Inhibition of the mitochondrial permeability transition by protein kinase A in rat liver mitochondria and hepatocytes

Abstract

NO and cGMP administered at reperfusion after ischaemia prevent injury to hepatocytes mediated by the MPT (mitochondrial permeability transition). To characterize further the mechanism of protection, the ability of hepatic cytosol in combination with cyclic nucleotides to delay onset of the calcium-induced MPT was evaluated in isolated rat liver mitochondria. Liver cytosol plus cGMP or cAMP dose-dependently inhibited the MPT, required ATP hydrolysis for inhibition and did not inhibit mitochondrial calcium uptake. Specific peptide inhibitors for PKA (protein kinase A), but not PKG (protein kinase G), abolished cytosol-induced inhibition of MPT onset. Activity assays showed a cGMP- and cAMP-stimulated protein kinase activity in liver cytosol that was completely inhibited by PKI, a PKA peptide inhibitor. Size-exclusion chromatography of liver cytosol produced a single peak of cGMP/cAMP-stimulated kinase activity with an estimated protein size of 180-220 kDa. This fraction was PKI-sensitive and delayed onset of the MPT. Incubation of active catalytic PKA subunit directly with mitochondria in the absence of cytosol and cyclic nucleotide also delayed MPT onset, and incubation with purified outer membranes led to phosphorylation of a major 31 kDa band. After ischaemia, administration at reperfusion of membrane-permeant cAMPs and cAMP-mobilizing glucagon prevented reperfusion injury to hepatocytes. In conclusion, PKA in liver cytosol activated by cGMP or cAMP acts directly on mitochondria to delay onset of the MPT and protect hepatocytes from cell death after ischaemia/reperfusion.

Figure 1. Inhibition of the MPT by liver cytosol

Mitochondria were incubated in swelling buffer in the presence of ATP (20 μM), 8-pCPT-cGMP (100 μM) and 0 (traces a and b), 25 μg/ml (trace c), 50 μg/ml (trace d) or 100 μg/ml (trace e) liver cytosol, as described in the Materials and methods section. In traces (b–e), 250 μM Ca²⁺ was added where indicated by the arrow. The absorbance at 540 nm was measured. Data are representative of at least three separate experiments.

Figure 2. Inhibition of the MPT requires hydrolysable ATP

Mitochondria were incubated in swelling buffer plus 8-pCPT-cGMP (100 μM), as described for Figure 1, and the absorbance was measured after the addition of 250 μM Ca²⁺ (indicated by arrows). In (A), additions to the incubation medium were: none (trace a), cytosol (60 μg/ml) (trace b), 10 μM ATP and cytosol (60 μg/ml) (trace c), 20 μM ATP (trace d) and 20 μM ATP and cytosol (60 μg/ml) (trace e). In (B), additions were: none (trace a), ATP (25 μM) (trace b), liver cytosol (60 μg/ml) (trace c), ATP (25 μM) and cytosol (60 μg/ml) (trace d), cytosol (60 μg/ml) and p[NH]ppA (25 μM) (trace e), and cytosol (60 μg/ml) and p[NH]ppA (500 μM) (trace f).

Figure 3. Lack of inhibition of mitochondrial Ca²⁺ uptake by liver cytosol

Mitochondria were incubated in the presence of Fluo5N (1 μM) and TMRM (1 μM) in swelling buffer, as described in the Materials and methods section, before addition of 250 μM Ca²⁺ (indicated arrows). Green fluorescence of Fluo5N (A), red fluorescence of TMRM (B) and absorbance at 540 nm (C) were then measured.

Figure 4. cGMP-stimulated PKA-like activities in cytosol and fractionated cytosol

(A) Isolated mitochondria were incubated in swelling buffer with 8-pCPT-cGMP (100 μM) and ATP (25 μM), and the absorbance was measured, as described in Figure 1. Other additions to the incubation buffer were: none (traces a and b), cytosol (60 μg/ml) (trace c), cytosol plus PKI (10 μM) (trace d), and cytosol plus DT-3 (1 μM) (trace e). In traces (b–e), 250 μM Ca²⁺ was added where indicated by the arrow. (B) Liver cytosol was loaded on to an S200 column and fractionated by FPLC, as described in the Materials and methods section. Shown are kinase activities from 10 μl of each fraction measured in the presence of 5 μM sp-cAMP (dotted line), 5 μM sp-cGMP (solid line) or no activator (dashed line). (C) Phosphorylation of Kemptide substrate (30 μM) was measured in the presence of 6.4 μg of the pooled fractions 14–17 in the presence of 0–10 μM cGMP and 500 μM IBMX (isobutylmethylxanthine), as described in the Materials and methods section. (D) Mitochondria were incubated with 100 μM 8-pCPT-cGMP and 30 μM MgATP, as described in (A), and the absorbance was measured. Additions to the incubation buffer were none (trace a), concentrated fractions 14–17 (20 μg/ml) (trace b), concentrated fractions plus 4 μM PKI (trace c) and concentrated fraction plus 1 μM DT-3 (trace d). In traces (b–d), 250 μM Ca²⁺ was added where indicated by the arrow.

Figure 5. Direct effect of the catalytic subunit of PKA on the MPT and phosphorylation of mitochondrial proteins

(A) Mitochondrial swelling was measured by absorbance during incubation in swelling buffer, as described in the Materials and methods section. Additions were none (trace a), 250 μM Ca²⁺ (arrow) plus 25 μM ATP (trace b), 250 μM Ca²⁺ plus 17 units/μl PKA (trace d), 250 μM Ca²⁺, 25 μM ATP and 8 units/μl PKA (trace e), and 250 μM Ca²⁺, 25 μM ATP and 17 units/μl PKA (trace f).(B) Mitochondrialswelling was measured with additions of none (trace a),250 μMCa²⁺ (arrow) plus 25 μM ATP (trace b), and 250 μM Ca²⁺, 25 μM ATP and 20 units/μl PKG (trace c). (C) Autoradiograms of phosphorylated proteins after incubation of 30 μM [γ-³²P]-labelled ATP were prepared with intact mitochondria (lane 1), mitochondria plus 17 units/μl catalytic PKA (lane 2) and PKA alone (lane 3), as described in the Materials and methods section. (D) Autoradiograms were prepared after incubation of mitochondrial outer membranes with [γ-³²P]-labelled ATP plus no further addition (lane 1), 17 units/μl catalytic PKA (lane 2) and PKA and 4 μM PKI (lane 3). In (C and D), the molecular mass in kDa is indicated on the right-hand side.

Figure 6. Protection by cAMP analogues against cell death after simulated I/R injury

Hepatocytes were incubated in anaerobic KRH buffer at pH 6.2 for 4 h to simulate ischaemia. (A) Hepatocytes were reoxygenated with aerobic KRH at pH 6.2 (●) or at pH 7.4 in the presence of 0 (○), 10 μM (▴), 100 μM (∎) and 200 μM (▾) 8-pCPT-cAMP, as described in the Materials and methods section. Results are means ± S.E.M. (B) Hepatocytes were reoxygenated with aerobic KRH at pH 6.2 (●) and at pH 7.4 in the presence of 0 (○), 10 μM (▴), 100 μM (∎), and 200 μM (▾) sp-cAMP. Values are from four or more hepatocyte isolations for each group.

Figure 7. Mitochondrial inner membrane permeabilization and depolarization after I/R of hepatocytes: protection by glucagon

Overnight cultured rat hepatocytes were subjected to 4 h of anoxia at pH 6.2 to simulate ischaemia followed by reoxygenation at pH 7.4 to simulate reperfusion, as described in the Materials and methods section. Hepatocytes were ester-loaded with green-fluorescing calcein into the cytosol to monitor inner membrane permeability and reperfused with red-fluorescing TMRM and propidium iodide to monitor mitochondrial polarization and loss of cell viability respectively. In (B), glucagon (10 nM) was added 15 min prior to and then continuously after reperfusion. In (A), note that mitochondrial voids in the green calcein fluorescence filled after reperfusion, which was followed by loss of nearly all cytoplasmic calcein fluorescence and nuclear labelling with propidium iodide after 40–45 min (arrows). Mitochondria did not take up TMRM. In panels marked ‘3x’, red fluorescence intensity was upwardly rescaled by a factor of 3 to illustrate better nuclear labelling with propidium iodide and the absence of mitochondrial TMRM uptake. In (B) in the presence glucagon, mitochondrial voids of calcein fluorescence were not lost, and mitochondria rapidly accumulated TMRM. Nuclear labelling with propidium iodide and loss of cytoplasmic calcein fluorescence did not occur even after 120 min of reperfusion. Each experiment is representative of three or more replicates. PI, propridium iodide.

Figure 8. cAMP and cGMP levels in NO-treated hepatocytes following simulated I/R injury

Hepatocytes were incubated in anaerobic KRH buffer at pH 6.2 for 4 h to simulate ischaemia. At reperfusion the hepatocytes were treated with DETA NONOate, and cAMP (A) and cGMP (B) were measured, as described in the Materials and Methods. Values represent the average of three separate cultures of 1×10⁶ hepatocytes.