In a 6-part series, we’re exploring the major challenges that are holding back progress in the field of brain tumour research. This fourth post explores how scientists study these complex diseases in the lab and the difficulties they face.
When scientists model cancer in the lab, they’re not sending cells down petri dish catwalks or crafting mini tumours out of clay. Rather, they’re creating a simplified version of the disease that they can probe and learn from.
These models can take many forms, from balls of cells to genetically modified flies and mice with cancer. As technology has advanced, some cancer models don’t even involve living cells at all – scientists can now make virtual tumours on computers and run huge numbers of simulations in a fraction of the time it would take to carry out similar studies on real tissue.
“What’s interesting is what different scientists mean by and expect from a model,” says Professor Steve Pollard, a Cancer Research UK-funded brain tumour expert at the University of Edinburgh.
“In my field of developmental biology, a model system is something that is reduced in complexity to the level where you can still use it quickly and effectively to study something of interest.”
The main issue with modelling brain tumours is that these diseases – like the organ itself – are incredibly complex. This makes translating discoveries from models to people incredibly hard. So much so that experts blame models – at least in part – for the high failure rate of experimental brain tumour treatments in clinical trials.
That’s why, as part of our inaugural Brain Tumour Awards, we’re calling out for scientists to develop better brain tumour models. They’re urgently needed to make the process of bringing new drugs into people faster and more reliable. Ultimately, this could mean patients are no longer presented with the same treatment options as those given decades ago.
Starting with simplicity
One of the major ways that scientists study brain tumours – and indeed any cancer – is by growing cells in plastic dishes. The cells might come from tissue taken during surgery, or biopsy samples if surgery isn’t possible. They offer a cellular snapshot of what’s going on in a patient. And that means scientists can start to unpick what has gone wrong to cause their disease, potentially highlighting ways to develop smarter, more precise treatments based on tumour biology.
But an issue with these models is that scientists are looking at a tumour that has already formed, which misses out early stages of the disease.
We’re not seeing the creation of the tumour. That’s what we need if we’re going to develop ways to detect and treat these diseases early.
– Dr Dan Tennant
“We’re not seeing the creation of the tumour,” says Dr Daniel Tennant, a Cancer Research UK-funded expert on cell models at the University of Birmingham.
“That’s what we need if we’re going to develop ways to detect and treat these diseases early.”
A way around this is make healthy brain cells turn cancerous in the lab by manipulating their DNA, copying genetic mistakes found in brain tumours. This can begin to unravel the changes that occur when tumours first develop. But doing so creates a colony of cells that are genetically identical, which is far from the reality of many brain tumours. Glioblastomas, for example, have been shown to be composed of a ‘patchwork’ of genetically distinct groups of cells.
“A major challenge is developing models that faithfully capture the genetic diversity of the cancer,” says Dr Michelle Monje, an expert in neuroscience from Stanford University in the US. Monje also points out that for childhood brain tumours, disease biology varies with age, adding yet another layer of complexity that needs to be considered when creating models.
Life in plastic, is not fantastic
Another prevailing issue with cell models is that the lab is poles apart from the real complex environment that cancers grow in. Tumours don’t naturally grow as flat, 2D sheets across plastic dishes. Scientists like Tennant are developing ways to more accurately mimic the architecture of brain tumours by growing 3D balls of cells called ‘neurospheres’, suspended in liquid. But he knows that this doesn’t go quite far enough.
“This still misses the brain environment, the blood brain barrier and interactions with immune cells,” he says, adding that a type of immune cell called microglia can make up a significant proportion of some brain tumours. “We’re missing a lot, but unfortunately it’s likely that we’ll never entirely capture the workings of a brain tumour in a cell model, regardless of how complex you make it.”
There is a balance between complexity versus simplicity… You need to have both in parallel.
– Professor Steve Pollard
Despite these pitfalls, Pollard highlights that making things complicated isn’t always the solution.
“There is a balance between complexity versus simplicity,” he says. “The simplistic model gives you quick answers, and they’re reliable and robust. The complex model gives you something that’s more relevant, but the danger is the variability is so great that you never make any significant discoveries.
“You need to have both in parallel.”
So, what does a more complex model look like? Mice are often the go-to as they give researchers the advantage of studying brain tumours in their natural environment. Just like cell models, scientists can take samples from brain tumour patients and grow these in mice. But to make sure the mouse’s immune system doesn’t reject the foreign, invading cells, the animals need to have their immune cells removed. That necessity takes away a crucial element of brain tumour biology.
“These patient-derived models address the issue of genetic diversity seen in brain tumours, but they need to be complemented with genetically engineered mice,” says Monje. She adds that advances in gene editing technologies, such as CRISPR, have made creating these genetically engineered mice faster and easier, so that scientists can mimic the genetic faults driving these tumours and study the consequences.
“These may not capture all aspects of human tumour biology, but they have an intact immune system,” she says.
Making models meaningful
Among all the buts and limitations, Pollard summarises one clear thread that emerges in this intricate field: “There’s no such thing as a single model that’s perfect for all questions.”
So just as brain tumours are too great a challenge for one scientist to solve, no one model will give all the answers. But by uniting researchers to share ideas and work together, hopefully the Cancer Research UK Brain Tumour Awards will help scientists find the ideal mix of methods to begin solving this problem.
“I think it’ll all coalesce,” says Tennant. “The cell work, mouse models; it’ll all come together.”