Self-Mutilating DNA

I already wrote about the research of my friend Olga Amosova who studied the sickle-cell anemia mutation. She and her colleagues needed to store short fragments of hemoglobin genes for their experiments. All the fragments were identical. They noticed that with time the fragments always broke down in the same place. It was a mystery. When good scientists stumble on a mystery, they start digging.

They found that one of the nucleotides rips off the DNA fragment at the site of the Sickle-cell mutation. That place on the DNA becomes fragile and later breaks down. These sites need to be repaired. The repair is very error-prone and often leads to a mutation.

When DNA strands are left unattended, they want to pair up. There are four types of nucleotides: A, C, G and T. So mathematically the fragment of DNA is a string in the alphabet A, C, G, T. These nucleotides are matched to each other. When two DNA strands pair up, A on one strand always matches T and C matches G. So it is logical that if there are two complementary DNA pieces on the same fragment, they will find each other and pair up. They form a hydrogen bond. For example, a piece AACGT matches perfectly another piece TTGCA. Suppose a substring of DNA consists of a piece AACGT and somewhere later the reverse of the match: ACGTT. Such a string is called an inverted repeat. The DNA fragment I mentioned contains a string AACGT****ACGTT. Two pieces AACGT and ACGTT are complementary and not too far from each other in space. So it is easy for them to find each other and to bond to form a so-called stem-loop or a hairpin structure. The site of Sickle-cell mutation falls into the loop.

Olga and her colleagues discovered that for some particular loops the orientation in space becomes awkward and one of the nucleotides rips off. Such a rip off is called depurination. In further investigation, Olga found examples of when depurination happens. The first sequence of the pair that will bond later has to have at least five nucleotides and has to end in T. Correspondingly the second part in the pair has to begin with A. In the middle there needs to be four nucleotides GTGG. The first G flies away. Enzymes rush like a first aid squad to repair it and introduce mistakes that lead to mutation and diseases like cancer.

DNA was thought to be simply a passive information storage system, not capable of any action. Now we see that DNA is capable of action. DNA can damage itself. Damage provokes a mutation. For all practical purposes it is self-mutilation. Olga and her colleagues scanned the human genome for other sequences that are capable of self-mutilation. They found that such sequences are overwhelmingly present. They are present in much higher numbers than would be expected statistically. The pieces that are capable of damaging themselves occur 40 times more often than would occur if the nucleotides were distributed randomly. They are especially overrepresented in genes linked to cancer.

Self-damaging shouldn’t happen in normal situations. It can be provoked by the environment, for example, the chemistry of the cell. That means, that our cancers are not only in our genes but also in our life-style. There was, for example, a suggestion in a recent NY Times article, Is Sugar Toxic?, that too much sugar in a diet might provoke cancer. If the rate of mutation depends on the environment, we can influence it and prolong our lives.

It is not clear why the ability to self-mutilate survives in the evolutionary process. It is quite possible that if something very bad happens to our planet, we need our genes to be able to mutate very fast in order to adjust to the environment so that humans can survive.

Though I never tried to donate my sperm to a sperm bank, because of my inability to produce it, I know that sperm banks look for people who have ancestors who lived for a very long time. Such sperm is in bigger demand as everyone wants their children to live longer. I wonder if this tendency is a mistake. Global warming is upon us. People with longevity genes might not be flexible enough for their children to survive the changing of the Earth.



  1. Peter Gerdes:

    In my experience the biggest barrier to entry as a sperm donor is the ability to produce sufficent sperm that survives the freezing process. In particular you are instructed to go without orgasm for the week or two between donations so that your sample will contain enough sperm to make it worth their while. I wasn’t serious enough about spreading my genes nor being offered enough cash to bother following those directions and for that (or perhaps naturally low/poorly frozen sperm) I never became a donor.

    However, I see no reason to believe that longevity genes correlate with inflexibility. Just the opposite. One would suspect that mutations which decrease cancer, heart attack, or alzheimers risks would likely grant increased protection against environmental factors that might exacerbate those diseases (diet changes, enviornmental toxins etc..). You might as well say smart people won’t be flexible enough. It makes about as much sense.

    Yes, it is true that certain DNA sequences (like genes for sickle cell) both cause harm in modern western society but grant a benefit in more perilous times and places. However, in general conditioning on the fact that your genes make you more healthy in our situation should increase the probability those genes would also be a boon to health in other circumstances.

    They are especially overrepresented in genes linked to cancer.

    Ohh c’mon you should know this means nothing. OF COURSE they are overrepresented in genes linked to cancer. After all these sequences are selected because they are especially likely to mutate and we know that mutations are necessery to give rise to cancer so if some segments are especially easy to mutate we should expect them to show up in cancers more frequently than the harder to mutate segments.


    As far as evolutionary adaptability goes I’m quite skeptical. We seem to have a much more nuanced and more powerful system in the form of promoters and inhibitors. Of course nature will use whatever happens to work but I suspect the overabundance may be the result of other factors like the incorporation of viral DNA (which often comes in loops) or some sort of symmetrical effect giving rise to such regions.

    Besides, adapting to global warming won’t be done in evolutionary time. It will be done in technological time. This may mean the people who get the most guns and hold out in the more temperate regions or it may be worldwide cooperation but it won’t be about mutations, the time scale is just too short.

  2. Dranorter:

    I definitely think they could be there for adaptability reasons. We have been through enough disasters in the past that we should have ‘evolved to evolve’, and I know other living things have been shown to have mechanisms for producing extra mutations before. Mutations are more powerful than promoters and inhibitors since they can add up to create totally new behavior.

    Any time a species undergoes a moderate amount of selection for some favorable trait, diversity and ability to adapt decrease, and take time to be built back up. Completely random traits which happened to co-occur with the useful mutation will be spread to a large part of the population, sometimes eliminating variations which were useful but only slightly useful compared with the favored allele. So sperm banks playing games with unnaturally increasing selection inevitably decrease adaptability, but I don’t think it’s prevalent enough to cause a problem.

  3. Arnett:

    Fascinating, Read something by Anne Fettner in Viruses Agents of Change ’91 perhaps, talking briefly about bits and pieces tagging along here and there. Fascinating research on Self Mutating DNA! Thanks for this.

Leave a comment