Legumain promotes tubular ferroptosis by facilitating chaperone-mediated autophagy of GPX4 in AKI

Abstract

Legumain is required for maintenance of normal kidney homeostasis. However, its role in acute kidney injury (AKI) is still unclear. Here, we induced AKI by bilateral ischemia-reperfusion injury (IRI) of renal arteries or folic acid in lgmn^WT and lgmn^KO mice. We assessed serum creatinine, blood urea nitrogen, histological indexes of tubular injury, and expression of KIM-1 and NGAL. Inflammatory infiltration was evaluated by immunohistological staining of CD3 and F4/80, and expression of TNF-α, CCL-2, IL-33, and IL-1α. Ferroptosis was evaluated by Acsl4, Cox-2, reactive oxygen species (ROS) indexes H₂DCFDA and DHE, MDA and glutathione peroxidase 4 (GPX4). We induced ferroptosis by hypoxia or erastin in primary mouse renal tubular epithelial cells (mRTECs). Cellular survival, Acsl4, Cox-2, LDH release, ROS, and MDA levels were measured. We analyzed the degradation of GPX4 through inhibition of proteasomes or autophagy. Lysosomal GPX4 was assessed to determine GPX4 degradation pathway. Immunoprecipitation (IP) was used to determine the interactions between legumain, GPX4, HSC70, and HSP90. For tentative treatment, RR-11a was administrated intraperitoneally to a mouse model of IRI-induced AKI. Our results showed that legumain deficiency attenuated acute tubular injury, inflammation, and ferroptosis in either IRI or folic acid-induced AKI model. Ferroptosis induced by hypoxia or erastin was dampened in lgmn^KO mRTECs compared with lgmn^WT control. Deficiency of legumain prevented chaperone-mediated autophagy of GPX4. Results of IP suggested interactions between legumain, HSC70, HSP90, and GPX4. Administration of RR-11a ameliorated ferroptosis and renal injury in the AKI model. Together, our data indicate that legumain promotes chaperone-mediated autophagy of GPX4 therefore facilitates tubular ferroptosis in AKI.

Fig. 1. Legumain deficiency attenuates acute tubular injury in a mouse model of IRI.

Lgmn^WT and lgmn^KO mice were randomized to 40 min of bilateral ischemia (n = 8). Kidney and serum samples were collected before and at day 1, 2, and 7 after IRI. A, B Renal functions assessed by serum creatinine (Cr) and BUN levels at the corresponding time points after IRI. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 versus lgmn^WT control; ^♯P < 0.05, ^♯♯P < 0.01 versus lgmn^KO control; ^§P < 0.05, ^§§P < 0.01 versus lgmn^WT after IRI at the corresponding time point. C, D Quantification of kidney KIM-1 and NGAL mRNA levels by qPCR. E Each pair of images includes a representative PAS-stained histologic photomicrograph of the kidney (top) from lgmn^WT and lgmn^KO mice on the indicated days after IRI and higher-magnification photomicrographs of the renal cortex (bottom; corresponding to the boxed area). Scale bar, 50 μm. F Quantification of histological renal damage. Tubular detachment, tubular dilation, and brush border damage were scored by the percentage area. G Histological ATN score. All data are expressed as the mean ± SD, ^*P < 0.05, ^***P < 0.001; two-way ANOVA.

Fig. 2. Legumain deficiency decreases inflammatory reactions in the mouse model of IRI.

Lgmn^WT and lgmn^KO mice were randomized to 40 minutes of bilateral ischemia (n = 8). Collection of kidney samples was performed as described in Fig. 1. A Representative images (left panel) and quantification (right panel) of CD3-positive cells in AKI kidneys of lgmn^WT and lgmn^KO mice. Scale bar, 50 μm. B Representative confocal images (left panel) and quantification (right panel) of kidney sections stained for F4/80 (red) and with DAPI (blue). Scale bar, 50 μm. C–F qPCR analysis of mRNA expression of inflammatory genes TNF-α, CCL-2, IL-33, and IL-1α in kidneys of lgmn^WT and lgmn^KO mice. G Protein levels of inflammatory factor TNF-α, CCL-2, IL-33 in kidney samples assessed by western blotting. All data are expressed as the mean ± SD. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001; two-way ANOVA.

Fig. 3. Legumain deficiency inhibits IRI-induced tubular ferroptosis.

Collection of kidney samples was performed as described in Fig. 1 (n = 8). A & B Quantification of kidney Acsl4 and Cox-2 mRNA levels by qPCR. C Level of intracellular ROS in kidneys measured by the fluorescent probe carboxy-H₂DCFDA. D Representative images of the fluorescence assay (left panel) and quantification (right panel) of ROS in kidneys using the fluorescent probe DHE (red), DAPI (blue) was used to stain the nuclei. Scale bar, 50 μm. E Levels of lipid peroxidation in kidneys were measured by MDA assay. F Quantification of kidney GPX4 mRNA levels by qPCR. G Protein levels of GPX4 in kidney samples assessed by western blotting. Statistical results are shown in the right panel. All data are expressed as the mean ± SD. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001; two-way ANOVA.

Fig. 4. Downregulation of legumain suppresses ferroptosis in tubular cells.

Primary mouse tubular epithelial cells were isolated from lgmn^WT and lgmn^KO mice. mRTECs were treated with 0.5, 1, and 2 μM erastin for 24 hours. A Cell survival was measured by CCK-8 assays. Data were collected from three independent experiments. B Quantification of Acsl4 and Cox-2 mRNA levels by qPCR. C Lipid peroxidation levels were measured by MDA assay. mRTECs were incubated under normoxic or hypoxic conditions for 6, 12, and 24 h. D Cell survival was measured by CCK-8 assays. E Cell lysates and supernatants were collected at the indicated time points for LDH release assays. F mRTECs were incubated under normoxic or hypoxic conditions for 6, 12, and 24 hours, with or without pretreatment of 2 hours of Fer-1 (1 μM), DFO (10 μM) or Nec1 (5 μM). Cell survival was determined 24 h thereafter using CCK-8 assays. G Quantification of Acsl4 and Cox-2 mRNA levels by qPCR. H Intracellular ROS levels were measured using the fluorescent probe carboxy-H₂DCFDA. I Lipid peroxidation levels were measured by MDA assay. Data were collected from three independent experiments. All data are expressed as the mean ± SD. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001; two-way ANOVA.

Fig. 5. Legumain promotes chaperone-mediated autophagy of GPX4.

Primary lgmn^WT and lgmn^KO mRTECs were treated with 1 μM erastin for 24 hours or incubated under normoxic or hypoxic conditions for 6, 12, and 24 hours. A Western blot assay of GPX4. B Quantification of GPX4 mRNA levels by qPCR. Primary lgmn^WT and lgmn^KO mRTECs were pretreated with MG132, CQ, and Baf A1 for 2 hours before treatment with 1 μM erastin for 24 h. C Western blot using an antibody against GPX4. D mRTECs were treated with 1 μM erastin for 24 h or incubated under hypoxic conditions for 24 h. Lysosomes and lysosome-free cell lysates were collected and the protein level of GPX4 was measured by western blotting. Lamp1 and β-actin were used as loading controls of lysosomal and non-lysosomal proteins, respectively. E, F Primary mRTECs from lgmn^WT and lgmn^KO mice were infected with control or legumain-overexpressing lentiviral constructs for 48 h and then treated with 1 μM erastin or incubated under hypoxic conditions. Western blotting was performed using an antibody against GPX4. G In cultured HK-2 cells, a co-IP assay was performed using an antibody against legumain or control IgG and western blotting for legumain, GPX4, Lamp-2a, HSC70, and HSP90 was performed. H HK-2 cells were incubated under normoxic or hypoxic conditions for 24 h. Co-IP assays were performed using an antibody against legumain or control IgG and then western blotting for legumain, GPX4, Lamp-2a, HSC70, and HSP90 was performed.

Fig. 6. Legumain deficiency attenuates tubular injury and ferroptosis in folic acid-induced AKI.

Folic acid was administrated intraperitoneally to lgmn^WT and lgmn^KO mice, and samples were collected at 24 and 48 hours after injection (n = 8) A Renal functions assessed by serum creatinine and BUN levels. B Quantification of kidney KIM-1 and NGAL mRNA levels by qPCR. C Representative images of PAS-stained kidney sections. Scale bar, 50 μm. D Quantification of histological renal damage. Tubular dilation and brush border damage are scored, and protein casts are presented as percentage areas. E Quantification of kidney Acsl4 and Cox-2 mRNA levels by qPCR. F Lipid peroxidation levels in kidneys were measured by MDA assay. G Protein levels of GPX4 in kidney samples assessed by western blotting. Right panel shows the statistical data. H Primary renal tubular epithelial cells were isolated from lgmn^WT and lgmn^KO mice after injection of folic acid and the protein level of GPX4 was quantified by western blotting. Left panels show representative images and the right panel shows the statistical data. All data are expressed as the mean ± SD. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001; two-way ANOVA.

Fig. 7. Legumain inhibitor RR-11a attenuates ferroptosis and tubular injury induced by IRI.

Primary lgmn^WT mRTECs were pretreated with RR-11a (100 nM) for 2 h before incubating under hypoxic conditions for 24 h. A Cell survival was measured by CCK-8 assays. B Western blot assay of GPX4. C Lipid peroxidation levels were measured by MDA assay. Lgmn^WT mice were i.p. injected with RR-11a or the vehicle control at 40 min before bilateral ischemia (n = 8). Kidneys and serum were collected before and at day 1, 2, and 7 after IRI. D, E Renal functions assessed by serum creatinine and BUN levels. ^**P < 0.01, ^***P < 0.001 versus vehicle control group before IRI; ^♯♯P < 0.01, ^♯♯♯P < 0.001 versus RR-11a group before IRI; ^§P < 0.05, ^§§P < 0.01, ^§§§P < 0.001 versus vehicle control group after IRI at the corresponding time point. F, G Quantification of kidney KIM-1 and NGAL mRNA levels by qPCR. H Each pair of images includes a representative PAS-stained histologic photomicrograph of the kidney (top) and higher-magnification photomicrographs of the renal cortex (bottom; corresponding to the boxed area). Scale bar, 50 μm. I Quantification of acute tubular necrosis. J Quantification of kidney Acsl4 and Cox-2 mRNA levels by qPCR. K Western blot assay using the antibody against GPX4. Right panel shows the statistical data. Data were collected from three independent experiments. All data are expressed as the mean ± SD. ^**P < 0.01; two-way ANOVA. All data are expressed as the mean ± SD. ^*P < 0.05^{, **}P < 0.01, ^***P < 0.001; two-way ANOVA.