Scientific Updates

Nature | A body map of somatic mutagenesis in morphologically normal human tissues

  On Aug. 25th, 2021, Fan Bai and Yanyi Huang from Biomedical Pioneering Innovation Center (BIOPIC) and Beijing Advanced Innovation Center for Genomics (ICG), collaborating with Jianbin Wang from the School of Life Sciences in Tsinghua University, Dongxin Lin, Chen Wu from Chinese Academy of Medical Sciences and Peking Union Medical College, published an article entitled “A body map of somatic mutagenesis in morphologically normal human tissues” in Nature. This article firstly depicted a body map of somatic mutations in different normal tissues from the same individual and revealed the patterns of somatic mutation accumulation and clonal expansion in normal human tissues with the same germline backgrounds and life histories, which lays a foundation for understanding the mechanisms of aging and cancer development.


  Normal cells inevitably acquire somatic mutations that mainly result from unrepaired or incorrectly repaired DNA replication errors during cell division. Although most somatic mutations in normal cells do not have any phenotypic consequence, mutations that affect essential genes, especially those related to cell proliferation and death, may trigger mutant clonal expansions, ultimately causing aging and disease, even cancer. Thus, investigations of somatic mutation accumulation and mutant clonal expansion in normal cells may derive new insights into early carcinogenesis.

  Unlike tumor cells arising from monoclonal expansion, normal cell populations usually accumulate somatic mutations independently, develop in the forms of polyclonal populations, and parallel evolution. Detecting such mutations in normal tissues by conventional genome-sequencing technologies based on large cell populations has been challenging, as the same mutation tends to occur in only a few normal cells.

  Firstly, the research group developed a low-input (600 cells) whole-genome/exon sequencing technology that enables the detection of few somatic mutations in normal tissues, and secondly, in 5 deceased organ donors aged above 85 years old, the group collected microbiopsies from 9 anatomic sites of autopsy samples, which included morphologically normal epithelia from the bronchus, oesophagus, cardia, stomach, duodenum, colon and rectum, and normal parenchyma from the liver and pancreas, and applied a consistent sampling strategy for sectioning and sequencing according to the spatial distribution of epithelial cells (Fig. 1). A considerable amount of somatic mutations was detected in normal tissues. Researchers found that the numbers of somatic mutations and distributions of variant allele frequency (VAF) varied greatly across tissues and donors.


  Fig. 1: Research strategy and summary of genomic alterations detected in normal tissues from five donors.

  After considering the physical tissue microstructures and making sensitivity corrections of detected mutations, the researchers compared mutations accumulation across different organs of the same donor. Pancreas parenchyma contained the fewest mutations, whereas the number of mutations in the liver was the greatest among all tissues-substantially higher than the number of mutations in epithelial cells from other organs. Combined with GTEx data, researchers further found that the number of mutations tended to decrease in highly expressed genes in different tissues, implying that transcription-coupled repair is more active in highly expressed genes.

  Sporadic copy number alterations (CNAs) could be detected in a number of samples. Of note, the samples with CNAs exhibited strong organ preferences. Normal oesophageal tissues were found to contain more CNAs that were enriched as whole-chromosomal amplifications of chromosomes 3, 5, and 7 (Fig. 1).

  Researchers deciphered seven mutational signatures across different organs to further examine the underlying mutational processes,. SBS1 and SBS5, two age-related endogenous mutational signatures, were found throughout all normal samples across organs and donors, indicating that the aging related mutational process is a significant driver of the accumulation of somatic mutations in normal tissues. The relative activities of SBS1 and SBS5 varied across tissues but exhibited a conserved tissue-specific pattern among donors. The duodenum, colon and rectum showed higher SBS1/SBS5 ratios compared with the bronchus, pancreas, oesophagus, and liver.

  SBS4 (associated with tobacco smoking) and SBS22 (associated with exposure to aristolochic acid), two exogenous mutational signatures exhibited mainly in liver samples are suggesting that the liver has a higher risk of exposure to environmental carcinogens than do other organs. The researchers also found considerable differences in the mutational signature spectra and relative activities, even between adjacent laser-capture microdissection (LCM) biopsies from the same individual and the same organ. This regional variation (both between and within tissue layers) in mutational signature activity may reflect regional activations of different mutagenic driving factors (Fig. 2).


  Fig. 2: Mutational signatures in normal tissues from five donors.

  32 potential driver genes, including canonical cancer drivers such as NOTCH1, TP53, ARID1A and ERBB2 were identified in normal tissues. Further analysis of driver mutations revealed that:

  (1) NOTCH1 was found to be the most frequently mutated gene;

  (2) Greatest number of hotspot mutations being detected in TP53;

  (3) Mutations in these potential driver genes were distributed heterogeneously across organs and donors and have an organ bias. For example, most driver mutations contained in oesophageal samples but less in pancreas; NOTCH1 and TP53 mutations showed enrichments in oesophageal tissues, while MUC6 was identified as a driver gene that is enriched in normal cardia and stomach;

  (4) The prevalence of MUC6 mutations in normal gastric tissues (cardia and stomach) was significantly higher than that in gastric cancers, suggesting that different molecular mechanisms underlie the clonal evolution of normal cells versus that of cancer cells (Fig. 3).


  Fig. 3: Mutational landscape of driver genes across organs.

  The researchers investigated how the accumulation of mutations and the expansion of mutant clones are coordinated in normal tissues. For each donor, they plotted the distribution of the number of mutations versus each samples’s the average mutant cell fraction (MCF). In the oesophagus and cardia, mutant clones tended to be significant, but the numbers of mutations were relatively low. By contrast, normal colonic and rectal tissues accumulated many mutations, although the degree of clonal expansion was low on the spatial scale. Some samples simultaneously had high mutational burdens in the liver and showed substantial clonal expansions.

  Combining the spatial distribution information and MCFs clustering results, the researchers reconstructed the spatial somatic clonal architecture with sub-millimetre resolution (Fig. 4). In the oesophagus, they identified large-scale clonal expansions that covered two to more than ten LCM biopsies and spread across two to three layers. Mutations occurred in NOTCH1, TP53 and ARID1A, CNAs occurred in chromosomes 3, 5 and 7 may have potentially driven the mutant clonal expansions in the oesophagus. Clonal expansions in the colon, rectum, and duodenum—which are constrained by tissue physical microstructures—exhibited low degrees of clonal expansions on the spatial scale. These results suggest that, in different organs and tissues, somatic mutant clonal expansions are constrained by various factors such as microanatomical structures in the tissue.


  Fig. 4: Estimation of somatic mutant clonal sizes and construction of spatial clonal expansion maps.

  Dr. Ruoyan Li, Ph.D. candidate Lin Di from Biomedical Pioneering Innovation Center (BIOPIC), Ph.D. candidate Jie Li from School of Life Sciences in Tsinghua University and Dr. Wenyi Fan from Chinese Academy of Medical Sciences and Peking Union Medical College are the co-first authors of the paper. Professor Fan Bai, Professor Yanyi Huang from Biomedical Pioneering Innovation Center (BIOPIC) and Beijing Advanced Innovation Center for Genomics (ICG), Professor Jianbin Wang from School of Life Sciences in Tsinghua University, Professor Chen Wu and Professor Dongxin Lin from Chinese Academy of Medical Sciences and Peking Union Medical College are the co-corresponding authors of the paper. This project was jointly supported by the National Natural Science Foundation of China, the Medical and Health Technology Innovation Project of the Chinese Academy of Medical Sciences, the National Key R&D Program of China, and the Beijing Advanced Innovation Center for Genomics.