Every cell in the body picks up changes to its DNA over a lifetime. For most of the history of cancer genomics, researchers could only read those changes once a tumor had formed, because mutations in normal tissue sit at frequencies too low for older sequencing methods to detect. That left basic questions unanswered, such as how many mutations a healthy cell carries by middle age, which of them give a cell a growth advantage, and what those expanding clones mean for cancer, aging and chronic disease.
In the most recent episode of The Genetics Podcast, host Patrick Short speaks with Dr. Inigo Martinconera, a Group Leader at the Wellcome Sanger Institute. Inigo trained as an evolutionary biologist, studying patterns of mutation across the E. coli genome during his PhD at the EBI. Halfway through, the first two cancer genomes convinced him that cancer evolution needed to be studied with formal evolutionary methods, and he joined Campbell's group as a postdoc to do that. Patrick, who began his own PhD at Sanger in 2015, remembers hearing about Inigo's "eyelid paper" around campus. The conversation covers that study, the esophagus work that followed, a recent discovery in autoimmune thyroid disease, and how his group thinks about turning somatic evolution into therapies.
During his postdoctoral research, Inigo sequenced roughly 200 to 300 small biopsies of eyelid skin left over from cosmetic surgery on four healthy, middle-aged people. Studies in the 1990s had found occasional clones with p53 mutations in normal skin, so the team expected some mutations. The scale of what they found was unexpected.
"What we found was that a square centimeter of normal skin would carry over 100 clones that had already acquired a cancer causing mutation," Inigo says. TP53 was one of the mutated genes, and NOTCH1 was far more frequent, mutated in about 30% of cells. Overall, he explains, "a third of all skin cells had acquired already a driver mutation by middle age."
Inigo expected skin to be an outlier, since ultraviolet light makes it the most heavily mutated tissue in the body. The follow-up study looked at the esophagus, which has no sun exposure, and found it as heavily colonized by driver-mutant clones as the skin, or more. Samples from nine people of different ages showed the process advancing with age. In one cancer-free 75-year-old, around 60 to 70% of esophageal cells carried a driver mutation, and about a quarter of all cells had lost p53.
He remembers receiving those results on a Friday afternoon and spending the weekend "with my head spinning" about what they meant for aging and for diseases that are understood by their phenotype, where nobody had been able to look for this kind of clonal expansion. The questions from that weekend set the direction of his group's work for the next ten years: what happens in other tissues, how much of the variation between people comes from genetics and how much from exposures, whether the work could scale with non-invasive biopsies, and what role these clones play in other diseases.
Patrick asks how so many cells can carry cancer drivers while most people never develop cancer. Inigo says he has always found the results consistent with existing knowledge of cancer. A cancer cell has typically acquired five or six driver mutations along with extensive chromosomal aberrations. Clones in normal tissue are common, but each one is usually small and carries one or at most two drivers.
"You can think of the landscape that we are seeing in normal tissues as the first one or two steps that cells are taking during life on the road to cancer," he explains. "You can think of it as a window into very early cancer development that we couldn't see before." The arithmetic supports this. Given measured mutation rates and clone sizes, the probability of five drivers accumulating in a single cell in normal tissue is very low, roughly in line with cancer incidence in the population.
Work across the field over the past few years has shown that somatic mutations accumulate in every tissue, increasing linearly with age even in cells that no longer divide. They build up in post-mitotic neurons, as Chris Walsh described in an earlier episode, and in cardiomyocytes in the heart. Mutation rates differ between tissues, though by less than researchers had expected.
Selection is more restricted, because a clone can only expand if its cells divide. The strongest evidence of clonal takeover comes from highly mitotic cell types: epithelial cells, B and T cells, hepatocytes and blood. In the brain, microglia divide and are drawing interest for pro-inflammatory clonal expansions, while neurons show no selection. Inigo sees these mitotic cell types as the best starting points when looking for clones that cause disease or protect against it.
From 2015 to around 2024, most studies focused on epithelial tissues, where clones are spatially organized and can be cut out and sequenced. Mobile immune cells such as lymphocytes and macrophages were out of reach. Inigo points to clonal hematopoiesis as an example of how measurability shapes a field: it has grown quickly partly because it can be detected with standard sequencing in biobank blood samples.
A large part of his group's work has gone into new technologies, including high-precision sequencing of microscopic areas of tissue. However, sample access remains a constraint. Cancers are routinely resected and biopsied, but tissue such as brain is hard to obtain.
Two developments are next. True single-cell whole-genome DNA sequencing has emerged in the last year or two, mainly through primary template amplification, and is now being paired with single-cell RNA sequencing and spatial methods, which will link genotype to phenotype at scale. The second is methylation. Methylation changes can arise stochastically and are inherited by a cell's descendants, and very little is known about how far somatic evolution runs through these epimutations.
Autoimmunity was on the group's list soon after the esophagus results, but lymphocytes move around the body and cannot be isolated as spatial clones, so the work had to wait for NanoSeq. Deep whole-exome NanoSeq on thyroid biopsies from three donors found hundreds of independent B-cell clones carrying mutations in immune-checkpoint genes. Two checkpoints, PD-L1 and HVEM, were recurrently mutated in tens to hundreds of independent clones in every individual. Each new method the team added, including single-cell DNA sequencing, spatial sequencing with microscopy and a larger cohort, turned up further results.
He is careful about what has been shown. The findings are consistent with a causal role, and the causes of autoimmune disease remain poorly understood at the molecular level. Whether these mutations are necessary or sufficient for disease has yet to be demonstrated, and testing that, along with checking other autoimmune diseases, is now a major focus for the group.
In unpublished work on aging lymphocytes from healthy people, the group is also seeing a rich landscape of mutations in key immune regulatory genes. These differ from the checkpoint mutations seen in autoimmunity and look less advanced. Inigo thinks this phenomenon may be more common than clonal hematopoiesis, having gone unnoticed because the clones are very small. Beyond autoimmunity, the team is seeing mutations in macrophages in osteoarthritis and findings in atherosclerosis.
Inigo co-founded Quotient Therapeutics to develop treatments from this biology, and the company works with two models. In the causal paradigm, somatic mutations cause or contribute to disease, and the aim is clonal suppression or deletion, applying an oncology approach to conditions such as autoimmunity. Treating clones before symptoms appear could make this preventive.
The protective paradigm starts from cells that survive disease. In chronic liver disease caused by alcohol or a high-fat diet, damaged hepatocytes die and the liver regenerates from hepatocytes carrying mutations that let them tolerate the toxicity. The goal would be a drug that copies the effect of those mutations.
Blood provides the clearest case of what Inigo calls somatic rescue. Some people are born with germline mutations that cause bone marrow failure. In a few, a single hematopoietic stem cell acquires a somatic mutation that reverses the effect, sometimes reverting the germline allele itself, and repopulates the whole blood system, curing the person.
Listen to the full episode below.