Maturation Block – Redefining Childhood Cancers

Over a lifetime, our cells are exposed to agents that can damage their DNA. UV radiation, certain chemicals and processes within a cell itself can all have an effect. Damage to a cell's DNA can accumulate. In some cases, altered DNA may enable a cell to escape the checks and balances that regulate it, and it may divide uncontrollably, forming a tumour. And so cancer is usually a disease of old age.

Childhood cancers are different. There hasn’t been a lot of exposure to anything. Some are caused by genetic changes inherited from a parent, but the vast majority are not. They form during pregnancy as a fertilised egg grows, and in many cases the cells which form tumours are seemingly ‘stuck’ in their development. There has been a long standing interest in the starting point of childhood cancers - uncovering their causes could hold the key to developing new treatments which push cells to mature, rather than kill them.

While it is not possible to directly access these cells as they develop, it is now possible to study the origin of childhood cancers in more detail than ever before. Sam Behjati, a Group Leader and Wellcome Intermediate Clinical Fellow at the Sanger Institute, is studying the origins of childhood cancer using cutting-edge genomic techniques.

What started as a “stab in the dark” a few years ago has become his major research programme which is now beginning to bear fruit and brings the possibility of new treatments.

The history of a cell

From the moment an egg fuses to a sperm, each time a cell divides it passes a copy of its genome to its descendants. As the 6 billion DNA letters of the genome are replicated, mistakes can be made. One letter might be replaced with another. Or, strings of letters might be missed out or swapped around. If these changes, or mutations, occur early in the embryo, they will be passed on to many different organs. Our bodies become mosaics, with different populations of cells harbouring ever so slightly different genomes, mostly with no ill-effects.

The precise DNA mutations in cell can serve as a ‘postcode’ that shows their developmental origins. In 2014, in a proof-of-principle experiment in mice, Sam showed it was possible to analyse genome data from mature cells to pinpoint their foetal beginnings. The team were able to use the ‘mutational postcodes’ to reconstruct embryo development and study the origins of tissues and organs¹.

How does this help us understand cancer?

Knowing the origins of a cancer cell can help identify the precise genetic mutations responsible for its growth. Understanding the genetic and cellular mechanisms that drive a disease may ultimately lead to new treatments.

“A lot of adult cancers are formed from precursors,” explains Sam. “Barrett's oesophagus is an example. This is a condition where some cells in the lining of the food pipe change shape. Mostly, they don’t cause any problems, but in about 10 per cent of cases, these cells can become cancerous.”

“There is a path from healthy cell, to pre-malignant cell, to cancer. There are cells which are sort of ancestors.”

Conventional wisdom said there hasn’t been time for this path to be the route for childhood cancers. “Conceptually, thinking about cancer in this way is not paediatrics,” says Sam. “Intuitively, it seems very unlikely that this could be a possibility in children and therefore people had not really looked into precursors of childhood cancers.”

Sam decided to sequence and anlyse childhood cancer cells to try and uncover their origins. He says he just ‘gave it go’. The technology was in place and there was freedom to go ahead. These are luxuries of being a Group Leader at Sanger that Sam is very thankful for.

“I have to admit I didn’t know what we would find, but I knew it was worth trying. A lot of our experiments are like that. You don’t fully understand them until you look at the data afterwards,” he adds.

The origins of kidney cancer

The first childhood cancer Sam’s team looked at was Wilm’s tumour. It is rare, with around 80 cases per year diagnosed in the UK. Nine out of ten cases are curable by surgery to remove the affected kidney, together with chemotherapy and sometimes radiotherapy.

The team studied tumour samples alongside normal kidney tissue samples from the same children. Comparing the genome sequences, they saw that in two-thirds of children, DNA changes associated with the disease were found in both normal kidney tissue and the tumour tissue. This allowed the researchers to identify patches of genetically abnormal cells which, when looked at down a microscope, appeared to be normal².

Tim Coorens, a bioinformatician in Sam’s group, led the analysis work. “A good analogy for what we found is that Wilms’ tumour is not just an isolated weed on an otherwise well-maintained field.” He describes the cell as having ‘roots’ in the kidney, which are the precursors to the tumour. They may look normal, but they carry cancer mutations in their DNA. To cure the cancer, those cells need to be completely removed. “Now we know we need to look for the patch of soil where the root has taken hold. If we remove that patch, the weed isn’t going to return.”

Sam didn’t believe it when he first saw the results. The tumours were arising from pre-malignant cells.

Sam says some of the reactions to their findings were predictable: “People said…..very interesting but that must be a Wilm’s tumour specific thing.”

So they moved on to look at malignant rhabdoid tumour (MRT), with similar results³. MRT is even rarer than Wilm’s tumour - just four or five cases a year are diagnosed in the UK. MRT is one of the childhood cancers with the poorest outcomes.

For this study, the team compared the genome sequences of two MRT tumours, alongside their corresponding healthy tissues. The researchers then did phylogenetic analyses of the mutations in the diseased and healthy tissue, to rebuild the timeline of normal and abnormal development.

The analyses confirmed that MRT develops from progenitor cells on their way to becoming Schwann cells, a cell type that insulates nerves. They found a mutation in the SMARCB1 gene that blocks the normal development of these cells, which can then go on to form MRT.

Collaborators at the Princess Máxima Center for Paediatric Oncology in the Netherlands created organoids using the MRT patients’ cells. These laboratory-grown clusters of cells represent the children’s tumours. The researchers added an intact version of the SMARCB1 gene into the organoid cells. Introducing the undamaged gene led the cells to develop normally.

Based on analyses of the cells’ functions, and predictions from the organoid experiments, the teams identified two existing medicines that overcome the maturation block. These could be used to treat children with MRT – a condition that is currently fatal.

“We have gone from hypothesis to discovery of origin to possible treatments for the disease in just over a year,” says Sam.

Arrested development

New applications of single-cell RNA sequencing (scRNA-seq) have been behind many of Sam’s team’s discoveries. scRNA-seq determines the quantity and sequences of all of the RNA in an individual cell. The amounts of each RNA molecule in a cell relates to how active individual genes are. The method was first established in the 1990’s, but is constantly being enhanced.

Single cells can now be analysed at scale, thousands at a time. Cells can also be analysed in place, in the tissues they exist in. scRNA-seq is giving researchers extraordinary details about the cells that make up our organs and bodies.

This technique is being used on a huge scale by the Human Cell Atlas project, which aims to characterise all the cell types in the body. The initiative was co-founded by Dr Sarah Teichmann, who heads up the Sanger Institute Cellular Genetics Programme where Sam is based. Hundreds of researchers and institutes across the globe are working towards cellular ‘maps’ of the human body. Each cell type and cell population are being characterised by the genes they are using, among other things.

Data from the Human Cell Atlas gives Sam’s team a ‘healthy’ reference to compare with their scRNA-seq data from children’s cancer cells. The team has confirmed that neuroblastoma (a childhood brain cancer), and seven types of childhood kidney cancer cells closely resemble specific types of foetal cells⁴. The cancer cells are somehow ‘stuck’ in development.

Their findings add weight to growing evidence that childhood cancers arise from cells that are blocked from maturing normally⁵. Understanding these cells is key, says Sam. They have features that don’t exist in mature cells, which may be possible to target with new or existing drugs. It is feasible that the cells could be ‘pushed’ to develop normally.

For example, his findings in neuroblastoma show that there is just one cell type responsible for the disease, despite the various ways the disease presents and progresses. The cell is not present in healthy tissue in either adults or children, making it an attractive drug target.

Towards prevention

For Wilm’s tumour, his findings of cancer precursors create the possibility of detecting tumours before they form and perhaps of preventing the disease altogether.

“Wilm’s tumour is more common than some of the conditions we screen all newborns for through the ‘heel prick test’,” Sam says. “Therefore, if we had methods to detect Wilms precursors, for example, we might be able to screen every child in the UK at birth for them.”

Sam says partnerships are the only way to put his research into practice. “As we continue with the genomic analysis, we are building collaborations with people who can take each finding forwards – for more research, or into clinical trials.”

Sam wants to continue doing fundamental research in genomics. Understanding our DNA, our genome, and how it works, continues to bring huge advances in medicine – not just in cancer. DNA or RNA codes for the proteins that build our cells, our bodies, and all life on Earth. Delving into genome sequences is going to be worthwhile, even when you’re not sure what you will find.