The ability to test mitochondrial bioenergetic function in a high-throughput and small scale manner will provide researchers with the ability to more rapidly conduct their experiments, allow for increased sample sizes and have tissue remaining for further experimentation. Additionally, this method requires significantly less mitochondrial protein such that different regions of the brain from the same animal can be tested individually alleviating the need to pool samples from multiple animals. In the present studies we have outlined a method which utilizes the Seahorse Biosciences XF24 flux analyzer for the profiling of brain mitochondrial bioenergetic function. We have utilized this method in order to compare differences in brain mitochondrial function and the susceptibility to mitochondrial inhibitors across brain regions. The results of these experiments demonstrate that our novel protocols for utilizing the XF24 can provide consistent results and allow for the simultaneous comparison of multiple samples.
An important factor when setting up the Seahorse Bioscience XF24 for use with isolated mitochondria was to determine the optimal amount of mitochondrial protein to use per well for a given experimental paradigm. For the present studies, determination of the optimal protein concentration was based upon determining the highest measurable rates since this would provide a larger range of oxygen consumption values to distinguish group differences. Previous work utilizing liver mitochondria at a concentration of 2.5 μg/well has shown the ability to measure oxygen consumption with the XF24, however, no comparison to other concentrations was made nor was brain mitochondria utilized (Gerencser et al., 2009
). Comparing the use of 2.5 μg or 5.0 μg of purified brain mitochondria revealed that the 5.0 μg amount provided optimal rates of oxygen consumption based upon our desired characteristics. The higher rate of Complex I dependent oxygen consumption () provides a greater measurable range for detecting group differences in future experiments. Additionally, it is important to know whether the XF24 is able to replenish the oxygen tension during an experiment so that there is not a progressive decrease in the starting oxygen tension for subsequent measurements. There was a slight drop in oxygen tension with both protein concentrations once mitochondrial respiration was started, however, comparison of 2.5 μg and 5.0 μg revealed that the XF24 was capable of replenishing the oxygen tension equally with both amounts of mitochondrial protein (). Comparison of measured rates of mitochondrial oxygen consumption between samples tested with a Clark-type oxygen electrode and samples tested with the XF24 revealed that both methodologies produce similar rates. Previously published data with the Clark-type oxygen electrode indicate that the cortex exhibits rates of oxygen consumption of approximately 300 pM O2
/min/μg of protein (Sullivan et al., 2004
). The present experiments with the Seahorse XF24 indicate that the cortex has a measured rate of 328.8 ± 64.69 pM O2
/min/μg of protein which matches our previously published results. Similar to the findings in the cortex, the present studies indicate that the striatum had measured rates of oxygen consumption of 352.4 ± 64.8 pM O2
/min/μg of protein and previous studies from our group have shown rates of approximately 350 pM O2
/min/μg of protein using the Clark-type oxygen electrode (Korde et al., 2005
), further indicating that the two methods can produce equivalent findings. Even though the two techniques provide similar measured rates of oxygen consumption, the Seahorse XF24 is limited by only being able to inject four different substrates during a single experiment. The initial characterization experiments reveal that 5.0 μg provides better experimental conditions since a higher starting oxygen consumption rate will provide greater experimental resolution. Additionally, since previous work has stated that RCRs greater than five indicate healthy mitochondria (Deng-Bryant et al., 2008
) and studies looking at various insult models show that only the injured samples have RCRs less than five (Gash et al., 2008
; Pandya et al., 2009
), it was concluded that the samples utilized in these experiments were all healthy and well-coupled.
Previous studies have shown that regional differences in the activities of mitochondrial enzymes such as malate dehydrogenase and creatine kinase exist (Ryder, 1980
; Gupta et al., 2000
). Additionally, neuronal activity has been shown to play a direct role in the expression of mitochondrial proteins (Wong-Riley et al., 1997
) and reductions in synaptic activity can lead to reduced cytochrome-C
levels (Nie and Wong-Riley, 1996
). Additionally, ischemic injury experiments have shown regionally different mitochondrial susceptibility to the injury (Sims, 1991
) which could be the result of mitochondrial or cellular differences. Multiple groups have further established that in addition to brain region differences in mitochondrial function there is a significant difference between synaptic and non-synaptic mitochondria (Davey et al., 1997
; Brown et al., 2006
; Naga et al., 2007
; Pathak and Davey, 2008
). Numerous factors are likely to play role in the differential susceptibility of mitochondria such as which region of the brain the mitochondria are from (Singh et al., 2010
), which portion of the CNS the mitochondria are from (Sullivan et al., 2004
), or the energy demands of the cell containing the mitochondria (Zeevalk et al., 1997
In order to further elucidate the role of the mitochondria in regionally specific toxin susceptibility, analysis of basal mitochondrial function was performed across various regions of the brain. We hypothesized that regions which appear more susceptible to Complex I or Complex II inhibition in vivo
may have reduced basal mitochondrial function. The results indicate that the greatest differences in Complex I and Complex II dependent oxygen consumption exist between the cerebellum and the striatum (). The finding of reduced basal oxygen consumption in the cerebellum compared to other brain regions has not been shown before and therefore the underlying reason for this effect is currently unknown. Previous work has shown a reduced activity of Complex II enzyme activity in the cerebellum (Fagundes et al., 2007
) and this could be an underlying reason for the reduced Complex II dependent oxygen consumption observed in these studies. It has recently been shown that resting aerobic glycolysis is lower in the cerebellum than other brain regions (Vaishnavi et al., 2010
) and this may be the result of different ratios of neuronal and non-neuronal cells in the cerebellum (Azevedo et al., 2009
). Interactions between neurons and astrocytes during in vivo
oxygen consumption has been shown to occur (Kasischke et al., 2004
) with the two cell types providing different roles during energy metabolism (for review see (Magistretti, 2006
)). Given the higher percentage of neurons to astrocytes in the cerebellum, the regional differences in the mitochondrial respiration observed in the present studies may be affected by different cell type ratios.
Basal Complex I and Complex II dependent mitochondrial oxygen consumption did not differ between the cortex, striatum, or hippocampus in these experiments which indicates that mitochondria from these regions have similar bioenergetic capacities and profiles. Interestingly, the cerebellum had a significantly slower rate of oxygen consumption in the presence of the Complex-V inhibitor oligomycin-A compared to the striatum and hippocampus (). The difference in oxygen consumption in the presence of oligomycin-A may be the result of lower levels of glutathione and superoxide dismutase in the striatum and hippocampus relative the cerebellum (Sanchez-Iglesias et al., 2009
) which could possibly allow for increased damage to the inner mitochondrial membrane and increased proton leak across the damaged membrane. The profiling of basal mitochondrial function further indicates that mitochondrial health and integrity is not detrimentally affected by analysis with the XF24. Calculation of the mitochondrial respiratory control ratio revealed extremely high levels of coupling between the pumping of protons across the inner membrane with the formation of ATP by Complex V (). These experiments indicate that not only is the XF24 capable of measuring mitochondrial function with only 5 μg of protein but also it can be done without damaging the mitochondria.
Differences in mitochondrial spare respiratory capacity have been documented which may provide insight into the regional heterogeneity of in vivo
mitochondrial toxin models. Previous reports utilizing the toxin trichloroethylene (Gash et al., 2008
; Liu et al., 2010
) or 3-nitroproionic acid (3-NP) (Beal et al., 1993
) have shown that different regions of the brain do appear more susceptible than other regions. Mirandola et al., 2010 showed that following in vivo
systemic exposure to the mitochondrial Complex II inhibitor 3-NP there are regional differences in the susceptibility to mitochondrial permeability transition in mitochondria isolated from the brain. However, in isolated mitochondria in vitro
treatment with 3-NP affected mitochondria equally across the same brain regions. It has been suggested that one of the major reasons certain brain regions are more susceptible than others to a particular toxin is not the mitochondria itself but the environment the cells are located in (Mirandola et al., 2010
). Analysis of the susceptibility of isolated mitochondria to Complex I or Complex II inhibition indicates that there are no differences between brain regions with regard to the effect these toxins have on enzyme activity ().
This finding is in agreement with results looking at Complex II activity and the susceptibility to 3-NP where the cortex, striatum, and cerebellum had similar basal enzyme activities and susceptibilities to inhibition (Mirandola et al., 2010
). It has previously been shown that the percentage inhibition of mitochondrial Complex I activity is correlated with the percentage reduction in mitochondrial oxygen consumption (Davey et al., 1998
). The current results support this previous finding since the dose dependent reduction in mitochondrial oxygen consumption in the presence of the Complex I inhibitor Rotenone () is similar to of the reduction in Complex I enzyme activity (). The present study further extends this effect to Complex II dependent oxygen consumption since the effect of the inhibitor malonate had a nearly identical concentration dependent effect on oxygen consumption () as it did on enzyme activity (). These results further indicate that at low concentrations of the inhibitor rotenone, differences in the percentage reduction of oxygen consumption does differ between brain regions (). These results further support the ability of the XF24 to measure small changes in oxygen consumption since significant and consistent changes in the dose dependent effect of the inhibitors can be reproducibly measured.
Previous unpublished experiments from our group indicated that a time dependent effect of rotenone on Complex I dependent mitochondrial function may exist at low concentrations of the inhibitor. To test if time of exposure was a factor, repetitive measurements were taken of mitochondrial samples treated with either 0, 10 fM, 10 pM, or 10 nM rotenone. As shown in our previous experiments (), the 10 nM rotenone produced an immediate significant reduction in oxygen consumption while the 10 pM rotenone did not produce a significant reduction in oxygen consumption (). Strikingly, continued exposure to the 10 pM rotenone produced a clear time dependent reduction in Complex I dependent oxygen consumption which became significant by 12 min post injection and reached a 75% reduction in oxygen consumption by 16 min (). This is an important finding to consider when attempting to understand human toxin exposure since chronic exposure to low doses of a toxin may result in a greater inhibition of mitochondrial function.
Fig. 5 Exposure to 10 pM rotenone produces a progressive decrease in mitochondrial oxygen consumption. Samples which were exposed to 10 nM rotenone produced an immediate reduction in oxygen consumption similar to previous finding. Similarly, exposure to 10 fM (more ...)