What is a genetic mutation, exactly?

My understanding of the process of evolution is that it is dependent on random genetic mutations that give an organism a relative survival advantage over other organisms. Fair enough, but why do they happen? Aren’t mutations basically an error in the copying of information? What would cause such an error? They’re obviously rare, since if they happened too often we wouldn’t have life. But if they didn’t happen at all, we’d have no evolution.

Mutations are not rare. Probably every gene in your body has some mutations in it. Most are harmless or neutral or lead to the same result. The gene that causes lactose tolerance - i.e. that never sends out the signal to stop production of the lactase enzyme so you can drink milk without symptoms as an adult - has at least 43 known variations.

That’s why there have been numerous articles recently saying that human evolution has speeded up in modern times (the last 10,000 years).

This amazing video is supposedly the best simulation of how DNA works in real time inside a cell. It shows how easily mistakes can be made.

Great video, Exapno Mapcase.

I think what the OP is asking, though, is what causes those mistakes to happen so easily?

This is a cool question, so I look forward to seeing what kind of responses it gets. I had always assumed that the errors were usually caused by external factors (e.g., temperature changes, etc.).

Here’s an interesting cite that might provide some helpful information:


From the cite:

Mutations in DNA sequences generally occur through one of two processes:
[li] DNA damage from environmental agents such as ultraviolet light (sunshine), nuclear radiation or certain chemicals[/li][li] Mistakes that occur when a cell copies its DNA in preparation for cell division.[/list][/li][/quote]

Not many things in nature are perfect. I believe there is actually quite a bit of activity that goes on to correct errors in replication, it’s probably pretty difficult to make it run as error-free as it all does.

Truly fascinating video. I’ve bookmarked that page. LilShieste’s link gets to my question.

What’s interesting is that evolution allows life to propagate in a changing environment, while the environment helps to cause the mutations that allow life to propagate.

There are a huge number of ways for an error to arise, and the vast overwhelming majority of them are correctly repaired by your cellular machinery. The mutation rate (that is, those that don’t get repaired), is around 1 per 10^8-10^9 nucleotides. There are balancing forces at work here. It’s good for a population to have new sequences arise, but they’re nearly always bad for the individual. So each of us is wanting someone else to play around with their genome, as long as we don’t have to do it. Then we can just mate with anyone with good genes and get them that way.

Anyway, you can have all kinds of spontaneous chemical changes. Photons and various chemicals are always smashing bits and pieces off your molecules, and some of those changes can be misread by your replication machinery if they’re not detected. There are whole classes of proteins whose whole job in life is to scan up and down your DNA, feeling for bulges caused by mismatches. Your replication machinery is also not perfect. It can slip, leading to duplications or deletions, especially in highly repetitive sequences. And it can occasionally stick the wrong nucleotide in. Again, the vast majority of the time, the machinery will detect a mistake, back up, chew up the messed-up region, and do it again, right. It’s truly a phenomenal thing, IMHO.

It may also be worth mentioning that the vast majority of mutations are perfectly harmless. As has been mentioned, you have a bunch in every cell, and don’t notice them (if you’re wondering how this is, given how accurate I just said the replication process is, well, it’s because you have a LOT of DNA). I’ll skip explaining why that is on a molecular level, as that would take a while, and just say that a copy of Moby Dick with the word “the” changed to “teh” is still Moby Dick. And even if you do have a significant mutation, unless it’s in a germline cell, and unless that cell happens to get to make a new baby, it will never really get a chance to be selected for or against. That is, if you have a liver cell that happens to pick up a super-hearing gene, well, too bad.

I teach Molecular Biology, and 3 or 4 days on DNA mutations and repair. Below is a very brief summary.

There are several types of genetic changes. The DNA can be altered by

  1. DNA damage, chromosome breakage - DNA can be damaged by mutagens, hydrolysis, deamination, alkylation, oxidation, radiation, base analogs, intercalating agents. Chromosomes can break as well, and switch tips.

  2. Mistakes during copying - tautomers are often mis-incorporated into the DNA, ones in the enol and imino form. Also, DNA polymerase can slip and lead to the addition or deletion of repeat regions. These actually happen 100,000 times more frequently than point mutations, and there is evidence these slippage events can lead to fast phenotypic changes (and adaptation) in organisms.

  3. Transposons, retrotransposons, unequal crossing over - these can lead to entire genes being duplicated.

  4. Histone/nucleosome modification (not technically altering the DNA) - histone modification can lead to parts of the genome being activated or silenced. They are the master control switches that allow access to the genes. In some cancer cells abnormal modifications have been observed.
    Statistically, there are errors constantly popping up, and quickly fixed by cellular enzymes. But some still make it through, and the site here http://www-personal.ksu.edu/~bethmont/mutdes.html states that there is about a 1 in 10 chance that each new zygote has a new mutation not contained by either parent.

There are many systems that seek out and find mutations. Tautomers can be degraded, DNA polymerase has a proofreading domain, there are entire mismatch repair systems in place, base excision, nucleotide excision, homologous recombination repair, nonhomologous end joining, and translesion repair.

Thanks for the answers. Way better than google!