News & Views | Published:

Cell biology

A clean energy programme

Naturevolume 444pages151152 (2006) | Download Citation


Mitochondria supply cells with energy, but in the process produce potentially damaging oxidants. It seems that a protein required to produce new mitochondria also protects against the resulting oxidative damage.

Climate scientists have taught us that the planet is getting warmer because of the intimate relationship between the production and consumption of energy and the consequent generation of toxic by-products. Cell biologists have long known that a similar struggle between product and by-product occurs continuously inside every cell. Ground zero for this intracellular battle is located in the mitochondria, tiny and evolutionarily ancient energy-producing organelles. Writing in Cell, Spiegelman and colleagues1 suggest that these structures might be able to teach modern cell biologists — and those concerned about global warming — a few lessons on handling the delicate balance between producing what we need, while not ruining what we have.

Much like any factory producing widgets, mitochondria consume carbon-based fuels. Their product is ATP, the energy currency of the cell. Nonetheless, just like factory smokestacks, mitochondria also release potentially harmful by-products into their environment. For mitochondria, these toxins come in the form of reactive oxygen species (ROS) that include superoxide and hydrogen peroxide. In turn, these oxidants can interact with other radical species or with transition metals to produce by-products that are even more damaging. To combat ROS production, the cell has evolved a number of sophisticated antioxidant defences, including enzymes such as superoxide dismutase to scavenge superoxide, as well as catalase and glutathione peroxidase to degrade hydrogen peroxide.

Spiegelman and colleagues1 show that treating cells with extra (exogenous) hydrogen peroxide stimulates the production of several antioxidant enzymes, including catalase, glutathione peroxidase and various forms of superoxide dismutase. This effect occurs through increased transcription, the initial step of gene expression, and it seems to require a nuclear protein called PGC-1α. Spiegelman and colleagues first described PGC-1α several years ago, and it is now clear that this protein belongs to a family of coactivators that includes PGC-1β and PRC (ref. 2). Coactivators are proteins that regulate gene transcription, usually in a tissue-specific fashion. Curiously, these factors don't actually bind to regulatory sequences in DNA. Instead, they often act as molecular scaffolds to turbo-charge gene expression. They do this by recruiting various DNA-modifying enzymes while simultaneously interacting with classical transcription factors that bind to DNA. Interest in PGC-1α escalated when it was realized that this particular coactivator has a major role in creating new mitochondria when they are needed by the cell, in a process known as mitochondrial biogenesis3.

In their latest experiments1, Spiegelman and colleagues show that simultaneously reducing the amounts of PGC-1α and PGC-1β essentially eliminates the cell's ability to increase its antioxidant defences in response to exogenous ROS. Their experiments suggest that although PGC-1β is involved in this response, most of the transcriptional drudgery is done by PGC-1α. Indeed, they show that cells from PGC-1α-deficient animals have increased basal levels of ROS and a reduced ability to withstand damage from these toxins. Similarly, overproduction of PGC-1α in neuronal progenitor cells allowed those cells to become more resistant to oxidative stress.

These in vitro results were mirrored by in vivo observations. For instance, treatment of mice lacking PGC-1α with two known inducers of neurological oxidative stress, MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) and kainic acid, caused significantly more neuronal damage than did similar treatment of control mice. In the 1980s, MPTP caused severe Parkinson's disease in a group of drug abusers who accidentally ingested this toxin. It is now known that in these unfortunate individuals MPTP was acting as a highly specific inhibitor of mitochondrial electron transport. Similarly, kainic acid treatment is widely used as a model of neuronal damage and epilepsy.

Together with previous work, these results suggest that PGC-1α can increase the number of mitochondria while also protecting the cell from subsequent mitochondrion-induced damage. In essence, it is as though PGC-1α is building the factory and starting the environmental clean-up at the same time. The observation that this protein can regulate the oxidative balance in cells might also provide insight into some unexplained traits observed in mice lacking PGC-1α. Such animals exhibit striking behavioural changes, including increased anxiety and hyperactivity4,5. Although these behavioural changes may relate to structural changes seen in the brains of PGC-1α-deficient mice, an independent group working on the genetic basis of anxiety has implicated ROS metabolism in this complex trait6. Are the behavioural changes in the PGC-1α-deficient mice also due to an absence of antioxidant defences? If so, does this imply that what cell biologists call oxidative stress and what social scientists call psychological stress might ultimately share a common mechanism? Further study is necessary, but people who fear public speaking may take some comfort in the notion that the problem might not be in their head but rather, more specifically, in their mitochondria.

The work of Spiegelman and colleagues1 brings up another interesting aspect of mitochondrial biology. It is well known that these organelles, whether they come from simple organisms or complex mammals, all leak measurable amounts of ROS. Experimental systems that simply increase the production of antioxidant proteins seem to be quite effective at reducing this leakage. If ROS synthesis is so bad, and a molecular solution so apparently straightforward, why has this 'design flaw' not been eradicated during the billions of years of evolution? There are many possible answers, but one is that the notion that ROS from the mitochondria are solely harmful could be incorrect. Indeed, substantial evidence exists that ROS generated in the cytoplasm could have vital signalling functions7, and this might also be true for oxidants derived from mitochondria8,9. The new study1 strengthens this possibility and suggests that a homeostatic loop exists between mitochondria and ROS and that this loop is, at least in part, orchestrated by PGC-1α (Fig. 1).

Figure 1: PGC-1α in mitochondrial biogenesis and antioxidant defence1,2,3.
Figure 1

Many external stimuli that signal an increase in the need for energy can activate PGC-1α. By acting on members of the nuclear-receptor superfamily — including NRF-1, NRF-2 and ERR-α — PGC-1α can induce mitochondrial biogenesis. The resulting production of reactive oxygen species (ROS; oxidative stress) may feed back through the CREB gene-regulatory factor, leading to further increases in PGC-1α. This oxidant-induced increase in PGC-1α is then necessary to stimulate expression of a host of antioxidant proteins, including forms of superoxide dismutase (SOD), glutathione peroxidase (GPx) and catalase.

Previous reports have suggested that certain life-extending strategies, such as calorie restriction, might work through the PGC-1α-induced mitochondrial biogenesis programme10. Spiegelman and colleagues' study suggests that PGC-1α could also alter our susceptibility to neurodegenerative conditions that are linked to mitochondrial dysfunction and oxidative stress, such as Parkinson's disease. Therefore, fine-tuning the activity of this resourceful coactivator might have a wide range of clinical benefits, including potentially allowing us to live longer and think more clearly. Not a bad set of objectives, especially if we are ultimately going to need to tackle really tricky problems like global warming.


  1. 1

    St-Pierre, J. et al. Cell 127, 397–408 (2006).

  2. 2

    Finck, B. N. & Kelly, D. P. J. Clin. Invest. 116, 615–622 (2006).

  3. 3

    Puigserver, P. et al. Cell 92, 829–839 (1998).

  4. 4

    Lin, J. et al. Cell 119, 121–135 (2004).

  5. 5

    Leone, T. C. et al. PLoS Biol. 3, e101 (2005).

  6. 6

    Hovatta, I. et al. Nature 438, 662–666 (2005).

  7. 7

    Finkel, T. Curr. Opin. Cell Biol. 15, 247–254 (2003).

  8. 8

    Nemoto, S., Takeda, K., Yu, Z. X., Ferrans, V. J. & Finkel, T. Mol. Cell Biol. 20, 7311–7318 (2000).

  9. 9

    Werner, E. & Werb, Z. J. Cell Biol. 158, 357–368 (2002).

  10. 10

    Lopez-Lluch, G. et al. Proc. Natl Acad. Sci. USA 103, 1768–1773 (2006).

Download references

Author information


  1. Cardiology Branch, NHLBI, National Institutes of Health, 10/CRC 5-3330, 10 Center Drive, Bethesda, 20892, Maryland, USA

    • Toren Finkel


  1. Search for Toren Finkel in:

About this article

Publication history


Issue Date


Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing