Somatic mosaicism, the co-existence of genetically distinct cell populations in a single individual as a result of post-zygotic genetic events that can occur at any time in the life cycle from fertilization to death, can be meaningless and clinically silent, but mosaicism is also potentially dangerous, contributing to a range of maladies from birth defects to cancer.1 Somatic mosaicism can result from errors in replication of segments of chromosomes or of whole chromosomes, leading to chromosomal aneuploidy. In addition, copy-neutral reciprocal gains and losses may result in acquired uniparental disomy or loss of heterozygosity (LOH), and somatic mosaicism can also be as simple as an acquired point mutation. The prevalence of somatic mosaicism in tissues in the general population is unclear.
In order to better estimate the prevalence of somatic mosaicism and explore a link to neoplasia risk, more than 100 investigators led by a team from the U.S. National Cancer Institute (NCI) analyzed blood or buccal cell-derived autosomes from 31,717 patients with cancer and 26,136 cancer-free controls. These 57,853 samples had been obtained during 13 genome-wide association studies from 48 epidemiologic analyses previously conducted by various groups in collaboration with the NCI’s Division of Cancer Epidemiology and Genetics (DCEG) and Core Genotyping Facility (CGF). In the current study, the investigators used a 2 megabase pair (2Mb)-size cutoff to define mosaicism in order to be conservative and to minimize “false discovery” of germline variants.
The major finding of the NCI analysis was that mosaic abnormalities > 2 Mb in size were found in 0.89 percent of study subjects overall, and the frequency of detectable somatic mosaicism increased with age, ranging from 0.23 percent in individuals under age 50 to 1.91 percent in patients between 75 and 79 years old. Mosaic abnormalities were also more common in the cancer cases (0.97%) compared with cancer-free persons (0.74%). The most frequent class of anomaly detected was copy-neutral LOH, representing 48 percent of mosaic events – most commonly LOH of chromosome 9p (where JAK2 is located) or of chromosome 14. Trisomies 8, 12, and 15 were also observed recurrently, as were deletions of chromosomes 13q and 20q; these are frequent findings in neoplastic cells from patients with hematologic malignancies.
A recent Swedish analysis of somatic alterations in 318 monozygotic twins and 296 single-born subjects also described age-related accumulation of copy-number variation, suggesting a mechanism for age-associated reduction in the diversity of blood cells and for accumulation of hematopoietic clones with structural abnormalities.2 Such clonal changes could lead to myelodysplasia, monoclonal B-cell lymphocytosis, or other aging-related blood dyscrasias. In fact, the Swedish study detected somatic aberrations characteristic of patients with myelodysplastic syndromes in three apparently healthy persons, who are likely to be at considerable risk for subsequent development of clinical disease.2
In Brief
Since cancer is an example of somatic mosaicism, one implication of these findings is that specific types of clonal mosaicism that randomly arise in a time-dependent fashion and confer a cellular survival advantage likely contribute to an increased risk of subsequent clonal shifts and eventually development of frank malignancy. Supporting this idea is the observation that among 43 individuals in the NCI study who subsequently developed a hematologic malignancy > 1 year after their DNA was obtained for analysis, clonal mosaicism was present in a whopping 20 percent of those who were ultimately diagnosed with myeloid leukemias and 22 percent with lymphocytic leukemias.
The most immediate application of the NCI-led consortium’s results is as a point of caution for molecular epidemiology studies comparing blood or buccal cells with tumor cells, but surveillance for age-related somatic mosaicism might one day serve as a biomarker for individuals at increased risk for subsequent development of hematologic malignancies or of other cancers.
References
Competing Interests
Dr. Steensma indicated no relevant conflicts of interest.
