Introduction

Sepsis is an organ dysfunction that is caused by an imbalanced response of the body to infection, and it has a high incidence and mortality rate [1]. The brain is the first affected organ during sepsis [2]. Sepsis-associated encephalopathy (SAE) is a diffuse brain dysfunction, mainly manifested as changes in consciousness and cognitive dysfunction [2]. The main parts of the brain affected by sepsis are the frontal cortex and hippocampus [3]. The latter plays a key role in the process of learning and memory [4]. SAE increases the mortality of sepsis in the acute phase, and causes cognitive dysfunction and loss of attention, memory and executive ability in the long-term [5, 6]. The pathogenesis of SAE is still unclear. It is considered that some processes including microglia activation, blood-brain barrier destruction, hypoxia, neurotransmitter imbalance, axon and neuron loss are involved [7, 8]. There is no effective treatment for SAE-associated cognitive dysfunction so far. Therefore, more basic research should be performed to understand the pathological process of SAE and explore new therapeutic targets.

Glutamate represents the main excitatory neurotransmitter in the brain. During sepsis, the release of glutamate increases and its reuptake decreases [9, 10]. N-methyl-D-aspartate receptor (NMDAR) is mainly located in the postsynaptic membrane and consists of 3 subunits, NMDAR1, NMDAR2 and NMDAR3 [11]. NMDAR1 is the main functional subunit and has all the properties of the NMDARs channel complex [11]. NMDARs belong to a family of ligand-gated ion channels and are closely related to synaptic plasticity as well as learning and memory, and there are two glutamate binding sites on each NMDAR [11]. Excessive glutamate activates the NMDAR, which produces excitotoxicity to cause nerve cell degeneration [12]. The transmission of the neurotransmitter depends on the fusion of synaptic vesicles with the plasma membrane, a process which is regulated by a variety of proteins [13]. The soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNARE) is a presynaptic membrane regulatory protein, which is composed of Synaptobrevin, Syntaxin1 and Synaptosomal-associated protein of 25kD (SNAP-25) [14]. Syntaxin1 is the core protein of SNARE, and mainly has two subtypes: Syntaxin1A and Syntaxin1B[14]. Syntaxin1A mainly mediates the docking and fusion of synaptic vesicles in the synaptic active area, and its expression level reflects the number of synaptic vesicles in the synaptic active area at a certain extent [14]. Munc18-1 (encoded by the STXBP1 gene) is a synaptic fusion protein binding protein, which interacts with Syntaxin1A to regulate synaptic vesicle anchoring and fusion to affect the neurotransmitter transmission [15]. A previous study has found that the decreased expression of Munc18-1 affects the number of released synaptic vesicles and supply of vesicles, which in turn affects the transmission of neurotransmitters, such as glutamate and gamma-aminobutyric acid (GABA), which impairs learning and memory [16]. Synaptophysin is a key synaptic vesicle protein that can correctly target the synaptic vesicles to the presynaptic membrane by regulating Synaptobrevin-II [17]. In addition, it plays an important role in the membrane transport, docking and fusion of synaptic vesicles [18].

Munc18-1 had been studied in various neurological diseases, including schizophrenia and autism [19, 20]. However, its role in SAE has not been investigated. In this study, a rat model of sepsis was established by cecal ligation and perforation (CLP), and the expression of Munc18-1 in the hippocampus of septic rats was investigated. Meanwhile, this study further explored the possible mechanism of Munc18-1 involvement in SAE.

Materials and Methods

Animals

Thirty-day-old clean and healthy male Sprague-Dawley rats (SD) rats (with the age equivalent to human 2–11 year [21, 22] and weight between 100 and 120 g) were purchased from Sichuan Jianyang Dashuo Animal Science and Technology Co., Ltd (Shu ICP 09003144). All the operations in this study are in lined with the Research Animal Care Committee of Sichuan University and all the experiments were performed in accordance with the guidelines. All the rats were housed in a 55–58% relative humidity environment with free access to water and food and a temperature of 22–25 °C on a 12 h light/dark cycle.

Rat Model of Sepsis and Experimental Groups

The rats (n = 187) were randomly divided into the following groups: Sham group (n = 29), CLP group (n = 78), Munc18-1 siRNA (Munc18-1 si) group (n = 40), and Munc18-1 siRNA control (Munc18-1 si-c) group (n = 40).

The dose of the administered drug in the same volume was verified during the pre-experimental phase. At 2 d before CLP, the Munc18-1 si group were injected with 5 ul Munc18-1 siRNA transfection complex in the lateral ventricle, and the Munc18-1 si-c group were injected with 5 ul control siRNA transfection complex. The Munc18-1 siRNA and control siRNA were synthesized by Ruibo Biotech (Guangzhou, China). The sequence chain of Munc18-1 siRNA was 5’-GCATCATCCTTCTCTACAT-3’. To prepare the transfection complex, the EntransterTM-in vivo(Engreen Biosystem, China) reagent was diluted with normal saline and added to a nucleic acid solution diluted to 1 ug/ul with non-toxic pure water (volume ratio of nucleic acid: transfection reagent = 2:1). After anesthetizing the rat with 100% O2 isoflurane (induction dose 2%, maintenance dose 1%), the rat’s head was fixed with a prone position on a stereotaxic device. Based on the left and right cochlear connection, the central axis of the head was defined as the direction, and the hair on the middle of the rat’s head was cut and sterilized, such that 1 cm of the skin and the subcutaneous fascia were cut with the blade to expose the skull. At 8 mm from the bregma and 2 mm from the right side of the midline, a dental drill was used to open the skull and pick up the dura mater. Then, a micro-pump syringe was used to insert the needle to a depth of 6 mm, and a bolus was injected at a speed of 1 ul/min. After finishing the injection, the needle was paused for 2 min and then slowly withdrew.

By performing the CLP, sepsis was established in the CLP, CLP + Munc18-1 si and CLP + Munc18-1 si-c groups. After the rats were deeply anesthetized by inhalation isoflurane (induction dose 2%, maintenance dose 1%) in 100% O2, the abdominal area was shaved and disinfected, and then an incision was made along the midline of the abdomen. The cecum was dissociated, ligated and then punctured twice with 18-gauge needle. A small amount of feces was extruded before replacing it into the abdominal cavity. Then, the cecum was returned and the abdominal cavity was sutured layer by layer. In order to compensate for the blood volume lost during surgery, the rats were treated with 0.9% (w/v) saline (50 mL/kg). The rats of the Sham group underwent the same procedure without ligation, puncture or any other treatment.

Vital Signs Monitoring

The rats (each group n = 6) were continuously anesthetized by inhalation isoflurane (induction dose 2%, maintenance dose 1%) in 100% O2. An indwelling catheter was placed in the femoral artery and connected to a biological signal recorder (iWorx Systems, Inc.), which was used to monitor the mean arterial pressure (MAP) and heart rate (HR).

Neurophysiological Assessment

The neurophysiological scoring (n = 6) was performed to observe their neural reflexes to confirm the development of SAE. The reflexes include the auricular reflex, corneal reflex, righting reflex, tail flick reflex and escape reflex over a specified time period. The scores were assigned as follows: in each reflex, 1 point for weakened reflexes (no reflexes within 10 s), 0 point for no reflexes, and 2 points for normal reflexes, according to a previous study [23].

Hematoxylin and Eosin (H&E) and Immunofluorescence Staining

After deep anesthesia, the rats (n = 6) were injected with 100 mL 0.9% (w/v) saline and 100 mL 4% (w/v) paraformaldehyde (PFA) into the left ventricle for cardiac perfusion. The brain tissue was fixed in 4% PFA, dehydrated with alcohol, cleared with xylene and embedded with paraffin. The sample was cut into 6-µm sections and dried. Afterwards, xylene was dewaxed and the sections were rehydrated through an alcohol series. Afterwards, two types of staining were performed. First, staining with H&E was performed, and the images were taken using the Leica inverted optical microscope (Leica, Mannheim, Germany).

Next, staining with immunofluorescence (n = 6) was performed. The slices were treated with 0.3% Triton 100 PBS solution for 30 min and fluorescent blocking solution for 1 h. The required primary antibodies, including the anti-Munc18-1 rabbit polyclonal antibody (1:50, ab3451, Abcam), anti-Syntaxin1A rabbit polyclonal antibody (1:25, ab41453, Abcam), anti-Synaptophysin mouse monoclonal antibody (1:50, ab8049, Abcam), anti-NMDAR1 rabbit polyclonal antibody (1:50, ab52177, Abcam), anti-NeuN mouse antibody (1:300, Abcam) and anti-NeuN rabbit antibody (1:300, Abcam) were incubated overnight in a refrigerator at 4 °C. After rinsing for 3 times in PBS solution on the next day, secondary antibodies were added, including the CY3 conjugated anti-rabbit IgG (1:500, Jackson ImmunoResearch, West Grove, PA, USA), CY3 conjugated anti-mouse IgG (1:500,Jackson ImmunoResearch, West Grove, PA, USA), Dy Light 488 conjugated anti-rabbit IgG (1: 500,Jackson ImmunoResearch, West Grove, PA, USA) and Dy Light 488 anti-mouse IgG (1:500,Jackson ImmunoResearch, West Grove, PA, USA), and then 4′,6-diamidino-2-phenylindole (DAPI) (1:500, Beyotime) was added followed by incubation for 1 h in the dark. The sections were sealed on a glass slide with an anti-fluorescence quencher (Beyotime). Finally, the images were taken using the Pannoramic MIDI digital section scanner (3DHistechxyl, Hungary).

Transmission Electron Microscopy

After deep anesthesia, the rats (n = 5) were injected with 100 mL 0.9% (w/v) saline and 100 mL 2% glutaraldehyde into the left ventricle for cardiac perfusion. Hippocampus tissues were placed with the size of rice grains in 2.5% glutaraldehyde solution, washed with PBS, fixed in a phosphate buffered osmium tetroxide solution containing 30 g/L glutaraldehyde and embedded in epoxy resin. The brain tissue was cut into 0.12 μm slices and stained with 2 g/L lead citrate and 10 g/L uranyl acetate. The synaptic structure of the hippocampal neurons and number of synaptic vesicles in the synaptic active area were observed using a transmission electron microscopy. On the inside of the presynaptic membrane, there is a fence-like structure, also called the synaptic active area, where the synaptic vesicles discharge neurotransmitters, and some receptor proteins gather under the postsynaptic membrane to form postsynaptic density (PSD) [24]. The typical excitatory synapse structure is shown as follows [24]: (1) The thickness of the postsynaptic membrane is much larger than that of the presynaptic membrane; (2) The PSD can be seen in the postsynaptic membrane; (3) Round and clear synapse vesicles are observed in the presynaptic membrane. A total of 15 typical excitatory synaptic structures were selected from each specimen. The Image J software (NIH, Bethesda, MD, USA) was used to quantify the distance of 150 nm from the presynaptic membrane of the synaptic active area, and then the number of vesicles in the synaptic active area was observed by HT7700 120 kV TEM (Hitachi, Tokyo, Japan).

Western Blotting (WB)

After deep anesthesia at a specific time point, the brains (n = 6) of the rats were quickly removed. The hippocampal tissue separated on ice was added to the freshly prepared solution consisting of Radio Immunoprecipitation Assay Lysis Buffer, Phenylmethanesulfonyl fluoride and Aprotinin for 30 min and was then broken by a tissue crusher. After placing the samples on ice for 30 min, they were centrifuged in a high-speed refrigerated centrifuge (4 °C, 14,000 rpm) for 30 min. Next, the supernatant was extracted, the loading buffer (Beyotime, Shanghai, China) was added in proportion, and the denaturation was performed at a high temperature of 100 °C for 10 min. To ensure the consistency of the protein content in each sample, the BCA analysis kit (Beyotime) was used to determine the protein content in the denatured sample. Afterwards, the deformed and cooled sample (Add 5ug to each well) was electrophoresed on a prepared SDS-polyacrylamide gel and transferred to a methanol-excited polyvinylidene fluoride (PVDF) membrane (Millipore, Billerica, MA, USA). After taking out the PVDF, it was immersed in 5% milk sealing liquid and sealed for 1 h. Then, the PVDF was incubated with primary antibodies, including the anti-Munc18-1 rabbit polyclonal antibody (1:1000, ab3451, Abcam), anti-Syntaxin1A rabbit polyclonal antibody (1:1000, ab41453, Abcam), anti-Synaptophysin mouse monoclonal antibody (1:500, ab8049, Abcam), anti-NMDAR1rabbit polyclonal antibody (1:1000, ab52177, Abcam), anti-β-actin mouse monoclonal antibody (1:5000, ZSGB-BIO, Beijing, China), anti-GAPDH mouse monoclonal antibody (1:5000, ZSGB-BIO, Beijing, China) and anti-β-tubulin mouse monoclonal antibody (1:5000, ZSGB-BIO, Beijing, China) in a refrigerator overnight at 4 °C. On the next day, the PVDF was rinsed with TBST for 3 times, then it reacted with the corresponding secondary antibodies, including the HRP-labeled anti-rabbit or anti-mouse IgG, and was incubated for 1 h and then rinsed with TBST for 3 times. The bands were visualized using enhanced chemiluminescence (Millipore) and imaged using a gel imaging analysis system (Bio-Rad, Hercules, CA, USA). The intensity of each band was quantitatively analyzed by the Image Lab software. The resulting density ratio represents the relative expression of each band against GAPDH, β-actin and β-tubulin.

Liquid Chromatograph Mass Spectrometer (LC/MS) Method

A sufficient amount of glutamate standard substance was accurately weighed, and a 1 mg/ml stock solution with methanol was prepared. The prepared stock solution was diluted into a standard curve working solution with a series of concentration gradients. The obtained concentration was separately 50 ug/ml, 20 ug/ml, 5 ug/ml, 0.5 ug/ml. Then, 500 ul solution (the hippocampal tissue solutions were prepared with using methanol plus 0.1% formate) was added to the hippocampus specimen (n = 6) and vortexed to mix for 5 min, centrifuged in a centrifuge (12,000 rpm, 10 min). The supernatant was diluted to 20 times and analyzed. The Xcalibur 3.0 software was used to process the chromatogram collection of glutamate (Column: Thermo Hypersil GOLD HILIC 100 × 2.1 mm, 1.9 μm; Flow rate: 0.45mL/min; Time: 5 min). The standard curve was obtained by regression with weighting coefficient (1/X2) with taking the peak area of glutamate as the ordinate and the concentration of glutamate as the abscissa.

Statistical Analysis

The values used in the experimental results were expressed as mean ± standard error of mean (SEM) for at least 5 rats. The SPSS version 23.0 software (SPSS, Inc., Chicago, IL, USA) was used for statistical analysis. The statistical significance of differences between groups was analyzed by one-way analysis of variance followed by the Student-Newman-Keuls post-hoc test. P < 0.05 was considered to be statistically significant.

Results

Establishment of a Septic Rat Model

To verify whether CLP successfully induced sepsis in rats, some assessments were conducted. The CLP group had a postoperative mortality rate of 9.3% at 12 h, 25.58% at 24 h and 32.55% at 48 h. The MAP (mmHg) gradually decreased, with the lowest point at 24 h (Sham vs. CLP: 99.83 ± 4.45 vs. 68.67 ± 9.75, P < 0.01), while the heart rate (beats/min) gradually increased, with peak point at 24 h (Sham vs. CLP: 379.33 ± 11.76 vs. 471 ± 20.54, P < 0.01) (Fig. 1a). At the same time, the CLP group showed less movement, erected hair, shaking and curling up. Moreover, neurophysiological scores of the CLP group decreased (Sham vs. CLP: 10 ± 0 vs. 5.17 ± 0.75, P < 0.01) (Fig. 1a). The pathological changes of the hippocampus in the CLP group showed disordered cell arrangement and abnormal morphology (Fig. 1b). These findings are in accordance with the findings by Kafa et al. [25], suggesting that the model of sepsis was successfully established.

Fig. 1
figure 1

The vital signs, neurophysiological scores and histopathological changes in the hippocampus in each group 

a the HR of CLP gradually increased, the MAP and neurophysiological scores gradually decreased, which was obvious at 24 h, compared with Sham; b Pathological changes of the hippocampus detected in each group at 24 h. Sham group were normal and orderly, with normal morphology, and a large number of clearly visible nucleoli. CLP group were degenerated and swollen, with irregular cell boundaries, disordered cell arrangement, deep staining of the cytoplasm (black arrow), and Pyknosis of the nucleus to form a vacuole structure (yellow arrow). scale bar = 50 μm. (n = 6 for each group. P* < 0.05 VS Sham, P** < 0.01 VS Sham)

Full size image

Abnormal Glutamate Levels in the Hippocampus of Septic Rats

We observed the changes of glutamate in hippocampal neurons of septic rats. First, using LC-MS, we found that the content of glutamate (ug/mg) in the hippocampus of the CLP group increased (Sham vs. CLP: 3.792 ± 0.513 vs. 4.638 ± 0.562, P < 0.05) (Table 1). Excitatory neuronal synapses were then visualized using transmission electron microscopy, quantified as the 150 nm distance to the active area of the presynaptic membrane, where an increased number of synaptic vesicles was found (Sham vs. CLP: 10.53 ± 0.47 vs. 18.76 ± 2.31, P < 0.05) (Fig. 2a). Finally, we monitored the expression of the postsynaptic membrane glutamate receptor NMDAR1 using WB and immunofluorescence and found that the expression increased (Sham vs. CLP: 0.46 ± 0.05 vs. 1.13 ± 0.21, 1.09 ± 0.2 vs. 1.63 ± 0.30, P < 0.05) (Fig. 2b, c). These findings suggest that glutamate level in the hippocampus of rats is increased during sepsis.

Fig. 2
figure 2

Number of synaptic vesicles in the synaptic active area of hippocampal neurons and the expression of NMDAR1 after CLP.

a The synaptic structure of the hippocampal neurons and number of synaptic vesicles in the synaptic active area were observed with transmission electron microscope. The results showed synaptic structures in Sham and CLP, including the presynaptic membrane, synaptic cleft, postsynaptic membrane, postsynaptic dense area, round and transparent synaptic vesicles in the presynaptic membrane. The number of synaptic vesicles in the CLP increased at 24 h after surgery. The data are represented by histograms. Scale bar = 200 nm, n = 5. b The levels of NMDAR1 in hippocampus harvested at different time points after CLP were normalized to the levels of β-tubulin, and displayed as relatively arbitrary units; The expression of NMDAR1 increased. The protein content is represented by histograms. c, the fluorescence of NMDAR1 in the hippocampus at 24 h after CLP; The fluorescence signal of NMDAR1 in the CLP was enhanced. The coincidence rate is represented by histograms. Scale bar = 20 μm, n = 6. (P* < 0.05 VS Sham)

Full size image

Increased Expression of Munc18-1, Syntaxin1A and Synaptophysin in the Hippocampus of Septic Rats

Munc18-1, Syntaxin1A and Synaptophysin regulate neurotransmitter transmission. The WB results showed that the expressions of Munc18-1, Syntaxin1A and Synaptophysin increased in the CLP group (Sham vs. CLP: 0.61 ± 0.79 vs. 1.53 ± 0.13, 2.17 ± 0.23 vs. 3.23 ± 0.33, 0.61 ± 0.08 vs. 0.90 ± 0.18, P < 0.05) (Fig. 3a, b, c). In addition, the immunofluorescence results at 24 h after surgery showed that the fluorescence signals of Munc18-1, Syntaxin1A, and Synaptophysin in the CLP group were enhanced (Sham vs. CLP: 39.67 ± 9.63 vs. 58.34 ± 5.13, 43.52 ± 6.19 vs. 62.46 ± 14.32, 25.6 ± 3 2.21 vs. 44.34 ± 7.88, P < 0.05) (Fig. 3a, b, c), which was consistent with the WB results. These findings suggested that sepsis-induced upregulation of Munc18-1, Syntaxin1A and Synaptophysin in the hippocampus of rats.

Fig. 3
figure 3

The expression of Munc18-1, Synataxin1A and Synaptophysin at different time points after CLP.

a The level of Munc18-1 in hippocampus harvested at different time points after CLP was normalized to the level of GAPDH and displayed as relatively arbitrary units; The expression of the Munc18-1 increased; The fluorescence signal of Munc18-1 at 24 h after CLP was enhanced. b The level of Syntaxin1A harvested at different time points after CLP were normalized to the level of β-actin and displayed as relatively arbitrary units; The expression of the Syntaxin1A increased; The fluorescence signal of Syntaxin1A at 24 h after CLP was enhanced. c, the level of Synaptophysin harvested at different time points after CLP were normalized to the levels of β-tubulin and displayed as relatively arbitrary units; The expression of the Synaptophysin increased; The fluorescence signal of Synaptophysin at 24 h after CLP was enhanced. The protein content is represented by histograms. The coincidence rate of fluorescence signals is represented by histograms. Scale bar = 50 μm. (n = 6 in each group; P* < 0.05 VS Sham group, P** < 0.01 VS Sham group)

Full size image

Munc18-1 Regulates the Expression of Syntaxin1A and Synaptophysin in the Hippocampus of Septic Rats

At the presynaptic membrane, Munc18-1 interacts with Syntaxin1A and Synaptophysin to regulate the synaptic vesicle anchoring and fusion. To explore the relationship between Munc18-1, Syntaxin1A and Synaptophysin in hippocampus of septic rats, we used Munc18-1 siRNA to intervene in septic rats. Interestingly, WB results showed that the expressions levels of Munc18-1, Syntaxin1A and Synaptophysin in the hippocampus of the CLP + Munc18-1 si group decreased (CLP vs. CLP + Munc18-1 si: 0.33 ± 0.14 vs. 0.23 ± 0.22, 0.38 ± 0.03 vs. 0.28 ± 0.03, 0.96 ± 0.12 vs. 0.57 ± 0.11, P < 0.05) (Fig. 4a, b). In addition, immunofluorescence results showed that the fluorescence signals of Munc18-1, Syntaxin1A and Synaptophysin in the hippocampus of rats in the CLP + Munc18-1 si group attenuated (CLP vs. CLP + Munc18-1 si: 55.13 ± 5.69 vs. 43.75 ± 2.75, 55.99 ± 0.85 vs. 44.03 ± 0.55, 41.54 ± 9.56 vs. 22.91 ± 3.76, P < 0.05) (Fig. 4a, b), which was consistent with the results of WB. These results show that the expressions levels of Syntaxin1A and Synaptophysin in the hippocampus of septic rats are regulated by Munc18-1.

Fig. 4
figure 4

Munc18-1siRNA down-regulated the protein levels of Syntaxin1A and Synaptophysin during sepsis

a The protein levels of Munc18-1 and Syntaxin1A in hippocampus harvested at 24 h after CLP were normalized to the level of β-actin and displayed as relatively arbitrary units; The expression of Munc18-1 and Syntaxin1A in the CLP and CLP + Munc18-1 si-c increased, compared with Sham; and decreased in the CLP + Munc18-1 si, compared with CLP. The fluorescence signals of Munc18-1 and Syntaxin1A in the CLP and CLP + Munc18-1 si-c were enhanced, compared with Sham; and were weakened in the CLP + Munc18-1 si, compared with CLP. b The protein level of Synaptophysin harvested at 24 h after CLP were normalized to the level of β-tubulinn and displayed as relatively arbitrary units; The expression of Synaptophysin in the CLP and CLP + Munc18-1 si-c increased, compared with Sham; and decreased in the CLP + Munc18-1 si, compared with CLP. The fluorescence signal of Synaptophysin in the CLP and CLP + Munc18-1 si-c group were enhanced, compared with Sham; and were weakened in the CLP + Munc18-1 si, compared with CLP. The protein content is represented by histograms; The coincidence rate of fluorescence signals is represented by histograms. Scale bar = 20 or 50 μm. Values are expressed as mean ± SEM. (n = 6 in each group; P* < 0.05 VS Sham, P# < 0.05 VS CLP).

Full size image

Munc18-1 Regulates Glutamate Levels in Septic Rats

To investigate whether glutamate level in the hippocampus of septic rats is regulated by Munc18-1, we conducted a series of observations after Munc18-1 siRNA intervention in CLP. Firstly, the LC-MS monitoring showed that the content of glutamate (ug/mg) in the hippocampus of the CLP + Munc18-1 si group was decreased (CLP vs. CLP + Munc18-1 si: 4.638 ± 0.562 vs. 3.713 ± 0.442, P < 0.05) (Table 1). Secondly, the excitatory synapses of the hippocampal neurons were observed using transmission electron microscope, and the results showed that the number of synaptic vesicles in the synaptic active area of the CLP + Munc18-1 si group decreased (CLP vs. CLP + Munc18-1 si:17.56 ± 1.50 vs. 11.11 ± 1.50, P < 0.05) (Fig. 5a). Finally, we monitored the expression of the NMDAR1, and WB showed that the expression of NMDAR1 decreased in the CLP + Munc18-1 si group (CLP vs. CLP + Munc18-1 si: 0.78 ± 0.08 vs. 0.52 ± 0.09, P < 0.05) (Fig. 5b). Immunofluorescence showed that the fluorescence signal of CLP + Munc18-1 si group was attenuated (CLP vs. CLP + Munc18-1 si: 1.71 ± 0.25 vs. 1.19 ± 0.25, P < 0.05) (Fig. 5c). These results suggest that glutamate levels in hippocampal neurons of septic rats is regulated by Munc18-1.

Fig. 5
figure 5

Munc18-1siRNA inhibited glutamate levels

a Number of synaptic vesicles in each group. The results showed round and transparent synaptic vesicles (red arrows) in the presynaptic membrane in each group. The number of synaptic vesicles of Munc18-1, Syntaxin1A and Synaptophysin in the CLP and CLP + Munc18-1 si-c increased, compared with Sham; and decreased in the CLP + Munc18-1 si, compared with CLP. The data are represented by histograms. Scale bar = 200 nm, n = 5. b The protein level of NMDAR1 harvested at 24 h after CLP were normalized to the level of β-tubulinn and displayed as relatively arbitrary units; The expression of NMDAR1 in the CLP and CLP + Munc18-1 si-c increased, compared with Sham; and decreased in the CLP + Munc18-1 si, compared with CLP. c The fluorescence signal of NMDAR1 in the CLP and CLP + Munc18-1 si-c were enhanced, compared with Sham ; and were weakened in the CLP + Munc18-1 si, compared with CLP. The protein content is represented by histograms; The coincidence rate of fluorescence signals is represented by histograms. Scale bar = 20 or 50 μm, n = 6. Values are expressed as mean ± SEM. (P* < 0.05 VS Sham, P# < 0.05 VS CLP).

Full size image

Inhibition of Munc18-1 Expression Can Improve Vital Signs and Hippocampal Pathological Damage in Septic Rats

To evaluate the effect of Munc18-1 expression levels on the vital signs and hippocampal pathological changes during sepsis, several evaluations were performed after Munc18-1 siRNA interference in septic rats. The vital signs monitoring results showed that MAP (mmHg) increased, HR (beats/min) decreased, and neurophysiological scores increased in CLP + Munc18-1 si group ( CLP vs. CLP + Munc18-1 si: 65.83 ± 5.30 vs. 91.83 ± 3.54, 478.33 ± 13.46 vs. 401 ± 14.66, 4.33 ± 0.82 vs. 8 ± 0.89, P < 0.05) (Fig. 6a, b, c). H&E results indicated that the pathological damage of hippocampus in CLP + Munc18-1 si group improved (Fig. 6d). The above-mentioned results indicate that targeted inhibition of Munc18-1 expression ameliorate the vital signs and the hippocampal injury of septic rats.

Fig. 6
figure 6

Inhibition of Munc18-1 expression improve vital signs and hippocampal pathological damage in septic rats

a HR; b MAP; c neurophysiological scores; the HR of CLP and CLP + Munc18-1 si-c gradually increased, the MAP and neurophysiological scores gradually decreased, which was obvious at 24 h, compared with Sham; the HR of CLP + Munc18-1 si group decreased, the MAP and neurophysiological scores increased, compared with CLP. d Pathological changes of the hippocampus detected in each group at 24 h. Sham group were normal and orderly, with normal morphology, and a large number of clearly visible nucleoli. CLP and CLP + Munc18-1 si-c group were degenerated and swollen, with irregular cell boundaries, disordered cell arrangement, deep staining of the cytoplasm (black arrow), and pyknosis of the nucleus to form a vacuole structure (yellow arrow); the CLP + Munc18-1 si group were normal and orderly, with a slightly normal morphology, with more clearly visible nucleoli, and occasionally deep staining of the cytoplasm (black arrow) and vacuole structure (yellow arrow). scale bar = 20 μm (n = 6 for each group. P* < 0.05 VS Sham, P** < 0.01 VS Sham, P# < 0.01 VS CLP, P## < 0.01 VS CLP).

Full size image

Discussion

Although sepsis has some commonalities in children and adults, some differences in the pathophysiological process still exist [26]. The children’s immune system is still developing, and the lack of anti-inflammatory ability makes inflammation less likely to be localized, so they are prone to infection [27]. Compared with adults, they have a lower rate of multiple organ dysfunction, and hypotension develops later in children because the increased cardiac output in children through a very rapid heart rate can be maintained for a long time without myocardial ischemia [27]. In this study, 30-day rats were selected to simulate the development of childhood with sepsis through CLP. Compared with the Sham group, the rats in the CLP group, with erecting hairs and shaking, had decreased MAP, increased HR and a certain mortality rate. The above-mentioned results indicate the successful establishment of the sepsis rat model through CLP. H&E staining showed that the hippocampus of the rats in the CLP group had aggravated pathological changes and decreased neurophysiological scores. The results suggest that hippocampal damage in rats can be induced by the establishment of a sepsis model by CLP. A previous study showed that the hippocampus of rats was injured during sepsis, and it was obvious at 12–24 h after surgery [28]. Interestingly, our results showed similar findings.

Altered neurotransmitter levels in the brain are a key factor in SAE, and inflammatory processes promote changes in many neurotransmitter systems, including glutamatergic, monoaminergic, and cholinergic [29]. Glutamate represents the main excitatory neurotransmitter in the central nervous system, and excessive levels of synaptic glutamate can over-activate NMDAR to produce excitotoxicity and damage the nerual cells [30]. NMDAR is an ionotropic glutamate receptor with an important role in the regulation of neuronal survival and synaptic plasticity, of which NMDAR1 is the main subtype constituting ion channels [31]. Excessive activation of NMDARs leads to Ca2+ overload and Na+ accumulation in neural cells, which induces the activation of various signaling pathways and enzymes including proteases, kinases, phosphokinases, lipoxygenases, NOS and ROS, leading to neuronal necrosis and apoptosis [31]. It has been found that transient forebrain ischemia under hyperthermic conditions accelerates memory impairment and neuronal death in the mouse hippocampus due to increased NMDAR1 expression, which is associated with increased hippocampal glutamate levels [32]. A previous study on sepsis-related hearing impairment found that the basal portion of the cochlear intimal cells in septic mice exhibited glutamate excitotoxicity [33]. In addition, another study showed that glutamate levels in brain tissue fluid increased during SAE and exerted excitotoxic effects on nerve cells by activating NMDARs [34]. In this study, we observed increased glutamate content and NMDAR1 expression in the hippocampus of septic rats. This suggests that the glutamate/NMDAR1 pathway is activated during sepsis. NMDARs are mainly located in the postsynaptic membrane [11]. Our results appear to suggest abnormal glutamate transmission during sepsis. However, It has been reported that NMDAR is also expressed in the presynaptic membrane [11]. Therefore, the determination of NMDAR1 expression at the postsynaptic membrane is required to further elucidate the role of abnormal glutamate transmission in SAE. In addition, although the specific mechanism of excitotoxicity is unclear, it was found that glutamate transporter-1 (GLT-1), the main glutamate transporter in brain tissue, affects glutamate excitotoxicity [35]. GLT-1 is a major transporter in the brain, which is expressed in both glial cells and neurons, but to a lesser extent in neurons [36]. A previous study has shown that the deletion of the GLT-1 gene abolishes 95% of the glutamate uptake activity in forebrain synaptosome [37]. During sepsis, brain tissue glutamate release is increased and uptake is decreased [9, 10]. Therefore, the determination of GLT-1 can further illustrate the role of glutamate excitotoxicity in hippocampal injury in septic rats. Unfortunately, due to limited technology equipment and the Covid-19 pandemic.

The anchoring and fusion of synaptic vesicles is the basic process of neurotransmitter transmission, and the increase in the number of synaptic vesicles in the synaptic active area is positively correlated with the amount of released neurotransmitters [38]. A previous study on a mouse model of heterozygous knockout of the STXBP1 gene found that the number of synaptic vesicles in the synaptic active area decreased with decreased expression of Munc18-1, and the transmission of glutamatergic and GABAergic signals was impaired [39]. As the expression of Munc18-1 increased, the number of synaptic vesicles increased, and its expression level reflected the change in the number of synaptic vesicles in the synaptic active area [16]. The activation of the diacylglycerol (DAG)/protein kinase C (PKC) pathway can promote the conversion of the Syntaxin1-Munc18-1 dimer to the assembled SNARE complex, enhancing the neurosynaptic transmission process [40]. In addition, this pathway also promotes the phosphorylation of Munc18-1 to transfer the synaptic vesicles to the active area [41]. Another study concluded that riluzole attenuated the neurological symptoms in septic rats by inhibiting glutamate release into the synaptic cleft [42]. Riluzole mainly inhibited the PKC pathway/phosphorylation of the Munc18-1 to reduce the level of vesicles in the synaptic active area and inhibit the presynaptic release of glutamate to play a protective role [43]. In this work, we observed an increase in the number of synaptic vesicles in areas of synaptic activity in neurons with excitatory synaptic structures. Interestingly, the down-regulation of Munc18-1 expression with Munc18-1 siRNA decreased the number of synaptic vesicles in the synaptic active area, as well as the glutamate content and NMDAR1 expression in the hippocampus of septic rats. Therefore, our findings suggest that Munc18-1 may be involved in neuronal damage by regulating the level of glutamate to induce excitotoxicity in the hippocampus of septic rats.

Munc18-1 mainly regulates the SNARE function by combining Syntaxin1 to affect the anchoring and fusion of the synaptic vesicles and the neurotransmitter transmission [44]. A previous study on patients with schizophrenia revealed that the expression of Syntaxin1A and Munc18-1 protein increased in their prefrontal cortex. Specifically, the expression of Syntaxin1A was upregulated by Munc18-1 [45]. A study on the heart regulation mechanism on sepsis reported that at 24 h after CLP, the expression of Syntaxin1A in the rat myocardial tissue increased, then decreased at 72 h [46]. In the experiment, WB and immunofluorescence detected an increased expression of Syntaxin1A of CLP group, which decreased after the intervention with Munc18-1 siRNA. The results showed that the expression of Syntaxin1A could be induced by CLP and regulated by Munc18-1. Synaptophysin plays an important role in regulating the neurotransmitter release, and its expression level reflects the number of synaptic vesicles released through exocytosis [47, 48]. In this work, the results of WB and immunofluorescence revealed an increase in the expression of Synaptophysin of CLP group, which decreased after the intervention with Munc18-1 siRNA. The above-mentioned results indicated that CLP induced an increase in the expression of Synaptophysin in the hippocampus of rats. Munc18-1 could regulate the expression of Synaptophysin of septic rats to affect the activity of the synaptic vesicles. However, in a previous study, WB of the hippocampus and frontal cortex of septic rats that survived at 30 d after CLP, showed increased β-amyloid protein expression and decreased Synaptophysin expression [49]. Besides, in a study on simvastatin and rats that survived sepsis with long-term cognitive deficits suggested that the expression of Synaptophysin in the hippocampus of the CLP group decreased at 10 d after CLP [50]. This seems inconsistent with our results, but we speculate that it may be related to the short observation time point of this experiment. Our results suggested that Synaptophysin level increased at 24 h after surgery and then started to decrease at 48 h after surgery. Whether Synaptophysin will continue to decrease after 48 h needs to be further explored.

Our data showed that the rats had aggravated pathological damage in the hippocampus, increased HR, decreased MAP and neurophysiological scores during sepsis. After intervening with Munc18-1 siRNA, improvements were seen in the pathological damage in the hippocampus, neurophysiological scores and deteriorating vital signs. The above-mentioned results indicated that the expression level of Munc18-1 played an important role in the brain injury of septic rats. In the acute phase of sepsis, the pathological damage of SAE may be improved by reducing the expression of Munc18-1. Based on a previous study, the view suggested that the increased expression of Munc18-1 could induce the excitotoxicity of glutamate to damage nerve cells, while the decreased expression could damage the synaptic plasticity to affect the balance of glutamate, GABA, dopamine, and other neurotransmitters to impair learning and memory [39]. In our study, in the acute sepsis phase, the expression of Munc18-1 increased to a peak at 24 h after surgery and began to decline after 48 h. However, the observation time of the experiment was short, and we did not explore the expression of Munc18-1 at the stage of sepsis after 48 h. Therefore, further research is needed to study whether the expression of Munc18-1 will show a continuous decline, and play a role in long-term cognitive and learning ability dysfunction of patients surviving SAE.

Prospects and Shortcomings

Taken together, our findings suggest that Munc18-1 is involved in the pathological process of SAE by affecting glutamate level via regulating the expression of Syntaxin1A and Synaptophysin. Targeted down-regulation of Munc18-1 ameliorated brain injury in septic rats, being a process that may be associated with suppressed glutamate excitotoxicity. This provides a new target for the clinical treatment of SAE. Although our findings are meaningful, they also reveal some deficiencies and limitations. The next step is to verify the direct relationship between Munc18-1 and the content of glutamate in the synaptic cleft through in vitro cell experiments, and to observe whether the overexpression of Munc18-1 in septic rats will aggravate hippocampal damage, further indicating that Munc18-1 is involved in target role in sepsis. In addition, the link between glutamate excitotoxicity and septic rat hippocampal neuronal cell death, such as apoptosis, needs to be further verified.

Table 1 The content of glutamate in hippocampus in each group.

Data are means ± SEM, n = 6 for each group. P* < 0.05 VS Sham, P# < 0.05 VS CLP.

Group

Glutamate content in the hippocampus(\(\bar {x} \pm s\))ug/mg

Sham

3.792 ± 0.513

CLP

4.638 ± 0.562*

CLP + Munc18-1 si

3.713 ± 0.442#

CLP + Munc18-1 si-c

4.548 ± 0.59*