So this paper came across my desk this weekend (yes it was published a month ago so I’m a bit late to the game here, shoot me). I’m having a hard time believing the central result isn’t due to a simple artifact.

The paper describes the ability of α-ketoglurate, an intermediate in the Krebs’ cycle, to extend lifespan in C. elegans. The authors let the worms swim around in a solution of 8 mM α-KG and they lived longer.  There are some interesting knockout experiments showing the importance of various downstream mediators (mTOR, ATPsynthase) in the effects, but the key problem here seems to be that the authors may have overlooked a potential alternative mechanism…

α-KETOGLUTARATE IS AN ANTIOXIDANT

Yeah, so side-stepping the issue that just about every small molecule has at some point in the past few years been tagged as an antioxidant, there might actually be some decent proof here, in the form of a chemical mechanism. You see, a few years ago I reviewed (and was subsequently asked to write an editorial on) a paper in Free Radical Biology & Medicine, showing that α-keto acids (e.g., α-KG, pyruvate, oxaloacetate) can be non-enzymatically decarboxylated in free solution by reaction with hydrogen peroxide.  The reaction is as follows:

α-KG + H₂O₂ → succinate + CO₂

As outlined in our editorial, what this amounts to is a short-cut in the TCA cycle… you can get from α-KG to succinate without α-KGDH, so if the enzyme is inhibited (coincidentally this may occur due to oxidative stress such as H₂O₂), there’s a way to bypass the inhibition and allow the cycle to keep turning. Furthermore, this property of α-keto acids is well known in experimental circles… if you’ve ever made up a solution of pyruvate you know it goes off pretty quickly. That’s oxidative decarboxylation in action (the product in this case would be acetate). We even showed that this reaction (oxaloacetate → malonate) could occur in perfused hearts and may play a role in ischemic preconditioning. So, it happens in real mammalian tissues and is not just a theoretical test-tube reaction of the type often wheeled out to show that something is a biologically relevant antioxidant.

So, overall it just strikes me as not a very big deal that adding 8 mM of a known antioxidant to a plate of worms has an effect on lifespan.

The dose response shows an effect on lifespan in the 1-10 mM range with an EC₅₀ of ~3 mM, which is right in line with a generic non-enzymatic effect, and not really compatible with the K_M values for most of the enzymes that handle these metabolites (typical K_M for α-KGDH is 150 μM in the presence of Ca²⁺, only going into the mM range in the complete absence of Ca²⁺ which never happens in-vivo). Moreover, Fig. 2 of the paper shows that many of the effects which are proposed to account for the lifespan phenotype at the whole organism level occur at concentrations of α-KG in the 100-200 μM range, i.e. 5-fold lower than the level used in live worms). Oh, and it’s probably worth mentioning that some of the proposed downstream mechanisms might be sensitive to oxidative stress.

Given all the kerfuffle in recent years about TCA cycle intermediates as hot sexy new signaling molecules, it’s hardly surprising that this study made it into Nature. However, it’s frustrating that some simple controls were not included… (i) Do other α-keto acids prolong lifespan? (ii) Does adding α-KG to media and letting it sit for a few days at room temp’ on the shelf diminish the effect? (ii) Are lifespan-extending effects of other antioxidants additive to the effect of α-KG? (iv) Are effects of α-KG greater in worms known to generate more H₂O₂?

I left a comment on this at PubMedCommons, and also at PubPeer. The latter notifies the authors automatically, so hopefully we can get some feedback on this.