Shape-shifting molecular cousins are the key to DNA repair

It’s taken years of frustration and dedication (not to mention countless hours spent in a small room roughly the temperature of a domestic fridge) but the hard work has finally paid off. Our researchers Martin Taylor and Simon Boulton at the Francis Crick Institute have solved a decades-old biological mystery, publishing the results of their molecular sleuthing in the journal Cell.

The story centres on a protein called RAD51 and its closely-related molecular ‘cousins’, known as paralogs, which are involved in helping cells repair DNA damage.

It’s been known for some time that people who inherit a single faulty version of one of the genes that makes these proteins are at higher risk of developing breast and ovarian tumours – for example, our researchers found that inheriting a mistake in a paralog gene called RAD51D increases a woman’s chances of developing ovarian cancer six-fold. Inheriting two broken copies causes a severe genetic syndrome known as Fanconi anaemia, which usually leads to leukaemia.

But until now, exactly how RAD51’s molecular cousins work – and how they contribute to cancer – was completely unknown.

Fixing the damage

To understand what these RAD51 paralogs are up to, we first need to take a look at the head of the family, RAD51 itself, and its role in repairing DNA damage inside cells.

When a strand of DNA is snapped in two, the damage needs to be spotted and repaired as quickly as possible. Failure to fix the problem correctly can lead to cell death, or even cancer.

Cells have two ways of repairing these kinds of mistakes. One approach is to just ‘glue’ the DNA back together again. It’s a simple, quick fix, but if there are a lot of DNA breaks or the two ends have become separated, then there’s a good chance that mistakes will be made.

A more accurate method, used widely across the animal kingdom, is known as homologous recombination, which can only take place once a cell has copied its DNA in preparation to divide – the process by which new cells are made. This means there are two copies of each of the cell’s DNA molecules (more technically known as sister chromatids).

As the diagram below shows, after trimming back the broken ends of the DNA double-helix to reveal single-stranded sections, molecular repair kits then use the other intact, duplicated DNA strand as a template to fill in the gaps around the break.

Double-stranded DNA breaks are repaired by homologous recombination

Into the cold room

Decades of hard graft have revealed the fundamental role RAD51 itself plays in this process. But what about its cousins, the paralogs?

“We know that the RAD51 paralogs are essential for homologous recombination,” Simon told us, “but for the last 30 years everyone’s attempts to find out what they’re doing have hit a brick wall, as we haven’t had the tools to understand what they actually do.”

It all boils down to biochemistry. In order to study protein molecules in the lab, researchers need to purify relatively large amounts for their experiments. Unfortunately, RAD51 paralogs from ‘regular’ sources – human and yeast cells – have proved to be particularly reluctant to behave.

“They’re very difficult to work with,” Simon explains. “They stick together and they stick to DNA, making it hard to do any kind of in-depth analysis.”

To overcome this problem he took a gamble, turning to an organism that he’s worked with for many years: a tiny worm known as C. elegans.

“Martin was at the beginning of his PhD and was aware that what we were embarking on was a real challenge – something that had foxed the entire field for decades. We took a leap of faith and turned to C. elegans because we thought it might offer some simplicity. Our previous work with worms has established in detail how certain DNA repair proteins function, so we came to it with an open mind. But we had no guarantees.”

It was a tough job, and Martin ended up spending the first three of the four years of his PhD trying to come up with ways of purifying two of C. elegans’ own RAD51 paralog proteins, called RFS-1 and RIP-1.

It was arduous work – proteins generally need to be kept cool as they’re purified, to prevent them disintegrating. As a result, he had to spend hours in the cold room at 4⁰ C, and even ended up staying overnight at the institute when things got really busy.

Eventually he made a breakthrough. The trick was to genetically modify the yeast to start producing both proteins at the same time. After a lot of tweaking and testing, Martin was able to grow around a hundred litres of soupy broth packed with yeast cells, all churning out the worm proteins.

Now the real experiments could start.

Putting together the puzzle

Painstaking research over many years has shown that RAD51 coats the single-stranded stretches of DNA before they get patched up. This creates a sturdy, stick-like filament that helps to protect the vulnerable strands from any further accidental damage, and helps them to pair up with the matching DNA template.

But there’s a problem: this RAD51-coated filament is stiff and inflexible – not the best shape for wriggling through the DNA inside the cell’s nucleus in search of its pair.

And as Simon and Martin have now discovered, it’s here that RAD51’s paralog cousins step in.

The first clue came when Martin mixed DNA and RAD51 filaments together with the paralog proteins purified from C. elegans. He noticed that they seemed to change the filaments’ physical properties, suggesting that the paralogs were having some kind of effect on their shape or structure. Other experiments suggested that the fibres were becoming more loosely packed when the paralogs bound to them.

To find out exactly what was going on he had to look closer. Much closer. And to do that, he needed some help.

RAD51 paralog proteins (black blob) binding to the end of a RAD51 and DNA filament.

First, he worked with Raffaella Carzaniga and Lucy Collinson in our Electron Microscopy Unit in London, to see whether the RAD51 paralogs were really making contact with these stiff little filaments.

Zooming in right down to the level of individual molecules, they saw that the RAD51 paralogs (the black blob in the image on the right) were firmly attached to one end of the stick-like RAD51 and DNA fibre.

To find out more about what was going on, they needed yet more specialist help. So Martin headed over to Brno in the Czech Republic to team up with a couple of researchers – Mário Špírek and Lumir Krejčí – who are experts in a technique called stopped flow analysis. This allowed them to look for changes in the RAD51 filaments’ structure over a few thousandths of a second as the paralogs bind to them.

The results of these experiments also hinted that adding the paralog proteins was making the DNA filaments more flexible and bendy, but they still weren’t 100 per cent definitive.

Further proof came from another collaboration, this time from a team at the MRC Clinical Sciences Centre at Imperial College London – Kathy Chaurasiya and David Rueda. They’re leading experts in a technique that can measure the mobility of biological molecules called FRET.

Again, their findings suggested that the RAD51 and DNA filaments become much more flexible and bendy when the paralog proteins bind to them, loosening up the previously compacted structure.

By piecing together all these different clues, the full picture started to emerge.

As Simon explains: “When we look at all the evidence together, we think that the RAD51 paralogs allow the filaments to become more open and flexible, so they can seek out their matching DNA pair and take part in the repair process more effectively.”

The best analogy is a child’s spring-like slinky toy: when it’s compacted together, it forms an inflexible, straight rod. But open it up and it can bend and move all over the place, as shown in the diagram below

RAD51 paralogs enable DNA repair to take place efficiently

Where now?

Figuring out the finer details of the molecular gymnastics involved in DNA repair in tiny worms is all very interesting, but the next step for Simon and his team is to find out whether the human RAD51 paralogs are working in the same way.

“We don’t yet know for sure that they’re important in humans,” he says, “but this is how science works. You have to start somewhere, and I think we have made a large conceptual jump in understanding how these molecules function. The next step is to ask whether this is true in the human situation, and can we exploit what we’ve learned to potentially treat cancer?”

Based on his previous discoveries, showing strong parallels between the worm world and our own, it’s likely that this will be the case, but it still needs to be proved. And then this knowledge needs to be turned into treatments.

“We know the RAD51 paralogs are implicated in human cancers,” says Simon, “and there are lots of opportunities to target them with drugs. Interestingly, they’ve also been implicated in the tropical disease sleeping sickness, so there could be potential for treating that too.”

One approach might be to use drugs that interfere with the RAD51 paralogs to prevent cancer cells repairing their DNA properly – potentially making chemotherapy more effective, at lower doses.

Following up on all these ideas is certainly keeping Simon and his team busy. But they might never have got here at all if it hadn’t been for Martin’s dogged efforts, including all those cold days and late nights. It’s clear that his boss is immensely proud of him, and we’re also happy to have supported his PhD studentship.

“I think this is an example of the importance of persisting,” Simon points out. “One of the things we have to do as scientists is know when we’ve hit a brick wall – when to drop a project and do something else. But this is an example of where persistence pays off.”

He’s also keen to point out the importance of collaboration between researchers, not just in the UK, but around the world.

“That’s just the way science is – it’s multidisciplinary,” Simon explains. “We didn’t know how to do all these different techniques ourselves, so my role was really to engage with our collaborators to make it happen.”

As Simon and his team prepare to move into the new Francis Crick Institute, packed full of researchers from all kinds of scientific fields, we hope there will be many more fruitful collaborations and important discoveries to come.

Kat

Reference

‘Rad51 paralogs remodel pre-synaptic Rad51 filaments to stimulate homologous recombination’ Taylor, M.R.G., Špírek, M., Chaurasiya, K.R., Ward, J.D., Carzaniga, R., Yu, X., Egelman, E.H., Collinson, L.M., Rueda, D., Krejčí, L. and Boulton, S.J. (2015). DOI: 10.1016/j.cell.2015.06.015