Science Education

Supplement for Biologists: The Mitochondrial DNA Mutation Rate

Warning: Not for the general public.

I make an exception to my usual rule of writing to the curious lay public for this supplement. I recently wrote about the use of mitochondrial DNA (or mtDNA) in identifying bodies more than a decade old. While researching the topic, I found a lot of seemingly contradictory information about how often the mtDNA genome mutates. To clear the air, I asked a former colleague of mine, Dr. Michael Coble, who provided me with a very detailed explanation of why there seemed to be so much confusion in the literature.

If this topic interests you, read on.

I first asked Dr. Coble if the mutation rate of mtDNA was indeed higher than that of nuclear DNA, which codes for most of the proteins in our bodies.

“Some early papers by Brown et al. (1979) PNAS 76:1967-1971 and Horai et al. (1995) PNAS 92: 532-536 established the fact that mtDNA mutates at about 10 times the rate of single copy nuclear DNA genes. Now as always, there are exceptions – some of the nuclear genes for immunity can mutate rapidly, creating diversity to fight the numerous viral and bacterial critters attacking our bodies all the time.

“The higher mutation rate in mtDNA makes sense because of the environment of the molecule – one major function of the mitochondria is to provide energy for the cell, and this occurs in a harsh environment where oxygen radicals are created – leading to damaged DNA. We know that DNA repair occurs in mtDNA, but the mechanism is probably not as efficient as the systems in the nuclear genome, so mutations can arise quicker there.”

So mtDNA does mutate at a higher rate than nuclear DNA, but can we accurately measure this rate?

“The next part of your question deals with the actual rate of mutation, and this is where we had the debate in the mid-to-late 1990s. Early studies used a phylogenetic approach [an approach one could compare to using a telescope] and found that on average a mutation arose on a phylogenetic tree about once in every 300 generations or so. Other studies used a pedigree approach [comparable to using a microscope and thus observing the phenomenon on a much smaller scale] where they looked at a lot of mother-child events and this led to the discovery that the rate was on average about once in every 30-40 generations (about 10-fold higher than the phylogenetic rate). The Icelandic paper came along later and showed a rate of about 1 in 200 or so generations.

“I think that for the most part, there is an acknowledgment that depending on how you are measuring the rate, you will see a difference, so there really isn’t much controversy on this now. More research has revealed the site-specific mutation rate – we know that not all sites are created equal in the mtDNA genome, some sites mutate much more rapidly than others. This can often explain the “hot-spots” we often see as heteroplasmy in mtDNA sequence data (but not always).

“I should mention that we are actually talking about substitution rates. We now know from the recent efforts from massive parallel sequencing that mutations in the mtDNA are rampant!! However, not all mutations will be passed on to the next generation (that is, visualized as a difference between mother and child) – forces such as selection and genetic drift are working here, so when we see a difference we are looking at what has been substituted (changed) and not necessarily the overall mutation rate.”

So why is mtDNA generally not sufficient for DNA typing in human identification (HID) cases?

“If the two rates are not dissimilar, is the use of mtDNA limited in HID because of how small it is (fewer mutations can arise in the short HVR [hypervariable regions])? I think that mtDNA is always limited given that you are looking at a very small region of a molecule that is shared among many, many relatives (some known, and others in the population unknown). Mutation rates do play an important role in those cases where you see a single base difference between the evidence and the person of interest. For example, we know that hair is a semi-clonal tissue – we are constantly growing hair from a limited number of constantly dividing follicle cells (although some of us are seeing less of this occurring as we get older). It is not infrequent to see a difference (or at least heteroplasmy) in evidentiary hair and the suspect (Mitotyping Technologies, a private mtDNA lab published a paper a few years ago where they reported heteroplasmy in about 10% of all of their hair cases). Having an idea on the site-specific mutation rate can be useful in these cases to show that the heteroplasmy may be at a known hot-spot, for example. Many of the “Innocence Project” cases that are worked now rely on hair evidence collected 30+ years ago before DNA testing.”

Big thanks to Dr. Michael Coble from the National Institutes of Standards and Technology for clearing the air on this mtDNA mutation rate confusion.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your 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 )

Google+ photo

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

Connecting to %s