Med1 is downregulated in ALF hepatocytes

The LPS/D-GalN-induced ALF model resulted in mortality begin at 5–6 h, consistent with our previous findings [25]. Notably, compared with livers from the normal control group, no significant differences were observed in the appearance of the liver harvested from the ALF group at 3 h. However, at 6 h, the harvested liver from the ALF group exhibited notable swelling and congestion (Fig. 1A). Additionally, at 6 h, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were significantly elevated (Fig. 1B). In addition, H&E staining revealed mild edema of hepatocytes at 3 h, whereas at 6 h, extensive liver damage (characterized by widespread hemorrhage and necrosis) was observed (Fig. 1C, D). These findings indicate the successful establishment of the ALF mouse model 6 h post-LPS/D-GalN exposure. To determine the role of Med1 in ALF, we evaluated Med1 expression in the livers of ALF mice. Our analysis revealed a decrease in both protein and mRNA levels of Med1 in the livers of ALF mice (Fig. 1E–G). Furthermore, we assessed Med1 expression in damaged hepatocytes in vitro. Remarkably, the protein levels of Med1 were diminished in hepatocytes following TNF-α/D-GalN or H2O2 stimulation (Fig. 1H–K).

Fig. 1

Med1 expression is downregulated in ALF hepatocytes. A Representative pictures of mice treated with PBS or LPS (60 μg/kg)/D-GalN (800 mg/kg) for 3 and 6 h (n = 5). B Serum levels of ALT and AST were significantly elevated in ALF mice (n = 5). C H&E staining of liver tissues from mice treated with PBS or LPS/D-GalN for 3 and 6 h (n = 5). Scale bar: 100 μm. D Quantification of necrotic area between three groups (n = 5). E Protein levels of Med1 in liver tissues from mice treated with PBS or LPS/D-GalN for 6 h (normalized to β-Tubulin) measured by Western blot and F relative grayscale analysis using Image J software (n = 3). G mRNA levels of Med1 in liver tissues from mice treated with PBS or LPS/D-GalN for 6 h (normalized to β-actin), assessed by RT-qPCR (n = 5). H–K Protein levels of Med1 in L02 and THLE2 cells following stimulation with TNF-α (100 ng/ml)/D-GalN (7.5 mg/ml) for 24 h or H2O2 (1.0 mM) for 2 h evaluated using Western blot and quantified using Image J software (n = 3–4). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, *p < 0.05, ***p < 0.001, ****p < 0.0001

Med1 overexpression alleviated LPS/D-GalN-induced ALF in mice

Since Med1 was downregulated in the liver of LPS/D-GalN-induced ALF mice, we aimed to investigate the potential protective effects of Med1 overexpression on liver injury. To establish the Med1 overexpression mouse model, Ad-Med1 particles were administered via the tail vein (Fig. 2A). Med1 overexpression in the liver of Ad-Med1 mice was confirmed through Western blot analysis (Fig. 2B, C). Subsequently, ALF was induced using LPS/D-GalN (Fig. 2A). In the ADM-FH group, mice began to die within 5 h of LPS/D-GalN exposure, with all mice succumbing to ALF within 10 h. In contrast, only one mouse in the Ad-Med1 group died within 24 h, suggesting that Med1 overexpression significantly reduced the 24-h ALF mortality rate (P < 0.0001). Additionally, Med1 overexpression led to a decrease in serum ALT, AST, and lactate dehydrogenase (LDH) levels (Fig. 2E–G). Furthermore, Med1 overexpression was associated with TNF-α and IL-6 downregulation in the liver (Fig. 2H, I). In contrast, following LPS/D-GalN exposure, liver sections from the Ad-Med1 group displayed reduced hepatic congestion and hemorrhage (Fig. 2J, K). These findings strongly indicate that Med1 overexpression effectively alleviates the progression of LPS/D-GalN-induced ALF.

Fig. 2

Med1 overexpression alleviated LPS/D-GalN-induced ALF in mice. A Wild-type mice injected with Ad-Med1 (3.6 × 109 pfu in 200 μl) or empty vector ADM-FH via the tail vein to establish the Med1 over-expression model or controls, respectively, 3 days later intraperitoneally injected with LPS (60 μg/kg)/D-GalN (800 mg/kg) to induce ALF. B, C Western blotting and relative grayscale analysis confirmed Med1 overexpression in the liver of Ad-Med1 mice (n = 3). D Med1 overexpression significantly improved the 24-h survival rate of ALF (n = 10). E–G Serum levels of ALT, AST, and LDH were elevated after LPS/D-GalN injection in the ADM-FH group, but significantly decreased in the Ad-Med1 group (n = 5). H, I mRNA levels of TNF-α and IL-6 in liver tissues were elevated after LPS/D-GalN injection in the ADM-FH group, but significantly decreased in the Ad-Med1 group (n = 5). J–K H&E staining of liver tissues and quantitation of necrosis area between groups (n = 5). Scale bar: 100 μm. Data are presented as mean ± SEM. Statistical analysis was performed using the Student’s t-test, *p < 0.05, **p < 0.01, ****p < 0.0001

Fer-1 inhibits ferroptosis but does not alleviate inflammation in ALF

Extensive research has demonstrated that ferroptosis plays a critical role in the cellular demise observed in liver failure models [4, 28]. To determine the involvement of ferroptosis in LPS/D-GalN-induced ALF in mice, we conducted qPCR analysis to assess the expression of ferroptosis-associated marker genes in the liver of ALF mice. PTGS2 and SLC7A11 gene expression was significantly upregulated (Fig. 3A, B), while GPX4 expression was noticeably downregulated (Fig. 3C). Subsequently, we determined the extent of lipid peroxidation in hepatic tissues. MDA and 4-HNE in the ALF group were significantly higher than that of the control group (Fig. 3D–F). To further explore the potential involvement of ferroptosis in LPS/D-GalN-induced ALF, Fer-1 was administered to determine the effects of ferroptosis inhibitors in ALF. Fer-1 effectively mitigated the elevation of AST and ALT (Fig. 3G, H). In addition, following 6 h of LPS/D-GalN exposure, Fer-1 alleviated liver congestion and hepatocyte necrosis (Fig. 3I, J). These findings suggest that ferroptosis plays a significant role in the development of LPS/D-GalN-induced liver injury in mice. However, it should be noted that Fer-1 administration did not result in TNF-α or IL-6 mRNA downregulation (Fig. 3K, L).

Fig. 3

Ferroptosis is an important driver of LPS/D-GalN-induced ALF. A, B mRNA levels of PTGS2 and SLC7A11 increased in liver tissues of the ALF group (n = 5). C GPX4 mRNA expression decreased in liver tissues of the ALF group (n = 5). D MDA elevated in liver tissues of the ALF group (n = 5). E, F 4-HNE immunohistochemistry staining and quantitative results showed that 4-HNE increased in liver tissues of the ALF group (n = 5). G, H Serum levels of AST and ALT were elevated in ALF mice, but significantly decreased in Fer-1 (10 mg/kg) treatment group (n = 5). I, J H&E staining of liver tissues and quantitation of necrosis area showed that Fer-1 alleviated liver damage (n = 5). K, L TNF-α and IL-6 mRNA expression between groups (n = 4–5). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, **p < 0.01, ***p < 0.001, ****p < 0.0001

Overexpression of Med1 alleviated LPS/D-GalN-induced ferroptosis in ALF

To elucidate the role of Med1 in ferroptosis, we evaluated the expression of multiple ferroptosis-linked markers in Ad-Med1 and ADM-FH mice co-injected with LPS/D-GalN. Med1 overexpression successfully alleviated GSH depletion in the mice liver (Fig. 4A). Furthermore, the expression of MDA and 4-HNE in the liver of Ad-Med1 mice decreased significantly (Fig. 4B–D). Subsequently, an electron microscopy analysis was conducted to examine the mitochondria in the liver. Compared with the control group, electron microscopy revealed that the mitochondrial structure in liver tissues obtained from the ADM-FH group were altered; mitochondrial swelling, outer mitochondrial membrane rupture, and disorganized cristae were observed. Med1 overexpression improved the aforementioned morphological phenotype (Fig. 4E). Additionally, we found that Med1 overexpression restricted LPS/D-GalN-induced SLC7A11 and PTGS2 upregulation (Fig. 4F, G). Furthermore, hepatic expression of TfR1 and ACSL4 was a notably reduced in the Ad-Med1 group (Fig. 4H–J). In contrast, GPX4 expression was diminished in the ADM-FH group (Fig. 4H, K). These findings suggest that Med1 overexpression effectively mitigated the occurrence of LPS/D-GalN-induced ferroptosis in ALF.

Fig. 4

Med1 overexpression alleviated LPS/D-GalN-induced ferroptosis in ALF. A Med1 overexpression alleviated the depletion of GSH in the liver of mice following LPS/D-GalN injection (n = 5). B Med1 overexpression inhibited the generation of MDA in the liver of mice after LPS/D-GalN injection (n = 5). C, D Immunohistochemistry staining and quantitative analysis showed that 4-HNE increased in the ADM-FH group, but not in Ad-Med1 group after LPS/D-GalN injection (n = 5). E Transmission electron micrographs of mitochondria from ALF mice pretreated with Ad-Med1 or ADM-FH, compared to normal controls, representative images from n = 3 mice per group, Scale bars: up, 2 µm; down, 500 nm. F, G mRNA levels of SLC7A11 and PTGS2 in liver tissues were elevated after LPS/D-GalN injection in the ADM-FH group, but significantly decreased in the Ad-Med1 group (n = 5). H Western blotting and I–K relative grayscale analysis by Image J software to assess the protein levels of TfR1, ACSL4, and GPX4 in the liver between groups (n = 5). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Med1 inhibits erastin-induced ferroptosis in hepatocytes via Nrf2 activation

To evaluate the potential protective effects of Med1 against ferroptosis, we established an in vitro hepatocyte ferroptosis model using erastin. Western blot analysis was used to confirm Med1 overexpression in two Lv-Med1-transfected cell lines (L02 and THLE2) (Fig. 5A–C). Upon erastin exposure, cell viability in Lv-Med1 cells was significantly higher than that in Lv-NC cells (Fig. 5D, E). Additionally, Med1 overexpression decreased MDA production (Fig. 5F, G), and increased GSH (Fig. 5H, I) in erastin-treated hepatocytes. These findings support the in vivo findings; Med1 plays an inhibitory role in erastin-induced ferroptosis in hepatocytes.

Fig. 5

Med1 inhibits erastin-induced ferroptosis in hepatocytes. A–C Western blotting and relative grayscale analysis by Image J software confirmed Med1 overexpression in L02 and THLE2 cells transfected with Lv-Med1 (n = 3–4). D, E Relative cell viability of L02 and THLE2 transfected with Lv-Med1 or Lv-NC confirmed using the CCK-8 assay (n = 6–8). F–G MDA increased following erastin exposure (10 μM, 24 h) in Lv-NC cells, but not in Ad-Med1 cells (n = 3–5). H, I The level of GSH decreased following erastin exposure in Lv-NC cells, Med1 overexpression inhibited GSH depletion (n = 3). J Western blotting and relative grayscale analysis by Image J software K–O revealed the protein level of Med1, Nrf2, HO-1, SLC7A11, and GPX4 in Lv-Med1 (L02) and Lv-NC (L02) cells treated with or without erastin (n = 3–4). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, *p < 0.05, **p < 0.01, ***p < 0.001

In order to gain a deeper understanding of the regulatory mechanism of Med1 in ferroptosis, we evaluated the expression of ferroptosis-related molecules. Interestingly, in Lv-Med1 cells, Nrf2 expression increased significantly regardless of the presence or absence of erastin stimulation (Fig. 5J–L). HO-1, a gene primarily regulated by Nrf2, exhibited a similar trend (Fig. 5J, M). However, there were no notable differences in SLC7A11 or GPX4 expression between Lv-Med1 cells and Lv-NC cells (Fig. 5J, N, O). Subsequently, two distinct siRNAs were used to inhibit Med1 expression in L02 cells. Western blot analysis revealed that the expression of both Med1 and Nrf2 in siMed1 cells were significantly reduced (Fig. 6A–C). However, SLC7A11, HO-1, and GPX4 protein expression was comparable between the siMed1 and siNeg groups (Fig. 6A, D–F). Furthermore, erastin administration stimulated Nrf2 and HO-1 expression in siNeg cells, but not in siMed1 cells (Fig. 6A, C, E). Moreover, upon erastin exposure, the expression of GPX4 was significantly diminished in siMed1 cells (Fig. 6A, F). The siMed1 experiment was replicated in primary mouse hepatocytes, revealing that the expression patterns of Nrf2 and GPX4 proteins were consistent with those of L02 cells (Fig. 6G–I, M). While HO-1 exhibited divergent behavior between the two cell lines (Fig. 6G, K), NQO1, another downstream antioxidant gene regulated by Nrf2, showed significant downregulation in siMed1 cells (Fig. 6G, J). Additionally, the expression of SLC7A11 was also unaffected by Med1 but decreased with erastin treatment (Fig. 6G, L). These results indicate that Med1 plays a crucial role in the activation of Nrf2 and its downstream targets.

Fig. 6

Med1 knockdown affects ferroptosis-related molecules. A Western blotting and relative grayscale analysis by Image J software verified siMed1 knockdown in L02 cells (B), and assessed Nrf2 (C), SLC7A11 (D), HO-1 (E), and GPX4 (F) protein expression in siNeg and siMed1 cells treated with or without erastin (n = 3–4). G Western blotting and relative grayscale analysis by Image J software verified siMed1 knockdown in primary mouse hepatocytes (H), and assessed Nrf2 (I), NQO1 (J), HO-1 (K), SLC7A11 (L), and GPX4 (M) protein expression in siNeg and siMed1 cells treated with or without erastin (n = 3). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

To investigate whether the inhibition of ferroptosis by Med1 is dependent on Nrf2, we utilized the Nrf2 inhibitor ML385 to suppress the expression of Nrf2 in Lv-Med1 cells. ML385 treatment inhibited the anticipated Nrf2 and HO-1 upregulation in Lv-Med1 cells (Fig. 7A–D), and GPX4 decreased significantly under these conditions (Fig. 7E). The Cell Counting Kit-8 (CCK-8) assay demonstrated that Med1 overexpression improved cell viability. However, this beneficial effect was counteracted by the presence of ML385 (Fig. 7F). Furthermore, ML385 reversed the protective ferroptosis effects exerted by Med1. This was evident through the observed increase in MDA levels and decrease in GSH levels (Fig. 7G, H). Additionally, the expression of Nrf2 downstream antioxidant genes HO-1, Glutamate cysteine ligase catalytic (GCLC), and NQO1 increased significantly in Lv-Med1 cells following erastin treatment. However, the effects were reversed upon ML385 administration (Fig. 7I–K). These findings suggest that Med1 promotes the expression of Nrf2 and its downstream genes.

Fig. 7

Ferroptosis inhibition by Med1 was reversed by ML385. A Western blotting and relative grayscale analysis by Image J software assessed the protein levels of Med1 (B), Nrf2 (C), HO-1 (D), and GPX4 (E), Lv-Med1 cells were treated with ML385 (10 μM for 12 h) before the treatment of erastin (10 μM for 24 h) in Lv-Med1 + ML385 + Erastin group (n = 3). F Relative Lv-NC and Lv-Med1 cell viability assessed using the CCK-8 assay (n = 13). G MDA and H GSH levels detected in Lv-NC and Lv-Med1 cells treated with or without eratstin and ML385 (n = 3). I–K mRNA levels of HO-1, GCLC, and NQO1 increased in Lv-Med1 cells following erastin treatment, but decreased upon the administration of ML385 (n = 3). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Consistent with the in vitro findings, upon LPS/D-GalN-exposure, the expression of Nrf2, NQO1, and HO-1 proteins were upregulated in the liver of the Ad-Med1 group (Fig. 8A–D). To explore how Med1 promoted Nrf2 expression, the mRNA levels of Nrf2 were assessed. Nrf2 mRNA levels decreased in the ADM-FH group following LPS/D-GalN treatment, while they increased in the Ad-Med1 group regardless of LPS/D-GalN stimulation (Fig. 8E), suggesting that Med1 overexpression facilitated the transcription of the Nrf2 gene. Nevertheless, the precise mechanism underlying the interaction between Med1 and Nrf2 remains elusive. The STRING database (https://string-db.org/) was utilized to analyze molecular interactions, revealing that among the top ten predicted functional partners of Nrf2 (also known as Nfe2l2) in mice with a confidence level of 0.700, only peroxisome proliferator-activated receptor gamma co-activator 1 alpha (Ppargc1α) exhibited interaction with Med1 (Fig. 8F). Furthermore, the mRNA expression of Ppargc1α was upregulated in Ad-Med1 mice regardless of LPS/D-GalN stimulation (Fig. 8G), suggesting a potential role of Med1 in enhancing Nrf2 transcriptional activity through Ppargc1α.

Fig. 8

Med1 inhibits ferroptosis via Nrf2 activation. A Western blotting and relative grayscale analysis by Image J software assessed the protein levels of Nrf2 (B), NQO1 (C), and HO-1 (D) in Ad-Med1 and ADM-FH groups after LPS/D-GalN injection (n = 5). E mRNA level of Nrf2 between groups (n = 3–5). F The interaction of top ten predicted functional partners of Nrf2 (Nfe2l2) and Med1 shown in the STRING database (https://string-db.org/). G mRNA level of Ppargc1α between groups (n = 3–5). Data are presented as mean ± SEM. Statistical analysis was performed using Student’s t-test, *p < 0.05, **p < 0.01, ****p < 0.0001