How Long Can DNA Survive?

An international team of scientists takes issue with recent claims that ancient DNA has been found in ice, amber, salt or rock many millions of years old (see 05/23/2002 entry, for instance).  They think such cases are due to contamination and have not been independently replicated.  They gathered samples in Siberian and Antarctic permafrost under ideal conditions for preservation, and found that DNA would become unrecognizable after millions of years due to increasing numbers of interstrand crosslinks.  Their report, published in the Jan. 6 issue of Current Biology,¹ puts an upper limit at 400,000 years on the durability of DNA.
    Unexpectedly, they found that non-spore-forming gram-positive Actinobacteria seems to survive the longest.  This was surprising, because endospores were assumed to be the hardiest of all cell types.  An endospore has no DNA-repair activity, so maybe that’s the reason.  They offer some other possibilities:

The mechanism behind the superior persistence of DNA from the non-spore-forming-GP Actinobacteria is currently unknown.  Slow but continuous metabolic activity and DNA repair at subzero temperatures is one possibility.  Adaptations connected to dormancy might be another explanation.  Finally, the DNA could simply originate from dead bacteria, whose DNA for some reason, e.g., structural features, survived better.

Willerslev et al., “Long-term persistence of bacterial DNA,” Current Biology, Vol 14, R9-R10, 6 January 2004.

Do any of those possibilities sound plausible to you?  Actually, DNA probably degrades much faster than they estimate.  There’s no way they could verify the 400,000 year figure; it’s a pure guess, based on certain assumptions.  At most, it is just an upper limit.  If DNA could survive the onslaughts of nature for so long, it would speak of designed mechanisms to fight the degrading pressure of entropy.