C57BL/6 (WT) mice were purchased from Dien Gene-Tech (Guangzhou, Guangdong, China). All animal programs were reviewed and approved by the Institutional Animal Care and Use Committees of Guangdong Medical University. Eight-week-old mice with a mean body weight of 25 g were used. Endotoxemia model was induced by intraperitoneal injection of LPS (20 mg/kg) (LPS group, n = 28). To evaluate the effect of remimazolam, mice were intraperitoneally injected (16 mg/kg) 30 min before and subcutaneous injected (8 mg/kg) 30 min after LPS with remimazolam (Re + LPS group, n = 28). Mice received only remimazolam were regarded as the Re group (n = 10). All reagents above were dissolved in normal saline at the indicated concentrations. NS was used as the blank vehicle in the mice of the control group. The mice were dislocated 12 h after the LPS injection, and serum and organ samples were collected at the time of execution. Each group had at least three biological replicates in vivo.

Liver and lung tissue were fixed overnight with 4% paraformaldehyde at room temperature, embedded in paraffin, and cut into 4 μm sections. The sections were stained with hematoxylin and eosin.

Cell Counting Kit-8 (CCK8) assay (Dojindo, Japan) was performed to evaluate cell viability. In brief, RAW264.7 cells or BMDM were inoculated in 96-well plates with a cell density of 0.8 × 10⁵/well for 24 h. Cells were pretreated with indicated concentrations of remimazolam for 20 min and then treated with LPS (1 μg/ml) for 24 h 10μL CCK-8 was added and incubated for another 2 h. Absorbance was measured at OD-450 nm using a spectrophotometer (ThermoFisher, United States).

TNF-α(88-7324-77), IL-6(88-7064-22), and IL-1β(88-7013-22) Mouse Uncoated ELISA kits (eBioscience, San Diego, CA, United States) were used to calculate TNF-α, IL-6, and IL-1β protein expression by setting a standard curve according to the manufacturer’s instructions. The measurements were standardized with cell numbers in vitro.

The mice were anesthetized with sevoflurane and died of dislocation. After the separation of bilateral femur and tibia, the bone marrow was flushed out with sterile precooling PBS, and the specific culture method was described above (Li et al., 2016a). In brief, after centrifugation, marrow rinse solution was resuspended in complete medium and cultured with 10 ng/ml M-CSF(#315-02, PEPRO TECH, Rocky Hill, United States) for 3 days, then M-CSF was added again to reach 20 ng/ ml, and the culture was mature on 5–7 days.

Macrophages collected from peritoneal lavage were labeled with BV421 Rat Anti-Mouse F4/80 (BD PharmingenTM, San Jose, CA, United States) at 4°C for 15 min. For measuring cell surface expression of TLR4, macrophages were stained with PE conjugated anti-mouse CD284 (TLR4) antibody (eBioscience affymetrix, San Diego, CA, United States) for 30 min. For detecting cell death, cells were incubated with PI and Annexin V (BD PharmingenTM, San Jose, CA, United States) for 15 min at room temp, and the double-stained cells were counted as dead cells. A total of 20,000 events were collected and the cells were analyzed using FACScalibur cytometer (BD Biosciences, San Jose, CA, United States). Peritoneal macrophages were identified as F4/80^high. The final data were analyzed by FlowJo-V10 software (Tree Star, Ashland, OR, United States).

Total RNA was isolated from BMDM or RAW264.7 cells using TRIzol RNA Isolation Reagents (TaKaRa, Janpan). Total RNA was reverse-transcribed into cDNA using a PrimeScript RT reagent Kit with gDNA Eraser (RR047A, TaKaRa, Japan). TB Green Premix Ex Taq II (RR820A, TaKaRa, Japan) was used for real-time fluorescent quantitative PCR (RT-PCR) to detect mRNA expression. ABI 7500 system (Applied Biosystems, United States) was used for performing real-time PCR. Primers sequences are in Supplementary information, Supplementary Table S1. The final results were analyzed and calculated according to an equation: 2^–ΔΔCt which provided the number of targets that were normalized to an internal reference. Ct value is the number of cycles required to reach the threshold during PCR process. The test was repeated three times for each biological sample.

RAW264.7 cells were grown in 35 mm specialized confocal dish (NEST, Wuxi, China). They were first fixed in 2% paraformaldehyde for 15 min, then fixed in 4% paraformaldehyde for 10 min, then permeated with 0.5% TritonX-100, followed by blocked with 5% BSA for 30 min at room temp, respectively. Next, the primary antibody was added and incubated with cells at 4°C overnight. After washing the cells with PBS for three times, appropriate fluorescent secondary antibodies were selected, added, and incubated with cells for 1 h at RT in the dark. After washing three times, the nuclei were stained with DAPI and incubated in the dark for 10 min, and then the plates were sealed with anti-fluorescence quenching agent, and the images were observed under a fluorescence microscope and collected.

Lipofectamine RNAiMAX Transfection Reagent (#13778150, Invitrogen, United States) was transfected with 100 pmol of mouse PBR siRNA or non-specific siRNA (RiboBio, Guangzhou, China) according to the manufacturer’s instructions. Western blot was used to confirmed the knockdown effect of corresponding gene after 48 h transfection.

Antibodies: AKT (4691), P38 (8690), ERK1/2 (4695), p-AKT (4060), p-P38 (4511), p-ERK1/2 (4370), IκBα (4814), p-IκBα (2859), p-IKKα/β (2697), NF-κB (8242), Rab5a (46449), horse anti-mouse IgG (H&L) (7076), and goat anti-rabbit IgG (H&L) (7074) were purchased from Cell Signaling Technology (Danvers, MA, United States). IKKβ (ab124957), PBR (ab109497) was from Abcam (CA, United States). Alexa Fluor 488 goat anti-rabbit (A0423) and Alexa Fluor 647 goat anti-mouse (A0473) secondary antibodies were obtained from Beyotime Biotechnology (Shanghai, China). TLR4 (66350-1-Ig), GAPDH (60004-1-Ig), and β-Tubulin (10068-1-AP) were purchased from Protein-Tech (Wuhan, Hubei, China). Lipopolysaccharides(L4391) was obtained from Sigma (Louis, MO, United States), Remimazolam Tosilate for Injection (Remimazolam) were obtained from Hengrui (Jiangsu, China).

All results were expressed as mean ± SD. Differences between two groups were analyzed using Student t test. One-way ANOVA was used for comparison of multiple groups. Post hoc analysis was made by using Tukey’s post-hoc test following ANOVA. Survival studies were analyzed by Chi-square test. p < 0.05 was considered statistically significant. Graphs and figures were made with Graphpad Prism 6.0 (GraphPad software, CA, United States).

Macrophages are crucial innate immune cells in the body, which play an significant immunoregulatory role in immune homeostasis (Epelman et al., 2014), and also play an essential role in host defense against a variety of pathogens and generation of inflammatory immune response in sepsis (Huen and Cantley, 2017). As a result, we will explore the protective effect of remimazolam on excessive inflammatory injury of the body with macrophages as the research object. To investigate the effect of remimazolam on LPS-induced cell death in macrophages, we first use CCK8 to assess cell vitality. As show in Figures 1A,B, different concentrations of remimazolam alone will not cause the change of cell vitality. At the same time, cell vitality decreases followed by LPS treatment, while pretreatment of remimazolam recovers cell vitality in a concentration-dependent manner. In addition, we also assess cell death by PI and Annexin V by flow cytometry in BMDM and RAW264.7 cells after LPS treatment for 24 h. As shown in Figures 1C–F, LPS increased cell death in RAW264.7 cells and BMDM, and this was attenuated with remimazolam pretreatment. These results suggested that remimazolam was resistant to LPS-induced cell death.

LPS stimulation cause cell inflammatory responses, including increased transcription and expression of inflammatory factors (Foster and Medzhitov, 2009). One of the most dramatic consequences of overwhelming acute inflammation is septic shock due to bacterial LPS, the most potent inducer of the inflammatory response (Hotchkiss and Karl, 2003). For this reason, the effects of remimazolam on LPS-induced macrophage inflammatory response were evaluated. Here, LPS (1 μg/ml) markedly enhanced the mRNA levels of TNF-α, IL-6, and IL-1β, and the phenomenon was partially reversed by remimazolam pretreatment in BMDM and RAW264.7 cells (Figures 2A,B). Subsequently, ELISA was performed to detect inflammatory cytokine expression. As expected, compared with the significantly increased protein levels of inflammatory factors TNF-α, IL-6, and IL-1β in the LPS group, preconditioning with remimazolam reversed the above factor protein expression level, therefore inhibiting the occurrence of inflammation (Figures 2C,D).

We next investigated a potential therapeutic role for remimazolam in LPS-induced endotoxicity in vivo. Mice that received an intraperitoneal injection of remimazolam were sedated (lightly anaesthetized), responded to pain, recovered, and began to move about 30 min later. After 12 h of LPS processing (20 mg/kg i. p.), blood serums were collected from mice. Blood serums inflammatory cytokines (TNF-α, IL-6, IL-1β) production from mice after LPS treatment (LPS group) alone were increased, and remimazolam pretreatment (Re + LPS group) markedly prevented LPS-induced production of serums inflammatory factors (Figure 2E). Histological analyses of mouse liver and lung showed that in the control group no inflammation and necrosis were found (Figure 2F). In liver tissue, Re group resulted in spotty coagulation necrosis in the lobules with mild inflammatory cells infiltration within the necrosis area, consisting of neutrophil (the cytoplasm contains fine reddish particles, and the nucleus was divided into leaves) and monocyte (black granular). However, necrosis and inflammation of liver tissue were greatly reduced in Re plus LPS group although the mice received the same dose of LPS. Similarly, in LPS group, the lung tissue displayed serious injury with the features of disrupted alveoli, hemorrhage, thickness of alveolar septum, and infiltration of inflammatory cells. Those pathohistological changes were resolved by treatment with remimazolam. These results revealed that remimazolam prevented the liver and lung injury caused by LPS and protected their function to some extent as well. Meanwhile, the survival rate of control, remimazolam alone, LPS, and LPS plus remimazolam group were studied. As shown in Figure 2G, the survival rate of mice in the LPS group was only 25%, while 75% of mice treated with remimazolam by intraperitoneal injection survived 72 h after LPS administration, much longer than mice injected with LPS alone. Remimazolam alone was not toxic. Collectively, these results indicate that remimazolam significantly prevented multiple organ injury caused by LPS and thus raised survival of endotoxemia mice. In addition, these results give the United States the hint that remimazolam, not just macrophage, may play a potential role to prevent or treat inflammatory diseases in other immune cells.

MAPK and PI3K signaling pathway, as an upstream signaling pathway of NF-κB, is activated by LPS and is involved in the production of LPS-induced inflammatory cytokines (Barton and Medzhitov, 2003; Ojaniemi et al., 2003). These results show that the phosphorylation of AKT, p38, and ERK1/2 was triggered at 30 min after LPS treatment in RAW264.7 cells, but this activation was substantially suppressed by remimazolam (Figures 3A–D). And then we evaluated the role of NF-κB in the protection of inflammatory response by remimazolam. NF-κB activation is directly related to a sequential cascade, including inhibitor kappa B kinase (IKK)-dependent inhibitor kappa Bα (IκBα) phosphorylation, ubiquitination, and proteolytic degradation, accompanying subsequent translocation of cytosolic NF-κB to the nucleus. As show in Figures 3F,G, LPS significantly upregulated the phosphorylations of IKKα/β and IκBα but degraded IκBα expression in RAW264.7 cells. However, pretreatment with remimazolam reversed these effects, indicating that remimazolam was able to inhibit LPS-induced IKKα/β activation and IκBα degradation. These findings were further supported by immunofluorescent staining showing that remimazolam blocked the nuclear translocation of NF-κB p65 when compared with that of LPS stimulation alone at 10 min (Figure 3E). Taken together, these results elucidate that remimazolam inhibited LPS-induced NF-κB activation.

Furthermore, to determine the effect of remimazolam on advanced inflammatory response, we measured the phosphorylation of AKT, P38, and ERK1/2 after LPS treatment in RAW264.7 cells. ERK1/2, P38, and AKT phosphorylation were measured 1, 3, or 6 h respectively after LPS treatment with or without remimazolam pretreatment. As shown in Figures 3K–M, LPS promoted the phosphorylation of ERK1/2, P38, and AKT at all time points and this effect could be inhibited by remimazolam in 3 or 6 h. Then, we verified this result in BMDM (Figures 3N–Q). Collectively, these results support the notion that the suppressive effect of remimazolam on LPS-induced macrophages activation is through preventing MAPK and PI3K signaling pathway.

TLR4 activation is a tightly regulated process. In addition to directly regulating different signaling pathways, the amount of TLR4/ MD-2 present on the cell surface also controls the LPS response. Based on this, we measured the effect of remimazolam pretreatment in cell surface TLR4 expression after LPS treatment by using flow cytometry. At 12 h after LPS treatment, TLR4 expression on the surface of RAW264.7 cells was increased ∼6 fold as compared with the control group, and this increase could be partially inhibited by remimazolam (Figures 4A,B). This alteration was observed in BMDM as well (Figures 4C,D).

Next, we decided to verify if this change also happened in vivo. As shown in Figures 4E,F, where in contract to LPS group, the amount of TLR4 on the surface of peritoneal macrophages from Re plus LPS group was markedly decreased than the former at 12 h after LPS treatment. These findings indicate that remimazolam inhibits LPS-induced upregulation of macrophage surface TLR4 expression.

Our previous study demonstrated that Rab5a-mediated internalization of TLR4 results in increased cell surface expression of the receptors (Tang et al., 2020). Pretreatment of RAW264.7 cells (Figures 5A,B) and BMDM (Figures 5C,D) with remimazolam given 20 min prior to LPS effectively inhibited the expression of Rab5a at protein level at 6, 12, and 24 h after LPS treatment. We further determined whether remimazolam modulates Rab5a in the progress of blocking macrophage activation. As shown in Figure 5E, remimazolam pretreatment attenuated the colocalization between early Rab5a and TLR4 at 1 h after LPS treatment in RAW264.7 cells by confocal immunoluorescence microscopy. Collectively, these results indicate that remimazolam regulates functions and expressions of Rab5a, and consequently enhanced late phase cell surface expression of TLR4 in response to LPS.

Therefore, we studied macrophage with PBR knockout. Knockdown efficiency is shown in Figures 6A,B. However, we found that the inhibition of remimazolam on MAPK and PI3K signaling pathway proteins in response to LPS did not reverse after PBR knockout in RAW264.7 cells (Figures 6C–F). These results indicated that remimazolam could not inhibit the inflammatory response by binding PBR in macrophage.