Mitochondrial respiration and respiration-associated proteins in cell lines created through Parkinson’s subject mitochondrial transfer


Address correspondence and reprint requests to Russell Swerdlow, MD, University of Kansas School of Medicine, MS 2012, Landon Center on Aging, 3901 Rainbow Blvd, Kansas City, KS 66160, USA. E-mail:


J. Neurochem. (2010) 113, 674–682.


Parkinson’s disease (PD) is associated with perturbed mitochondrial function. Studies of cytoplasmic hybrid (cybrid) cell lines containing mitochondria from PD subjects suggest complex I dysfunction in particular is a relatively upstream biochemical defect. To evaluate potential downstream consequences of PD mitochondrial dysfunction, we used a cybrid approach to model PD mitochondrial dysfunction; our cybrid cell lines were generated via transfer of PD or control subject platelet mitochondria to mtDNA-depleted NT2 cells. To confirm our PD cybrid mitochondria did indeed differ from control cybrid mitochondria we measured complex I Vmax activities. Consistent with other PD cybrid reports, relative to control cybrid cell lines the PD cybrid cell line mean complex I Vmax activity was reduced. In this validated model, we used an oxygen electrode to characterize PD cybrid mitochondrial respiration. Although whole cell basal oxygen consumption was comparable between the PD and control cybrid groups, the proton leak was increased and maximum respiratory capacity was decreased in the PD cybrids. PD cybrids also had reduced SIRT1 phosphorylation, reduced peroxisome proliferator-activated receptor-γ coactivator-1α levels, and increased NF-kB activation. We conclude mitochondrial respiration and pathways influenced by aerobic metabolism are altered in NT2 cybrid cell lines generated through transfer of PD subject platelet mitochondria.

Abbreviations used:

Dulbecco’s modified Eagle’s medium


electron transport chain


fetal bovine serum


carbonyl cyanide-p-trifluoromethoxyphenylhydrazone


green fluorescent protein


maximum respiratory capacity


Parkinson’s disease


peroxisome proliferator-activated receptor-γ coactivator-1α


reactive oxygen species

Multiple lines of evidence suggest mitochondria play a role in Parkinson’s disease (PD) neurodegeneration (Parker and Swerdlow 1998). Alteration of the electron transport chain (ETC) enzyme complex I appears particularly relevant to this disease. Toxin-mediated complex I inhibition causes Parkinsonism and induces PD-relevant neuropathology in humans and animal models (Langston et al. 1983; Betarbet et al. 2000). Multiple tissues from PD patients show reduced complex I Vmax activity (Swerdlow 2007a).

Genetically altered mice are commonly used to model PD (Moore and Dawson 2008). These murine lines are created through transgenic expression of mutant genes that cause autosomal dominant PD variants or knock-out of genes that cause autosomal recessive PD variants. Mitochondrial physiology can be altered in these mice (Gispert et al. 2009), which is consistent with the view that mitochondria are relevant to PD. Most PD patients, though, do not show obvious Mendelian inheritance, do not carry currently identifiable PD-related nuclear gene mutations, and are felt to have a sporadic disease. It is unclear how rigorously mitochondrial alterations in the Mendelian models recapitulate sporadic PD patient mitochondrial alterations.

Human sporadic disease research has arguably been slowed by limitations in sporadic disease modeling. When it comes to the specific modeling of mitochondrial function in human sporadic diseases, although, approaches such as the cytoplasmic hybrid (cybrid) approach may prove useful (Swerdlow et al. 1997). Cybrids are cell lines created when the cytoplasmic contents of one cell are transferred to another cell. To date, multiple investigators have used different cybrid platforms to study sporadic PD mitochondrial function. Most studies of sporadic PD subject mitochondria show reduced complex I activity, and in most PD cybrid studies transfer of sporadic PD subject platelet mitochondria to mtDNA-depleted cells creates cell lines with persistently reduced complex I Vmax activities. The PD cybrid complex I defect is further associated with and may alter other physiologic parameters including oxidative stress, calcium homeostasis, toxin susceptibility, stress pathways and synuclein aggregation (Swerdlow 2007a; Esteves et al. 2008, 2009). In this study, we evaluated mitochondrial respiration in PD cybrid cell lines, as well as the status of certain proteins [SIRT1, peroxisome proliferator-activated receptor-γ coactivator-1α (PGC1α), and NF-kB] whose functions are influenced by cell aerobic-anaerobic balances.

Materials and methods

Chemicals and cell media

Carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), oligomycin, rotenone and myxothiazol were obtained from Sigma-Aldrich. Optimem and Dulbecco’s modified Eagle’s medium (DMEM) media were obtained from Gibco-Invitrogen. Non-dialyzed and dialyzed fetal bovine serum (FBS) was obtained from Sigma. NT2 ρ0 cell growth medium consisted of Optimem supplemented with 10% non-dialyzed FBS, 200 μg/mL sodium pyruvate, 150 μg/mL uridine and 1% penicillin–streptomycin solution. NT2 cybrid selection medium consisted of DMEM supplemented with 10% dialyzed FBS and 1% penicillin–streptomycin solution. Cybrid growth medium consisted of Optimem supplemented with 10% non-dialyzed FBS and 1% penicillin-streptomycin solution. Prior to experiments, cell lines were maintained in the cybrid growth medium.

Human subjects

Subject participation was approved by the University of Kansas School of Medicine’s Institutional Review Board. After obtaining informed consent, sporadic PD (n = 9) and age-matched control (n = 5) subjects underwent a 10 mL phlebotomy using tubes containing acid-citrate-dextrose as an anticoagulant. The age of the PD subjects who participated in this study was 64 ± 12.8 years, and for the control subjects it was 74.3 ± 5.5 years. The PD subjects were followed regularly in a tertiary referral movement disorders clinic at the Kansas University Medical Center and met criteria commonly used to diagnose PD in clinical and research settings (Litvan et al. 2003). The control subjects were participants of a longitudinal ‘normal aging’ cohort that is characterized serially by the Brain Aging Project at the University of Kansas School of Medicine. These control subjects have not been diagnosed with a neurodegenerative or pre-neurodegenerative disease condition.

Creation of cybrid cell lines and cell culture

Relative to control subjects, platelets from PD subjects are known to have reduced complex I activity (Parker et al. 1989). We used platelet mitochondria to generate cybrid cell lines from PD and control subjects. Platelets were isolated from the individual blood samples and briefly co-incubated with NT2 ρ0 cells in polyethylene glycol diluted with S-minimum essential medium (Swerdlow et al. 1997). The products of this co-incubation were plated on T75 flasks, maintained for 1 week in ρ0 growth medium, and then switched to cybrid selection medium for 6 weeks. NT2 ρ0 cells lack intact mtDNA, do not possess a functional ETC, and are auxotrophic for pyruvate and uridine. Maintaining cells in selection medium removes ρ0 cells that have not repopulated their mtDNA with platelet mtDNA. ‘Mock fusions’ in which NT2 ρ0 cells not co-incubated with platelets were plated and maintained in selection medium were performed in parallel with the true fusions, and during the selection period all cells from the mock fusions died. After selection was complete, the resultant cybrid cells were changed to cybrid expansion medium. Flasks were maintained in this medium at 37°C, 5% CO2 for 24 h prior to harvesting.

NADH-ubiquinone oxidoreductase assay

Complex I activity was determined by a modified version of Ragan et al. (1987) which follows the decrease in NADH absorbance at 340 nm that occurs when ubiquinone (CoQ1) is reduced to form ubiquinol. The reaction was initiated by adding CoQ1 (50 μM) to the 30°C reaction mixture. After 5 min, rotenone (10 μM) was added and the reaction was followed for another 5 min. Complex I activity was expressed both as nanomoles per minute per milligram of protein, as well as the ratio of complex I activity per citrate synthase activity.

Citrate synthase assay

Citrate synthase activity was determined by the method of Coore et al. (1971), which spectrophotometrically follows the formation of 5-thio-2-nitrobenzoate at 412 nm. The assay was initiated by the addition of 100 μM oxaloacetate at 30°C. Results were calculated as nanomoles per minute per milligram of protein.

Respirometric analyses

Cell oxygen consumption was measured using an Oroboros Oxygraph-2k high resolution respirometer in a standard configuration (Hutter et al. 2006). Flasks were harvested in trypsin, and after washing four million cells were suspended in 2.1 mL of assay buffer (unbuffered DMEM base media supplemented with 200 mM Glutamax-1, 1 mM pyruvate, 25 mM Glucose), added to each chamber of the respirometer, and maintained during the assay at 37°C and 750 rpm stirrer speed. In addition to determinations of cell basal respiratory rates, the maximal respiratory capacity was determined by uncoupling respiration with 1 μM FCCP and the mitochondrial proton leak rate was determined by inhibiting ATP synthase with 4 mg/mL oligomycin. A complex I-independent flux was obtained by adding 0.5 μM rotenone to the chambers. Non-mitochondrial oxygen consumption was measured through addition of 1 μM of the complex III inhibitor myxothiazol.


Cells were harvested with trypsin, centrifuged at 500 g for 5 min, resuspended in phosphate-buffered saline, and centrifuged again. To prepare whole cell lysates, the washed cell pellet was suspended in M-PER Mammalian Protein Extraction Reagent (Pierce, Rockford, IL, USA) supplemented with Halt Protease Inhibitor Cocktail (Pierce) and gently shaken for 5 min as directed. To prepare pure nuclear and cytosolic lysates, the washed cell pellet was processed using an N-PER Nuclear and Cytoplasmic Extraction Reagent kit (Pierce) as directed. Protein concentrations for all lysates were determined using a DC Protein Assay kit (Bio-Rad Laboratories, Hercules, CA, USA). For western blot analyses, samples were boiled, diluted 1 : 5 in sample buffer, resolved by electrophoresis in pre-cast 4–12% gels (Bio-Rad), and transferred to polyvinylidene difluoride membranes. Non-specific binding was blocked by gently agitating the membranes in 5% non-fat milk and 0.1% Tween in Tris-buffered saline for 1 h at 22°C. Blots were subsequently incubated in buffer containing a designated primary antibody [phospho-sirtuin-1 antibody from Cell Signaling Technology (Beverly, MA, USA) at a 1 : 1000 dilution; sirtuin 1 antibody from Cell Signaling at a 1 : 1000 dilution; PGC1α antibody from Santa Cruz Biotechnology (Santa Cruz, CA, USA) at a 1 : 1000 dilution; TATA binding protein antibody from Abcam (Cambridge, MA, USA) at a 1 : 2000 dilution; and actin antibody from Santa Cruz Biotechnology at a 1 : 1000 dilution) overnight at 4°C with gentle agitation. Blots were washed with Tris-buffered saline containing 0.1% Tween three times (each time for 10 min) and incubated with the appropriate horseradish peroxidase-conjugated secondary antibody for 1 h at 22°C with gentle agitation. After three washes, the blots were treated with SuperSignal West Femto Maximum Sensitivity Substrate. Chemiluminescence signals were detected using a Bio-Rad ChemiDoc Imager and band densities determined using Quantity One Software.

NF-kB activity

NF-kB activity was determined using a lentivirus construct in which a green fluroescent protein (GFP) reporter is located downstream of a modified cytomegalovirus promoter and an NF-kB binding region (pTRF1-NF-kB-dscGFP reporter vector from System Biosciences, Mountain View, CA, USA). Cybrid cell lines were transduced with this construct and sorted by GFP fluorescence intensity using a FACSDiva Version 6.1.2 cell sorter. Because for cells transduced with this construct GFP expression is dependent on NF-kB transcriptional activity, GFP fluorescence intensity provides an indicator of NF-kB activity.

Data analysis

For all experimental endpoints in which a particular cybrid line was assayed more than once, the mean value for all the assays was calculated. Overall group mean ± SEM for the PD and control cybrid groups were calculated. p-values were calculated by two-way Student’s t-tests. p-values less than 0.05 were considered significant.


Complex I activity is reduced in PD cybrids

Complex I (NADH-ubiquinone oxidoreductase) Vmax activities were measured in PD (n = 9) and control (n = 5) cybrid cell lines. Consistent with other PD cybrid studies (Swerdlow 2007a), the mean complex I activity of the PD cybrid lines was lower than that of the control cybrids. When the mean complex I activity was normalized according to mg protein the PD cybrid complex I mean was 41.5% lower, and when normalized by citrate synthase activity it was 43.6% lower (Fig. 1).

Figure 1.

 Parkinson’s disease (PD) cybrid mean complex I activity is less than the control cybrid mean complex I activity. This relationship is seen when the Vmax activity is corrected for mg protein (a) and citrate synthase activity (b). **p < 0.01 difference between PD and control cybrid groups.

PD cybrids have normal basal respiration but reduced respiratory reserve capacity

Oxygen consumption in our PD (n = 9) and control (n = 5) cybrid cell lines was assessed using an Oroboros Oxygraph II respirometer. In one experimental paradigm (Fig. 2a), we first determined basal oxygen consumption rates. Comparing basal oxygen consumption between cell lines requires that equal numbers of cells for each line are assayed. While we attempted to include 4 million cells per chamber, to more rigorously insure equal numbers of cells were always assayed we further normalized oxygen consumption rates to mg of protein in the chamber. We also corrected the basal oxygen consumption rates by subtracting out the non-mitochondrial oxygen consumption that was present in the chamber after mitochondrial oxygen consumption was entirely shut down using the complex III inhibitor myxothiazol. Basal oxygen consumption rates were comparable between CT and PD cybrids (Fig. 2b).

Figure 2.

 Basal and maximum respiratory capacity analyses of cybrid whole cells. (a) Representative experiments in which cells were suspended in Oroboros oxygraph chambers are shown. After the basal oxygen consumption is determined, carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) is added to the chambers to uncouple the mitochondria and establish the maximum respiratory capacity (MRC). Next, rotenone and myxothiazol are sequentially added to determine the rotenone sensitive and insensitive contributions to the MRC. (b) Parkinson’s disease (PD) and control cybrid cell lines show equivalent basal oxygen consumption. (c) Under MRC conditions the PD cybrids increase their oxygen consumption to a lesser degree than the control cybrids, indicating that under basal conditions PD cybrids are respiring closer to their MRC than are control cybrids. (d) Under MRC conditions, we were unable to demonstrate intergroup differences in the rotenone-sensitive flux through complex I. (e) Under MRC conditions, we were unable to demonstrate intergroup differences in the rotenone-sensitive flux through complex II. ***p < 0.001 difference between PD and control cybrid groups.

As shown in Fig. 2a, after establishing a basal oxygen consumption rate we injected FCCP into the chamber. FCCP is a protonophore that disrupts the integrity of the inner mitochondrial membrane. Following FCCP treatment, protons pumped from the mitochondrial matrix to the intermembrane space as a consequence of ETC electron transfer can re-access the mitochondrial matrix independent of ATP synthase. This uncouples electron transport from ATP production, allows ETC electron flux and mitochondrial oxygen consumption to proceed at their greatest possible rates, and quantifies the maximum respiratory capacity (MRC). To eliminate differences in cell number or mass as a potential confounder, we determined the percent increase that occurred as an FCCP treated cell line moved from its basal oxygen consumption rate to its MRC rate. We found that in switching from a basal respiratory rate to an MRC rate, the PD cybrid cell lines increased their oxygen consumption to a much smaller degree than the control cybrid cell lines (Fig. 2c). The mean increase in the PD cybrid line oxygen consumption was 27% of what it was in the control cybrid lines. This suggests the PD cybrid cell lines have less respiratory reserve than the control lines, and that under baseline conditions the PD cybrids are respiring closer to their MRC.

Because the complex I Vmax activity is reduced in these PD cybrid lines we assessed the relative contributions of complex I and complex II-dependent fluxes to oxygen consumption under MRC conditions. As shown in Fig. 2a, we did this by following our FCCP injections with first rotenone and then myxothiazol injections. Rotenone eliminates the complex I-dependent oxygen consumption, leaving just the complex II-dependent mitochondrial oxygen consumption. By inhibiting complex III, myxothiazol eliminates residual complex II-dependent oxygen consumption. After myxothiazol is added to the chamber, only non-mitochondrial oxygen consumption remains. To insure differences in cell number did not confound this analysis, for each line we determined the percent of the MRC that was complex I or complex II dependent. We found that in both groups rotenone eliminated about 95% of the mitochondrial oxygen consumption present under MRC conditions (Fig. 2d). Only a small amount of mitochondrial oxygen consumption was complex II-dependent, and the complex II dependent fluxes were comparable between groups (Fig. 2e). Taken together, our results suggest that under these experimental conditions flux through complex II was not increasing to compensate for the change in the complex I Vmax, and further suggests under basal conditions compensation by complex II was not required to maintain bioenergetic integrity.

PD cybrids have an increased mitochondrial proton leak

Electron transfer by the ETC is accompanied by pumping of mitochondrial matrix protons to the intermembrane space. Because of the electrochemical gradient across the mitochondrial inner membrane, if allowed to intermembrane space protons will re-access the matrix. Protons may access the matrix through the ATP synthase (complex V), and proton movement from the intermembrane space to the matrix via the ATP synthase is accompanied by phosphorylation of ADP to ATP. Depending on the intrinsic resistance of the inner membrane to proton transit, protons may also ‘leak’ into the matrix in a way that proton movement is not coupled to ADP phosphorylation. We used our Oroboros respirometer to assess the leakiness of the mitochondrial inner membrane to protons. This was accomplished by injecting the chambers with oligomycin, an ATP synthase inhibitor. In the presence of oligomycin, protons cannot re-access the matrix through the ATP synthase, the potential across the inner membrane hyperpolarizes, and the mitochondrial oxygen consumption rate declines. The only residual mitochondrial oxygen consumption that can occur in this state is directly related to the mitochondrial proton leak; the greater the leak, the greater the ongoing oxygen consumption. By following the oligomycin injection with injection of myxothiazol, the non-mitochondrial oxygen consumption in the chamber is obtained. Subtracting the post-myxothiazol rate from the basal rate provides the total basal mitochondrial respiration rate, and subtracting the post-myxothiazol rate from the post-oligomycin rate provides the mitochondrial leak rate. A representative experiment is shown in Fig. 3a. We calculated the percent of mitochondrial oxygen consumption that was due to inner membrane proton leak (Fig. 3b), and the percent of oxygen consumption that was oligomycin-sensitive and therefore coupled to ATP production (Fig. 3c). Under basal respiratory conditions, the amount of oxygen consumption that was due to proton leak was slightly higher in the PD cybrids (36.7%) than it was in the control cybrids (30.3%).

Figure 3.

 Parkinson’s disease (PD) cybrids have an increased proton leak. (a) Representative experiments in which cells were suspended in Oroboros oxygraph chambers are shown. After the basal oxygen consumption is reached, oligomycin is added to the chambers to inhibit proton movement through the ATP synthase. Myxothiazol is then added to quantify the non-mitochondrial oxygen consumption present in the chamber. (b) The ATP synthase-independent proton leak across the mitochondrial inner membrane was slightly but significantly increased in the PD cybrids. (c) Mitochondrial oxygen consumption that is not related to proton leak is presumed to associate with proton movement through the ATP synthase, and thus provides an estimate of how tightly electron flux is ‘coupled’ to ATP production. *p < 0.05 difference between PD and control cybrid groups.

SIRT1, PGC1α, and NF-kB alterations are seen in PD cybrids

SIRT1 is an NAD+-dependent lysine deacetylase that helps determine cell aerobic activity levels (Finkel et al. 2009). Phosphorylation of SIRT1 reportedly correlates positively with its activation (Sasaki et al. 2008). We therefore assessed the SIRT1 activation state in our cybrid cell lines by quantifying its phosphorylation status. Relative to control cybrid cell the lines, SIRT1 phosphorylation in PD cybrids was reduced (Fig. 4a and b). Because SIRT1 can shuttle between nuclear and cytosolic compartments (Tanno et al. 2007), we further evaluated the status of cybrid cell line SIRT1by assessing its nuclear to cytosolic ratios. In both sets of cybrids there appeared to be more nuclear than cytosolic SIRT1, and the nuclear : cytosolic ratios in the PD and control cybrid cell lines were equivalent. (Fig. 4c).

Figure 4.

 SIRT1 analyses. The Parkinson’s disease (PD) cybrids showed a slight but significant reduction in SIRT1 phosphorylation (a, b). We did not observe any changes in total SIRT1 levels or in SIRT1 compartmentalization; representative immunochemistry data are shown in (c). *p < 0.05 difference between PD and control cybrid groups.

PGC1α is a transcriptional coactivator that participates in the up-regulation of aerobic metabolism (Puigserver and Spiegelman 2003; Lin et al. 2005; Handschin and Spiegelman 2006). It facilitates transcription of genes required for mitochondrial biogenesis, transcription of genes encoding antioxidant enzymes, and its own expression. PGC1α is de-acetylated by SIRT1, and de-acetylation of PGC1α increases its activity. Relative to control cybrid cell lines, PGC1α levels were reduced in PD cybrid cell lines (Fig. 5). Because PGC1α participates in its own expression, lower PGC1α levels in PD cybrids suggests PGC1α activity in PD cybrids is itself reduced.

Figure 5.

 Peroxisome proliferator-activated receptor-γ coactivator-1α (PGC1α) levels are decreased in Parkinson’s disease (PD) cybrids. (a) Representative PGC1α bands. (b) Relationship of the PD cybrid mean PGC1α density to the control cybrid mean PGC1α density. **p < 0.01 difference between PD and control cybrid groups.

SIRT1 is also known to de-acetylate NF-kB; this is an inhibitory de-acetylation (Yeung et al. 2004). We used a GFP reporter system to evaluate NF-kB activity. We found that NF-kB was more transcriptionally active in PD cybrid cell lines than it was in control cybrid lines (Fig. 6).

Figure 6.

 NF-kB activity is increased in Parkinson’s disease (PD) cybrids. Each cybrid cell line was transduced with a GFP reporter construct in which GFP expression is dependent on binding of NF-kB to its response element. Each transduced cell line was then subjected to FACS sorting. The GFP fluorescence intensity from the live cells of each line was determined. The relative mean GFP fluorescence of the PD cybrid cell lines (as compared with the mean GFP fluorescence of the control cybrid cell lines) was increased approximately three-fold. ***p < 0.001 difference between PD and control cybrid groups.


In this study, a cybrid approach was used to model sporadic PD mitochondrial function and evaluate its impact on certain proteins that are influenced by mitochondrial respiration. In these cell lines, it was found that the complex I defect seen in sporadic PD patients associates with a reduced MRC, increased mitochondrial proton leak, reduced SIRT1 phosphorylation, reduced PGC1α protein, and increased NF-kB activation. We conclude mitochondrial respiration and pathways influenced by aerobic metabolism are altered in NT2 cybrid cell lines generated through transfer of PD subject platelet mitochondria.

The cybrid approach was first applied to the study of PD mitochondria in 1996 (Swerdlow et al. 1996). A major intent of that study was to address the origin of the systemic complex I Vmax defect seen in sporadic PD subjects. Prior to that study, the cybrid approach had mostly been used to study the functional consequences of known mtDNA mutations (Swerdlow 2007a). A conclusion reached in the 1996 PD cybrid study was that mtDNA was likely to be at least partly responsible for the sporadic PD complex I defect. To date, however, no single unequivocal mtDNA signature has emerged as the definitive source of the complex I defect observed in that and subsequent other PD cybrid studies. One possible interpretation of this is the PD cybrid complex I defect arises independent of mtDNA. Another interpretation is simply that although mtDNA is responsible no single mtDNA mutation or polymorphism solely accounts for the defect. Reports that mtDNA deletions or microheteroplasmic mutations in the ND5 gene are found to a much greater degree in PD subject brains than in control subject brains are more consistent with the latter possibility, although studies looking for such mtDNA mutations in PD subject platelet mtDNA have not been published (Smigrodzki et al. 2004; Parker and Parks 2005; Bender et al. 2006; Kraytsberg et al. 2006). Our current study does not help to address this longstanding debate.

A second PD cybrid question relates to whether PD cybrids let investigators model sporadic PD mitochondrial dysfunction. Since 1996, over a dozen different PD cybrid studies have reported PD cybrid mitochondria or physiology directly affected by mitochondrial function differ between cybrids containing mitochondria transferred from PD subjects as compared to cybrids containing mitochondria transferred from control subjects (Swerdlow 2007a; Esteves et al. 2008, 2009). These ‘positive’ PD cybrid studies have been performed using SH-SY5Y neuroblastoma, A549 lung carcinoma, and NT2 teratocarcinoma nuclear backgrounds. To date, only one ‘negative’ cybrid study has been reported (Aomi et al. 2001). This study was performed using a HeLa cervical carcinoma nuclear background, and interestingly the negative conclusions of this study were partly based on a whole cell basal oxygen consumption analysis. In this study, we also found whole cell basal oxygen consumption between NT2 PD and control cybrids was equivalent. However, we also found the PD cybrid complex I Vmax was decreased, the PD cybrid maximum respiratory capacity was decreased, the mitochondrial proton leak was increased, and differences in SIRT1, PGC1α, and NF-kB levels or function were present. Therefore, although we confirm the respiratory analysis of Aomi et al., we conclude PD cybrids do facilitate modeling of sporadic PD subject mitochondrial function.

Because cybrids lines are generated using tumor cell lines it is perhaps not surprising that basal whole cell oxygen consumption is equivalent between PD and control cybrids. Tumor cells in general are subject to the Warburg Effect in which mitochondrial oxygen consumption is actively maintained at limited, relatively low levels (Feron 2009). A more intensive analysis of mitochondrial respiration did, however, reveal differences. Most importantly, we found the MRC of cybrid cell lines containing mitochondria transferred from PD subject platelets was reduced.

Whole cell respiratory analyses are becoming increasingly popular as the technology for assessing whole cell oxygen consumption improves. Along with improvements in respiratory analysis instruments comes refinement in whole cell respiratory analysis protocols. Some investigators recommend using a gradual FCCP titration to induce MRC states (Hutter et al. 2006) although when serial FCCP injections are not practical single FCCP injections are used (Amo et al. 2008). In deference to this technical issue, we compared the effects of FCCP titration to a single FCCP bolus. Both approaches produced similar results (data not shown). Using oligomycin to measure proton leak across the mitochondrial membrane is also becoming more common. With this approach, we found the PD cybrid proton leak rate was relatively high. An increased proton leak would predictably depolarize the mitochondrial membrane potential. Other PD cybrid studies report PD cybrid mitochondria are relatively depolarized (Esteves et al. 2008). We wonder whether the increased proton leak we observed is at least partly responsible for this.

We further extended our analyses to proteins such as SIRT1 that influence or are influenced by cell respiratory function. SIRT1 helps determine cell respiratory set points and because it is dependent on cytosolic NAD+/NADH ratios, SIRT1 activity is itself determined by the cell redox state (Yang and Sauve 2006; Arduino et al. 2009; Finkel et al. 2009). Although cytosolic and mitochondrial NAD+/NADH pools are not in direct contact, the two pools can influence each other through shuttles that permit the exchange of reducing equivalents. For example, if the mitochondrial NAD+/NADH ratio is high, cytosolic NADH can be converted to NAD+ via the transporter-mediated transfer of reducing equivalents from cytosolic NADH to mitochondrial NAD+. In general, high cytosolic NAD+/NADH ratios are associated with SIRT1 activation. We found no evidence that SIRT1 activity was increased and evidence consistent with a possible activity decrease. This suggests the cytosolic NAD+/NADH ratio in PD cybrids is either comparable to or slightly less than that of control cybrids. Although a complex I lesion could contribute to a low NAD+/NADH ratio, it is difficult to know whether this is an artifact of studying SIRT1 in a tumor cell model and we are unsure of the significance of this finding. Regardless, because SIRT1 activates PGC1α and PGC1α facilitates its own expression decreased SIRT1 activity is functionally consistent with the observed reduction in PD cybrid PGC1α levels (Nemoto et al. 2005; Handschin and Spiegelman 2006). In related studies, we have found increasing mitochondrial reactive oxygen species (ROS) production in neuroblastoma cells associates with reduced SIRT1 activity (in preparation). Multiple studies report PD cybrid mitochondria have increased ROS production. We speculate that in these tumor cell lines increased mitochondrial ROS production may actually stimulate the Warburg Effect, which could potentially manifest as a reduction in mitochondrial biogenesis signals.

ROS are known to activate NF-kB (Kaltschmidt et al. 1999; Kabe et al. 2005). Cells maintain a cytosolic pool of NF-kB and activation of this transcription factor associates with its nuclear translocation. Nuclear NF-kB translocation is observed in PD subject dopaminergic neurons (Hunot et al. 1997). Relevant to this, Onyango et al. (2005) previously reported immunochemistry data indicating NF-kB activity was increased in SH-SY5Y PD cybrids. Our current data obtained from studies of NT2 PD cybrids and using a transcription reporter system confirm and extend this report. Because SIRT1-mediated NF-kB deacetylation is known to inhibit NF-kB (Yeung et al. 2004), and our PD cybrids have a potential reduction in SIRT1 activity, it is possible that in addition to increased ROS decreased SIRT1 activity may also mediate NF-kB activation.

Whether or not the cybrid approach facilitates modeling of sporadic PD mitochondrial function is not simply an academic question. Although animal models based on rare Mendelian PD variants are increasingly relied upon by the PD research community, the ability to extrapolate findings from Mendelian PD model studies to sporadic PD may be limited. In Alzheimer’s disease and amyotrophic lateral sclerosis numerous therapies that showed efficacy in Mendelian disease models subsequently failed in human clinical trials, underscoring the importance of developing valid sporadic disease models (Swerdlow 2007b; Yamamoto et al. 2008). Our data argue that although the cybrid approach has limitations, it can facilitate the study of sporadic PD subject mitochondrial function and provide valuable insight into sporadic PD molecular physiology.


This work was supported by the Parkinson’s Disease Foundation of the Heartland and the Portuguese Foundation for Science and Technology.