Liver mitochondrial membrane crosslinking and destruction in a rat model of Wilson disease

Abstract

Wilson disease (WD) is a rare hereditary condition that is caused by a genetic defect in the copper-transporting ATPase ATP7B that results in hepatic copper accumulation and lethal liver failure. The present study focuses on the structural mitochondrial alterations that precede clinical symptoms in the livers of rats lacking Atp7b, an animal model for WD. Liver mitochondria from these Atp7b–/– rats contained enlarged cristae and widened intermembrane spaces, which coincided with a massive mitochondrial accumulation of copper. These changes, however, preceded detectable deficits in oxidative phosphorylation and biochemical signs of oxidative damage, suggesting that the ultrastructural modifications were not the result of oxidative stress imposed by copper- dependent Fenton chemistry. In a cell-free system containing a reducing dithiol agent, isolated mitochondria exposed to copper underwent modifications that were closely related to those observed in vivo. In this cell-free system, copper induced thiol modifications of three abundant mitochondrial membrane proteins, and this correlated with reversible intramitochondrial membrane crosslinking, which was also observed in liver mitochondria from Atp7b–/– rats. In vivo, copper-chelating agents reversed mitochondrial accumulation of copper, as well as signs of intra-mitochondrial membrane crosslinking, thereby preserving the functional and structural integrity of mitochondria. Together, these findings suggest that the mitochondrion constitutes a pivotal target of copper in WD.

Figure 1. Liver copper accumulation in Atp7b^–/– rats is accompanied by organ failure and animal death.

(A) In Atp7b^–/– animals, hepatic copper progressively accumulates, reaching a plateau around 70 days of age, in contrast to control rats (i.e., Atp7b^+/– and Atp7b^+/+ rats). Boxes display the middle 50% of the data, restricted by the 25th and 75th percentiles. The line within the box represents the median. Maximum and minimum values are represented by the whiskers. ^†P < 0.001 versus respective control age group; *P < 0.05, **P < 0.01, ***P < 0.001 versus the youngest Atp7b^–/– age group. Liver copper was normalized to liver wet weight (w.w.). (B) Liver damage, as assessed by an increase in AST activity in serum, occurs abruptly in Atp7b^–/– rats around 90 days of age (n = numbers of animals in the different age groups). ***P < 0.001. (C) Untreated Atp7b^–/– rats (strain LPP) die between 90 and 120 days of age, with a median survival of 106 days.

Figure 2. Structural alterations coincide with increasing copper load in Atp7b^–/– mitochondria.

Electron micrographs of liver mitochondria isolated from (A) control rat, displaying normal mitochondria; (B) Atp7b^–/– rat (age 50 days) with altered morphology characterized by enlarged cristae and slightly altered organelle shapes; (C) Atp7b^–/– rat (age 70 days); (D) Atp7b^–/– rat (age 90 days) with severely condensed mitochondria characterized by greatly enlarged intermembrane spaces and shape alterations. Scale bars: 0.5 μm. (E) Structural alterations in isolated Atp7b^–/– mitochondria increase with age relative to control mitochondria. Quantification at each time point was based on electron micrographs showing around 200 mitochondria. (F) Copper steadily increases in isolated Atp7b^–/– mitochondria, whereas zinc levels remain relatively constant. Note the logarithmic scale, demonstrating an approximately 10-fold increase in copper values in 91- to 100-day-old Atp7b^–/– mitochondria and approximately 200-fold increase in the pellet fraction in comparison to control mitochondria. Mitochondria were isolated from control animals at age 73–509 days (n = 11); from Atp7b^–/– animals at age 31–50 days (n = 2), 60–73 days (n = 13), 85–89 days (n = 6), 91–100 days (n = 17); and from the pellet fractions of clinically apparent Atp7b^–/– animals at age 88–163 days (n = 11). *P < 0.05, ***P < 0.001 versus the preceding age group; ^†P < 0.001 versus control; t test.

Figure 3. Clinically apparent Atp7b^–/– livers contain severely damaged mitochondria.

Electron micrographs of pellet fractions isolated from Atp7b^–/– livers as WD becomes clinically apparent. This pellet fraction contained mitochondria (A) with greatly enlarged intermembrane spaces and (B) mitochondria with matrix remnants gathering at the membrane (arrows), which were frequently associated with massive electron-dense deposits (C). Scale bars: 0.5 μm.

Figure 4. Oxidative damage in Atp7b^–/– mitochondria occurs at a late disease stage.

(A) No early mitochondrial oxidative damage occurs at the level of the redox-sensitive mitochondrial aconitase in Atp7b^–/– animals in comparison to control animals (the higher activity in the age group 60–61 days may be due to the younger animal age). Mitochondria from clinically apparent Atp7b^–/– animals showed markedly lower activities. Mitochondria were isolated from controls (86–114 days, 4 measurements) and Atp7b^–/– rats (60–61 days, 3 measurements; 81–89 days, 5 measurements; clinically apparent, 106–113 days, 2 measurements; D-PA–treated, 121–122 days, 4 measurements). (B) Unsaturated mitochondrial fatty acids are depleted only at a late stage of WD progression in Atp7b^–/– rats. The major abundant fatty acids, accounting for 94% of the typical mitochondrial fatty acid composition (values from control mitochondria given in parentheses) are displayed. Fatty acids were isolated from mitochondria from Atp7b^–/– animals at the ages of 60–61 days (n = 3), 72–73 days (n = 3), and 85–86 days (n = 3) and from clinically apparent Atp7b^–/– animals at age 86–100 days (n = 8). Fatty acid amounts are given relative to values obtained from control mitochondria (n = 3) set to 100%. (C) Enzymatic activities in mitochondria isolated from controls (85–86 days, n = 6) were highly similar to activities in mitochondria from Atp7b^–/– animals (88–89 days, n = 5) and only slightly altered in mitochondria from clinically apparent animals (92–99 days, n = 5). A significant decrease in activities occurs in mitochondria from the pellet fraction of clinically apparent animals (88–103 days, n = 3).*P < 0.05, **P < 0.01, ***P < 0.001.

Figure 5. Simulation of pathological mitochondrial alterations in a cell-free system.

(A) Copper and calcium induce mitochondrial swelling as assessed by optical density measurements of mitochondrial suspensions at 540 nm. In contrast to calcium, copper-triggered swelling could be inhibited by reduced GSH and dithiols such as DTT and DTE. Each data curve represents the average of 4 individual measurements for copper and duplicates for calcium. (B) In the presence of DTT, repetitive copper dosing induced a staircase-like increase in light refraction, indicative of mitochondrial contraction. This increase could be reversed by repetitive equimolar doses of the copper chelator BCDS. Measurements of mitochondrial suspensions were performed at 540 nm in a fluorescence spectrometer. For clarity, fluorescence peaks due to additions made during the measurement were removed from the displayed graph. (C–G) Electron micrographs of control mitochondria treated with copper in the presence of DTT. Note that copper was added repeatedly to the same mitochondrial suspension and samples withdrawn after the additions: (C) untreated starting control, (D) 100 μM copper/1 mM DTT, (E) 200 μM copper/1 mM DTT, (F) 300 μM copper/1 mM DTT, (G) 400 μM copper/1 mM DTT. Mitochondria with progressively enlarged cristae and intermembrane spaces were found. At high copper doses matrix remnants gathered at the mitochondrial membranes (arrows). Scale bars: 0.5 μm.

Figure 6. Increasing copper/DTT ratios impair mitochondrial function and membrane proteins.

(A) Increasing copper/DTT ratios compromise Δψ_m, as assessed by fluorescence quenching of Rh123, and induce a complete immediate Δψ_m loss at ratios close to parity (Cu/DTT 10:10). DTT alone had no effect on Δψ_m (Cu/DTT 0:10). Different amounts of copper were added to reach the indicated increasing copper/DTT ratios. Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP) served as an internal control for complete Δψ_m dissipation. Each data curve represents the average of 4 individual measurements. (B) Mitochondrial ATP synthesis capacity is impaired at an increasing copper/DTT ratio and completely lost at copper/DTT ratios close to parity. Data represent the average of 5 independent experiments. (C) Upon consecutively increasing copper/DTT ratios, liver mitochondria gather at the border of the higher-density solution. Further copper addition leads to mitochondrial accumulation in this solution and the pellet fraction. This pellet fraction is also observed in clinically apparent Atp7b^–/– animals, but not in control rats. (D) The mitochondrial membrane proteins VDAC, ANT, and ATP synthase B react with copper/DTT. These proteins were found to be altered in thiol status in 3 independent replicates. Mitochondria were either incubated with copper/DTT or left untreated as controls. Mitochondrial membrane proteins were enriched by carbonate extraction, and thiol residues were stained with the fluorophore BODIPY FL. Protein extracts were separated by BAC/SDS-PAGE, scanned for determination of the protein-bound fluorescence, and finally silver stained for protein amount comparisons. Proteins were identified by mass spectrometry.*P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7. Intramitochondrial membrane crosslinking can be detected in Atp7b^–/– mitochondria.

(A) Mitochondria from Atp7b^–/– animals and controls were subjected to hypo- and hyperosmotic conditions with subsequent shearing, resulting in the formation of mitoplasts. Control mitochondria but not their counterparts from Atp7b^–/– livers lost parts of their outer membrane, as demonstrated by a marked depletion of the outer membrane VDAC. The inner membrane markers ANT and ATP5B served as loading controls. (B) Electron micrograph of mitoplasts generated from Atp7b^–/– mitochondria as described in A. Outer membrane relics were attached to these structures (arrows; scale bar: 1 μm). (C) ZE-FFE analysis of mitochondria from control and Atp7b^–/– animals subjected to Ca²⁺-induced MPT. Control mitochondria treated with calcium presented two major peaks, M3 and M4, deflecting strongly toward the anode in comparison to untreated mitochondria. In contrast, calcium-treated Atp7b^–/– mitochondria presented a much less pronounced anodal shift in ZE-FFE, with M1 mitochondria as major fraction. (D) Immunoblot analysis of the separated mitochondrial subpopulations of C. M3 and M4 mitochondria had lost large amounts of their outer membrane, as indicated by depletion of VDAC and losses of cytochrome c (CYT C). M1 mitochondria did contain almost equal amounts of VDAC as compared with untreated mitochondria yet demonstrated a partial depletion of cytochrome c, indicative of outer membrane rupture. The major fraction of Ca²⁺-treated control mitochondria was M3 and that of Atp7b^–/– mitochondria was M1 (marked in bold).

Figure 8. Copper chelation therapies restore mitochondrial structure and function in Atp7b^–/– rats.

(A) The copper chelators D-PA and methanobactin decrease the copper content in liver compartments of Atp7b^–/– animals (n = 4 and n = 3, respectively), preferentially in mitochondria. Copper amounts in liver subfractions are expressed relative to values from untreated clinically apparent Atp7b^–/– animals (n = 4) set to hundred percent. Electron micrographs of liver mitochondria isolated from (B) a D-PA–treated and (C) a methanobactin-treated healthy Atp7b^–/– animal (122 days and 120 days old, respectively; treatment started at day 85) and (D) an Atp7b^–/– rat with initial liver damage (129 days old, D-PA treatment started at day 92). Whereas structural integrity was observed in B and C, slight enlargements of the intermembrane spaces were observed in D, indicating incomplete mitochondrial recovery (arrows). Scale bars: 0.5 μm. (E) D-PA and methanobactin treatments restore mitochondrial functional activity, as assessed by the succinate-driven respiratory control ratio (RCR_S) in Atp7b^–/– animals. A significantly higher RCR_S was detected in mitochondria from Atp7b^–/– animals that positively responded to these treatments versus clinically apparent Atp7b^–/– rats.*P < 0.05, **P < 0.01, ***P < 0.001.