Epimedin B Through Enhancing NLRP3 Inammasome Activation To Induces Idiosyncratic Hepatotoxicity

Background: The NLRP3 inammasome plays a crucial role in the pathogenesis of various human diseases, also idiosyncratic drug-induced liver injury (IDILI). Epimedii Folium (EF) is commonly used for treating bone fractures, joint diseases and some chronic illness, but the EF also could induce IDILI. Several studies have conrmed that EF may induce liver injury by upregulating the activity of the NLRP3 inammasome. However, the major active constituents of EF have not been well-studied. Results: In the present study, we showed that epimedin B, a major active ingredient of EF, induced the development of IDILI by promoting the activation of the NLRP3 inammasome. Synergistic induction of mitochondrial reactive oxygen species was a crucial contributor to the promoting effect of epimedin B observed on nigericin- or ATP-induced NLRP3 inammasome activation. Importantly, epimedin B induced liver injury in the LPS-mediated susceptibility mouse model of IDILI(cid:0)while specical NLRP3 inhibitor MCC950 pretreatment completely abrogated the Caspase-1 activation and IL-1β secretion then couldn't induce liver injury. Conclusions: Epimedin B specically facilitated nigericin- or ATP-induced NLRP3 inammasome activation and the development of IDILI, which is responsible for EF-induced liver injury. These ndings suggest that epimedin B is one of the key constituents of liver injury caused by EF(cid:0)the content of epimedin B in EF may be a risk factor for IDILI, especially in patients with diseases related to nigericin- or ATP-induced NLRP3 inammasome activation.

Activation of the NLRP3 in ammasome may be a critical mechanism underlying the development of IDILI [14,15,17].
Epimedii Folium (EF) is a widely used herbal medicine in many countries and has been extensively used as a tonic or antirheumatic agent in clinics. Nevertheless, EF and its preparations have garnered signi cant interest because they can induce liver injury [10,[17][18][19]. The major active constituents of EF are avonoids. More than 60 kinds of avonoids have been identi ed, of which epimedin A, B, C, and icariin are considered major bioactive components that constitute more than 52% of the total avonoids in EF [20][21][22]. However, the underlying molecular mechanisms of the major constituents of EF about liver injury remains unclearly. In this study, we have demonstrated that epimedin B, which is one of the major constituents derived from EF, speci cally promotes the activation of the NLRP3 in ammasome to induce IDILI.

In ammasome activation
To induce in ammasome activation, BMDMs were seeded at a density of 5×10 5 cells/well in 0.5 mL of the medium in 24-well plates and were incubated overnight. Next, the medium was replaced with the fresh medium the following day, and BMDMs were subjected to stimulation with 50 ng/mL LPS or 1 μg/mL Pam3CSK4 for a duration of 4 h. The cells were subjected to treatment with Epimedin B for 1 h and were then stimulated as follows: 5 mM ATP for 1 h, 7.5 μmol/L nigericin for 30 min, or 250 μg/mL silicon dioxide (SiO2) for 6 h. Cells were transfected with poly (I:C) (2 μg/mL), poly(dA:dT) (2 μg/mL), or LPS (1 μg/mL) for 6 h using Lipofectamine 2000 according to the manufacturer's instructions.

Western blotting
Cell extracts and precipitated supernatants were subjected to lysis using 1×loading buffer containing radioimmunoprecipitation assay buffer. The samples were subjected to denaturation at 105 °C for 15 min. Equal amounts of protein samples were separated by performing 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis and protein bands were transferred onto 0.2-mm polyvinylidene uoride membranes. The membrane was blocked with 5% non-fat milk for 1 h at room temperature. Next, the indicated primary antibodies were added and the bands were incubated at 4 °C overnight and then subjected to treatment with anti-goat IgG (1:3000), anti-mouse IgG (1:5000), or anti-rabbit IgG (1:5000) for 1 h at room temperature. The signals generated thereafter were analyzed using an enhanced chemiluminescence reagent.
Caspase-1 activity assay The Caspase-Glo® 1 In ammasome Assay (Promega, Madison, WI, USA) was used to assess caspase-1 activity in cell culture supernatants according to the manufacturer's instructions.
Alanine aminotransferase (ALT) and aspartate transaminase (AST) Serum ALT and AST levels were determined according to the Nanjing Jiancheng Bioengineering Institute (Nanjing, China) and the Nanjing Jiancheng Bioengineering Institute (GOT) assay kit instructions [23].

Lactate dehydrogenase (LDH) assay
The release of LDH into the culture supernatant was assessed using the CytoTox 96® 1 Non-radioactive Cytotoxicity Assay (Promega, Madison, WI, USA) according to the manufacturer's instructions.

ASC oligomerization
Cells were subjected to lysis using Triton buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.5% Triton X-100, and EDTA-free protease inhibitor cocktail). The samples were then centrifuged at 6,000 × g at 4°C for 15 min. The supernatant was referred to as Triton X-soluble, and the pellet fractions were referred to as Triton X-insoluble fractions. To enable ASC oligomer cross-linking, the Triton X-100-insoluble fractions were subjected to washing steps and were resuspended in 200 μL of PBS, followed by the establishment of cross-linking at 37°C with 2 mM disuccinimidyl suberate (DSS) for 30 min. The pellets were centrifuged at 6,000 × g for 15 min, after which they were collected and dissolved in 1×SDS loading buffer for immunoblot analysis.

Intracellular K + measurement
BMDMs were seeded in 12-well plates overnight and were primed with 50 ng/mL LPS for 4 h. The cells were subjected to treatment with epimedin B and were then stimulated with nigericin for 30 min. The culture medium was thoroughly aspirated and subjected to washing steps thrice using potassium-free buffer. Ultrapure HNO3 was added to perform lysis of the cells. Samples were collected in glass bottles and boiled for 30 min at 100 °C. Intracellular K+ measurements were performed via inductively coupled plasma mass spectrometry.

Measurement of intracellular Ca 2+ levels
BMDMs were seeded in a 384-well plate at a density of 2.5×10 4 cells/mL overnight. Then, the cells were primed with LPS for 4 h, followed by stimulation with ATP for 45 min with or without epimedin B. A trace showing ATP-induced Ca 2+ ux was analyzed using the FLIPRT Tetra system (Molecular Devices, San Jose, CA, USA).

Mitochondrial reactive oxygen species assay
BMDMs were seeded at a density of 1×10 6 cells/mL in culture dishes with a diameter of 100 mm and primed with LPS (50 ng/mL) for 4 h. The cells were added in a test tube, subjected to washing steps with Opti-MEM, and were stimulated as per methods described previously. For mitochondrial reactive oxygen species (ROS) measurements, BMDMs were subjected to staining procedures using 4 μM MitoSOX for 20 min at 37 °C and then washing steps were conducted twice with HBSS, followed by assessments using ow cytometry. After the completion of staining and washing procedures, ow cytometry was performed to measure mtROS levels.
Hepatic mRNA expression RNA extraction from the liver tissue was performed using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Subsequently, total RNA samples were reversetranscribed into rst-strand cDNA using the RevertAid First Strand cDNA Synthesis Kit (K1622, Thermo Fisher Scienti c). A total of 5 μL of the PowerUp SYBR Green Master Mix (A25742, Invitrogen), 3 μL DEPC water, 0.5 μL F primer, 0.5 μL R primer, and 1 μL cDNA were introduced into each well of a 384-well plate, and each sample was analyzed in triplicate via reverse-transcriptase quantitative PCR (RTqPCR).
Assessment of the effects of LPS/epimedin B cotreatment-induced DILI in vivo C57BL/6 mice (6-8-week-old female) were subjected to starvation for 24 h and were administered with LPS (2 mg/kg) or saline vehicle via tail vein (i.v.). Following an observation period of 2 h, Epimedin B (20 mg/kg, 40 mg/kg, and 80 mg/kg) or its vehicle was administered via intraperitoneal injection. Mouse serum and a fraction of liver tissues were collected 6 h after epimedin B treatment. Serum ALT, AST, IL-1β, and TNF-α levels were measured. Histopathological analysis was performed via hematoxylin and eosin (H&E) staining. F4/80-positive macrophages in the liver were also estimated.
In the second experiment, female C57BL/6 mice (age: 6-8 weeks) were administered with MCC950 (40 mg/kg) or saline vehicle through intraperitoneal injection. After 1 h, LPS (2 mg/kg) or saline vehicle was administered intravenously via tail vein. Epimedin B (40 mg/kg) was administered via an intraperitoneal injection after 2 h. Mouse serum and a fraction of liver tissues were collected after 6 h. H&E staining was performed, and the serum IL-1β, TNF-α, ALT, and AST levels were determined.

Statistical analyses
Prism 6 and SPSS statistics (version 21.0) were used for statistical analysis. All experimental data are expressed as mean ±SD. A two-tailed unpaired Student's t-test was conducted to evaluate signi cant differences between the two groups. Statistical signi cance was set at P < 0.05.

Epimedin B accelerates NLRP3 in ammasome activation
In the present study, seven major active constituents of EF (epimedin A, epimedin A1, epimedin B, epimedin C, icariin, icaritin, and anhydroicaritin) were analyzed for their ability to activate the NLRP3 in ammasome. The results showed that only epimedin B signi cantly promoted the activation of caspase-1 and IL-1β production induced by nigericin in LPS-primed BMDMs (Fig. 1A-C) but did not affect the production of TNF-α (Fig. 1D). Therefore, only one major active constituent of EF could enhance nigericin-induced NLRP3 in ammasome activation.
To ascertain whether epimedin B could accelerate NLRP3 in ammasome activation, we examined the impact of epimedin B on caspase-1 activation and IL-1β secretion. Epimedin B exhibited dose-dependent active effects on caspase-1 cleavage, IL-1β secretion ( Fig. 2A, B, D), and lactate dehydrogenase (LDH) induced by nigericin in LPS-primed BMDMs (Fig. 2C) but exerted no effect on in ammasome-independent cytokine TNF-α production (Fig. 2E). We also assessed the effect of epimedin B on ATP-induced NLRP3 in ammasome activation. LPS-induced caspase-1 activation and IL-1β secretion in BMDMs could also be induced by epimedin B. However, it exerted no effect on TNF-α production ( Fig. 2F and G, Supplementary  Fig. 1A-C). Additionally, we assessed the impact of epimedin B on nigericin-induced NLRP3 in ammasome activation in THP-1 cells. The results indicated that epimedin B enhanced caspase-1 maturation, IL-1β secretion, and LDH release in a dose-dependent manner in response to nigericin in PMAprimed THP-1 cells (Fig. 2H-K).
Pretreatment with epimedin B promoted nigericin-induced caspase-1 cleavage and IL-1β release in wildtype (WT) BMDMs but not in NLRP3-knockout (NLRP3-/-) BMDMs ( Supplementary Fig. 3A). MCC950 is a small-molecule inhibitor of the NLRP3 in ammasome [24]. We further evaluated whether the activation of the NLRP3 in ammasome induced by epimedin B could be inhibited by MCC950. Our results indicated that epimedin B accelerated NLRP3 in ammasome activation. However, it could also be inhibited by MCC950 (Supplementary Fig. 3B).
Moreover, we explored the effect of epimedin B on NLRP3 in ammasome activation initiated in response to other stimuli. Unexpectedly, treatment with epimedin B exerted no effect on caspase-1 cleavage and IL-1β secretion stimulated by other NLRP3 agonists, including SiO2 and poly(I:C) (Fig. 3A, C-E). Epimedin B did not affect cytosolic LPS, NLRC4, or AIM2 in ammasome activation (Fig. 3B, F-H). Thus, these results indicate that epimedin B is a speci c promoter that increases nigericin-and ATP-induced NLRP3 in ammasome activation.
We then examined whether epimedin B affected LPS-induced priming for in ammasome activation.
When BMDMs were stimulated with epimedin B at doses of 10-40 µM before or after LPS treatment, epimedin B did not activate LPS-induced NLRP3 expression, IL-6, and TNF-α production. Figure 4A-C suggests that epimedin B does not enhance LPS-induced priming at the doses that are effective for NLRP3 activation, suggesting that epimedin B exerts a robust effect on NLRP3 in ammasome activation. Epimedin B promotes nigericin or ATP-induced ASC oligomerization but does not block K + e ux and Ca 2+ ux We investigated the mechanism underlying the activation of NLRP3 by epimedin B. First, our studies showed that epimedin B could activate nigericin-induced ASC oligomerization (Fig. 4D), an essential step for NLRP3 activation, suggesting that epimedin B acted upstream of ASC oligomerization to exacerbate nigericin-induced NLRP3 activation. Second, epimedin B promoted ASC oligomerization induced by ATP (Fig. 4E). However, epimedin B demonstrated no impact on ASC oligomerization induced by SiO2, poly (I:C), poly (dA:dT), Salmonella typhimurium, or cytosolic LPS ( Fig. 4E; Supplementary Fig. 2). This also indicated that epimedin B acted upstream of ASC oligomerization to exacerbate ATP-or nigericin-induced NLRP3 in ammasome activation. Next, we investigated whether epimedin B affected K + e ux during NLRP3 in ammasome activation. The results indicated that epimedin B exhibited no effect on K + e ux triggered by nigericin (Fig. 4F), suggesting that K + e ux did not contribute to the enhancement effect of epimedin B on nigericin-induced NLRP3 in ammasome activation. Moreover, Ca 2+ ux is an extremely important event in the upstream signaling of NLRP3 in ammasome activation. Epimedin B could not block ATP-induced Ca 2+ ux (Fig. 4G). Thus, Ca 2+ ux may not be responsible for the enhanced effect of epimedin B on ATP-induced NLRP3 in ammasome activation.
Epimedin B facilitates NLRP3 in ammasome activation by increasing mitochondrial ROS production Mitochondrial ROS play a crucial role in NLRP3 in ammasome activation [25]. Next, we studied whether epimedin B-mediated mitochondrial ROS was involved in NLRP3 in ammasome activation. Mitochondrial ROS production was not induced after epimedin B treatment alone. Epimedin B successfully potentiated mitochondrial reactive oxygen species production induced by nigericin and ATP but not by SiO2 (Fig. 5A-E). We focused on the ROS scavenger N-acetylcysteine (NAC), which is an inhibitor of mitochondrial ROS production. NAC was selected to evaluate whether the nigericin-dependent activity of epimedin B on NLRP3 activation was mediated by ROS mitochondrial production. Mitochondrial ROS production was suppressed by NAC treatment. As expected, when stimulated with nigericin, NAC treatment reversed epimedin B-induced caspase-1 maturation or IL-1β production ( Fig. 5F-G). These results indicated that epimedin B increased mitochondrial ROS production to facilitate nigericin-induced NLRP3 in ammasome activation.

Epimedin B induces the development of IDILI by promoting NLRP3 in ammasome activation in vivo
We next examined whether epimedin B could induce the development of IDILI by promoting NLRP3 in ammasome activation in vivo. Co-exposure of animals to a non-hepatotoxic dose of LPS and drugs could mimic IDILI. First, we investigated whether epimedin B, which could activate the NLRP3 in ammasome, could also induce liver injury in an LPS-mediated susceptibility mouse model of IDILI.
The results showed that treatment with epimedin B alone did not alter the levels of plasma ALT and AST compared with those of control mice. As expected, in the LPS-mediated mouse model, epimedin B increased the levels of ALT and AST and also induced an increase in the production of IL-1β and TNF-α compared to those in the LPS group (Fig. 6A-D). Similar results were observed in the mRNA expression of the pro-in ammatory genes IL-1β and IL-18 (Fig. 6E, F). Additionally, liver histology analysis showed that the combination of LPS and epimedin B treatment resulted in a trend of hepatocyte focal necrosis and in ammatory cell in ltration in the liver tissue (Fig. 6G). To further explore the effects of epimedin B on the immunological reaction in the liver tissue, immunohistochemical (IHC) analysis of liver samples was performed. IHC staining of liver sections revealed that epimedin B increased the in ltration of F4/80positive macrophages in the liver (Fig. 6H). Our results demonstrated that epimedin B could induce liver injury in the LPS-mediated susceptibility mouse model.
MCC950 is a potent selective NLRP3 inhibitor that is deemed a useful tool to mimic the consequences of NLRP3 in ammasome knockdown in mice [23]. To verify the relationship between the NLRP3 in ammasome and liver injury induced by epimedin B, mice treated with MCC950 were included. Figure 7 (A-D) indicates that the combination of LPS and epimedin B led to an increase in the levels of ALT, AST, IL-1β, and TNF-α but not in cotreated MCC950 mice. Moreover, mRNA expression of pro-in ammatory genes IL-1β, IL-18, and TNF-α was increased in the LPS group and the group subjected to treatment with LPS in combination with epimedin B but not in the MCC950 cotreatment group (Fig. 7E-G). As shown in Fig. 7H, MCC950 treatment suppressed caspase-1 activation in the liver tissues co-treated with epimedin B and LPS. Liver histology analysis showed that epimedin B induced hepatocyte focal necrosis and in ammation in the LPS-mediated mouse model but not in other groups (Fig. 7I). These results con rmed that epimedin B could activate the NLRP3 in ammasome, leading to liver injury in vivo.

Discussion
IDILI represents a major global health issue. For instance, in North America, it has already surpassed viral hepatitis as a major cause of acute liver failure [26,27]. However, in recent years, IDILI caused by TCM has also been broadly recognized, especially in traditional non-toxic Chinese medicines. Thus, there is a lack of understanding of the possible IDILI mediated by TCM globally. Few traditional nontoxic Chinese medicines, such as Dictamni Cortex, Epimedii Folium, Psoraleae Fructus, and Polygoni Multi ori Radix have been reported to cause liver injury. Particularly, the Chinese Food and Drug Administration has reported that liver injury can be caused by two Chinese patent medicines, namely Zhuangguguanjiewan pills and Xianlinggubao capsules, which both contain Epimedii Folium. Furthermore, our previous studies have demonstrated that Epimedii Folium may induce hepatotoxicity in an LPS-mediated susceptibility mouse model of IDILI. Interestingly, in the present study, we demonstrated that epimedin B which is one of the constituents about EF could speci cally reinforce the activation of the NLRP3 in ammasome, then to induce liver jury.
The NLRP3 in ammasome is one of the major contributors to in ammation and possesses the ability to sense both endogenous and exogenous danger signals through intracellular NLRs [12,13]. NLRP3 in ammasome activation is also responsible for the development of several liver-related in ammatory diseases, especially chronic hepatitis C, nonalcoholic steatohepatitis, alcoholic liver disease, and IDILI [17,[28][29][30][31][32][33]. EF is a commonly used herbal medicine for promoting the functionality of liver and kidney. Our previous studies con rmed that EF could induce liver injury in an LPS-mediated susceptible mouse model of IDILI. However, the mechanisms underlying NLRP3 in ammasome-mediated hepatic in ammation of the major active constituents of EF remain elusive. In the present study, we demonstrated that epimedin B, which is one of the major bioactive components, could facilitate nigericin-or ATP-induced NLRP3 in ammasome activation, leading to the development of IDILI. This suggests that epimedin B contributes to EF-induced idiosyncratic hepatotoxicity via enhancement of NLRP3 in ammasome activation.
In the present study, epimedin B could speci cally reinforce the activation of the NLRP3 in ammasome induced by nigericin or ATP but not by SiO2, poly (I:C), and cytosolic LPS. Additionally, epimedin B exerted no effect on the activation of AIM2 and NLRC4 in ammasomes. These data indicate that epimedin B could speci cally enhance nigericin-or ATP-induced NLRP3 in ammasome activation. We also examined the effects of epimedin B on upstream and downstream signaling, which is associated with NLRP3 in ammasome activation, and evaluated the mechanism underlying the enhancement of nigericin-or ATP-induced NLRP3 in ammasome activation by epimedin B. ASC oligomerization is a key event in NLRP3 in ammasome activation. Epimedin B promoted ASC oligomerization triggered by nigericin and ATP. Therefore, we noted that epimedin B acted on the upstream signaling events of ASC oligomerization to exacerbate nigericin-or ATP-induced NLRP3 in ammasome activation. K + e ux is considered one of the main upstream events of NLRP3 in ammasome activation. However, our study indicated that epimedin B did not alter the K + e ux triggered by nigericin. Ca 2+ ux is also deemed one of the upstream mechanisms associated with NLRP3 in ammasome activation, but our results demonstrated that epimedin B did not alter Ca 2+ ux triggered by ATP. Moreover, mitochondrial damage and the release of mitochondrial reactive oxygen species are key upstream events of NLRP3 in ammasome activation. This study showed that nigericin and ATP could induce the production of mitochondrial reactive oxygen species. Notably, epimedin B speci cally ampli es the production of mitochondrial reactive oxygen species triggered by nigericin and ATP, but not by SiO2, suggesting that epimedin B may facilitate nigericin-or ATP-induced NLRP3 in ammasome activation dependent on mitochondrial ROS production. Next, we evaluated whether ROS played an important role in the enhanced effect of epimedin B on nigericin-or ATP-induced NLRP3 in ammasome activation. The results also indicated that NAC could inhibit NLRP3 in ammasome activation triggered by nigericin. We concluded that the effect of epimedin B was dependent on mitochondrial ROS production for facilitating nigericin-induced NLRP3 in ammasome activation.
We have previously reported that EF combined with non-hepatotoxic doses of LPS can induce liver injury. Some studies have also indicated that co-exposure to a non-hepatotoxic dose of LPS and drugs in animals could mimic IDILI. Therefore, we investigated whether epimedin B acting as an NLRP3 in ammasome activation promoter could cause liver injury in an LPS-mediated susceptible mouse model of IDILI. Our results indicated that epimedin B could induce liver injury in vivo in the LPS-mediated susceptibility mouse model. Thereafter, MCC950 was used to explore the relationship between liver injury induced by epimedin B and the NLRP3 in ammasome. Taken together, the combination of epimedin B and LPS induced liver injury but not in mice with MCC950 pretreatment. These data clearly demonstrated that epimedin B induced IDILI by promoting NLRP3 in ammasome activation in vivo.

Conclusions
In conclusion, our study demonstrated that epimedin B could induce NLRP3 in ammasome activation triggered by nigericin and ATP. Mitochondrial ROS are crucial contributors to the enhancement of the activation of the NLRP3 in ammasome stimulated by epimedin B. Treatment with the combination of nonhepatotoxic doses of LPS and epimedin B increased the production of ALT, AST, IL-1β, and TNF-α, resulting in hepatocyte necrosis. However, the results were not observed in mice co-treated with LPS and MCC950. Our study indicates that epimedin B, as a major risk component, is also responsible for EFinduced IDILI.   release of LDH (J), and ELISA of IL-1β(K) in Sup from samples described in H. Data are represented as mean ± SD from at least three biological samples. The signi cance of the differences was analyzed using unpaired Student's t-test: *P <0.05, **P <0.01, ***P <0.001, NS; not signi cant, RLUs; the relative light units.  -1β (D, G), and TNF-α E, H in Sup derived from samples described in A and B. RLUs, the relative light units. Data are represented as mean ± SD from at least three biological samples. The signi cance of the differences was analyzed using unpaired Student's t-test: *P <0.05, **P <0.01, ***P <0.001, NS: not signi cant. Figure 4 Epimedin B promotes ATP or nigericin-induced ASC oligomerization but does not block K+ e ux and Ca2+ ux. (A) Western blot analysis of whole-cell lysates from BMDMs subjected to treatment with epimedin B for 1 h; thereafter, they were stimulated with LPS (50 ng/mL) for 3 h or BMDMs were stimulated with LPS (50 ng/mL) for 3 h and then subjected to treatment with epimedin B for 1 h. (B, C) ELISA of TNF-α (B) and IL-6 (C) in Sup derived from samples described in A. (D) Western blot analysis of ASC oligomerization from LPS-primed BMDMs subjected to treatment with various doses of epimedin B before nigericin stimulation. (E) Western blot analysis of ASC oligomerization from LPS-primed BMDMs subjected to treatment with epimedin B, following which they were stimulated with nigericin, ATP, SiO2, and poly (I:C). (F) Quanti cation of potassium e ux in LPS-primed BMDMs subjected to treatment with various doses of epimedin B and were then stimulated with nigericin. (G) A trace of ATP-induced Ca2+ ux was measured using the FLIPRTETRA system in LPS-primed BMDMs subjected to treatment with epimedin B. Data are represented as mean ± SD from at least three biological samples. The signi cance of the differences was analyzed using unpaired Student's t-test: *P <0.05, **P < 0.01, ***P < 0.001, NS: not signi cant. positive cells in LPS-primed BMDMs subjected to treatment with epimedin B that were either not stimulated or stimulated with nigericin, ATP, or SiO2. (F) Western blot analysis of supernatants and wholecell lysates derived from LPS-primed BMDMs subjected to treatment with epimedin B, NAC, or epimedin B plus NAC before stimulation with nigericin or without stimulation. (G) Caspase-1 activity in samples described in F. Data are represented as mean ± SD derived from at least three biological samples. The signi cance of the differences was analyzed using unpaired Student's t-test: #P <0.05, ##P < 0.01, ###P < 0.001, *P <0.05, **P <0.01, ***P <0.001, NS: not signi cant. Figure 6 Epimedin B promotes early liver injury and in ammatory mediator production in vivo (A-H) Female C57BL/6 mice (age: 6-8 weeks) subjected to starvation for 24 h were administered with 2 mg/kg of LPS or its saline vehicle via the tail vein (i.v.). After an observation period of 2 h, various doses of epimedin B (20 mg/kg, 40 mg/kg, 80 mg/kg) or its vehicle were administered through intraperitoneal injection for 6 h. The signi cance of the differences was analyzed using unpaired Student's t-test: *P <0.05, **P <0.01, ***P <0.001, NS: not signi cant.