Although the epidemiological links between tobacco smoking and lung cancer and the benefits of stopping smoking are clear, the molecular mechanisms that explain these links are not. For example, the linear dose–response relationship between numbers of cigarettes smoked and cancer risk and the decreases in cancer risk from stopping smoking cannot be explained by a model in which tobacco carcinogenesis occurs solely through increasing mutational burden.

Credit: Lara Crow/Springer Nature Limited

In order to survey early stage changes in the lungs that might explain some of these observations, Yoshida, Gowers et al. examined somatic mutations in normal lung epithelium from individuals who underwent bronchoscopies for conditions that were eventually diagnosed as benign. Single bronchial epithelial cells were isolated from 16 people (3 children, 4 ‘never-smokers’, 3 current smokers and 6 former smokers) and were grown into single-cell-derived colonies on mouse feeder cells.

Whole genome sequencing data was compiled for 632 single cells. There was an expected increase of the mean number of single-base substitution mutations with ageing and smoking (in both current and former smokers). However, in current or former smokers, there was marked heterogeneity in the mutational burden between cells of the same individual, and in some cases cells were observed that had a mutational burden similar to that in cells from never-smokers. Intriguingly, these ‘near-normal’ cells were fourfold more frequent in former smokers compared with current smokers. Given that lung cancers that arise in former smokers have a high mutational burden and mutational signatures linked to tobacco exposure, and the epidemiological data on reduced cancer risk in former smokers, it seems likely that these near-normal cells can protect from cancer development in some way. Furthermore, these cells had longer telomeres than more mutated cells from the same individual, which supports the hypothesis that these near-normal cells have undergone fewer cell divisions and could have recently arisen from quiescent stem cells that were protected from tobacco-induced mutagenesis (either in a physically protected niche or protected through lack of DNA replication).

Comparing the mutational signatures observed in these 16 individuals with those reported previously, the authors found two known and one novel signature that seemed to increase linearly with age in their cohort (endogenous signatures). Three further signatures were observed primarily in current or former smokers: one was expected as it is common in lung cancers found in smokers (SBS-4), one had been correlated with hepatocellular carcinomas linked to smoking or alcohol use, and one not previously described signature contained a large number of T>A and T>C mutations. In former smokers, the SBS-4 signature was absent from near-normal cells but was present in cells that had a high mutational burden. Near-normal cells had a prevalence of endogenous signatures, similar to cells from never-smokers. Moreover, these near-normal cells were polyclonal in origin, indicating that they did not all arise from a single cell in a given individual.

Several mutations in bronchial epithelial cells appeared to be under positive selection. NOTCH1, TP53 and ARID2 were more frequently mutated than expected when all coding genes were considered, and FAT1, PTEN, CHEK2 and ARID1A were more frequently mutated when genes known to be mutated in lung cancers or non-lung squamous tissues were considered. Adult never-smokers had such driver mutations in 4–14% of cells, whereas 25% or more cells from current smokers had one of these mutations, with some cells containing as many as three.

it seems likely that these near-normal cells can protect from cancer development

One caveat of this study is that it analysed only a few individuals, so whether this data will hold up in larger cohorts is unknown. It also raises many interesting questions related to our molecular and cell biological understanding of lung cancer development. For example, what is the role of within-patient heterogeneity in cell competition and clonal evolution, how does tobacco exposure (and cessation of smoking) affect the different populations of cells within the lung, and how might near-normal cells in former smokers protect from cancer development?