An increase in mitochondrial superoxide extends lifespan in C. elegans but not by hormesis?

Edit: I was forwarded a convincing skeptical argument that because C. elegans lacks dividing cells and DNA repair genes that function like higher organisms likely because it doesn’t live as long, it might have a greater protective response to superoxide. Longer-living organisms have cancer to worry about so cells may allow senescence to occur to freeze cells in a non-dividing state.

Rate of ageing theories have evolved significantly over the years but now frequently centralize around the involvement of the mitochondria.  Of recent (and seemingly on the way out) is the oxidative stress theory of ageing (formally the free radical theory of ageing), which is largely supported by correlational data.  Recent genetic manipulation studies on the endogenous antioxidant network has caste considerable doubt over this theory.

Yang and Hekimi discuss in a new paper in PLoS Biology that in the nematode C. elegans, the mutations in the genes mev-1 and gas-1 reduce lifespan and are associated with an increase in mitochondrial oxidative stress, and those in clk-1, isp-1, lrs-2, and nuo-6 extend lifespan and are associated with reduced oxidative stress; this seems to suggest that mitochondrial oxidative stress may be a candidate in the ageing process.  Additionally, knockdown of isp-1 and nuo-6 (encode subunits of mitochondrial respiratory complexes), among other genes by RNA interference (RNAi) also prolong lifespan, but independently of mutations as described above.  For example, a mutation in isp-1 lowers oxidative damage to proteins, results in an increased cytoplasmic SOD-1 (super oxide dismutase, an endogenous antioxidant enzyme) and mitochondrial SOD-2, and increased resistance to the prooxidant paraquat (an herbicide).  Paraquat is toxic at a high enough concentration, primarily at the mitochondria.  Interestingly however, knocking down SOD by RNAi, which returned oxidative damage to normal or higher levels has no effect on lifespan in these mutations, suggesting that a resistance to oxidative damage was not mediating the increase in lifespan. Hekimi and Lapointe reviewed a number of other studies and concluded earlier this year that the mitochondrial theory of oxidative stress has been thoroughly refuted.  So onto this new study, but first some of my own interjections:

Rethinking ROS; hormesis and health

Reactive oxygen species (ROS) are not just “toxic” products that are cells cannot tolerate, unlike most supplement and fruit juice salespeople will claim.  They are in fact important signaling molecules in the body and mediate many functions.  They seem to mediate adaptations to exercise for example, some studies suggest anti-inflammatory drugs or vitamin-antioxidant supplementation can prevent some of the beneficial adaptations to exercise (though both areas still require further study).  The theory of “hormesis,” that low level stressors provoke beneficial adaptations (this study however suggests that superoxide extends lifepan in a non hormetic fashion; read on) also extends into the calorie restriction field (the only consistent life extension method in the literature), to phytochemicals, and even pesticides and a number of other possible nutrients and dietary substances (though this seems fairly speculative at this point).  Then there is the epidemiological evidence that suggests supplementing certain vitamin-antioxidants, in general, may even increase mortality risk; this of course is not without controversy, especially of recent (there are other studies on this as well).

The point is, ROS have earned perhaps an undeserved reputation, and their relationship with our health and longevity is very complex.  They seem especially abused by nutritionists and non alike- food or product selections should not be made off claims based on out-dated and disproved theories.

New findings

Back to the new study- Yang and Hekimi looked at C. elegans with mutations that increase longevity in isp-1, nuo-6, clk-1, eat-2, daf-2 (interferes with insulin signaling), and sod-2 (enzyme that converts superoxide into its products) with a relatively new technique that they developed to measure ROS generation and levels in the living specimens.  They measured total mitochondrial ROS as well as a specific ROS, superoxide in the mutant models and found that mutant isp-1, nuo-6, and daf-2 had elevated superoxide generation (and lifespan; these models are known to have elevated mitoSOD though as well and no increase in oxidative damage) but not overall ROS.  All of these comparisons were made against a wild type model without enhanced lifespan.

NAC, an antioxidant that reacts with all ROS, was given to the worms; it had no effect on the longevity of the wild type model, but completely removed the life extension enjoyed by the nuo-6 mutant and significantly decreased that of the isp-1 mutant model.  Treatment with vitamin C (ascorbic acid; also an antioxidant) mirrored these results.  In daf-2 mutants, NAC treatment only had a small effect on lifespan, suggesting independent influences on longevity in this model.  In the sod-2 mutants, NAC also completely removed the increase in lifespan although superoxide generation or total ROS is not elevated based on their measurements, suggesting superoxide or other ROS does mediate their lifespan.  clk-1 mutants were not affected by NAC.

Next, they treated the models with the herbicide paraquat, which generates mitochondrial superoxide.  Paraquat is highly toxic at high concentrations (greater than 0.2 mM), but at a low concentration (0.1 mM) in this experiment, it increases oxidative damage in proteins (measured by protein carbonylation and increased SOD-1 & SOD-2) without killing the specimens.  Interestingly, they found that when the wild type worms were treated with paraquat at concentrations of 0.05 mM, 0.1 mM, and 0.2 mM, mean and maximum lifespan increased, the best results occurring at 0.1 mM (they speculate because the higher concentration likely started to become toxic).  They confirmed that this effect was not exclusive to paraquat by treating the wild types with benzyl-viologen as well, which is structurally different but also increases mitochondrial superoxide.  They note that a prooxidant called juglone is noted to do this in a previous study as well.  They found that paraquat extended lifespan if administered during adulthood or only during development.  Administering paraquat did not increase the lifespan of nuo-6 and isp-1 mutants, suggesting that mitochondrial superoxide is the common link between the experiments. It was determined that this was not because of a resistance to paraquat in the models because 0.2 mM paraquat shortened their lifespans.  Further, treatment of the sod-2 mutants with paraquat increased their lifespan to match the lifespans of the wild types treated with paraquat (suggesting no additive effect), but in the clk-1 and eat-2 mutants (recall that these mutations do not effect coding for mitochondrial complexes and do not have elevated superoxide) the combination of the mutation and paraquat extended the lifespans beyond what the wild type/paraquat specimens, or the mutants alone, enjoyed.  Since these phenotypes are influenced by other variables than just superoxide, they and superoxide are “mechanistically distinct and additive.” The daf-2 model has elevated superoxide and experienced a small increase in lifespan with paraquat treatment, suggestive that superoxide still mediates some of the effect in that model.

To be sure that superoxide was responsible for the lifespan extensions, they sought out evidence of other common changes on mitochondrial function brought on by the mutations and paraquat treatment.  Existing research has found that mitochondrial defects stimulates mitochondrial biogenesis, and indeed they found that in isp-1 and nuo-6 mutants this was the case, but in the wild types treated with paraquat this was not found.  Subsequent experiments eliminated oxygen consumption and ATP levels as they were different with various treatments.

Next, they explored some of the potential gene pathways that mediate superoxide’s effects.  In the daf-2 mutants, recall that they had elevated superoxide, yet still experience an (small) increase in lifespan with paraquat treatment, suggesting that the level of superoxide is insufficient for the maximal life extending effect with the mutation alone.  Other research has found that at least 3 other genes are needed to be expressed in the daf-2 model for the full lifespan extension: daf-16, aak-2, and hsf-1.  Knocking these out one at a time, the investigators found that paraquat treatment still prolonged the lifespan of each of these mutants, though not as much as the wild types, suggesting that at least part of the lifespan extension mediated by superoxide requires these genes.

Finally, they treated a variety of specimens with mutations that are required for some aspect of superoxide signaling with paraquat: jnk-1 (stress responses, + regulator of DAF-16) experienced a large increase in lifespan; skn-1 (mobilizes phase II detox response and protects against oxidative stress- can delay ageing) mutants still experienced an increase as well; wwp-1 (necessary for lifespan extension by calorie restriction) and hif-1 (implicated in protective pathways) too lived longer.  All of these experiments further confirm that the effect of superoxide on longevity is distinct from many other previously explored mechanisms, as increasing mitochondrial superoxide with paraquat is able to further extend the lifespans of the above mutants models who already experience an increase in lifespan from their mutations.

Perhaps most interesting is that these effects are likely not mediated through an elevated protection from the endogenous antioxidant enzymes SOD-1 and SOD-2, which when reduced by RNAi as discussed previously did not shorten lifespan in the isp-1 or nuo-6 mutants.  As best summed up by the authors:

…our observations suggest that the lifespan effect we observed is not hormetic, as neither superoxide-detoxifying enzymes, nor the regulatory factors that are involved in protection from oxidative stress, are crucially implicated.


In the discussion section, the authors make comparisons between superoxide potentially triggering signal transduction pathways that alter nuclear gene expression to another highly reactive molecule: nitric oxide.  They note that a number of other studies in C. elegans show sub-lethal but deleterious acute treatment of various compounds can prolong lifespan by a hormetic effect, but this study involved a longer treatment time with paraquat (though treatment only during development or only during adulthood still increased lifespan).

Importantly, the findings here establish superoxide signaling as independent of mechanisms that underly the lifespan extending effects of glucose restriction and calorie restriction, at least with the models they utilized.

In this study, there was no exploration of specific tissues to see if mitochondrial superoxide effects specific cell types or in all tissues; other ageing research shows that changes in specific cells may influence the whole phenotype of the organism.

The authors give their interpretation of this research in light of other findings in a new framework:

The oxidative stress theory of aging has been one of the most acknowledged theories of aging for the simple reason of the strikingly good correlation between the levels of oxidative stress and the aged phenotype [8]. A number of recent results in worms and in mice, however, have suggested that oxidative stress cannot be the cause of aging [24],[25],[26],[30]. Our findings suggest a conceptual framework for why oxidative stress and the aged phenotype are so tightly correlated [31]. In this model mitochondria, like the rest of the cell, sustain a variety of age-dependent insults (not only and not even principally from oxidative stress) that trigger an increase in superoxide, which acts as a signal that induces general protective and repair mechanisms. However, aging in most animals is clearly irreversible, indicating that the protective mechanisms, which must have evolved to control damage in young organisms, are unable to fully prevent the accumulation of age-related damage. Thus, as superoxide is a reactive molecule as well as a signal, and as age-dependent damage cannot be fully reversed, it is possible that at high ages the chronically elevated superoxide will participate in creating some of the damage itself. This could explain the strong tendency for aged animals to have high oxidative stress and high oxidative damage, although it does not imply that ROS cause aging or even that they are a major source of age-dependent damage. In this model, the nuo-6 and isp-1 mutations lead to increased longevity because they turn on the stress signal prematurely and thus slow down the entire process.

Of note: I’ve summarized a study previously on the blog that marks mitochondrial superoxide as a unifying cause of insulin resistance, which may in fact be a protective response by the body.  It is interesting to consider this new finding in context of that one, the clear negative consequences of prolonged insulin resistance, yet the paradoxical findings in the ACCORD trial discussed there which found an increase in mortality with intensive glucose lowering treatments.  There may appear to be a connection but I am hesitant to go there yet.

Relevance to nutrition?

There isn’t any direct relevant to nutrition at this point, but this seems to further suggest that (certain) antioxidant supplementation is unwarranted and may have negative consequences, especially if you have one of these favorable mutant genotypes.  Most antioxidant regimens seem to be based on the oversimplified theories that oxidative stress is a bad thing and we should sop up free radicals with antioxidants to fight ageing.  This is certainly wrong in the general sense.


In summary, through a series of elaborate experiments, Yang & Hekimi demonstrate that elevated mitochondrial superoxide is the mechanism that extends lifespan in the genetic mutation models in the nematode C. elegans: isp-1 and nuo-6, that superoxide can extend lifespan in non-mutants as demonstrated by treatment with paraquat and benzyl-viologen (both which increase mitochondrial superoxide), and that the life extension by superoxide is distinct from many other mutant models of ageing.  This adds an interesting twist into what relationship oxidative stress has with the ageing process; the classic theories that oxidative stress/free radicals are a primary cause of ageing do not hold up to recent evidence.  Precisely how superoxide is able to extend lifespan in C. elegans through non hormetic mechanisms which will need to be explored in future research.


Yang W, & Hekimi S (2010). A Mitochondrial Superoxide Signal Triggers Increased Longevity in Caenorhabditis elegans PLoS Biology : 10.1371/journal.pbio.1000566

  • Liochev

    The rumors that the oxidative stress theory of aging is dead are premature. I found it in good health.
    Stefan Liochev