Somatic mutations in cancer development

Abstract

The transformation of a normal cell into a cancer cell takes place through a sequence of a small number of discrete genetic events, somatic mutations: thus, cancer can be regarded properly as a genetic disease of somatic cells. The analogy between evolution of organisms and evolution of cell populations is compelling: in both cases what drives change is mutation, but it is Darwinian selection that enables clones that have a growth advantage to expand, thus providing a larger target size for the next mutation to hit. The search for molecular lesions in tumors has taken on a new dimension thanks to two powerful technologies: the micro-arrays for quantitative analysis of global gene expresssion (the transcriptome); and 'deep' sequencing for the global analysis of the entire genome (or at least the exome). The former offers the most complete phenotypic characterization of a tumor we could ever hope for--we could call this the ultimate phenotype; the latter can identify all the somatic mutations in an individual tumor--we could call this the somatic genotype. However, there is definitely the risk that while we are 'drowned by data, we remain thirsty for knowledge'. If we want to heed the teachings of Lorenzo Tomatis, I think the message is clear: we ought to take advantage of the new powerful technologies--not by becoming their slaves, but remaining their masters. Identifying somatic mutations in a tumor is important not because it qualifies for 'oncogenomics', but because through a deeper understanding of the nature of that particular tumor it can help us to optimize therapy or to design new therapeutic approaches.

Figure 1

Headings of one of the first and on of the last publications by Lorenzo Tomatis.

Figure 2

Pedigree of a family with a high rate of breast cancer and ovarian cancer: the increased tendency to developing cancer shows a Mendelian autosomal dominant pattern of inheritance, suggesting that a single gene is largely responsible.

Figure 3

In this extended family there were 3 cases of hairy cell leukaemia (HCL): their co-existence can be hardly a coincidence, since HCL is one of the rarest forms of B cell leukaemia. Here the pattern is not Mendelian, suggesting that several genes and/or environmental factors are involved.

Figure 4

Three patients with hairy cell leukaemia in the same family.

Figure 5

Meta-analysis of the quantitative effect of a polymorphic allele of the TGF b receptor gene on the frequency of some types of tumors.

Figure 6

A genetic polymorphism in the coding sequence of the TGF-β receptor gene influences the risk of cancer.

Figure 7

A cartoon illustrating current views of the origin of cancer, which is consequent on n successive somatic mutations. The final result is a clonal population of cells with highly disregulated growth. It can be presumed that in fact each one of the mutational steps entails a growth advantage, even if small: this increases the number of cells that can be targeted by the next mutation. The term n-1 is used to indicate the penultimate step in the pathway, because the number n is not fixed: it is estimated that it may range, for the majority of tumors, from 3 to 6 or even more.

Figure 8

Figure 9

Inherited mutations can increase cancer proneness through different mechanisms. The top section of the cartoon is a schematic of the process outlined in Fig. 5. The middle section illustrates how an increased rate of somatic mutations can produce an accelerated rate of the oncogenic pathway: this is the case for instance for patients with Fanconi anemia, who have a serious defect in DNA repair and often develop cancer at a young age. The bottom section illustrate that the number of steps for a normal cell to become a cancer cell is cut by one if the first mutation is an inherited (germ-line) mutation rather than an acquired somatic mutation: this is the case for instance for patients who have an APC mutation and present with familial adenomatous polyposis.

Figure 10

A graphic representation (current referred to as a cyclo-plot) of the multiple defects detected in the genome of a tumor (a small-cell lung cancer cell line) by deep sequencing. Individual chromosomes are depicted on the outer circle followed by concentric tracks for point mutation, copy number and rearrangement data relative to mapping position in the genome. Arrows indicate examples of the various types of somatic mutation present in this cancer genome. From Stratton et al., 2009 (ref. 29).

Figure 11

Figurative depiction of the landscape of somatic mutations present in a single cancer genome.

Figure 12

A cartoon illustrating the central role of chance in cancer formation, based on the fact that somatic mutations are stochastic events. Inherited factors (see text and Fig. 6) can modulate the process, but they somatic mutations are still needed for the onset of cancer; and environmental factors work in large measure by increasing either the mutation rate (mutagenic agents) or the number of cell divisions (e.g. with an inflammatory process, such as one caused for instance by Helicobacter pylori).

Figure 13

A new methodology makes it relatively easy to measure in any individual the intrinsic rate of somatic mutation. In two conditions known to be associated with cancer proneness the rate of somatic mutation is markedly increased over that observed in a control group.