Parkinson's disease brain mitochondrial complex I has oxidatively damaged subunits and is functionally impaired and misassembled

Abstract

Loss of mitochondrial complex I catalytic activity in the electron transport chain (ETC) is found in multiple tissues from individuals with sporadic Parkinson's disease (PD) and is a property of some PD model neurotoxins. Using special ETC subunit-specific and complex I immunocapture antibodies directed against the entire complex I macroassembly, we quantified ETC proteins and protein oxidation of complex I subunits in brain mitochondria from 10 PD and 12 age-matched control (CTL) samples. We measured nicotinamide adenine dinucleotide (NADH)-driven electron transfer rates through complex I and correlated these with complex I subunit oxidation levels and reductions of its 8 kDa subunit. PD brain complex I shows 11% increase in ND6, 34% decrease in its 8 kDa subunit and contains 47% more protein carbonyls localized to catalytic subunits coded for by mitochondrial and nuclear genomes We found no changes in levels of ETC proteins from complexes II-V. Oxidative damage patterns to PD complex I are reproduced by incubation of CTL brain mitochondria with NADH in the presence of rotenone but not by exogenous oxidant. NADH-driven electron transfer rates through complex I inversely correlate with complex I protein oxidation status and positively correlate with reduction in PD 8 kDa subunit. Reduced complex I function in PD brain mitochondria appears to arise from oxidation of its catalytic subunits from internal processes, not from external oxidative stress, and correlates with complex I misassembly. This complex I auto-oxidation may derive from abnormalities in mitochondrial or nuclear encoded subunits, complex I assembly factors, rotenone-like complex I toxins, or some combination.

Figure 1.

Western blot of CTL and PD brain mitochondrial ETC subunits. The same mitochondrial preparations from frontal cortex samples were used for all figures. The blot images are shown for CTL and PD mitochondrial samples. The table at the bottom gives the actual mean band densities with the SDs underneath. p values for comparing the CTL and PD groups did not demonstrate any significant differences. See Materials and Methods for details. CI–CV, Complex I–V.

Figure 2.

Western blot of CTL and PD mitochondria for complex I subunits. The same mitochondrial preparations from frontal cortex samples were used for all figures. The blot images are shown for CTL and PD mitochondrial samples, and localization of the complex I subunit bands was performed as described previously (Triepels et al., 2001; Murray et al., 2003). The table at the bottom gives the actual mean band densitites with the SDs underneath. p values indicated highly significant differences for the 20 and 8 kDa subunits. See Materials and Methods for details.

Figure 3.

Protein carbonyl levels in immunocaptured complex I samples. The same mitochondrial preparations from frontal cortex were used for all figures. Shown are immunoblots derived from 10 μg protein samples of immunocaptured complex I from three CTL (Ctl) and six PD brains after derivatization with DNPH, SDS-PAGE, transfer, and immunostaining for DNPH. The band on the far left (“DNPH”) shows the location and relative intensity of the DNPH protein standards. The arrow points to DNPH-trypsin inhibitor band, which was used as a standard to normalize all the other detected bands. MW marker bands are not shown. See Materials and Methods for details.

Figure 4.

Top, Examples of NADH-driven AmplexRed oxidation experiments. Shown are typical results from incubation of 0.4 mg/ml mitochondrial protein isolated from human frontal cortex and incubated at 37° with the indicated additions. Results from a CTL brain are on the left, and those from a PD brain are on the right. Bottom, Diagram of electron flow through complex I. mitos, Mitochondria; IM, OM, inner and outer mitochondrial membranes; FMN, flavin mononucleotide; N-(x), Fe-S clusters; Q, coenzyme Q. Adapted from Genova et al. (2004).

Figure 5.

Mitochondrial rates of NADH-driven AmplexRed oxidation are inversely correlated with protein carbonyl levels in complex I bands. Shown are plots of rates of NADH-driven mitochondrial AmplexRed oxidation (expressed as % CTL values) versus optical densities of protein carbonyls (as different DNPH bands from SDS-PAGE gels) in immunocaptured complex I. CTL samples are circles and PD samples are squares; R values for linear regression fitting and p values for the regression fit are shown.

Figure 6.

NADH plus rotenone increases oxidative damage to complex I subunits. Mitochondria from a CTL frontal cortex samples were incubated for 75 min with constituents added as indicated, washed, and then complex I was immunocaptured and analyzed for protein carbonyls as described in Materials and Methods. A, With [NADH] = 125 μm, rotenone (10 μm) appears to increase DNPH reactivity in bands at 18, 21, 26, 37, 50, and 94 kDa. Hydrogen peroxide only increased DNPH reactivity in the ∼20 kDa band. B, In two other CTL brain mitochondrial fractions incubated with a 10-fold higher [NADH] of 1.25 mm (lanes 1 and 3), the increased DNPH density of the ∼50 kDa band from immunocaptured complex I is not further increased with rotenone (10 μm, lanes 2 and 4), but that of the ∼26 kDa band is clearly increased by rotenone. Lanes 1 and 3 are derived from separate CTL brain mitochondrial preparations incubated with 1.25 mm NADH. Lanes 2 and 4 are derived from the CTL mitochondrial preparations used to generate lanes 1 and 3, but after addition of 10 μm rotenone.

Figure 7.

Correlations between levels of 8 kDa complex I subunit and NADH-driven AmplexRed oxidation rates (left) and protein carbonyl levels in complex I (right) in PD (square) and CTL (round) brain mitochondria. R values for linear regression fitting and p values for the regression fit are shown.