Evidence for Physical Association of Mitochondrial Fatty Acid Oxidation and Oxidative Phosphorylation Complexes

Preparation of Mitochondria from Rat Liver

Freshly isolated rat liver was immediately homogenized in a tissue blender at high speed for 20 s at 4 °C in a buffer containing 25 mm Tris-HCl, pH 7.5, 100 mm KCl, 0.4 m sucrose, and protease inhibitor mixture (Sigma). The homogenate was centrifuged at 900 × g for 10 min at 4 °C, and the pellet was discarded. The supernatant containing mitochondria was subjected to centrifugation at 14,000 × g for 15 min, the pellet was washed once with the homogenization buffer, and the mitochondria were collected by centrifugation under the same conditions and resuspended in the same buffer.

BNGE

Isolated mitochondria (100 μl, 1 mg of protein) were lysed by the addition of 200 μl of high purity digitonin solution (1:4 g of protein/g of digitonin) prepared as follows. 5 mg of digitonin (MP Biomedicals, Solon, OH) was dissolved in 200 μl of 30 mm HEPES buffer, pH 7.4, containing 150 mm potassium acetate and 10% glycerol, heated at 95 °C, then cooled on ice. The mitochondria/digitonin solution (final protein concentration 3 mg/ml) was incubated for 20 min on ice, a Coomassie Blue solution (5% Coomassie Blue G250 in 750 mm 6-aminocaproic acid) was added (1/20 v/v), and the reaction mixture was centrifuged at 14,000 × g for 30 min at 4 °C. The supernatant was directly loaded onto a 4–15% Tris-HCl Ready Gel (Bio-Rad), pH 8.8, and subjected to electrophoresis in 25 mm Tris-glycine buffer (without SDS) at 80 v constant voltage for 4 h, at 4 °C. Following electrophoresis, gels were stained with Bio-Safe Coomassie Blue G250 (Bio-Rad) for 30 min and destained with water. For second dimension separation, a strip of the gel corresponding to a single sample well was rotated 90 degrees and placed on a 12% SDS-polyacrylamide gel and subjected to electrophoresis as described (19). Following electrophoresis, the gel was visualized either with silver staining or Western blotting as below.

Western Blotting

Following BNGE or SDS-PAGE, proteins in the gel were electrophoretically transferred to PVDF membrane for 2 h as described (20). The membrane was blocked for 1 h at room temperature in 20 ml phosphate buffered saline containing 0.1% Tween 20 and 7% powdered milk. The primary antiserum was diluted 1:500–1000 in the same buffer, whereas the secondary antisera were diluted 1:2000–5000 (goat anti-rabbit IgG conjugated to alkaline phosphatase) and 1:5000 (goat anti-rabbit IgG conjugated to horseradish peroxidase).

Sucrose Gradient Centrifugation

Approximately 1.2 ml of digitonin-lysed mitochondria were loaded onto a 10–50% sucrose gradient in 50 mm phosphate buffer, pH 7.2, containing 0.2% digitonin. Samples were then subjected to centrifugation at 40,000 rpm (∼120,000 × g) using a Beckman Ti70 rotor for 19 h. The gradient was collected starting from the bottom in 12 2-ml fractions using a peristaltic pump.

In Situ Gel Staining for ETC Enzyme Activity

Strips corresponding to the sample wells were excised from the blue native gels and incubated with reaction buffer systems specific for ETC complex I, II, IV, or V. The activity staining procedures were essentially as described (21).

Enzyme Activity Assays

ACAD activity was measured with the anaerobic ETF fluorescence reduction assay using an LS50B fluorescence spectrophotometer from PerkinElmer Life Sciences with a heated cuvette block set to 32 °C as previously described (21–23). For standard reactions, the final substrate concentration was 50 μm. One unit of activity is defined as the amount of enzyme necessary to completely reduce 1 μmol of ETF in 1 min.

ETC complex I activity was measured in 50 mm phosphate buffer, pH 7.4, 50 μm NADH, 1 mm KCN, and 50 μm oxidized CoQ as described (24, 25). The reaction was detected by monitoring the decrease in absorbance at 340 nm and quantified using a molar extinction coefficient of 6.2 mm⁻¹·cm⁻¹. Complex III activity (defined as cytochrome c reduction) was measured in 50 mm phosphate buffer, pH 7.4, 1 mm KCN, 30 μm cytochrome c, 20 μm DBH₂, monitoring the increase in absorbance at 550 nm (26, 27). Complex IV activity was measured in 50 mm phosphate buffer, pH 7.4, 6 μm ferrocytochrome c, 100 mm KCl. The reaction was monitored at 550 nm (28).

FAO-ETC Bridging Assay

This reaction reflects the interaction of FAO and ETC as measured by reduction of cytochrome c in response to the addition of an acyl-CoA substrate to the reaction mixture. The basic reaction scheme was as follows: an aliquot of sample (usually the sucrose gradient supercomplex fraction) was added to reaction buffer (50 mm phosphate, pH 8.2, containing 50 μm acyl-CoA substrate, 3 μm ETF, 20 μm oxidized CoQ, 30 μm cytochrome c, 1 mm KCN) to give a final volume of 0.7 ml. The reaction was started by the addition of ETF and monitored for the reduction of cytochrome c as indicated by an increase in absorbance at 550 nm using a Beckman DU7500 spectrophotometer (29). Stearyl-CoA, palmitoyl-CoA, octanoyl-CoA, or butyryl-CoA was utilized as substrate to test the long, medium, and short chain specificity of the assay, respectively. A variety of enzymatic inhibitors were added to the base reaction to characterize the contribution of reducing equivalents from reduced ETF and NADH to cytochrome c reduction. Myxothiazol and antimycin A inhibit complex III. Rotenone is a specific inhibitor of complex I. 3-Mercaptopropionic acid (MPA) inhibits long chain ACADs (30). Kinetic parameters for cytochrome c reduction were calculated using a nonlinear regression algorithm as previously described (29, 31).

High Pressure Liquid Chromatography (HPLC) FAO Flux Assay

The ability of a sample to perform all of the functions of the entire FAO pathway was tested by following depletion of an acyl-CoA substrate and accumulation of various cycle intermediates via HPLC. Sucrose gradient fractions (2 μg of protein) or purified recombinant ACAD (2–5 ng of protein) was incubated with 50–300 μm butyryl-, octanoyl-, or palmitoyl-CoA at room temperature for 0–50 min. The reaction contained 1 mm pig ETF, 1.5 mm NAD, 0.3 mm CoASH, and 0.2 mm ATP. The reaction was stopped at 5 min intervals by the addition of 40 ml of 2 n HCl, and then immediately neutralized with 40 ml of 2 n NaOH in 200 mm MES. Following clearing of the sample in a bench top microcentrifuge for 10 min, 150 μl of sample was injected into a Luna (Phenomenex, Torrance, CA) 3 μm C18(2), 100 Å, 4.6 × 150-mm, 3-μm column (for palmitoyl-CoA reactions) or a SunFire (Phenomenex) C18, 3.5-μm column (for octanoyl- and butyryl-CoA reactions) equilibrated with 50 mm NH₄PO₄, pH 5.5, with 9% acetonitrile. The CoA esters that bound to the columns were eluted with a 9–45% linear gradient of acetonitrile in 50 mm NH₄PO₄, pH 5.5, at 0.9 ml/min over 50 min (Luna column) and 25 min (SunFire column). The presence of CoA ester intermediates in the eluent was monitored by absorbance at 254 nm and compared with the retention times of acyl-CoA ester standards eluted under identical conditions.

Characterization of the Association of ETC and FAO Pathway Components Using BNGE

To examine the association of ETC supercomplexes with FAO components, we separated extracts from digitonin-permeabilized rat liver mitochondria by BNGE and tested for the presence of FAO enzymes by Western blotting. Permeabilization conditions were selected to emphasize preservation of the ETC supercomplexes (1, 6). Following BNGE, a Coomassie Blue banding pattern characteristic of the ETC complexes and supercomplexes was observed (Fig. 1A). A complex I stain identified the top three Coomassie-stained bands as containing complex I activity (Fig. 2A). The two highest relative molecular mass bands (∼1200–2000 kDa) correspond to the previously described mobility of ETC supercomplexes, whereas the third band (∼1000 kDa) is the appropriate size for individual complex I (6). The remaining bands were readily identified by migration and specific activity stains as the other individual ETC complexes V (∼750 kDa), III (∼500 kDa), IV (∼230 kDa), and II (∼120 kDa) in order of migration (1, 6).

The ETC complexes and supercomplex bands separated by BNGE were transferred by electrophoresis to a nylon membrane and immunoblotted individually with antisera to the FAO enzymes. Variable patterns of staining of the supercomplex and complex I bands were observed (Fig. 2B). In contrast, antiserum to isovaleryl-CoA dehydrogenase, a member of the structurally homologous ACAD family that also uses ETF as its electron acceptor but is active in leucine metabolism, did not react with the three high molecular mass ETC bands. Thus, the FAO enzymes, but not closely related metabolic enzymes from other pathways, co-migrate with complex I and ETC supercomplexes on BNGE. Two-dimension gel electrophoresis with separation by SDS-PAGE following BNGE gave a pattern consistent with those previously published (Fig. 2C). Western blotting of this gel confirmed the association of very long ACAD (Fig. 2D) and long ACAD (data not shown) with supercomplexes and complex I. Because IVD antiserum was not found to react with the supercomplex fractions following single dimension BNGE, two-dimensional gel experiments were not pursued.

Characterization of the Association of ETC and FAO Pathway Components Using Sucrose Gradient Centrifugation Fractionation

Digitonin-permeabilized rat liver mitochondria were subjected to sucrose gradient centrifugation, and the gradient fractions were analyzed by BNGE (Fig. 1B). The three highest density fractions containing ETC supercomplexes and complex I were combined for further experiments. The proteins present in the pooled supercomplex fraction were separated by SDS-PAGE and subjected to Western blotting with antisera to the various FAO proteins. Very long chain ACAD, long chain ACAD, medium chain ACAD, and ETF along with short chain ACAD were identified in the pooled supercomplex fraction, and all migrated parallel to their corresponding purified recombinant proteins (Fig. 3). The membrane was also tested with antiserum to isovaleryl-CoA dehydrogenase, which detected no antigen in the high M_r fractions, confirming that the other highly homologous ACADs in the supercomplex fractions were not simply contamination with mitochondrial matrix proteins.

High levels of ACAD activity were detected in the supercomplex containing fractions with straight chain acyl-CoA substrates ranging between 4 and 20 carbons in length (Fig. 4). In contrast, activity toward the short branched chain substrates isovaleryl-CoA, isobutyryl-CoA, 2-methylbutyryl-CoA, and glutaryl-CoA was very low. Again, the absence of all of the FAO ACADs but not those associated with branched chain amino acid metabolism confirms the specificity of the association of the former with supercomplexes.

Evidence for Metabolic Channeling of FAO in the M_r Gradient Fractions

Because a variety of FAO enzymes were identified in the ETC supercomplex gradient fractions, we investigated the ability of these fractions to support coordinated flux through all four reactions necessary for complete β-oxidation. A HPLC-based assay was used to monitor the utilization of palmitoyl-CoA and the appearance of FAO reaction intermediates. In this assay, purified recombinant very long chain ACAD converted palmitoyl-CoA (Fig. 5A, curve a) to its enoyl-CoA product (Fig. 5A, curve b). When the supercomplex containing fractions were substituted for purified enzyme in the reaction, the palmitoyl-CoA substrate peak still decreased with time, but minimal accumulation of β-oxidation intermediates occurred (Fig. 5B, curves a and b). When MPA, a specific inhibitor of the long chain ACADs, was added to the reaction mixture, the palmitoyl-CoA concentration was stable throughout the reaction time (Fig. 5B, curve c). The supercomplex-containing fractions likewise oxidized octanoyl-CoA and butyryl-CoA. Thus, not only are the enzymes of FAO all present in association with the ETC complexes, they appear to be functionally organized to allow substrate channeling from one enzyme to the next in analogy to that seen with OXPHOS.

Evidence of Electron Transfer from FAO to ETC in the High Molecular Mass Sucrose Fractions

We postulated that the physical association of the multifunctional FAO complex and the ETC supercomplexes in the sucrose gradients might represent a functional unit that promotes metabolic channeling of electrons directly from FAO to the ETC. To test this, we utilized a bridging assay that starts with an FAO substrate and measures reduction of cytochrome c in ETC (Scheme 1 and Fig. 6). When palmitoyl-CoA was used as substrate in the reaction mixture in the presence of excess cytochrome c and CoQ (along with KCN to inhibit electron transfer from cytochrome c to complex IV), the sucrose gradient fractions containing ETC supercomplexes efficiently promoted reduction of cytochrome c (Fig. 6). Addition of exogenous NAD greatly enhanced the reaction. Cytochrome c reduction was dependent on the addition of ETF and was inhibited completely by myxothiazol, an inhibitor of ETC complex III (Fig. 6A). Adding rotenone (a complex I inhibitor) to the reaction mixture reduced the rate of cytochrome c reduction by about half, indicating that only part of the electron flow to cytochrome c was through complex I when acyl-CoA substrate was in excess (Fig. 6B). Thus, the remaining reducing equivalents flowing to cytochrome c must be a result of the ACAD reaction with entry of electrons to the ETC through ETF and ETF-CoQ-oxidoreductase. In keeping with this finding, bridging activity was completely inhibited by the addition of the long chain ACAD inhibitor MPA (Fig. 6C). As a control, the activity of complex I in the presence of MPA was shown to be intact by the ability of NADH to reduce cytochrome c (Fig. 6C). The bridging activity was of similar magnitude when octanoyl-CoA was used as substrate. ACAD activity toward palmitoyl-CoA was most prominent in the high density fractions that contain the ETC supercomplexes (first four fractions; Fig. 7A). The lower density fractions corresponding to the expected molecular mass of the individual ACADs (∼140–170 kDa) exhibited approximately one-tenth of the ACAD activity of the ETC supercomplex fractions. The high density fractions also exhibited the highest bridging activity (Fig. 7B). These results, together with the data presented in Figs. 2 and 3, indicate that fatty acid oxidation and the ETC are both physically and functionally linked.

To examine the kinetic of the bridging reaction, assays were performed with variable palmitoyl-CoA concentrations. The apparent K_m of the bridging reaction was calculated to be 14.3 μm, only ∼5 times higher than the in vitro K_m of purified recombinant human very long chain ACAD and thus is in a range that is physiologically relevant.