Scientists figure out the ‘LinkedIn profile’ of cancer cells

We can learn a lot about a person from his or her LinkedIn profile:

Who do they regularly associate with? What are their interests – and special interests? How proactive are they in seeking contact with their regular connections? And which other, more peripheral, networks have tentacles stretching into the profile?

For the most part, the same applies to cancer cells: The genes which control the growth of cancer cells and their devastation in human beings are part of an extremely complex biochemical network where information is constantly being exchanged – a bit like LinkedIn profiles.

A Danish-American research team has, to put it simply, now hacked into the depths of some of these profiles. Using an ingenious calculation method, their work – recently published in science journal Nature Methods – has added 62 candidates to the list of genes linked to the development of cancer in humans.

“So, the list has grown from 219 to 281,” says Danish system biologist Kasper Lage, head of the research project that received funding from, among others, the Lundbeck Foundation and the American Cancer Society. He runs a bioinformatics programme at the American Harvard Medical School in Boston and, for a number of years, has worked on hacking into the ‘LinkedIn profiles’ of disease genes.

Cancer cells are the main target of Kasper Lage’s hacking attacks, but he is also attempting to apply the same method to figuring out the genetically determined causal drivers behind a range of psychological disorders, including schizophrenia. This part of his research also received funding from the Lundbeck Foundation and is conducted in collaboration with colleagues at the Institute of Biological Psychiatry, which is linked to the medical faculty at the University of Copenhagen and is also under the auspices of the mental health services of the Capital Region of Denmark.

Patients seek answers
When doctors treat cancer – regardless of the type – a certain percentage of patients will end up in the ‘inexplicable’ category. There is no doubt that they have cancer, but the doctors have no idea why they became ill.

“Of course, we could say that the cause of the illness is irrelevant to the individual patient. But it’s not actually that simple. For instance, if you’re diagnosed with lung cancer but have never smoked a cigarette and aren’t aware that you’ve been exposed to other chemical agents proven to cause this type of cancer, you will wonder why you contracted the disease and will, naturally, give it a lot of thought. Being able to give these patients more precise answers would have value in itself, but new knowledge about the causes of the cancers that we, today, categorise as ‘inexplicable’ may also hold the key to new forms of treatment. So, our research activities in this field actually have more than one goal,” Kasper Lage explains.

The percentage of inexplicable causes varies from one form of cancer to another. In the case of lung cancer, for example, it is around 30. And when Kasper Lage and his American colleagues decided to hack as deeply as possible into the genetic ‘LinkedIn profiles’, they used data from lung cancer patients as their starting point.

The scientists received DNA profiles and disease data from a total of 660 patients, 242 of which fell into the ‘inexplicable’ category. And then they switched on their powerful computers – it was time to calculate!

Mathematical model – and mice
Human beings have around 20,000 genes, and Kasper Lage and his colleagues had already mapped the ‘LinkedIn profiles’ of each and every one of these genes. But they now wanted to dig even deeper into these networks and look for connections – other genes – which could prove to be harmful. Kasper Lage explains:

“Over the course of the past few decades, science has developed a number of methods that have identified 219 genes with a high degree of certainty which, on further scrutiny, we can safely say are carcinogenic. When diagnosing cancer, doctors largely look for one of these 219 ‘usual suspects’. But in the case of cancers termed inexplicable, there is also some kind of genetic pattern. In an attempt to identify this pattern in the 242 patients with inexplicable lung cancer, we began to compare their data with the ‘LinkedIn profiles’ of all human genomes.”

Both powerful computers and persistent researchers were needed to go through around 20,000 genes with a fine-toothed comb. But thanks to NetSig – a special calculation model developed by Kasper Lage and his colleagues – they were successful.

“We identified 62 genes which, with very high likelihood, will cause cancer if they mutate,” says Kasper Lage. “These genes have not previously been linked to cancer and, on closer inspection, we discovered that they could explain around 10% of the so-called inexplicable cases of lung cancer we had encountered in our study data.”

To test the tenability of this discovery, Kasper Lage and his colleagues performed a number of tests on animals. They injected cells with mutated versions of some of the 62 genes into mice. They initially tested a total of 23 of the 62 genes, and nine of these quickly proved to cause lung cancer in the laboratory animals.

“It’s a lot of work to test cells in this way so we couldn’t get through all 62 – we’ll have to tackle the rest in stages,” says Kasper Lage. “I’ve no doubt that most of the 62 genes we identified by combing through ‘LinkedIn profiles’ will cause cancer if they mutate. But we don’t know what causes them to mutate in some people and not in others.”

NetSig is now available on the internet and doctors all over the world can download the system and use it for their research – free of charge.

“And the obvious next step is to study all types of cancer using NetSig. The more we know about the genetic factors underlying a type of cancer the easier it will be to design medicine for it because we discover new potential sites of attack,” says Kasper Lage.

The Lundbeck Foundation has twice funded Kasper Lage’s research: a PhD scholarship for DKK 1.6 million in 2014 and a general grant of DKK 10 million in 2016.