As we age, the DNA in all our cells accumulates changes, or mutations. The mutations that cause cancer are well studied: Which are dangerous? What effects do they have? How do they accumulate? How fast? Genome sequencing has been helping scientists answer these questions for decades. But for the mutations in healthy cells, the picture was much more hazy.
In 2014, a conversation between Mike Stratton, Director of the Sanger Institute, Peter Campbell, head of cancer research at Sanger, and Luiza Moore, researcher at the Institute and Consultant Pathologist at Addenbrooke’s Hospital, changed that. Studies in Peter’s lab at the time were among the first in the world to show that our genome – the DNA sequence that contains the instructions to build and run our body - isn’t a fixed entity. Investigations in specific tissues, including the skin of the eyelid and the colon, found we are made up of different populations of cells, each with a slightly different genome to its neighbours1. Many healthy cells were found to have DNA changes associated with cancer – prompting researchers to question what defines a cancer mutation. There was an urgent ambition to understand what DNA mutations looked like in healthy tissues across the whole body.
While the aim was clear, the route to finding answers was not. It took six years, combining expertise in histology, pathology, bioinformatics, and the development of a new genome sequencing technique, but the team has now published the most comprehensive study ever of DNA mutations in healthy tissues. They have used the data to help understand how we develop from a single cell, and how our cells, tissues and organs age2. Their data will form the basis of a new understanding of how our genomes change over our lifetimes, and influence health and disease.
An incredible gift
Around 2014, several research groups across the globe were looking at DNA mutations in individual, non-cancerous, tissues. The Sanger team decided that the best way to asses genetic mutations across the whole body would be by analysing many tissues from one individual person. This would mean that cells would be the same age and would have been in the same environment – essentially living the same life. The samples would need be taken post-mortem, in what is known as a warm autopsy, from an individual who had decided to donate their body to research after their death.
Working with colleagues at the Medical Research Council (MRC) Cancer Unit, it took a year to find someone. At 5am one Sunday morning in 2015, Luiza got a call from the hospice nurse to say that the donor had just died. Together with colleagues, they arranged everything to perform the autopsy immediately.
Luiza recalls the experience: “As a clinical pathologist and forensic scientist, I have done hundreds of autopsies before. But this was something different. I’m not a religious person, but this was like crossing the line between life and death. The donor had died so, so recently. It was completely surreal.” The autopsy took a team of 10 people over five hours. Every tissue sample collected was carefully documented and preserved.
“We are incredibly grateful to the donor and their family,” added Luiza.
With the tissues frozen and stored, the next step was to sequence the genomes of the cells to determine the order of the 6 billion letters of DNA in them. The latest technologies mean that in general, it is cheap to sequence DNA, with robust and well established methods. The Sanger Institute has one of the largest genome sequencing facilities in the world. But, sequencing DNA from such tiny populations of cells presented the biggest challenge of the project.
Technical limitations at the time meant that sequencing one cell wasn’t possible – the amount of DNA would be too small to get a reliable read-out. Pooling cells from a big section of tissue wouldn’t work either, as there would be too many different populations of cells with different mutations mixed in together.
The answer was to combine laser capture microscopy with low-input DNA sequencing. Using laser capture microscopy, Luiza cut out minute sections of tissue, each containing just a few hundred cells. She took these from tiny anatomical regions where they knew that the cells would likely all be one population with the same genome.
“With huge thanks to our histopathology team who helped prepare the samples onto microscope slides, I was able to cut out thousands of micro-biopsies for sequencing,” said Luiza. “I think it was probably more than 5,000.”
The Sequencing Research and Development team at Sanger developed a new workflow to sequence the genomes of the laser-captured tissue taken from the microscope slides. They refined the method over the course of two years, optimising it for different tissues types to get robust results. The team then fully automated the process enabling tens of thousands of genomes to be swiftly sequenced3.
“The difficulty was using such tiny amounts of DNA without introducing errors that would obscure real mutations – something that we had never done before. The method has now been used to process more than 40,000 laser capture samples. It has also been adapted for other projects where only minimal DNA is available, such as sequencing parasitic blood flukes. It’s a powerful new technique.”
Dr Peter Ellis
Senior Staff Scientist at the Wellcome Sanger Institute, now working at Inivata
Tim Coorens, a PhD student, working on mathematical genomics in medicine, analysed the genome data. He said: “Sometimes a tissue sample was nicely clonal – that is all cells came from one ancestor and had the same mutations. That was a dream for data analysis! Other times, there was no structure whatsoever, so you sequence a bit of skin or muscle and you just hope for the best. And it didn’t always work out.”
“As a group, we have learned a lot about different tissues and their architecture,” says Luiza. “We have discovered that most of the tissues in our bodies are polyclonal – we are a patchwork of cells, each with a slightly different genome.”
Different questions, same solutions
Other teams at the Sanger Institute were also sequencing body tissues to understand their mutation patterns. Raheleh Rahbari moved into the cancer department in 2019 as a Cancer Research UK Career Development Fellow.
She had previously been working on developmental diseases that occur in children as a result of DNA mutations in the father’s sperm. These mutations were inferred from sequencing blood samples from the child, mother and father and looking at the differences.
Raheleh and her team were interested in understanding the origin of mutations in the germcells (sperm and egg), and how they contribute to disease in the next generation. “One of my main interests is to understand how different mutations in sperm and egg might cause cancer predisposition in the offspring,” she said.
Together with Alex Cagan (co-first author of the paper) and colleagues, they generated and analysed genome data from testicular and colonic tissues from 13 individuals. As the study progressed, they discovered extensive variation in number of somatic mutations that accumulate in these two tissues from the same individual. “The results were astonishing. We saw germcells in testes accumulate about 27-fold less mutations compared to cells from the colon in the same person. But we were not sure how this variation compared to rest of the tissues in our body”.
We are molecular mosaics
We start life as a single cell, just 0.1 millimetres wide - the result of a fusion of an egg and a sperm. That one cell divides, its descendants divide, as do theirs, and so on, until eventually we are comprised of 40 trillion cells, of hundreds of different types. Each cell carries the 6 billion letters of our genetic code, our genome, which is copied during division. Some cell types divide rapidly, while others are more slow growing. When our DNA code is copied, errors can creep in. Other changes to DNA may come from external sources, like ultraviolet light or tobacco smoke. Some mutations are corrected by a cell’s repair mechanisms, but some changes to the letters of DNA become fixed, and are passed on as a cell divides.
It has long been thought that sperm cells are relatively protected from mutations – in order to preserve the genome that creates the next generation. The team’s analysis confirmed, for the first time, that this is the case. They found that sperm precursor cells have a much lower rate of mutation compared to other tissues in the body – and this is despite their rapid rates of division.
“If we can uncover how these cells protect themselves from mutation – that would be a dream come true for me,” says Raheleh. “If we understand how they are doing it, for example it might be via a better DNA repair mechanism, then we might be able to use that and apply it to other cells. Maybe we could use it to design stem cell therapy or other sorts of cancer treatments. That would be fantastic.”
Mutational landscapes and lineages
The fact that all of our cells descend from one fertilised egg enabled the team to asses the ancestry of different cells to establish where they came from during development.
“You can use mutations as markers of a cell’s past, so you can deduce which cells are related to which other cells. You can build a whole family tree, all the way back up to the first cell division. It’s absolutely mind boggling. Especially when you consider the first donor in the study was nearly 80 years old. And we can see in their DNA what happened in the womb of their mother,” said Tim.
“I don’t think when the human genome was first sequenced anyone thought we might be able to answer developmental questions like this,” he added.
The team was able to infer how different cells of the embryo contributed to different organs of the body – something that it is not possible to study in the laboratory.
The other line of analysis was into the mutational rates and ‘signatures’ of the different tissues. Mutational signatures are patterns of DNA change, each with a distinct cause. They were first identified by Mike Stratton and his group at Sanger back in 2013, as they were investigating the causes of cancer4. There are now over 50 recognised signatures, of which half have a known cause. Bioinformatic analysis in the new study detailed for the first time the rates of mutation, and the mutational signatures, in healthy tissues across the whole body.
The team were surprised by the variation they saw, both within, and between, individuals. “It was a black box,” says Raheleh, “We just didn’t know what we were going to find.”
The next steps for the project are already underway, with the team assessing why different tissues have different mutation patterns, and how that relates to our risk of disease. New techniques have been developed by others in the department to sequence even smaller amounts of DNA5.
Luiza said: “What we’ve got is the landscape of mutations across the whole body – we can peer back in time and see how our cells, tissues and organs develop and age. Before now, we’ve had the final snapshot if you like – that is what happens when a cell accumulates so many mutations it becomes cancerous.
“Now, we can begin to understand what normal, healthy DNA mutations look like in our bodies. On top of that, we can lay the knowledge of the last decade of cancer research. We have a more detailed picture than ever before about how our cells age in health and disease.”