A few years ago, Tomasetti and Vogelstein (2015) published a paper where they noted a correlation between rates of cancer and the number of cell divisions. They concluded that a lot of cancers could be attributed to bad luck. This conclusion didn't sit well with most people for two reasons. (1) There are many well-known environmental effects that increase cancer rates (e.g. smoking, radiation), and (2) there's a widespread belief that you can significantly reduce your chances of getting cancer by "healthy living" (whatever that is). The first objection is based on solid scientific evidence but the second one is not as scientific.Some of the objections to the original Tomasetti and Vogelstein paper were based on the mathematical models they used to reach their conclusions. The authors have now followed up on their original study with more data. The paper appears in the March 24, 2017 issue of Science (Tomasetti and Vogelstein, 2017). If you're interested in the debate over "bad luck" you should read the accompanying review by Nowak and Waclaw (2017). They conclude that the math is sound and many cancer-causing mutations are, in fact, due to chance mutations in somatic cells. They point out something that should be obvious but bears repeating.
An understanding of cancer risk that did not take bad luck into account would be as inappropriate as one that did not take environmental or hereditary factors into account...This brings me to the topic for today. What is the somatic cell mutation rate?
Cancer is a by-product of the fact that we are made of cells that are individual replicators. Mutations destroy their cooperative program and elicit unwanted replication (defection). Mutants arise whenever cells divide. These normal mutations are due to "bad luck."
I did a series of posts on the human mutation rate a few years ago. I calculated the average number of mutations per generation using a standard error rate of DNA replication of 10-10 per base. This means approximately 0.32 mutations every time the haploid genome is replicated (genome size = 3.2 × 109 bp). Given the number of germline cell divisions between generations this equates to 138 mutations per generation [Estimating the Human Mutation Rate: Biochemical Method]. This value is roughly consistent with the estimates from phylogeny [Estimating the Human Mutation Rate: Phylogenetic Method]. The estimates from direct sequencing of parent and child genomes tend to be a bit lower [Estimating the Human Mutation Rate: Direct Method].
There's a lot of debate about which estimates are correct but keep in mind that we're dealing with only a two-fold effect (Ségurel et al., 2014) [see also Human mutation rates; Human mutation rates - what's the right number?]. The consensus seems to be around 100 new mutations per generation. This corresponds to about 0.25 mutations per genome replication and a mutation rate of 0.8 × 10-10 per base pair per replication.
If that rate is the same in somatic cells then there should be about 0.5 (2 × 0.25) new mutations for every cell division.
Tomasetti and Vogelstein disagree with this estimate in their latest paper. They say,
It has been extensively documented that approximately three mutations occur every time a normal human stem cell divides (Lynch, 2010; Tomasetti et al., 2013).That's six times the expected mutation rate based on germline cell divisions. Why should there be a difference in DNA replication error rates in somatic cells compared to germline cells?
The first reference they give is to a paper by Michael Lynch in 2010. He claims the germline mutation rate is about 0.6 × 10-10 per bp per replication, which is close to the value I calculated above (0.8 × 10-10) (Lynch, 2010; Lynch et al., 2016). His estimate is based on the mutation rate determined from frequencies of disease-causing mutations.
Lynch (2010) also looked at mutations in somatic cells and estimated that the mutation rate in somatic cells is higher than the rate in germline cells. He figures it might be 4-25 times higher. His best estimate of the rate is 7.7 × 10-10 per bp per replication. That's about ten times higher than the germline mutation rate. It corresponds to about 5 mutations every time a cell divides. That's what Tomasetti and Vogelstein are referring to in their 2017 paper.
Lynch thinks this somatic cell mutational load must be enormous. So large, in fact, that there's little hope we can survive much past current life expectancies in industrialized nations. As an example, he notes that the intestinal epithelial cells of an average 60 year-old will each harbor thousands of mutations (4,000 - 40,000) and if you take all these cells into account the tissue will contain more than one million independent mutations.1
The second reference in the Tomasetti and Vogelstein paper is Tomasetti et al. (2013). They looked at whole genome sequences of various cancers and estimated that the somatic cell mutation rate is 6.4 × 10-10 per bp per replication. This is similar to Lynch's value and almost ten times higher than the germline mutation rate.
There are a lot of assumptions and estimates in the Lynch (2010) and the Tomasetti et al. (2013) paper but they seem reasonable to me. However, it doesn't make sense to me that the DNA replication error rate would be ten times higher in typical somatic cells compared to typical germline cells.
The issue gets even more confusing with a paper in one of the latest (March 30, 2017) issues of Nature. In that issue, Ju et al. (2017) looked at whole genome sequences of blood samples from 279 individuals. They were looking for sites where there were multiple variants and one of the variants (alleles) was in the frequency range of 10% to 35%. These are likely due to mutations that occurred early on in embrygenesis so that a substantial fraction of the descendants inherit the mutation. For example, if a mutation occurs at the first cell division of the zygote then half the somatic cells of the adult will carry the mutations and half will have the original allele. Since the cells are diploid, this means that the overall frequency of the mutant allele will be 25%.
Using this technique, the authors calculate that the somatic cell mutation rate during early embryogenesis is about three (2.8) mutations per cell division. This works out to a DNA replication mutation rate of 5 × 10-10 and that's similar to the estimates above for somatic cells. The difference is that this rate is estimated for early embryogenesis creating an awkward situation were germline cells in the same embryo are replicating DNA with an error rate ten time lower (fewer mutations) than the rate in somatic cells. I can't imagine how this happens. How can the embryo "know" which cells are destined to become the germline, especially in the first few divisions?
(Ju et al. don't think there's a problem since they believe the germline mutation rate may be as high as 2 × 10-10. This is not consistent with most studies.)
There's a problem here. I'm skeptical of the somatic cell mutation rates, especially when extrapolated back to embryos. The literature is confusing but most papers conclude that the somatic cell mutation rate is, indeed, higher. In some cases, it could be due to an increase in oxidative damage in some tissues. In other cases, it may be due to increases in methylation that have consequences in repair. I'm not sure these effects are very important in real somatic tissues.
1. Most of Lynch's paper is devoted to concerns about the average germline mutation rate and the genetic load on the species. He is not optimistic about our chances for long-term survival as a species.
Lynch, M. (2010) Rate, molecular spectrum, and consequences of human mutation. Proc. Natl. Acad. Sci. (USA), 107:961-968. [doi: 10.1073/pnas.0912629107]
Lynch, M. (2016) Mutation and Human Exceptionalism: Our Future Genetic Load. Genetics, 202:869-875. [doi: 10.1534/genetics.115.180471]
Ju, Y. S., Martincorena, I., Gerstung, M., Petljak, M., Alexandrov, L. B., Rahbari, R., Wedge, D. C., Davies, H. R., Ramakrishna, M., and Fullam, A. (2017) Somatic mutations reveal asymmetric cellular dynamics in the early human embryo. Nature 543:714-718. [doi: 10.1038/nature21703]
Nowak, M. A., and Waclaw, B. (2017) Genes, environment, and “bad luck”. Science, 355:1266-1267. [doi: 10.1126/science.aam9746 ]
Tomasetti, C., and Vogelstein, B. (2015) Variation in cancer risk among tissues can be explained by the number of stem cell divisions. Science, 347:78-81. [doi: 10.1126/science.1260825 ]
Tomasetti, C., Li, L., and Vogelstein, B. (2017) Stem cell divisions, somatic mutations, cancer etiology, and cancer prevention. Science, 355:1330-1334. [doi: 10.1126/science.aaf9011]
Tomasetti, C., Vogelstein, B., and Parmigiani, G. (2013) Half or more of the somatic mutations in cancers of self-renewing tissues originate prior to tumor initiation. Proc. Natl. Acad. Sci. (USA) 110:1999-2004. [doi: 10.1073/pnas.1221068110 ]
Ségurel, L., Wyman, M.J., and Przeworski, M. (2014) Determinants of mutation rate variation in the human germline. Ann. Rev. Genomics and Human Genetics, 15:47-70. [doi: 10.1146/annurev-genom-031714-125740]