, we used 30 μM as an intermediate value and conducted gradient exploration using Western blot experiments. Finally, we found that DCA at concentrations of 10-20 μM could inhibit the protein expression of NLRP3 (Fig. S1a). CCK8 experiment shows that DCA concentration in the range of 10-100 μM has no effect on the activity of BV2 cells (Fig. S1c). Therefore, we chose 10 μM as the concentration for DCA treatment in BV2 cells. Western blot experiments showed that compared to the L + ATP group, the DCA group reduced the expression levels of NLRP3 and its downstream proteins in BV2 cells (Fig. 3d). Cell fluorescence results were consistent with the above findings (Fig. 3e, statistical charts are provided in Fig. S1b). To determine if DCA affects the NLRP3-mediated pyroptosis pathway through the PDK4 pathway, we transfected BV2 cells with siPDK4 and observed that knockdown of PDK4 also inhibited the NLRP3-mediated pyroptosis pathway (Fig. 3f-g). To further verify whether DCA mediates neuroinflammation through the PDK4/NLRP3 axis, this study utilized the NLRP3 activator BMS-986299. Western blot experiment and ELISA experiments showed that when the NLRP3 activator was used, DCA’s original inhibitory effect on NLRP3 and its downstream proteins, including GSDMD-N, IL-1β and Cl-caspase-1, was significantly offset. Based on the above research findings, this study indicates that DCA regulates neuroinflammation by mediating the NLRP3 pathway through PDK4 (Fig. 3h-i).

DCA Inhibits the Activation of NLRP3 in BV2 Cells by Regulating Reactive Oxygen Species and Mitochondrial Membrane Potential

Current research suggests that ion flux, mitochondrial dysfunction, autophagy, production of reactive oxygen species (ROS) and lysosomal damage can activate the NLRP3 inflammasome [52]. Studies have shown that DCA has neuroprotective potential in improving energy metabolism disorders and oxidative stress. In neurological diseases, researchers have found that DCA can regulate changes in reactive oxygen species in the body, thereby reducing the damage caused by oxidative stress [53–56]. Therefore, we comprehensively explored how DCA activates NLRP3. Our experiments found that both the DCA group and the siPDK4 group reduced the production of ROS by pyroptosis (Fig. 4a). To further demonstrate that DCA activates NLRP3 by regulating ROS, this study treated BV2 cells with H₂O₂ and/or DCA. Western blot and ELISA experiments showed that, compared to the DCA group, the protein levels of GSDMD-N, IL-1β, and cleaved Caspase-1 were increased in the DCA + H₂O₂ group. The inhibitory effect of DCA on pyroptosis was influenced by H₂O₂, indicating that DCA activates NLRP3 through the PDK4/ROS pathway (Fig. 4b-c). In addition, changes in intracellular ROS may be related to mitochondrial membrane potential. This study explores the effects of DCA on mitochondrial membrane potential [57, 58]. Experiments have shown that both DCA group and siPDK4 group can alleviate the decrease in mitochondrial membrane potential caused by pyroptosis (Fig. 4d). In summary, in the cell pyroptosis model induced in BV2 cells, DCA can inhibit the activation of NLRP3 by suppressing reactive oxygen species.

Fig. 4.

DCA inhibits the activation of NLRP3 in BV2 cells by regulating reactive oxygen species and mitochondrial membrane potential. A Intracellular ROS levels were analyzed through inverted fluorescence microscopy after different treatments (scale bar = 20 μm); N ≥ 3. B Western blot images and quantitation showing protein levels of NLRP3, GSDMD-N and Cl-caspase1 in BV2; N = 3. C ELISA experiment was conducted to detect IL-1β and Caspase1 in the supernatant of different BV2 cell groups; N = 4. D The MMP of cells was analyzed using an inverted fluorescence microscope after different treatments (scale bar = 20 μm); N ≥ 3. ^*P < 0.05, ^**p < 0.01 and ^***p < 0.001.

DCA Can Rescue Neuronal Death and Reduce Cognitive Dysfunction Caused by SAE

To verify whether DCA treatment can improve neuronal survival, we used FJB to stain the damaged neurons. The results showed that the number of green fluorescent neurons in the cortex and hippocampus increased in the LPS group, while the number of green fluorescent neurons decreased significantly after DCA treatment (Fig. 5e). Flow cytometry experiments indicated that compared to the LPS group, the DCA group reduced the level of neuronal apoptosis (Fig. 5d). We simply constructed a co-culture environment of BV2-HT22 and Western blot and immunofluorescence experiments showed that compared to the LPS group, the DCA group reduced the protein and fluorescence levels of Cl-caspase3 (Fig. 5b-c). In summary, these findings indicate that DCA treatment can reduce neuronal death in the brains of SAE mice.

Fig. 5.

DCA can rescue neuronal death and reduce cognitive dysfunction caused by SAE. A After undergoing various treatments, the supernatant from BV2 cells was transferred to HT22 cells and incubated for 12 hours. B Western blot images and quantitation showing protein levels of Caspase3 and Cl-caspase3 in each group; N = 4. C Immunofluorescence revealed the average fluorescence intensity levels of Cl-caspase3 in various groups of HT22; N ≥ 4. D Flow cytometry was used to analyze the apoptosis of neurons in different groups of the brain tissue; N ≥ 4. E FJB staining in the cerebral cortex and hippocampus DG region in each group; N=3. F The percentage of time spent exploring the novel object relative to the total object exploration time; N = 6. G Escape latency; N = 8. H-I Statistical chart of swimming speed after modeling in each group; Statistical chart of the proportion of target phase in each group; Statistical chart of the number of crossings the platform in each group; Representative water maze presentation of mice in each group; N = 6 ~ 8. ^*P < 0.05, ^**p < 0.01 and ^***p < 0.001.

SAE is often accompanied by cognitive dysfunction after prognosis. To verify whether DCA has any impact on cognitive function, we divided mice into Sham group, LPS group and LPS + DCA group and used novel object recognition test and Morris water maze to evaluate cognitive function in each group. In the novel object recognition test, the DCA intervention group can effectively prevent the decrease of the novel object discrimination index (NODI) in LPS mice, which refers to the percentage of times of contact with novel objects (Fig. 5f). In the Morris water maze test, we trained the mice in each group for four days and it was observed that the cognitive ability of the mice to reach the target platform improved throughout the training tests. Among them, the cognitive learning ability of LPS mice was significantly impaired, while DCA alleviated the learning impairment induced by SAE (Fig. 5g). On the last day, a probe trial without a platform was conducted. The experiment found that compared to the control group, the proportion of the platform quadrant and the number of target platform crossings in the LPS group were reduced. Compared to the LPS group, the DCA group alleviated the cognitive impairment caused by SAE and the proportion of the platform quadrant and the number of target platform crossings in the DCA mice increased (Fig. 5h). This study has already demonstrated that DCA can reduce the level of pyroptosis in SAE, so we took NLRP3 as the target to investigate whether NLRP3 knockout mice could also improve cognitive function. Experiments have shown that compared to the LPS group, the NLRP3 knockout group can alleviate the cognitive defects caused by LPS (Fig. 5f, i). In summary, DCA improves cognitive function in LPS-induced septic mice.

DISCUSSION

SAE is one of the severe complications of sepsis, associated with high incidence and mortality. Clinical guidelines indicate that some sepsis survivors experience long-term cognitive impairment after prognosis [59, 60]. Therefore, how to identify and intervene early in SAE has become a key issue. Pyroptosis is a form of regulated cell death (RCD) and plays a crucial role in the occurrence and progression of sepsis [61]. There have been studies reporting the role of pyroptosis in SAE, especially in microglia pyroptosis [62, 63].

This study took PDK4 as the starting point for research and used single-cell transcription analysis to show that there were differences in the expression of PDK4 in peripheral mononuclear cells between the normal group and sepsis patients. PDK4 is believed to play a role in macrophage polarization and PDK2/4 deficiency can block the transition of macrophages to the M1 proinflammatory phenotype [26]. Based on the relationship between PDK4 and macrophage polarization, we used the PDK4 inhibitor DCA to analyze the M1 polarization ratio of peripheral and central macrophages. Considering that DCA has adverse effects such as reversible peripheral neuropathy, dose-dependency and individual variability [64], the animal dose of DCA in SAE has not been explored. The conventional dose range of DCA in animal experiments is 50–200 mg/kg [21, 41]. Therefore, we used 50 mg/kg and 100 mg/kg to explore the role of DCA in SAE. Both 50 mg/kg and 100 mg/kg can reduce the M1 polarization ratio of peripheral macrophages and central microglia. Finally, we used the lower dose of 50 mg/kg as the DCA dose for subsequent animal experiments.

Using the GSE112595 dataset, we obtained differential genes related to PDK4 and M1 macrophages and began to explore the role of DCA in microglia pyroptosis based on GO and GSEA analysis. The NLRP3 inflammasome is a cytoplasmic multiprotein complex composed of the innate immune receptor protein NLRP3, the adaptor protein ASC, and the inflammatory protease caspase-1, which can respond to microbial infections, endogenous danger signals, and environmental stimuli [65]. The assembled NLRP3 inflammasome can activate the protease caspase-1 to induce gasdermin D-dependent pyroptosis and promote the release of IL-1β and IL-18, which contributes to innate immune defense and homeostasis maintenance [52, 66]. Both in vivo and in vitro experiments have demonstrated that DCA can inhibit the NLRP3-mediated pyroptosis pathway. In addition, since DCA is a key target of PDK4, we transfected BV2 cells with siPDK4 and the results showed that DCA regulates the NLRP3 pathway through PDK4. There are many factors that activate NLRP3. Studies have shown that in different disease models, DCA can regulate reactive oxygen species and mitochondrial function [41], which are NLRP3 activation conditions. The results showed that DCA or siPDK4 can reduce the increase in ROS caused by LPS + ATP stimulation and improve MMP.

SAE can lead to direct neuronal death and increased inflammatory factors and cytokines [67, 68]. We explored the role of DCA in neuronal death and demonstrated through in vivo and in vitro experiments that DCA can effectively improve hippocampus neuronal death caused by SAE. After determining that DCA can regulate NLRP3-mediated microglia pyroptosis, we finally explored the effect of DCA on cognitive function in SAE. The Morris water maze test and novel object recognition test showed that compared with LPS group, DCA + LPS group and NLRP3 KO + LPS group could alleviate cognitive deficits. Therefore, DCA can reduce neuronal death and improve cognitive function in SAE and NLRP3 may be a potential target for SAE cognitive deficits.

However, our study has some limitations that cannot be ignored. This study only explored the role and mechanism of DCA in microglia activation in SAE, but did not further analyze other immune cells in the periphery and central nervous system. In the GO and GSEA analysis, we only selected the Nod-like pathway to carry out research on pyroptosis and should explore other inflammatory and immune aspects in the future. In this study, BV2 cells were used as substitutes for primary microglia, which may not fully and accurately reflect the specific effects of DCA on microglia pyroptosis to some extent. We will continue to study this relationship at the primary microglia in the future. In exploring the factors that activate NLRP3 by DCA, we only described the characteristics and did not further investigate the mechanism that mediates the NLRP3 pathway. DCA has dose-related neurological adverse effects. In subsequent studies, we should combine nanomaterials with drugs to increase drug efficacy and reduce drug toxicity.

CONCLUSION

DCA can reduce the polarization ratio of peripheral and central M1 macrophages in SAE. Further research has found that DCA can inhibit pyroptosis and reduce neuronal damage through the PDK4/ROS/NLRP3 axis in microglia. Targeting PDK4 or using DCA intervention may provide new strategies for the prevention and treatment of SAE.

Author Contribution

XH and YZ designed the entire study, analyzed the results and wrote the manuscript. NW participated in most of the experiments. MZ was responsible for animal and cell modeling. YZ was responsible for writing, reviewing and revising the papers. JL provided biological information services for experiments. YL, XF and WL helped to perform the experiments. JW and YM conceptualized the research, directed the study and prepared the manuscript. All the authors have read and approved the final paper.

Funding Information

This study was supported by the Natural Science Foundation of China (Nos. 82,104,622), Zhejiang Medical and Health Science and Technology Project (Nos. 2024KY1244, 2022KY891, 2023RC045), Wenzhou Science and Technology Bureau Project (Nos. Y20220216, Y2021Y0713), Zhejiang Natural Science Foundation (Nos. LY22H150003, LQ23H090006) .

Data Availability

No datasets were generated or analyzed during the current study.

Declarations

Ethics Approval and Consent to Participate

All animal care and experiments were conducted according to the Institutional Animal Care and Use Committee of Wenzhou Medical University, and all experiments were designed to minimize animal suffering. The study protocol was approved by the Animal Experiment Committee of Wenzhou Medical University (WYYY-IACUC-AEC-2024-048).

Competing Interests

The authors declare no competing interests.