In a landmark paper published today (1 May, 2024) by the Mutographs team in Nature, they identify new mutational signatures at play in the development of clear cell renal cell carcinoma, of unknown origin, which contribute towards geographical differences in the incidence of this cancer. These findings are in stark contrast to their previous work in esophageal squamous cell carcinoma, which found no geographical difference in mutational profiles.
The Challenge
In 2015 Cancer Grand Challenges, then known as CRUK Grand Challenge, set the Unusual Mutations challenge, to discover how unusual patterns of mutation are induced by different cancer-causing events, and in 2017 funded the Mutographs team, led by Professor Sir Mike Stratton, then Director of the Wellcome Sanger Institute. By utilising paired whole genome sequencing of tumour and normal tissue, they set out to combine mutational signature analysis with thorough epidemiology, in order to comprehensively catalogue the mutational processes leading to cancer, and their causes, at scale.
One of the team’s goals was to explore whether geographical differences in the incidence of some cancers could be attributed to distinct mutational signatures, explained by differential exposure to unknown carcinogens, whose identification would ultimately aid cancer prevention. Critical to this goal would be collection of patient samples on an international level, always including countries with differing incidence rates for the cancer type being studied. Enabled by Cancer Grand Challenges, this would be made possible by the strategic alliance developed between the International Agency for Research on Cancer (IARC), France, and the Wellcome Sanger Institute, UK, and their collaboration with Cancer Centres across the globe. Together the team have collected over 5,000 patient samples, from over five continents. They focused on five tumour types, where known risk factors do not explain geographical differences in incidence rates.
So the question was clear- are there additional carcinogens at play which explain global differences in cancer incidence rates and could the team identify them via mutational epidemiology, a field the team have pioneered? Initial findings published by the team, showed the answer might not be what they expected.
Revisiting the promotional hypothesis
In 2021, the team published their results from esophageal squamous cell carcinoma (ESCC) in Nature Genetics. In this work, led by Mike at the Sanger and Paul Brennan at IARC, they found that mutational profiles of the 552 genomes studied were in fact similar across all eight countries, despite the differences in incidence. So in ESCC at least, the team found no evidence to suggest an exogenous mutagenic exposure could explain the geographical variation in incidence.
Work from other members of the team led by David Adams at the Sanger and Allan Balmain at the University of California, San Francisco, had also given them pause for thought. Published in Nature Genetics the year prior, they had shown that when they sequenced lung and liver tumours of mice exposed to suspected human carcinogens, the majority did not actually cause distinct mutational signatures or an increase in mutational load. For those that did (only 3/20), driver mutations were still rather caused by endogenous processes.
These findings made the team take a step back, and think, what really causes cancer?
In the classical view, carcinogens cause mutations, and these mutations cause cancer, but their work in ESCC and with mice, together with growing evidence that normal tissues are riddled with oncogenic mutations, caused a revival in thinking around the promotional hypothesis of carcinogenesis, and how these carcinogens are accelerating tumour formation, if not via mutagenesis.
This got the team and the wider community thinking about non-mutagenic events and how normal clonal expansion contributes to carcinogenesis. At Cancer Grand Challenges, this informed our setting the Normal Phenotypes challenge in 2020, to understand how cells and tissues maintain “normal” phenotypes whilst harbouring oncogenic mutations and how they then transition to become a tumour. We subsequently funded the PROMINENT team, with their name coming from ‘Discovering the molecular signatures of cancer PROMotion to INform prevENTion’.
But the Mutographs team were far from done, although their initial findings did make them wonder if they would find the same for the other cancer types they were investigating.
The holy grail of Mutographs
Next the team set their sights on clear cell renal cell carcinoma (ccRCC), where obesity, hypertension and smoking tobacco are known risk factors, however these lifestyle factors cannot explain the geographical variance in incidence. Here the team sequenced 962 ccRCCs, from 11 countries, across four continents.
First, they found a known mutational signature, single base substitution (SBS) SBS22, previously associated with aristolochic acids, derived from Aristolochia plants. It was known aristolochic acids were mutagenic, and that exposure, presumed to be via herbal remedies containing Aristolochia extracts, had occurred in a localised fashion in Romania. However, here the team found exposure was much more widespread than previously thought, spreading to Serbia, meaning potentially millions of people in the region had been exposed. In addition, with their sequencing resolution, they were able to identify a highly related signature, which the team thinks could be due to different aristolochic acids or their metabolites, with the possibility that different people metabolise them differently. At this point the team cannot say whether SBS22 is causal, this will be a critical focus of future work, but they do know it appears in conjunction with driver mutations. In addition, the team want to follow up on how widespread the exposure is geographically - does it spread into Hungary, Greece, Ukraine? From a public health perspective, it will be important to determine if exposure is still occurring, and what exactly is the source.
But could the Mutographs approach be used to identify new mutational signatures, pointing to as yet unidentified exposures that would explain differences in incidence rates on a global level? In short - yes...
Strikingly, they identified a previously unseen mutational signature of unknown cause, found in all 11 countries. This signature, SBS40b, potentially explains the variation in incidence across the world, with mutation loads correlating with incidence rates. The team performed initial metabolite studies indicating that the signature associates with decreased kidney function, but could not pinpoint the source further, or whether decreased kidney function leads to the SBS40b mutational signature, or vice versa. Wider geographical epidemiology will be required to determine the source, and the right questions will need to be asked of patients.
The team then went on to identify another mutational signature of unknown cause, SBS12, which was found in approximately 70% of cases of ccRCC in Japan. This signature displayed high transcriptional strand bias, with mutations found more often on the transcribed strand of protein-coding genes, which points to an exogenous exposure. The team saw no association with a polymorphism in aldehyde dehydrogenase 2, which is known to be common in Japan, and impairs alcohol metabolism. Reanalysis of hepatocellular carcinoma samples from the liver found SBS12 was also enriched in patients from Japan, but the team now want to investigate whether SBS12 is present in other tumour types. They also want to determine if the exposure extends outside of Japan, and are now contacting colleagues in China, Singapore and South Korea in order to collect patient samples from the surrounding countries.
As for the known risk factors, whereas tobacco correlated with the known tobacco mutational signature, the team found no signature associated with obesity or hypertension, again suggesting the importance of non-mutagenic mechanisms of cancer promotion.
The immediate question that arises from all three elements of this seminal work, is whether the mutational signatures are causal in cancer development. Followed by the key questions of what is the exact source, is the population still being exposed, and how does exposure lead to mutation.
The team are keen to see this work published, so that the field can start to answer some of these questions, and better understand the potential implications and actions needed for public health.
Towards a global survey of mutational signatures, enabled by team science
Cancer Grand Challenges, co-founded by the National Cancer Institute in the US and Cancer Research UK, aims to bring together the best minds around the world, to solve the biggest questions in cancer research, with diversity in thinking being key. This paper embodies these values at all levels. The work is a collaboration between 55 institutes, with researchers in 19 different countries. The work was driven by co-first authors, Sergey Senkin on the epidemiology side and Sarah Moody on the genomics side, at IARC and the Sanger respectively. They come to the core of the problem from different angles, and have very different scientific backgrounds, with Sarah transitioning from wet lab work to computational biology, and Sergey transitioning from particle physics to genomic epidemiology! Despite being in different countries, there has been no barrier to teamwork, with the two talking frequently via Slack, and the broader team holding virtual meetings every Wednesday to keep momentum going. The vast number of genomes sequenced, across 11 countries, enabled by collaborations with cancer centres in these locations, has allowed the resolution required to uncover these patterns in mutational signatures. Imagine what the team can reveal with even higher resolution, or even a global survey.
By identifying a rich landscape of exposures accounting for geographical variance in ccRCC, Sergey, Sarah and the team have demonstrated the power of the Mutographs approach. But they’ve also changed the way we think about cancer along the way, highlighting the importance of both mutagenic and non-mutagenic processes, and their interplay.
References
Riva et al Nat Genetics 2020 https://www.nature.com/articles/s41588-020-0692-4
Moody, Senkin et al Nat genetics 2021 https://www.nature.com/articles/s41588-021-00928-6
Senkin, Moody et al Nature 2024 https://www.nature.com/articles/s41586-024-07368-2
Photo: Sequencing laboratory at the Sanger institute, photo by David Levene / Wellcome Sanger Institute
Article written by Rebecca Eccles, with thanks to Sergey Senkin, Sarah Moody, Mike Stratton and Paul Brennan for their input.