Autophagic clearance of mitochondria in the kidney copes with metabolic acidosis

Abstract

Metabolic acidosis, a common complication of CKD, causes mitochondrial stress by undefined mechanisms. Selective autophagy of impaired mitochondria, called mitophagy, contributes toward maintaining cellular homeostasis in various settings. We hypothesized that mitophagy is involved in proximal tubular cell adaptations to chronic metabolic acidosis. In transgenic mice expressing green fluorescent protein-tagged microtubule-associated protein 1 light chain 3 (GFP-LC3), NH4Cl loading increased the number of GFP puncta exclusively in the proximal tubule. In vitro, culture in acidic medium produced similar results in proximal tubular cell lines stably expressing GFP-LC3 and facilitated the degradation of SQSTM1/p62 in wild-type cells, indicating enhanced autophagic flux. Upon acid loading, proximal tubule-specific autophagy-deficient (Atg5-deficient) mice displayed significantly reduced ammonium production and severe metabolic acidosis compared with wild-type mice. In vitro and in vivo, acid loading caused Atg5-deficient proximal tubular cells to exhibit reduced mitochondrial respiratory chain activity, reduced mitochondrial membrane potential, and fragmented morphology with marked swelling in mitochondria. GFP-LC3-tagged autophagosomes colocalized with ubiquitinated mitochondria in proximal tubular cells cultured in acidic medium, suggesting that metabolic acidosis induces mitophagy. Furthermore, restoration of Atg5-intact nuclei in Atg5-deficient proximal tubular cells increased mitochondrial membrane potential and ammoniagenesis. In conclusion, metabolic acidosis induces autophagy in proximal tubular cells, which is indispensable for maintaining proper mitochondrial functions including ammoniagenesis, and thus for adapted urinary acid excretion. Our results provide a rationale for the beneficial effect of alkali supplementation in CKD, a condition in which autophagy may be reduced, and suggest a new therapeutic option for acidosis by modulating autophagy.

Figure 1.

MA activates autophagic flux exclusively in kidney PTCs. (A) LC3 dot formation increased in the PTCs of GFP-LC3 transgenic mice under MA, caused by NH₄Cl feeding for 30 days (n=3, respectively; left), and in immortalized PTCs stably transfected with GFP-LC3 and cultured in acidic medium (pH 6.5) (n=5, respectively; right). Kidney sections are coimmunostained for megalin, a marker of proximal tubules in red (left) and for DAPI in blue (right). (B) Autophagic flux is estimated by the SQSTM1/p62 degradation assay using PTCs (n=3, respectively). (C and D) Accumulation of SQSTM1/p62-positive (C) and ubiquitin-positive (D) aggregates in proximal tubules of Atg5^F/F:KAP and control mice with or without acid loading (n=5, respectively). The number of SQSTM1/p62-positive or ubiquitin-positive dots is counted in at least 10 high-power fields. All values are given as the mean±SEM. *P<0.05. DAPI, 4',6-diamidino-2-phenylindole; F/F, Atg5^F/F mice; F/F:KAP, Atg5^F/F:KAP mice. Bar, 50 μm in A (left), C, and D; 5 μm in A (right).

Figure 2.

Autophagy deficiency blunts the compensation by PTCs against MA. (A) Arterial blood pH (left) and bicarbonate concentration (right) of control and Atg5^F/F:KAP mice with or without acid loading (n=5, vehicle-treated Atg5^F/F mice; n=6, other groups). (B and C) Ammoniagenesis in kidney sections from control and Atg5^F/F:KAP mice with or without acid loading (n=4, respectively). Kidney sections are obtained as indicated with the dotted line in B. (D) Ammoniagenesis in Atg5-positive and Atg5-negative PTCs cultured under normal and acidic media (n=3, respectively). All values are given as the mean±SEM. *P<0.05. Atg5(+), Atg5-positive PTC; Atg5(−), Atg5-negative PTC; F/F, Atg5^F/F mice; F/F:KAP, Atg5^F/F:KAP mice.

Figure 3.

Autophagy deficiency reduces mitochondrial respiration activity under MA. (A and B) COX staining (left) and SDH staining (right) of kidney sections of control and Atg5^F/F:KAP and control mice with or without MA. (C) Tracing of the OCR of Atg5-positive PTCs cultured under normal pH and acidic conditions (left). Measurements are performed in triplicate under basal status and after addition of 1 μM of oligomycin (point a), 0.5 μM of FCCP (point b), and 0.1 μM of rotenone and antimycin A (point c) (n=18, pH 7.4; n=20, pH 6.5). The ATP-linked OCR represents the difference of the OCR between the basal level (point a) and after treatment with 1 μM of oligomycin (point b). The maximum OCR represents the value after treatment with 0.5 μM of FCCP (point c) (n=18, Atg5(+) pH 7.4; n=20, other groups; right) (also see Supplemental Figure 4). The former indicates the OCR for generating ATP, and the latter indicates the latent OCR independent of the mitochondrial membrane potential, which is another type of mitochondrial function. (D) The ATP-linked OCR in Atg5-positive and Atg5-negative PTCs with or without acid loading (n=18, Atg5(+) pH 7.4; n=20, other groups). All values are given as the mean±SEM. *P<0.05. F/F, Atg5^F/F mice; F/F:KAP, Atg5^F/F:KAP mice; Atg5(+), Atg5-positive PTC; Atg5(−), Atg5-negative PTC; FCCP, carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone. Bar, 500 μm in A and B.

Figure 4.

Autophagy deficiency distorts the mitochondrial morphology of the PTCs under MA. Electron microscopic images of kidney sections of control and Atg5^F/F:KAP mice with or without MA are presented. F/F, Atg5^F/F mice; F/F:KAP, Atg5^F/F:KAP mice. Bar, 5 μm in top row; 500 nm in bottom row.

Figure 5.

Autophagy deficiency reduces the mitochondrial membrane potential in the PTCs under MA. The mitochondrial membrane potential of Atg5-positive and Atg5-negative PTCs cultured under normal and acidic media is assessed by TRME (n=5, respectively). All values are given as the mean±SEM. *P<0.05. Atg5(+), Atg5-positive PTC; Atg5(−), Atg5-negative PTC. Bar, 50 μm.

Figure 6.

Acid loading induces mitophagy in the PTCs. (A) PTCs isolated from wild-type mice stably expressing GFP-LC3 are stained with MitoTracker Red FM after treatment with 200 nM of bafilomycin A1 2 hours before harvesting (n=5). Yellow dots in merged images represent colocalization of mitochondria and autophagosomes. (B) Colocalization is assessed by Pearson’s correlation. (C) PTCs isolated from wild-type mice stably expressing GFP-LC3 and mStrawberry-ubiquitin are stained with MitoTracker Deep Red FM after treatment with 200 nM of bafilomycin A1 2 hours before harvesting (n=5). White dots in merged images represent colocalization of ubiquitinated mitochondria and autophagosomes. All values are given as the mean±SEM. *P<0.05. MTR Red FM, MitoTracker Red FM; MTR Deep Red FM, MitoTracker Deep Red FM. Bar, 5 μm.

Figure 7.

Restoring autophagic activity regains ammoniagenesis via quality control of mitochondria in acidic conditions in vitro. (A) Schematic illustration of the generation of a transmitochondrial cybrid. (B and C) Mitochondrial membrane potential as assessed by TMRE staining (B) and ammoniagenesis (C) of the rescue cybrid and control cybrid cultured under normal or acidic media (n=3, respectively). The cybrid generated by fusion of the mitochondrial donor cells derived from Atg5-negative PTCs and the nuclear donor cells derived from Atg5-positive PTCs is defined as a rescue cybrid, whereas the cybrid generated by fusion of the mitochondrial donor and the nuclear donor, both of which are derived from Atg5-negative PTCs, is defined as a control cybrid. All values are given as the mean±SEM. *P<0.05. mtDNA, mutant mitochondrial DNA.

Figure 8.

The proposed mechanism of autophagic renovation of mitochondria in PTCs under acidic conditions. MA-induced damaged mitochondria are eliminated adequately in the autophagy-competent condition (left), whereas damaged mitochondria accumulates in cytoplasm in the autophagy-deficient condition (right), which leads to decompensation for MA and triggers a vicious cycle.