It’s all talk

In my previous blog posts I have mentioned the importance of a good relationship between science and the media a couple of times. The MMR jab scare perfectly exemplifies the danger of a breakdown in this line of communication. With measles cases still rising, the papers are full of articles about where it all went wrong; who was to blame, and how such a fiasco could be prevented from happening again.  In thinking about this, it’s very easy to point the blame squarely at the media. The journalists reporting the scare failed to understand the science, or lack thereof, behind the claims. But scientists must also take a more active role in maintaining a good relationship with the media.

The thing about a lot of science is, it’s quite complicated. Often researchers themselves struggle to get to grips with the concepts behind scientific disciplines other than their chosen field. To explain such work clearly to a non-scientific audience is even harder. Many’s the time I’ve tried to explain what I do to a friend only to be met with  blank face and a speedy change of subject.  But, if scientists are capable of explaining research clearly and concisely to the general public, the risks of another scare like MMR are bound to decrease. The key to improving any skill is practise.  With that in mind, I recently entered a science communications competition.

The aim of the wordily-named ‘access to understanding science writing competition 2013’ is to encourage scientists to practise communicating complex topics in a way that a member of the general public can understand. Entrants had to choose a scientific paper from a list and explain the concepts in simple terms.  Although easy to understand, the information must not be dumbed down. The entrant also had to explain why the research was interesting, and, importantly, whether there were pitfalls in the work. This is particularly significant; if journalists had been able to understand the flaws in the MMR-autism research, this scare might have been avoided.

So, into the fold I went, discussing a paper from the journal Nature about the difference between treatable and drug-resistant breast cancers. I didn’t win.  I’m not surprised. The winning article was an excellent explanation of some research into osteoporosis. What did surprise me was just how hard it was to explain all the research clearly, without missing out complex ideas or methods. I have a long way to go, but I’m determined to keep practising. If scientists can learn to step outside so their specialist bubble and engage the public, maybe we can decrease the likelihood of scares like MMR in the future.

I’ve included my article below. Can anyone follow it?


Scientists have discovered important differences between how the protein oestrogen receptor (ER) acts in breast cancers that respond well to drugs and those that become resistant, which may allow doctors to better predict outcomes.

Understanding ER is key because most breast cancers are ‘hormone-responsive’ – the cancer cells grow in response to oestrogen, which acts through ER. Breast cancers of this type are treated with drugs that stop ER working. Most patients can be treated with these drugs (referred to as a ‘good’ outcome), but certain patients become resistant to the drugs used (a ‘poor’ outcome). Although doctors can usually tell which patients will become drug-resistant, their predictions are not 100% accurate. Drug-resistant breast cancers are likely to be fatal, as alternative drugs have not been found.

Jason Carroll and colleagues wanted to understand the differences between how ER works in breast cancers with ‘good’ and ‘poor’ outcomes by studying how the protein interacts with DNA. Many proteins sit on the DNA and give signals to nearby genes. This is called DNA binding. ER binds to many sections of DNA and tells nearby genes to make protein. The group reasoned that by studying these sections of DNA in patients, they would better understand how ER works. They chose to work with human samples to give the most accurate results possible. However, human samples are limited, so the experiments used only a small sample number. They categorised the samples as ‘good’ or ‘poor’, depending on whether the patient that gave the sample had treatable or drug-resistant cancer, respectively.

The team used a 2-step technique known as ChIP-seq, for studying protein-DNA binding in a cell. First, they removed the ER protein from the cell, bringing with it small sections of DNA on which it is sitting. Then, they studied this DNA to determine to which parts of the DNA ER binds. This allowed them to see which genes are close to ER, and so are likely to be receiving signals to make protein.

When comparing ‘good’ and ‘poor’ cancers, the team realised that ER binds to different areas of DNA in each group, telling different genes to make protein. This is called an alternative ‘binding signature’. In each group, ER was likely to be close to genes that are already known to make many proteins in that group. So, for example, in the ‘poor’ group, ER sits close to a gene making the protein ERBB2. In ‘poor’ cancers, more ERBB2 is made than in ‘good’ cancers. The team looked at this signature in a much larger group of patients and realised that it can accurately predict which patients have drug-resistant cancer. This means that doctors may be able to use the new binding signatures to increase the accuracy with which they can predict outcomes.

Why is ER binding in different places in the ‘poor’ group? To answer this, the team looked in more detail at the DNA. The ChIP-seq experiments can show what other proteins might bind close by to help ER act, as proteins often work together. They found that a protein called FOXA1 was more likely to bind close to ER in the ‘poor’ cancers. FOXA1 is involved in uncoiling sections of DNA, which is tightly packaged in a cell, to allow other proteins (including ER) to bind to the unwound sections.

Carroll’s team have previously published research showing that FOXA1 helps ER to continue to bind to DNA in drug-resistant breast cancers. Now, they took this work further by predicting that FOXA1 was making ER bind in different places in the ‘good’ and ‘poor’ cancers, and so send signals to different genes. They showed that this was likely to be true by performing an experiment using cells grown in a Petri dish that behave like human breast cancer cells. They put a substance on the cells that is known to change the binding signature of ER protein. They then showed that where ER moves, FOXA1 also moves. The team believe that in the ‘poor’ cancers, FOXA1 is telling ER to bind to different sections of DNA, although they do not understand why.

These experiments are important for several reasons. Firstly, the scientists have used human breast cancer samples in ChIP-seq experiments to study ER, which has not been done before. Secondly, they show that in ‘poor’ cancers, ER has a different binding signature. This signature could be used to give breast cancer patients a more accurate prognosis. Thirdly, the changes in ER binding signature might come about due to movement of FOXA1. Could we design a new drug to stop FOXA1 changing how ER acts? If successfully repeated on a larger group of samples, the work done by this group could indicate that preventing FOXA1 working is a potential way to treat drug-resistant breast cancers.

2 thoughts on “It’s all talk

  1. I like the ‘could’s and ‘might’s in your conclusion, but they are really the problem. The media likes will, which is why, according to the Daily Mail, most substances both cause and prevent cancer!

    • I couldn’t agree more, it comes back again to newspapers failing to portray the science properly. Any caveats of the research are ignored; the papers prefer a simple, attention-grabbing story.

What do you think?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s