The cell fate regulator DACH1 modulates ferroptosis through affecting P53/SLC25A37 signaling in fibrotic disease

Abstract

Background:

Dachshund homolog 1 (DACH1) is widely acknowledged for its involvement in regulating diverse cell fates, but its precise regulatory mechanism in ferroptosis remains elusive. In this study, we investigated whether DACH1 modulates ferroptosis through affecting P53/solute carrier family 25 member 37 (SLC25A37) signaling in hepatic fibrogenesis.

Methods:

CRISPR-Cas9 system was used to knockout DACH1 in HSC to determine the effect of DACH1 on ferroptosis. Immunoprecipitation, pulldown, and mouse model of hepatic fibrogenesis were used to analyze the potential molecular mechanism of ferroptosis regulation by DACH1.

Results:

We found that ferroptosis inducers increased the protein expression of DACH1 by suppressing the ubiquitin-proteasome signaling. DACH1 knockout can resist ferroptosis, whereas DACH1 knockin can enhance it. Interestingly, the upregulation of DACH1 resulted in the mitochondrial translocation of p53 by inducing phosphorylation at serine 392. The mutation of serine 392 can prevent the combination of DACH1 and p53, the mitochondrial translocation of p53, and DACH1-mediated ferroptosis. Moreover, SLC25A37 was identified as a candidate target for mitochondrial p53. The binding of p53 to SLC25A37 can enhance the iron uptake capacity of SLC25A37, which may cause an overload of iron in the mitochondria and hyperactive mitochondrial electron transport chain. Knockdown of SLC25A37 can impair p53-mediated mitochondrial iron overload and ferroptosis. Furthermore, treatment with erastin can induce HSC ferroptosis and relieve fibrotic lesion damage in the mouse model of hepatic fibrogenesis. HSC-specific knockdown of DACH1, p53, and SLC25A37 can abolish the induction of HSC ferroptosis and reversal of hepatic fibrogenesis by erastin treatment.

Conclusions:

Our findings suggest that the DACH1/P53/SLC25A37 signaling pathway is a promising target for fibrotic disorders and reveals new regulatory mechanisms of ferroptosis.

INTRODUCTION

Hepatic fibrogenesis involves the excessive deposition of the extracellular matrix due to chronic damage repair, leading to potential progression to cirrhosis.¹ The pivotal mechanism in hepatic fibrogenesis is the transdifferentiation of quiescent HSCs into extracellular matrix–producing myofibroblasts.² Consequently, targeting the removal of activated HSCs emerges as a promising therapeutic strategy for reversing hepatic fibrogenesis. Our previous studies demonstrated that inducing HSC apoptosis,³ necroptosis,⁴ senescence,⁵ and lipocyte phenotype⁶ can alleviate hepatic fibrogenesis. Notably, ferroptosis is a recently identified form of programmed cell death, which presents a novel approach for eliminating HSCs.⁷ Our recent findings highlight the involvement of RNA-binding proteins ELAV like RNA binding protein 1 and zinc finger protein 36 in HSC ferroptosis through the regulation of autophagy signaling.^8,9 Additionally, we reported that m⁶A methylation is required for dihydroartemisinin in alleviating liver fibrosis by inducing HSC ferroptosis.¹⁰ The current study aims to better understand the critical function of ferroptosis in the removal of activated HSCs.

Ferroptosis is a regulated form of cell death initiated by iron-dependent peroxidation of polyunsaturated fatty acids.¹¹ In contrast to other cell death mechanisms, ferroptosis is characterized by unique morphological and genetic features. Ferroptosis can be induced by physiological factors such as elevated extracellular glutamate or by small molecules such as sorafenib and erastin, which inhibit system X_c ⁻-mediated cysteine import.¹² Genetic deletion or functional inhibition of the glutathione-dependent antioxidant enzyme glutathione peroxidase 4 also serves as a trigger for ferroptosis.¹² Conversely, iron chelators, lipophilic antioxidants, inhibitors of lipid peroxidation, and depletion of polyunsaturated fatty acyl phospholipids can prevent ferroptotic cell death.¹² Mechanistically, ferroptosis is regulated by several pathways, including the p53, glutamine, autophagy, and p62-Keap1-nuclear factor erythroid-2-related factor 2 pathways.^11,12 Importantly, mitochondria play a critical role in regulating iron homeostasis and energy metabolism, and mitochondrial injury has been linked to lipid peroxidation and iron dyshomeostasis.¹³ Subsequent research is imperative to explore the potential contribution of mitochondrial iron metabolism to ferroptosis.

Dachshund homolog 1 (DACH1) is recognized as a crucial cell fate determination factor that regulates many important cellular life activities.^14,15 DACH1 was reported to induce cell apoptosis and arrest the cell cycle in the G2/M phase by upregulating p53 expression in HCC.¹⁴ Moreover, DACH1 was showed to inhibit the proliferation and migration of papillary thyroid carcinoma via interacting with various cytokines such as C-X-C motif ligand8 and C-X-C motif ligand12.¹⁵ Noteworthy, ferroptosis acts as a cell death modality that is characterized by the iron-dependent accumulation of lipid peroxidation, but whether DACH1 participates in regulating ferroptosis has not been reported. Importantly, many studies found abnormal expression of DACH1 in various fibrotic diseases, suggesting that DACH1 may be involved in the occurrence and development of fibrotic diseases. DACH1 in type II alveolar epithelial cells contributes to the progression of pulmonary fibrogenesis and may be a viable target for the treatment of pulmonary fibrosis.¹⁶ Moreover, transcriptome-wide association analysis identifies DACH1 as a kidney disease risk gene that contributes to fibrogenesis.¹⁷ These findings make us realize that DACH1 may be involved in regulating hepatic fibrogenesis.

In this study, we found that the cell fate regulator DACH1 could trigger HSC ferroptosis through affecting P53/solute carrier family 25 member 37 (SLC25A37) signaling in hepatic fibrogenesis. This study illuminates the regulatory mechanisms of ferroptosis in hepatic fibrogenesis, highlighting the significance of DACH1 as a potential therapeutic target.

METHODS

Animal experiments

A total of 60 male C57BL/6 mice were purchased from Nanjing Medical University (Nanjing, China) and randomly assigned to 6 groups with 10 mice in each group. The groups were composed of sham, bile duct ligation (BDL)+vitamin A (VA)-Lip-Control- short hairpin RNA (shRNA), BDL+VA-Lip-Control-shRNA+erastin, BDL+VA-Lip-P53-shRNA+erastin, BDL+VA-Lip-DACH1-shRNA+erastin, BDL+VA-Lip-SLC25A37-shRNA+erastin. To induce hepatic fibrogenesis, a midline open procedure was performed and the common bile duct was ligated with a 3-0 surgical filament immediately adjacent to the liver bifurcation. Sham group received a sham procedure. After the BDL operation, VA-Lip-Control-shRNA, VA-Lip-DACH1-shRNA, VA-Lip-P53-shRNA, and VA-Lip-SLC25A37-shRNA were administered intravenously 3 times a week for 2 weeks, followed by i.p. administration of erastin for 2 weeks at a dose of 30 mg/kg once every other day. At the end of the experiment, liver tissues were collected under general anesthesia for histopathological studies. All animal experiments were approved by the institutional and local committee of Nanjing University of Chinese Medicine. All animals received humane care according to the National Institutes of Health (USA) guidelines. We complied with the ARRIVE reporting guidelines.

Histological analysis

According to a reported experimental protocol,¹⁸ the liver tissue samples were fixed in a fixed buffer for 2 days, followed by embedding in paraffin after transferring them to different concentrations of ethanol for histopathological analysis. Hematoxylin-eosin, Sirius red, and Masson staining were used to prepare 4-μm sections for histological observation. The quantification of staining fields from 10 random regions was performed using ImageJ software.

Isolation and characterization of primary mouse HSCs

Primary mouse HSCs were isolated from livers, according to previous reports.^8–10 Briefly, the livers were irrigated with a solution containing DNase to prevent gelatinous substances and undigested fragments and then filtered to remove any remaining debris. The filtrate was centrifuged at 50g for 5 minutes at 4°C, and the primary HSCs were separated using gradient centrifugation with 25% Nycodenz. The purified cells were placed on a 60-mm tissue culture dish and characterized by the detection of alpha-smooth muscle actin.

Reagents and antibodies

Liproxstatin-1, Z-VAD-FMK, necrostatin-1, erastin, (1S,3R)-RSL3 (RSL3), ferroptosis inducer (FIN56), 20S proteasome inhibitor, cycloheximide, deferoxamine, eugenol, and necrosulfonamide were obtained from Selleck Chemicals. Rotenone, diethyl butylmalonate, antimycin, and NaN₃ were purchased from Santa Cruz Biotechnology. Anti-DACH1 antibody, anti-p53 antibody, anti-p53 (phospho S392) antibody, anti-SLC25A37 antibody, anti-Hsp90 antibody, anti-Voltage dependent anion channel protein antibody, anti-lamin B antibody, and anti-beta actin antibody were obtained from Abcam Technology. Anti-mouse IgG and anti-rabbit IgG were bought from Cell Signaling Technology.

Cell culture and drug treatment

The rat cell line HSC-T6 and human cell line HSC-LX2 were obtained from BeNa culture collection and cultured in DMEM supplemented with 10% fetal bovine serum and 1% antibiotics in a humidified incubator at 37°C with 5% CO₂ and 95% air. Drugs were prepared in DMSO at a concentration of 10 mM and adjusted as needed by dilution with DMSO. Cells were treated with drugs at various doses for different durations when they reached 70% confluence.

Construction and transfection of plasmids

DACH1 shRNA, p53 shRNA, and SLC25A37 shRNA were obtained from Santa Cruz Biotechnology. The eukaryotic expression vector3 (pcDNA).1-DACH1 plasmid, pcDNA3.1-P53 plasmid, and pcDNA3.1-SLC25A37 plasmid were purchased from Hanbio (Shanghai, China). P53 plasmids with different domains were cloned and inserted into the pcDNA 3.1-myc vector. VA-Lip-Control-shRNA, VA-Lip-SLC25A37-shRNA, VA-Lip-P53-shRNA, and VA-Lip-DACH1-shRNA were prepared according to previous reports.^8–10 According to the manufacturer’s protocols, shRNA and plasmids were transfected by Lipofectamine 3000. Stable clones were selected in 2 µg/mL puromycin for 4 weeks, and a single clone was isolated using a limited dilution technique.

CRISPR/Cas9-mediated knockout of DACH1 and S392A mutant of p53

We used the CRISPR/Cas9 system to interfere with gene expression as described.¹⁹ To knockout (KO) DACH1 in rat and human HSC cells, we prepared CRISPR/Cas9 KO plasmids consisting of a pool of 3 plasmids, each encoding the Cas9 nuclease and a target-specific 20 nt guide RNA designed for maximum KO efficiency. Additionally, we sequenced regions of the p53 gene from HSC-T6 cells to construct an S392A mutant. Each selected guide RNA was inserted into the pSpCas9 (BB)-2A-Puro (PX459) V2.0 vector obtained from Addgene. We used a 200-base single-stranded DNA oligonucleotide purchased from integrated DNA to insert the mutation into the genome by homologous recombination. According to the manufacturer’s protocols, plasmids were transfected by Lipofectamine 3000. Puromycin (0.5 μg/mL) was used for screening the infected cells after transfection.

Cell viability assay

The cell viability assay was performed using the Cell Counting Kit 8 as described.^8,9 Briefly, cells were seeded into a 96-well plate. After incubation, 10 μL of Cell Counting Kit 8 reagent was added to each well, and the plates were incubated for 4 hours at 37°C with 5% CO₂. The absorbance was measured at 450 nm using a plate reader.

Iron assay

Following previous research,^8,9 we utilized the Iron Assay Kit to determine the concentration of iron. Briefly, cells were seeded onto a 96-well plate. After a 5-minute incubation period in iron assay buffer, the insoluble impurities were eliminated via centrifugation at 16,000g for 10 minutes. The collected supernatants were incubated with assay buffer at 37°C for 30 minutes and then with 100 μL of iron probe for 1 hour. The concentration of iron was determined utilizing the standard calibration curve approach by measuring the absorbance at 593 nm.

Lipid peroxidation assay

To evaluate lipid peroxidation, we used the Lipid Peroxidation Assay Kit to determine the levels of intracellular malondialdehyde (MDA) as described.^8–10 Briefly, cells were seeded onto a 96-well plate. After a 5-minute incubation period in MDA Assay Buffer on ice, insoluble contaminants were removed by centrifugation at 13,000g for 10 minutes. The collected supernatants were treated with TBA solution at 95°C for 1 hour, followed by cooling to room temperature. Subsequently, the absorbance was measured at 532 nm, and the levels of intracellular MDA were detected using the standard calibration curve method.

Detection of lipid ROS

To detect levels of lipid reactive oxygen species (ROS), we used the peroxide-sensitive fluorescent probe C11-BODIPY, as reported.^8–10 Briefly, cells were seeded onto a 6-well plate. After treatment, cells were incubated with 50 μmol/L C11-BODIPY at 37°C for 1 hour. Subsequently, HSC cells were collected and washed with PBS containing 1% bovine serum albumin before detecting lipid ROS levels using a flow cytometer.

Glutathione assay

In accordance with previous studies,^8,9 the concentration of glutathione (GSH) was examined by Glutathione Analysis Kit. Briefly, cells were seeded in a 96-well plate. After treatment, cells were lysed in 50 mM 4-Morpholineethanesulfonic acid buffer on ice for 5 minutes. Insoluble contaminants were removed via centrifugation at 10,000g for 15 minutes. The collected supernatants were treated with GSH detection working solution for 25 minutes at room temperature. The absorbance was measured at 412 nm, and the concentration of GSH was calculated using the standard calibration curve method.

Detection of SLC25A37 activity

SLC25A37 activity was measured as the rate of mitochondrial iron uptake according to our reported protocol.¹⁹ Briefly, 1 μCi of ⁵⁵FelC3 was incubated with 2 μg of iron-free enterobactin at room temperature for 3 hours. Purified mitochondria from cells were added to the samples and incubated for 4 hours at room temperature. The amount of ⁵⁵Fe in lysed mitochondria was determined by liquid scintillation.

Western blot analysis and immunoprecipitation assay

Western blot analysis and immunoprecipitation (IP) assay were conducted according to established methods.^8–10 For western blot, cells were washed with PBS and lysed in mammalian lysis buffer. Protein concentrations were measured using the Pierce Bicinchoninic Acid Protein Assay Kit. Protein extracts were separated by SDS-PAGE and transferred onto polyvinylidene fluoride membranes. The membranes were probed with primary and secondary antibodies, and protein bands were visualized using a chemiluminescence system. For IP, cells were lysed with IP assay buffer, and the total proteins were collected by centrifugation at 12,000g at 4°C. Protein G agarose beads were incubated with cell lysates to clear the same amount of proteins at 4°C for 3 hours and incubated with designated antibodies. The proteins bound to the agarose beads were collected and separated by SDS-PAGE.

RNA isolation and real-time PCR

The RNA isolation and real-time PCR were performed as described.^8,9 Briefly, cells were seeded in a 96-well plate. Total RNA was isolated using Trizol Reagent and reverse transcribed into cDNA using SuperScript III Platinum One-Step quantitative real time-PCR Kit. Real-time PCR was carried out using nucleic acid gel stain green dye on the StepOne sequence detection system. The primer sequences can be provided upon request. The relative abundance of genes was calculated by using the 2^−△△CT method.

Transmission electron microscopy

Transmission electron microscopy was performed following previously described methods.^8–10 Briefly, cells were seeded onto 4-well Chambered Coverglass at a density of 2 × 10⁴ cells/mL. The cells were fixed with 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M phosphate buffer at 4°C for 2 hours. Postfixation was carried out with 1% osmium tetroxide in the same buffer for 1 hour at room temperature, followed by dehydration in a graded series of ethanol and embedding in Epon resin. Ultrathin sections were cut with an ultramicrotome and examined using an Olympus EM208S transmission electron microscope for image acquisition.

Immunofluorescence analysis

Immunofluorescence analysis was performed as described.^8–10 Briefly, cells were seeded onto coverslips in a 24-well plate. Cells were fixed with 4% paraformaldehyde for 15 minutes at room temperature, permeabilized with 0.1% Triton X-100 in PBS for 10 minutes, and blocked with 1% bovine serum albumin in PBS for 30 minutes. Cells were incubated with primary antibodies at 4°C overnight, followed by incubation with secondary antibodies conjugated with fluorescent dyes for 1 hour at room temperature. 2-(4-Aminophenyl)-6-indoleformamide was used to stain the nucleus. Images were captured using a fluorescence microscope.

Statistical analysis

The statistical analysis was performed using the SPSS program and included the Student t test for 2-group comparisons or one-way ANOVA, followed by the Student-Newman-Keuls test for more than 2 groups. The data were expressed as mean±SEM and statistical significance was indicated at p<0.05.

RESULTS

The cell fate regulator DACH1 levels are increased during HSC ferroptosis

We first investigated whether several classic ferroptosis stimulators can cause ferroptosis in HSCs. Interestingly, we found that the inhibition of cell viability induced by erastin, RSL3, and FIN56 in rat and human HSC lines and primary mouse HSC cells was completely prevented by liproxstatin-1 (a potent ferroptosis inhibitor) but not by caspase inhibitor (a potent apoptosis inhibitor) or necrostatin-1 (a potent necroptosis inhibitor) (Figure 1A) (Supplemental Figure S1A, http://links.lww.com/HC9/A818). Furthermore, trypan blue staining and Prussian blue staining demonstrated that erastin, RSL3, and FIN56 induced HSC cell death via ferroptosis (Supplemental Figure S2A, B, http://links.lww.com/HC9/A819). It is widely known that redox-active iron buildup, lipid peroxidation, and lipid ROS formation all occur simultaneously and mutually amplify during ferroptosis.²⁰ As expected, treatment with erastin, RSL3, and FIN56 led to the overload of redox-active iron (Figure 1B) (Supplemental Figure S1B, http://links.lww.com/HC9/A818), the production of the lipid peroxidation MDA (Figure 1C), and the generation of lipid ROS (Figure 1D) in HSC-T6, HSC-LX2, and primary mouse HSC cells. Notably, ferrostatin-1, but not caspase inhibitor or necrostatin-1, reversed the classical ferroptotic events (Figure 1B–D) (Supplemental Figure S1B, http://links.lww.com/HC9/A818). These results confirm that erastin, RSL3, and FIN56 could induce ferroptosis in vitro.

FIGURE 1.

DACH1 protein levels are increased during HSC ferroptosis. HSC cells were treated with erastin (10 μM), RSL3 (2.5 μM), and FIN56 (10 μM), along with indicated inhibitors liproxstatin-1 (10 μM), ZVAD-FMK (10 μM), and necrostatin-1 (10 μM) for 24 hours. (A) Cell viability was detected using CCK8 kit (n=3, *p<0.05). (B–D) The levels of iron, MDA, and lipid ROS were measured using commercial kits (n=3, ^* p<0.05). (E, F) HSC cells were treated with erastin (10 μM), RSL3 (2.5 μM), and FIN56 (10 μM) for 24 hours. The expression of DACH1 mRNA and protein was determined by real-time PCR and western blot, respectively (n=3). (G) HSC cells were treated with erastin (10 μM) with MG-132 (5 μM) or CHX (20 μg/mL) for 24 hours. The expression of DACH1 protein was detected by western blot (n=3, *p<0.05, **p<0.01, #p<0.05). HSC-T6 cells were exposed to CHX (20 μg/mL) with or without erastin (10 μM) for 12 hours. (H) The expression of DACH1 protein was detected by western blot at 0, 4, 8, and 12 hours (n=3, *p<0.05, **p<0.01, ***p<0.001). Abbreviations: CCK8, Cell Counting Kit 8; CHX, cycloheximide; DACH1, dachshund homolog1; FIN56, ferroptosis inducer; MDA, malondialdehyde; ROS, reactive oxygen species; RSL3, ferroptosis inducer; Z-VAD-FMK, caspase inhibitor.

Several studies showed that DACH1 plays a significant role in the regulation of cell fate.^14,15 Could DACH1 be involved in the regulation of ferroptosis? To address this possibility, we first evaluated DACH1 mRNA and protein expression during HSC ferroptosis. Interestingly, the induction of HSC ferroptosis markedly elevated DACH1 protein expression, but not mRNA expression (Figure 1E, F) (Supplemental Figure S1C, http://links.lww.com/HC9/A818). This result implies that the rise in DACH1 protein may occur in a transcription-independent manner. To reinforce this hypothesis, we employed a targeted 26S proteasomal inhibitor, MG-132, to impede DACH1 protein degradation. Additionally, a chemical protein synthesis inhibitor, cycloheximide, was utilized to suppress DACH1 protein synthesis. Consistently, MG-132 treatment increased DACH1 expression, whereas cycloheximide treatment abolished the increase in DACH1 expression during HSC ferroptosis (Figure 1G). Besides, we explored the stability of DACH1 protein and observed that HSC ferroptosis greatly prolonged the half-life of the DACH1 protein (Figure 1H). A number of studies have found that sineoculis homeobox homolog1 can regulate DACH1 expression levels.¹⁴ As expected, the expression of sineoculis homeobox homolog1 was decreased in erastin-treated HSCs (Supplemental Figure S3A, http://links.lww.com/HC9/A820). sineoculis homeobox homolog1 overexpression can decrease the expression of DACH1 (Supplemental Figure S3B, http://links.lww.com/HC9/A820) and impaired DACH1-mediated ferroptotic events (Supplemental Figure S3C, D, http://links.lww.com/HC9/A820). Overall, these findings suggest that the cell fate regulator DACH1 levels were increased during HSC ferroptosis.

KO of DACH1 confers resistance to HSC ferroptosis

To investigate the role of DACH1 in HSC ferroptosis, DACH1 KO and knockin HSC lines were generated using a modified inducible CRISPR/Cas9 system and a specific DACH1 plasmid (Figure 2A). The results showed that DACH1 knockin heightened the sensitivity of HSCs to ferroptotic death, whereas DACH1 KO conferred resistance to ferroptosis in HSCs (Figure 2B). Moreover, DACH1 knockin facilitated the typical ferroptotic events such as iron accumulation (Figure 2C) (Supplemental Figure S1D, http://links.lww.com/HC9/A818), lipid peroxidation (Figure 2D), lipid ROS production (Figure 2E), and GSH depletion (Figure 2F), whereas DACH1 KO impaired these events in response to erastin or RSL3 treatment (Figure 2C–F) (Supplemental Figure S1D, http://links.lww.com/HC9/A818). Ferroptotic cells are commonly characterized by mitochondrial crumpling and a reduction in ridges.²⁰ Transmission electron microscope images revealed that erastin-treated cells exhibited crinkled mitochondria and the absence of mitochondrial cristae structures (Figure 2G). However, DACH1 KO prevented the morphological alterations of mitochondria in ferroptotic cells (Figure 2G). Importantly, ferroptosis inhibitors, deferoxamine, and eugenol reinstated erastin-induced growth suppression in DACH1 knockin HSC lines, while the growth inhibition remained unaffected by necroptosis inhibitors necrosulfonamide and necrostatin-1, or the apoptosis inhibitor caspase inhibitor (Figure 2H). These data suggest that the elevation of DACH1 protein may facilitate HSC ferroptosis but not necroptosis and apoptosis.

FIGURE 2.

Upregulation of DACH1 protein levels facilitates HSC ferroptosis. (A) DACH1 knockout and knockin HSC lines were created using the modified inducible CRISPR/Cas9 system and the DACH1 plasmid, respectively. Transfection efficiency was verified by real-time PCR (n=3, ^*** p<0.001). DACH1 knockout and knockin HSC lines were treated with erastin (10 μM) or RSL3 (2.5 μM) for 24 hours. (B) Cell viability was detected by CCK8 kit (n=3, *p<0.05). (C–F) The levels of iron, MDA, lipid ROS, and GSH were detected by commercial kits (n=3, *p<0.05). DACH1 WT and KO HSC-T6 cells were treated with erastin (10 μM) for 24 hours. (G) The ultrastructure of mitochondria was observed by transmission electron microscope. Scale bars are 500 nm (n=3). DACH1 WT and knockin HSC-T6 cells were exposed to erastin (10 μM) with the indicated inhibitors deferoxamine (100 µM), eugenol (2 μM), ZVAD-FMK (10 μM), necrostatin-1 (10 μM), necrosulfonamide (5 μM) for 24 hours. (H) Cell viability was determined by CCK8 kit (n=3, ***p<0.001). Abbreviations: CCK8, Cell Counting Kit 8; DACH1, dachshund homolog1; GSH, glutathione; KO, knockout; MDA, malondialdehyde; pcDNA, eukaryotic expression vector; ROS, reactive oxygen species; WT, wildtype; ZVAD-FMK, caspase inhibitor.

P53 mitochondrial translocation is required for DACH1 to promote HSC ferroptosis

It should be emphasized that the biological function of DACH1 is tightly correlated with the subcellular localization of p53.¹⁴ Does elevated DACH1 expression facilitate HSC ferroptosis in a p53-dependent way? To assess this hypothesis, we defined the subcellular localization of p53 during HSC ferroptosis. Interestingly, DACH1 KO decreased the protein levels of p53 in mitochondria but not in the nucleus and cytoplasm, whereas DACH1 knockin enhanced the protein expression of p53 in mitochondria rather than in the nucleus and cytoplasm (Figure 3A) (Supplemental Figure S1E, http://links.lww.com/HC9/A818). Moreover, confocal imaging confirmed that DACH1 knockin-induced colocalization of p53 and mitochondria (Figure 3B). The phosphorylation of p53 at serine 392 is widely established to be a key step in its mitochondrial translocation.²¹ As expected, DACH1 KO reduced the serine 392 phosphorylation of mitochondrial p53 protein, but DACH1 knockin increased the phosphorylation of mitochondrial p53 at serine 392 (Figure 3C). To further confirm the role of serine 392 phosphorylation, the S392A mutant, a nonphosphorylatable alanine, was substituted for serine 392.²¹ Notably, treatment with the S392A mutant decreased the mitochondrial translocation of p53 (Figure 3D) and inhibited cell viability (Supplemental Figure S4A, http://links.lww.com/HC9/A821) during HSC ferroptosis. Furthermore, ferroptotic events such as iron accumulation, lipid peroxidation, lipid ROS production, and GSH depletion were reduced by S392A mutant (Figure 3E–G) and p53 knockdown (Supplemental Figures S1F, http://links.lww.com/HC9/A818 and S4B–E, http://links.lww.com/HC9/A821). Importantly, we also investigated the interaction between p53 and DACH1. IP assay revealed that erastin, RSL3, and FIN56 promoted the combination of p53 and DACH1 (Figure 3H). The p53 protein is composed of 5 functional domains, including the C-terminal basic domain, tetramerization domain, DNA binding domain, pro-rich domain, and N-terminal transactivation domain (TAD).²² To identify the essential p53 protein domain for the biological functions of DACH1, myc-tagged plasmids were constructed, including full-length p53, TAD, pro-rich domain, DNA binding domain, tetramerization domain, and basic domain. These plasmids were cotransfected with DACH1 into HSC cells, followed by IP and western blot analyses using specific antibodies to detect the myc-tag and DACH1, respectively. The results revealed that DACH1 can bind to p53 through its TAD domain but not the pro-rich domain, DNA binding domain, tetramerization domain, or basic domain domains (Figure 3I). Overall, these findings confirm that the mitochondrial translocation of p53 is necessary for DACH1 to promote HSC ferroptosis.

FIGURE 3.

The mitochondrial translocation of p53 is essential for DACH1 to promote HSC ferroptosis. DACH1 WT and KO HSC-T6 cells were treated with erastin (10 μM) for 24 hours. (A) The protein expression of p53 in mitochondria, cytoplasm, and nucleus was detected by western blot (n=3, ^*** p<0.001). DACH1 wild type and knockin HSC-T6 cells were treated with erastin (10 μM) for 24 h. (B) The colocalization of mitochondria and p53 was examined by laser confocal. Scale bars are 100 μm (n=3). HSC-T6 cells with DACH1 knockout and knockin were treated with erastin (10 μM) for 24 hours. (C) The protein expression of phosphorylated p53 in mitochondria was determined by western blot (n=3, ***p<0.001). Serine 392 was substituted by a nonphosphorylatable alanine (S392A mutant) in HSC-T6 cells with DACH1 knockin. The indicated cells were then treated with erastin (10 μM) for 24 hours. (D) The protein expression of mitochondrial p53 was determined by western blot (n=3, ^*** p<0.001). (E–G) The levels of iron, lipid ROS, and MDA were detected by commercial kits (n=3, **p<0.01, ***p<0.001). HSC-T6 cells were treated with erastin (10 μM) for 24 hours. (H) The binding of DACH1 to p53 was determined by immunocoprecipitation (n=3). (I) The full-length and different domains of p53 with myc tag and pcDNA3.1-DACH1 were transfected into HSC-T6 cells. The protein expression of p53 was determined by western blot (n=3). Abbreviations: BD, basicdomain; DACH1, dachshund homolog1; DBD, DNA binding domain; IB, Immunoblotting; IP, immunoprecipitation; KO, knockout; LMNB1, lamin B1; MDA, malondialdehyde; N.S., not significant; pcDNA, eukaryotic expression vector; PRD, pro-richdomain; ROS, reactive oxygen species; TAD, trans-activationdomain; TD, tetramerizationdomain; VDAC, mitochondrial loading control; WT, wildtype.

Mitochondrial p53 interacts with SLC25A37 to disrupts iron homeostasis in DACH1-enhanced HSC ferroptosis

It is widely believed that mitochondria are the primary hub for iron utilization and storage.²³ Whether mitochondrial translocation of p53 disrupts iron homeostasis? To verify this possibility, the effect of S392A mutant and p53 knockdown on iron metabolism was examined using the mitochondria isolated from ferroptotic HSC cells. Interestingly, we found that DACH1 knockin increased the levels of mitochondrial iron, whereas S392A mutant and p53 knockdown impaired DACH1-enhanced mitochondrial iron buildup (Figure 4A). To further explore the possible targets of mitochondrial p53, we carefully carried out an unbiased screen, including mitochondrial iron usage genes, iron stockpile genes, iron transfer genes, and iron uptake genes. Importantly, we have identified a potential target, SLC25A37, among iron uptake genes, which was suppressed by S392A mutant and p53 knockdown (Figure 4B) (Supplemental Figure S1G, http://links.lww.com/HC9/A818). Additionally, IP analysis clearly indicated that DACH1 knockin facilitated the binding of mitochondrial p53 and SLC25A37, but not ABCB10 and HSP90 in HSC ferroptosis (Figure 4C, D). Confocal imaging further confirmed that DACH1 knockin promoted the binding of mitochondrial p53 to SLC25A37 in erastin-treated HSC cells (Figure 4E). Does mitochondrial p53 enhance the protein stability and activity of SLC25A37? We found that p53 knockin extended the half-life and increased the activity of the SLC25A37 protein (Figure 4F, G). Furthermore, SLC25A37 knockdown eliminated mitochondrial iron buildup and the typical ferroptotic events induced by DACH1 knockin (Figure 4H–J). Collectively, these results showed that mitochondrial p53 may trigger mitochondrial iron overload by increasing SLC25A37 activity in DACH1-enhanced HSC ferroptosis.

FIGURE 4.

Mitochondrial p53 disrupts iron homeostasis by regulating SLC25A37 activity in DACH1-enhanced HSC ferroptosis. HSC-T6 cells with mutant p53 or p53 KD were transfected with DACH1 plasmid. The cells were then treated with erastin (10 μM) for 24 hours. (A) The levels of mitochondrial iron were detected by iron assay kit (n=3, **p<0.01). (B) The mRNA expression of SLC25A37 was detected by real-time PCR (n=3, **p<0.01, ***p<0.001). DACH1 WT and knockin HSC-T6 cells were treated with erastin (10 μM) for 24 hours. (C) The binding of p53 to SLC25A37 was determined by immunocoprecipitation (n=3). (D) The binding of p53 to ABCB10 or HSP90 was examined by immunocoprecipitation (n=3). (E) The colocalization of p53 and SLC25A37 in mitochondria was detected by laser confocal. Scale bars are 100 μm (n=3). HSC-T6 cells were transfected with p53 plasmid. The cells were treated with erastin (10 μM) with CHX (20 μg/mL) for 12 h. (F) The protein levels of SLC25A37 were determined by western blot at 0, 4, 8, and 12 hours (n=3). HSC cells were transfected with p53 plasmid. The cells were then treated with erastin (10 μM) for 24 hours. (G) The activity of SLC25A37 protein was determined as the rate of mitochondrial iron uptake (n=3, *p<0.05). HSC-T6 cells were transfected with DACH1 plasmid and SLC25A37 shRNA. The cells were then treated with erastin (10 μM) for 24 hours. (H–J) The levels of mitochondrial iron, lipid ROS, and MDA were determined by commercial kits (n=3, *p<0.05, **p<0.01, ***p<0.001). Abbreviations: CHX, cycloheximide; DACH1, dachshund homolog 1; IB, Immunoblotting; IP, immunoprecipitation; KD, knockdown; MDA, malondialdehyde; N.S., not significant; pcDNA, eukaryotic expression vector; ROS, reactive oxygen species; shRNA, short hairpin RNA; SLC25A37, solute carrier family 25 member 37; VDAC, mitochondrial loading control; WT, wildtype.

Accumulation of mitochondrial iron interferes with the electron transport chain and triggers lipid peroxidation in DACH1-enhanced HSC ferroptosis

It is generally known that the electron transport chain (ETC) requires mitochondrial iron for enzyme participation (Figure 5A). Does mitochondrial iron buildup affect ETC in DACH1-enhanced HSC ferroptosis? To verify this possibility, we initially assessed the effect of SLC25A37 on mitochondrial biogenesis. Our results revealed that SLC25A37 knockdown and knockin did not alter mitochondrial mass during HSC ferroptosis (Figure 5B). Moreover, we observed no significant changes in the quantity of mitochondrial DNA following either SLC25A37 knockdown or knockin (Figure 5C). We then investigated whether SLC25A37 affected the electron transport capacity of enzyme complexes anchored in mitochondrial inner membranes. As predicted, SLC25A37 knockdown resulted in a reduction in electron transport activities of these complexes, whereas SLC25A37 knockin increased these activities in HSC ferroptosis (Supplemental Figure S5A, http://links.lww.com/HC9/A822). The function of the ETC was suppressed by DACH1 KO (Supplemental Figure S5B, http://links.lww.com/HC9/A822), p53 knockdown (Supplemental Figure S5C, http://links.lww.com/HC9/A822), and S392A mutant (Supplemental Figure S5D, http://links.lww.com/HC9/A822), but the electron transport activities of complexes were enhanced by DACH1 knockin (Supplemental Figure S5B, http://links.lww.com/HC9/A822) and p53 knockin (Supplemental Figure S5C, http://links.lww.com/HC9/A822). Furthermore, we investigated the impact of SLC25A37 knockdown and knockin on the subunits of the mitochondrial complex. Interestingly, our data showed that NADH:ubiquinone oxidoreductase subunit B8 knockin upregulated the expression of NDUFB8 (complex I), succinate dehydrogenase complex (complex II), MTCO1 (complex III), and UQCRC2 (complex IV) in HSC ferroptosis (Figure 5D). Conversely, the expression of these subunits was reduced when DACH1 was knocked out during HSC ferroptosis (Figure 5D). Subsequently, we determined the necessity of ETC in DACH1-enhanced HSC ferroptosis using various ETC inhibitors. We found that inhibitors of mitochondrial complex I (rotenone), complex II (diethyl butylmalonate), complex III (antimycin), and complex IV (NaN3) attenuated DACH1 knockin-induced, p53 knockin-induced, and SLC25A37 knockin-induced ferroptotic cell death and classical ferroptotic events (Figure 5E–G) (Supplemental Figure S1H and I, http://links.lww.com/HC9/A818). Altogether, these data suggested that the accumulation of mitochondrial iron may interfere with the ETC function and trigger lipid peroxidation in DACH1-enhanced HSC ferroptosis.

FIGURE 5.

Accumulation of mitochondrial iron interferes with electron transport chain (ETC) and triggers lipid peroxidation in DACH1-enhanced HSC ferroptosis. (A) Schematic diagram indicated mitochondrial ETC complexes and inhibitors. HSC-T6 cells were transfected with SLC25A37 shRNA or SLC25A37 plasmid. The cells were then treated with erastin (10 μM) for 24 hours. (B) Mitochondria mass was detected by MitoTracker (n=3). (C) mtDNA copy number was determined by real-time PCR (n=3). HSC-T6 cells with DACH1 KO or knockin were treated with erastin (10 μM) for 24 hours. (D) The mRNA expression of mitochondrial complexes was examined by real-time PCR (n=3, **p<0.01, ***p<0.001). HSC-T6 cells were transfected with DACH1 plasmid. The cells were then exposed to erastin (10 μM) with or without ETC inhibitors rotenone (10 mM), DBM (2 mM), antimycin (50 mM), and NaN₃ (15 mM) for 24 hours. (E) The cell viability, the levels of mitochondrial iron, lipid ROS, and MDA were detected by commercial kits (n=3, ***p<0.001). HSC-T6 cells were transfected with p53 plasmid. The cells were then exposed to erastin (10 μM) with or without ETC inhibitors rotenone (10 mM), DBM (2 mM), antimycin (50 mM), and NaN₃ (15 mM) for 24 hours. (F) The cell viability, the levels of mitochondrial iron, lipid ROS, and MDA were detected by commercial kits (n=3, ***p<0.001). HSC-T6 cells were transfected with SLC25A37 plasmid. The cells were then exposed to erastin (10 μM) with or without ETC inhibitors rotenone (10 mM), DBM (2 mM), antimycin (50 mM), and NaN₃ (15 mM) for 24 hours. (G) The cell viability, the levels of mitochondrial iron, lipid ROS, and MDA were detected by commercial kits (n=3, ***p<0.001). Abbreviations: DACH1, dachshund homolog 1; DBM, diethyl butylmalonate; KD, knockdown; KO, knockout; MDA, malondialdehyde; NDUFB8, NADH:ubiquinone oxidoreductase subunit B8; N.S., not significant; pcDNA, eukaryotic expression vector; ROS, reactive oxygen species; SDHB, succinate dehydrogenase complex, subunit B; shRNA, short hairpin RNA; SLC25A37, solute carrier family 25 member 37; WT, wildtype.

The DACH1/P53/SLC25A37 signaling regulates HSC ferroptosis in murine hepatic fibrogenesis

Does DACH1/P53/SLC25A37 signaling regulate HSC ferroptosis in vivo? To validate this potential mechanism, we synthesized vitamin A–coupled liposomes containing DACH1 shRNA (VA-Lip-DACH1-shRNA), p53 shRNA (VA-Lip-P53-shRNA), and SLC25A37 shRNA (VA-Lip-SLC25A37-shRNA) to disrupt DACH1-P53-SLC25A37 signaling in BDL-induced murine hepatic fibrogenesis. We first investigated the effect of DACH1/P53/SLC25A37 signaling on hepatic fibrogenesis pathology. Severe fibrotic liver injury was observed in the liver of the model group mice, but treatment with erastin ameliorated BDL-mediated fibrotic pathological changes (Figure 6A). Interestingly, the therapeutic effect of erastin on hepatic fibrogenesis was impaired by treatment with VA-Lip-DACH1-shRNA, VA-Lip-P53-shRNA, or VA-Lip-SLC25A37-shRNA (Figure 6A). Furthermore, histopathological analysis indicated collagen deposition in the liver of the model group, whereas i.p. injection with erastin substantially reduced fibrotic scar formation (Figure 6B). VA-Lip-DACH1- shRNA, VA-Lip-P53-shRNA, or VA-Lip-SLC25A37-shRNA reversed the impact of erastin on collagen deposition (Figure 6B). Moreover, IHC staining and real-time PCR showed that erastin treatment reduced the levels of hepatic fibrogenesis markers, including alpha-smooth muscle actin, collagen, fibronectin, and desmin, but VA-Lip-DACH1-shRNA, VA-Lip-P53-shRNA, or VA-Lip-SLC25A37-shRNA impaired the inhibitory effect of erastin (Figure 6B, C). Which specific cells undergo ferroptosis following erastin treatment? To address this question, we isolated primary hepatocytes, macrophages, liver sinusoidal endothelial cells (LSECs), and HSCs from fibrotic livers. Notably, the levels of the ferroptosis marker PTGS2 were increased in primary HSCs but not in primary hepatocytes, macrophages, and LSECs after erastin treatment (Supplemental Figure S6A, http://links.lww.com/HC9/A823). Besides, treatment with erastin resulted in iron accumulation (Supplemental Figure S6B, http://links.lww.com/HC9/A823), lipid ROS generation (Supplemental Figure S6C, http://links.lww.com/HC9/A823), and MDA production (Supplemental Figure S6D, http://links.lww.com/HC9/A823) in primary HSCs, but not in primary hepatocytes, macrophages, and LSECs. These findings suggest that erastin may mitigate fibrotic injury by inducing HSC ferroptosis. We further investigated the role of DACH1/P53/SLC25A37 signaling in regulating HSC ferroptosis. Consistent with the in vitro results, treatment with erastin upregulated the protein expression levels of DACH1 (Figure 6D), the combination of DACH1 and p53 (Figure 6E), the mitochondrial translocation of p53 (Figure 6F), the binding of p53 to SLC25A37 (Figure 6G), the accumulation of mitochondrial iron (Figure 6H), the hyperfunction of ETC (Supplemental Figure S7A–D, http://links.lww.com/HC9/A824), and ferroptotic cell death (Figure 6H). Importantly, treatment with VA-Lip-DACH1-shRNA, VA-Lip-P53-shRNA, or VA-Lip-SLC25A37-shRNA attenuated the regulatory effects of DACH1/P53/SLC25A37 signaling on HSC ferroptosis (Figure 6D–H). Collectively, these results confirm that DACH1/P53/SLC25A37 signaling may play a critical role in regulating HSC ferroptosis in hepatic fibrogenesis.

FIGURE 6.

The DACH1/P53/SLC25A37 signaling regulates HSC ferroptosis in murine hepatic fibrosis. (A) The morphological changes of the liver were detected by macroscopic examination. Scale bars are 1 cm (n=6). (B) The pathological changes of the livers were detected by H&E staining, Masson staining, and Sirius red staining. The expression of α-SMA was determined by immunohistochemistry. Scale bars are 50 μm (n=6, ***p<0.001). (C) The expression of ACTA2, COL1A1, fibronectin, and desmin was determined by real-time PCR in fibrotic liver tissues (n=6, ***p<0.001). Primary HSCs were isolated from fibrotic livers. (D) The expression of DACH1 was determined by western blot (n=6). (E) The binding of DACH1 to p53 was determined by immunocoprecipitation (n=6). (F) The expression of mitochondria p53 was determined by western blot (n=6). (G) The binding of p53 to SLC25A37 was determined by immunocoprecipitation (n=6). (H) The levels of mitochondrial iron, lipid ROS, MDA, and cell death were examined by commercial kits (n=6, *p<0.05, **p<0.01, ***p<0.001). Abbreviations: BDL, bile duct ligation; DACH1, dachshund homolog1; H&E, hematoxylin-eosin; IP, immunoprecipitation; MDA, malondialdehyde; N.S.,not significant; ROS, reactive oxygen species; shRNA, short hairpin RNA; SLC25A37, solute carrier family 25 member 37; α-SMA, alpha-smooth muscle actin; VA, vitamin A.

DISCUSSION

Ferroptosis, a recently identified form of programmed cell death marked by the accumulation of iron-dependent lipid peroxides, represents a novel approach to effectively eliminate activated HSCs.^7,24 Li et al⁷ reported that ellagic acid–induced ferroptosis can retard hepatic fibrogenesis by impairing the formation of SNARE complexes. Moreover, Liu et al²⁴ found that wogonoside can attenuate hepatic fibrogenesis by triggering HSC ferroptosis through the SOCS1/P53/SLC7A11 pathway. In line with these findings, the current study showed that ferroptosis inducers can inhibit HSC activation by inducing ferroptosis in vitro. Furthermore, erastin treatment can improve the pathological damage of hepatic fibrogenesis by inducing HSC ferroptosis in vivo. Although studies have shown that ferroptosis may resist the progression of hepatic fibrogenesis, controversial roles of ferroptosis in hepatic fibrogenesis have also been reported. Cho et al²⁵ reported that ferroptosis may contribute to HSC activation and hepatic fibrogenesis. Ferroptosis inducer RSL3 increased the expression of plasminogen activator inhibitor-1, c-JUN phosphorylation, and activator protein 1 luciferase activity but did not alter Smad phosphorylation and Smad-binding element luciferase activity.²⁵ These contradictory results suggest that more studies are necessary to define the relationship between ferroptosis and HSC activation as well as subsequent hepatic fibrogenesis.

DACH1 is a transcription factor that plays a critical role in regulating various cellular processes, including cell proliferation, differentiation, and apoptosis.^26,27 Li et al²⁶ recently found that knockdown of DACH1 inhibits the proliferation, migration and invasion of FaDu cells via Akt/NF-κB/MMP2/9 signaling. Moreover, Aman et al²⁷ reported that DACH1 inhibits breast cancer cell invasion and metastasis by downregulating the transcription of matrix metalloproteinase 9. In the present study, we investigated for the first time the role of DACH1 in the regulation of HSC ferroptosis. On induction of ferroptosis, DACH1 protein expression was evidently increased by inhibiting the ubiquitin-proteasome signaling. CRISPR/Cas9-mediated DACH1 KO conferred resistance to ferroptosis, whereas specific plasmid-mediated DACH1 overexpression contributed to classical ferroptotic events. Although definitive experimental evidence is still necessary to establish the role of DACH1 in ferroptosis regulation, our study presents a novel perspective on DACH1 research.

Numerous studies have reported that p53 plays a crucial role in the regulation of ferroptosis-related genes and molecules.^28–30 Nuclear p53 protein can affect the expression of ferroptosis-related genes, while cytoplasmic p53 protein can regulate the biological function of ferroptosis-related molecules.^31,32 However, the regulation of ferroptosis by mitochondrial p53 is poorly understood. Additionally, Chen and colleagues reported that endogenous DACH1 colocalized with p53 in a nuclear and extranucleolar location and shared occupancy of 15% of p53-bound genes in ChIP Seq. The carboxyl terminus of DACH1 was necessary for direct p53 binding, contributing to the inhibition of colony formation and cell cycle arrest.³³ In the current study, we investigated the role of mitochondrial p53 in DACH1-mediated ferroptosis. Our results showed that upregulated DACH1 promotes p53 mitochondrial translocation by directly binding with the p53 N-terminal TAD. Site-directed mutations of serine 392 impaired the binding of DACH1 to p53, blocked the mitochondrial translocation of p53, and prevented DACH1-enhanced HSC ferroptosis. These findings contribute to the understanding of the molecular mechanisms underlying ferroptosis regulation and suggest that targeting p53 may have therapeutic potential for ferroptosis-related diseases.

Interestingly, 2 isoforms of mitoferrin, SLC25A28 and SLC25A37, have been identified on the inner mitochondrial membrane, which are essential for the enter closes of iron for mitochondrial iron-sulfur and heme clusters.³⁴ Wang Y et al. indicated that SLC25A37 is necessary for mitochondrial glutathione import in mammalian cells.³⁵ In our previous research, we found that the upregulation of bromodomain-containing protein 7 (BRD7) may trigger p53 mitochondrial translocation.¹⁹ Mitochondrial p53 interacted with SLC25A28 to form complex and enhanced the activity of SLC25A28, which could lead to the abnormal accumulation of redox-active iron and ferroptosis.¹⁹ Of note, mitochondrial p53 has the potential to bind with both SLC25A28 and SLC25A37.³⁶ BRD7 upregulation can promote the binding of p53 to SLC25A28 rather than SLC25A37. Which key factor can promote the binding of p53 to SLC25A37 but not SLC25A28 has sparked our research interest. In the current study, we found that DACH1 may have the potential to become a candidate molecule. DACH1 knockin facilitated the combination of mitochondrial p53 and SLC25A37 in HSC ferroptosis. SLC25A37 knockdown fully eliminated mitochondrial iron buildup and the typical ferroptotic events by DACH1 knockin. This study highlights the essential role of SLC25A37 in HSC ferroptosis and suggests that it may serve as a potential therapeutic target for the treatment of iron-related diseases.

In summary, our study found that DACH1 plays a critical role in regulating HSC ferroptosis by modulating the p53/SLC25A37 signaling (Figure 7). By shedding light on the molecular mechanisms underlying ferroptosis in HSCs, our findings provide valuable insights for developing new strategies to prevent hepatic fibrogenesis.

FIGURE 7.

The DACH1/P53/SLC25A37 signaling regulates HSC ferroptosis in hepatic fibrosis. DACH1 upregulation results in p53 mitochondrial translocation, the binding of p53 to SLC25A37, the accumulation of mitochondrial iron, the hyperfunction of electron transport chain, lipid peroxidation, and HSC ferroptosis. Abbreviations: DACH1, dachshund homolog 1; ROS, reactive oxygen species; SLC25A37, solute carrier family 25 member 37.

Supplementary Material

hc9-8-e0396-s001.eps ^{(1.8MB, eps)}

hc9-8-e0396-s002.eps ^{(1.6MB, eps)}

hc9-8-e0396-s003.eps ^{(2.4MB, eps)}

hc9-8-e0396-s004.eps ^{(762.8KB, eps)}

hc9-8-e0396-s005.eps ^{(1.2MB, eps)}

hc9-8-e0396-s006.eps ^{(937.3KB, eps)}

hc9-8-e0396-s007.eps ^{(873.8KB, eps)}