MondoA and AKI and AKI-to-CKD Transition

Abstract

Key Points:

1. The expression of MondoA was decreased in the renal tubules of patients with CKD.

2. Genetic ablation of MondoA in proximal tubules inhibited autophagy and increased vulnerability to AKI through increased expression of Rubicon.

3. MondoA ablation during the recovery phase after ischemia-reperfusion aggravated kidney injury through downregulation of the transcription factor EB-peroxisome proliferator-activated receptor-γ coactivator-1α axis.

Background: Elderly individuals and patients with CKD are at a higher risk of AKI. The transcription factor MondoA is downregulated in the kidneys of aged individuals or patients with AKI; however, its roles in AKI development and the AKI-to-CKD transition remain unknown.

Methods: We investigated the expression of MondoA in human kidney biopsy samples, ischemia-reperfusion–injured (IRI) mouse kidneys, and cultured proximal tubular epithelial cells under hypoxia/reoxygenation. The role of MondoA during the initial and recovery phases after IRI was evaluated using proximal tubule–specific MondoA knockout mice and MondoA-deficient proximal tubular epithelial cells. Furthermore, we explored the involvement of Rubicon and transcription factor EB (TFEB), both of which are downstream factors of MondoA.

Results: MONDOA expression was decreased in the renal tubules of patients with CKD. In mouse kidneys, MondoA expression was decreased under ischemia, whereas its expression was increased during reperfusion. Genetic ablation of MondoA in proximal tubular epithelial cells inhibited autophagy and increased vulnerability to AKI through increased expression of Rubicon. Ablation of Rubicon in MondoA-deficient IRI kidneys activated autophagy and protected mitochondrial function. MondoA ablation during the recovery phase after ischemia-reperfusion aggravated kidney injury through downregulation of the TFEB-peroxisome proliferator-activated receptor-γ coactivator-1α axis. Pharmacological upregulation of TFEB contributed to maintaining mitochondrial biogenesis and increased peroxisome proliferator-activated receptor-γ coactivator-1α transcription.

Conclusions: Our findings demonstrate that MondoA protected against vulnerability to AKI by maintaining autophagy and subsequently supporting mitochondrial function to prevent progression to CKD.

Keywords: AKI; CKD; fibrosis; hypoxia; ischemia-reperfusion; kidney biopsy; kidney tubule; mitochondria; molecular biology; transcription factors.

Graphical abstract

Figure 1

Nuclear MONDOA is decreased in proximal tubular epithelial cells of CKD patients. Correlation analysis between eGFR, age, nuclear MONDOA, and other downstream factors, using human kidney biopsy samples. (A) Correlation of eGFR with age (left) and the percentage of proximal tubular epithelial cells with nuclear MONDOA (right) (n=23). (B) Representative images of immunohistochemical staining for MONDOA, RUBCN, and TFEB in kidney biopsy samples of patients with CKD, each with the indicated age, sex, and eGFR (ml/min per 1.73 m²). Specimens were counterstained with hematoxylin. Magnified images are shown in the lower panels. Bars: 50 μm. Correlation of the percentage of proximal tubular epithelial cells showing nuclear MONDOA translocation with the percentages of proximal tubular epithelial cells with each of the following: RUBCN in cytoplasm (left), nuclear TFEB (middle), and nuclear PGC1α (right). (A and B) Relationships were examined using Pearson correlation and the corresponding P values. PGC1α, peroxisome proliferator-activated receptor-γ coactivator-1α; RUBCN, run domain beclin-1–interacting and cysteine-rich domain-containing protein; TFEB, transcription factor EB.

Figure 2

Transcriptional activity of MondoA decreases under ischemic conditions but increases during reoxygenation in proximal tubular epithelial cells. (A and B) scRNA-seq transcriptomic analysis was performed by using data of mice subjected to bilateral IRI in a previous study. (A) Dot plot showing the scaled expression of Mlxip/MondoA in the proximal tubule segments at different time points after bilateral long (30 minutes) IRI. PT S1 and PT S3 represent the S1 and S3 segments of the proximal tubule, respectively. The diameter of the dots corresponds to the percentage of cells expressing Mlxip/MondoA, and the density of the dots corresponds to the average expression for all proximal tubular epithelial cells in the dataset. (B) Differential gene expression analysis between long (30 minutes) IRI versus short (23 minutes) IRI in the S1 and S3 segments of the proximal tubule across time course. (C and D) The dynamics of MondoA transcriptional activity in cultured proximal tubular epithelial cells after H/R. The protein level of TXNIP (C) and the mRNA levels of Mlxip, Rubicon, Txnip, and Arrdc4 (D) in cultured proximal tubular epithelial cells subjected to normoxia, hypoxia, or H/R for the indicated periods (n=3). (C) A representative western blot image. (D) mRNA levels normalized by those of cells cultured under normoxia for 0 hour. (E) Representative images of immunohistochemical staining for MONDOA in kidney biopsy samples of patients with CKD and allograft biopsy samples of donors, each with the indicated age, sex, and eGFR (ml/min per 1.73 m²). Specimens were counterstained with hematoxylin. Magnified images are shown in the lower panels. Bars: 50 μm. Data are provided as means±SEM. Statistically significant differences: *P < 0.05 versus kidneys at 0 hour or proximal tubular epithelial cells cultured under normoxia for 0 hour (D, one-way ANOVA followed by Dunnett test). Arrdc4, arrestin domain containing 4; H/R, hypoxia/reoxygenation; IRI, ischemia-reperfusion injury; MGA, minor glomerular abnormalities; MLXIP, MLX interacting protein; PT, proximal tubule; scRNA-seq, single-cell RNA sequencing; TXNIP, thioredoxin-interacting protein.

Figure 3

Loss of MondoA increases Rubicon and impairs autophagic flux in proximal tubular epithelial cells during IR. The effect of MondoA deficiency on autophagic flux and cell injury during IR or H/R was investigated using MondoA-deficient mice and cultured proximal tubular epithelial cells. (A and B) Wild-type and MondoA-deficient proximal tubular epithelial cells were subjected to H/R. (A) Western blot images of TXNIP, Rubicon, and SQSTM1/p62. Arrow indicates Rubicon-specific band. (B) Western blot images of MAP1LC3B in proximal tubular epithelial cells incubated with or without bafilomycin A1 are presented. The autophagic flux index was assessed (n=3–7 per group). (C–E) MondoA^F/F and MondoA^F/F; KAP mice were subjected to a sham operation or unilateral IRI. Images are shown for PAS staining (C), mRNA levels of Havcr1/Kim-1 and Lcn2/Ngal (D), and western blot analysis of Rubicon and SQSTM1/p62 (E) in the kidney cortical regions of IRI mice 2 days after unilateral IRI. (C) The tubular injury score is shown (n=4–6 per group) and (D) n=7–10 per group. (E) Arrow indicates SQSTM1/p62-specific band. (F) Autophagic flux was assessed by counting the number of GFP-positive dots in the proximal tubules of GFP-MAP1LC3 transgenic MondoA^F/F or MondoA^F/F; KAP mice with or without CQ administration; analysis was performed either after sham-operation or 12 hours after IRI (n=3–4 in the sham-operated group and 3–5 in the IRI group). The number of GFP-positive dots per proximal tubule under each condition was counted in at least ten high-power fields (original magnification, ×600), with each high-power field containing 10–15 proximal tubules. (G) Images of electron micrographs of the kidneys of IRI mice 2 days after unilateral IRI. Arrows indicate autophagosomes. (A, B, and E) ACTB was used as loading control. (A, B, C, F, and G) Representative images are presented. Magnified images are shown in the insets. Bars: 100 µm (C), 10 µm (F), and 5 µm (G). Data are provided as means±SEM. Statistically significant differences: *P < 0.05 versus treatment matched wild-type proximal tubular epithelial cells or MondoA^F/F control littermates; ^#P < 0.05 versus sham-operated kidney in each group; **P < 0.05 versus mice with no CQ treatment (B–D, one-way ANOVA followed by Tukey–Kramer test; F, Student t test). Control, normoxia-treated controls; MondoA (+), wild-type proximal tubular epithelial cells; MondoA (−), MondoA-deficient proximal tubular epithelial cells; MondoA F/F, MondoA^F/F mice; MondoA F/F; KAP, MondoA^F/F; KAP mice. ACTB, actin beta; CQ, chloroquine; DAPI, 4′,6-diamidino-2-phenylindole; GFP, green fluorescent protein; IR, ischemia-reperfusion; KAP, kidney androgen-regulated protein-Cre; PAS, Periodic acid–Schiff; SQSTM1, sequestosome 1.

Figure 4

Loss of MondoA exacerbates H/R or IR-induced mitochondrial dysfunction in proximal tubular epithelial cells. The mitochondrial membrane potential (A) and the ROS production (B) of wild-type or MondoA-deficient proximal tubular epithelial cells subjected to H/R were assessed by TMRE and MitoSOX Red staining, respectively. Quantitative data of relative intensity are shown. More than 150 cells were quantified under each condition. (C) Representative images of SDH staining in the kidney cortical regions of MondoA^F/F control and MondoA^F/F; KAP mice 2 days after sham operation or unilateral IRI (n=4–6 in each group). The relative staining intensity is shown. (A and C) The mean values for wild-type proximal tubular epithelial cells without H/R and for the sham-operated kidney cortical regions of MondoA^F/F mice are expressed as 1. Magnified images are shown in the insets. Bars: 10 µm (A and B) and 200 µm (C). Data are provided as means±SEM. Statistically significant differences: *P < 0.05 versus treatment-matched wild-type proximal tubular epithelial cells or MondoA^F/F control littermates; ^#P < 0.05 versus normoxia-treated proximal tubular epithelial cells or sham-operated kidneys in each group (A–C, one-way ANOVA followed by Tukey–Kramer test). Control, normoxia-treated controls; MondoA (+), wild-type proximal tubular epithelial cells; MondoA (−), MondoA-deficient proximal tubular epithelial cells; MondoA F/F, MondoA ^F/F mice; MondoA F/F; KAP, MondoA ^F/F; KAP mice. ROS, reactive oxygen species; SDH, succinate dehydrogenase; TMRE, tetramethylrhodamine ethyl ester.

Figure 5

Loss of Rubicon in MondoA-deficient mice alleviates IRI by restoring autophagic activity. The effect of genetic ablation of Rubicon on IRI was investigated in MondoA-deficient kidneys. (A) Wild-type or Rubicon-deficient proximal tubular epithelial cells were knocked down for MondoA by shRNA, followed by H/R. Western blot images of MAP1LC3B in proximal tubular epithelial cells incubated with or without bafilomycin A1 are presented. The autophagic flux index was assessed (n=4 per group). (B–E) MondoA^F/F, MondoA^F/F; KAP, and MondoA^F/F;Rubicon^F/F; KAP (DKO) mice were subjected to unilateral IRI. (B) Representative western blot images of Rubicon and SQSTM1/p62. Arrow indicates SQSTM1/p62-specific band. The relative amount of Rubicon protein was determined by densitometric analysis. Representative images of PAS (C) and SDH (D) staining and electron micrographs (E) of kidney cortical regions 2 days after unilateral IRI (n=7–11 per group). The tubular injury score (C) and the relative staining intensity (D) are shown. (E) Arrows indicate autophagosomes. Magnified images are shown in the insets. Bars: 100 µm (C), 200 µm (D), 5 µm (E, left), and 1 µm (E, right). Data are provided as means±SEM. (B and D) The mean value for IRI kidney cortical regions of MondoA^F/F mice is expressed as 1. Statistically significant differences: *P < 0.05 versus MondoA-deficient and Rubicon-expressing proximal tubular epithelial cells or MondoA^F/F; KAP mice (A, one-way ANOVA followed by Tukey–Kramer test; B–D, Student t test). Rubicon (−), Rubicon-deficient proximal tubular epithelial cells; MondoA F/F, MondoA^F/F mice; MondoA F/F; KAP, MondoA^F/F; KAP mice; MondoA F/F; Rubicon F/F; KAP, MondoA^F/F; Rubicon^F/F; KAP (DKO) mice. DKO, double knockout; shRNA, short hairpin RNA.

Figure 6

MondoA reactivation counteracts the AKI-to-CKD transition via the TFEB-PGC1α axis. To genetically inhibit MondoA reactivation, MondoA^F/F and MondoA^F/F; NDRG1 mice were subjected to a sham operation or unilateral IRI, followed by intraperitoneal tamoxifen injection 1 and 4 days after the operation. (A) Experimental protocol. Kidneys were harvested and assessed at the indicated time points after IRI. (B) The mRNA levels of Mlxip/MondoA in kidney cortical regions are shown at the indicated time points after IRI (n=3–4 [sham, 2 and 5 days], and 7–9 [1 and 3 weeks]). (C–G) Representative images of SDH staining (C), Masson trichrome staining (D), and immunostaining for COL1A1 (E), TFEB (red) (F), and PGC1α (G) in kidney cortical regions either 1 (F and G) or 3 (C–E) weeks after the operation (n=8–9 in each group (C–F), n=5–6 in each group (G). (C) The relative intensity of SDH was assessed in at least five high-power fields (×400). The mean value for the sham-operated kidney of MondoA^F/F mice is expressed as 1. (E) The COL1A1-positive area was assessed as a percentage in at least five high-power fields (×200). (F) Kidney sections were counterstained with DAPI (blue). The percentage of proximal tubular epithelial cells with TFEB nuclear translocation was counted in at least ten high-power fields (×600). (G) Sections were counterstained with hematoxylin. Magnified images are shown in the insets. Bars: 200 µm (C), 100 µm (D, E, and G), and 40 µm (F). Data are provided as means±SEM. Statistically significant differences: *P < 0.05 versus treatment-matched MondoA^F/F control littermates; ^#P < 0.05 versus sham-operated kidney in each group (B, Student t test; C–G, one-way ANOVA followed by Tukey–Kramer test). MondoA F/F, MondoA^F/F mice; MondoA F/F; NDRG1, MondoA^F/F; NDRG1 mice. NDRG1, N-myc downstream-regulated gene 1.

Figure 7

Pharmacological upregulation of the TFEB-PGC1α axis mitigates the AKI-to-CKD transition in MondoA-deficient kidneys. (A and B) Wild-type and MondoA-deficient proximal tubular epithelial cells were cultured with or without the indicated concentrations of trehalose. Representative images of western blot analysis (A) and immunofluorescence staining (B) of TFEB. (C–G) MondoA^F/F and MondoA^F/F; NDRG1 mice were subjected to a sham operation or unilateral IRI, followed by injections of tamoxifen 1 and 4 days after the operation and trehalose three times a week. Representative images of immunostaining for TFEB (red) (C) and PGC1α (D), Masson trichrome staining (E), immunohistochemical staining for COL1A1 (F), and SDH staining (G) in kidney cortical regions were obtained 1 (C and D) or 3 (E–G) weeks after the operation with or without trehalose treatment (n=4–6 per group). (C) Kidney sections were counterstained with DAPI (blue). The percentage of proximal tubular epithelial cells with TFEB nuclear translocation was counted in at least ten high-power fields (×600). (D and F) Sections were counterstained with hematoxylin. (F) The COL1A1-positive area was assessed as a percentage in at least five high-power fields (×200). (G) The relative intensity of SDH was assessed in at least five high-power fields (×400). The mean value for the IRI kidney cortical regions of MondoA^F/F mice is expressed as 1. Magnified images are shown in the insets. Bars: 10 µm (B), 40 µm (C), 100 µm (D–F), and 200 µm (G). Data are provided as means±SEM. Statistically significant differences: ^#P < 0.05 versus sham-operated kidney in each group; ^##P < 0.05 versus IRI kidney in each group (C, F, and G, one-way ANOVA followed by Tukey–Kramer test). MondoA (+), wild-type proximal tubular epithelial cells; MondoA (−), MondoA-deficient proximal tubular epithelial cells; control, without trehalose; Tre, trehalose; MondoA F/F, MondoA^F/F mice; MondoA F/F; NDRG1, MondoA^F/F; NDRG1 mice. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 8

MondoA ameliorates AKI by increasing autophagic flux and prevents transition to CKD by promoting mitochondrial biogenesis. Schematic illustration of this study. The transcriptional activity of MondoA is suppressed in CKD patients and under hypoxic conditions. It is enhanced after reoxygenation: (1) MondoA activity induces autophagy because of decreased Rubicon and (2) MondoA reactivation counteracts the AKI-to-CKD transition through mitochondrial biogenesis via the TFEB-PGC1α axis. In elderly individuals or patients with CKD, the excessive expression of Rubicon because of decreased MondoA levels suppresses autophagy and exacerbates AKI on CKD after the acute phase of kidney injury. On the other hand, the reactivation of MondoA during the recovery phase of kidney injury leads the TFEB-PGC1α axis to sustain mitochondrial biogenesis. The use of a TFEB activator to target the MondoA-TFEB-PGC1α axis in elderly individuals or patients with CKD may hold promise in attenuating the progression to CKD during the recovery phase of kidney injury.