Skip to content

On spotting a very odd claim, and poking it

November 8, 2012

I was recently engaged in a conversation about the ballistic properties of Justin Beiber when a friend mentioned that he’d heard someone claim that “DNA lasts for 521 years in solid bone”. This claim, I thought, was a much finer topic of conversation than whether it was acceptable to model the world’s latest pop idol as a sphere of uniform density, so I decided to find out where on earth such a bizarrely specific testable claim had come from.

Mammoth M25

One of the mammoths with a sequenced genome has the museum designation M25. This is what you see if you Google “Mammoth M25”. I presume it to be a dramatisation.

On its face, the 521 years claim seemed pretty unlikely. After all, I knew that wooly mammoth genomes, albeit a little damaged, have been gaining people Nature papers in recent years, and those are in the 18,500-year ballpark. I also knew of the T. rex soft tissue that has recently been extracted – not DNA, sure, but something with enough protein that we could probably infer DNA sequences from it [incidentally, watch that TED talk about T. rex soft tissue. It is fantastic].

So I wanted to take a look at the data (if there were any) and see whether there was actually any support for the “DNA lasts 521 years” claim. If there were any good data, I wanted to know what was meant by “lasts”, and get a handle on which DNA lasts that long, and why.

The advantage of having a very specific claim like “521 years” is that if there is a real scientific source for it, you’ll be able to find it with tools like Google Scholar. So I very easily found the original research paper and its figure of 521 years. The paper, “The half-life of DNA in bone: measuring decay kinetics in 158 dated fossils” is the work of a 14-author collaboration across Australia, New Zealand, Denmark, the UK, Portugal, and China.

Moa left tibiotarsus

Moa left tibiotarsus. Source: Wikipedia (red box by author)

The authors took DNA samples from the left tibiotarsi of 158 different moa specimens collected from a combination of three sites within a five-kilometre radius of each other in North Canterbury in the South Island of New Zealand. The sites were limestone caves with average internal temperatures around 13.1 °C. For each tibiotarsus, they obtained a radiocarbon date to establish when the bird died. They also noted how long each tibiotarsus had been in storage for before DNA was extracted from it – some were freshly dug, while others had been held in museums for as long as 70 years.

The authors then tested their samples to see how badly degraded they were. The samples were analysed for DNA degradation in two ways: first, they used qPCR on a 242 bp mtDNA fragment. This technique estimates the amount of degradation by counting the number of intact copies of a known chunk of DNA there are in a sample.

Secondly, two tibiotarsi were selected for ‘shotgun sequencing‘ to see if breaks in DNA happen at random locations in the DNA molecule. This was important because the authors used a specific chunk of DNA to estimate overall degradation rates for all an animal’s DNA. If breaks don’t occur randomly – if, say, breaks happen more often near the ends of chromosomes and the chunk they looked at was near the end of a chromosome – they could seriously mess up their estimate of the DNA degradation rate.

What they found was that breaks in DNA appear to happen randomly along the length of the DNA molecule, so it is probably OK to use the degradation they measured to estimate the overall degradation rate. And, based on the number of intact chunks they managed to get from their 158 known-age moa tibiotarsi and the length of the chunk they were looking for, the authors estimated a 521-year half-life for 242 bp-long chunks of DNA. Because DNA breaks randomly, it is very easy to take those numbers and work out the probability that any given intact link in their samples would break in any given year. Which they did – giving us the ability to say that for their moa samples, DNA degraded at a rate of 0.0000055 degradations per intact base per year.

So, where does that leave us on the question of how long DNA lasts?

The short answer is ‘at the start of a long process’. We now know we can characterise DNA degradation in terms of its half-life, and we have a case study to give us a feel for how fast we might expect DNA to degrade under known conditions.

Lab freezer

“…what about in my lab freezer?”

From here, we can start asking questions like “What happens if, instead of 13.1 °C like the moa samples, we had a bone in 30 °C tropics, or -10 °C permafrost?”, “What about in an acidic environment, not the limestone-buffered caves that the moa were in?”, “What if the bones were smaller?”, and “How is the rate different for plants, and is it different for woody plants versus fleshy plants?”. The authors of the moa study make estimates of the effects of temperature on decay rates, which they base on extrapolation from lab-bench studies by other authors, who looked at the rate of DNA decay between 45 °C and 80 °C. As a staunch empiricist, I would be a lot happier if questions about the decay rate of DNA in cold bones were answered by looking at DNA in cold bones, rather than in solution on a hot lab bench – but I’ll have to wait for that research.

So, where does that leave us on the question of how long DNA lasts?

Not for “521 years in bone” without giving context, because that is simplifying so much that the message loses any resemblance to the truth. Not even “it has a half-life of 521 years”, because there are a host of other variables that will affect its decay rate. It might be best to stick with “it depends”. As with so much in science.


From → Science

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: