Login to Your Account

Excited States of RNA Could Offer Targeting Possibilities

By Anette Breindl
Science Editor

Understanding the structure of a biological molecule, whether it's a protein, DNA or RNA, is often critical for understanding its function.

But for a biological molecule, unlike a bridge or a building, its structure is not an unchanging thing.

Biological molecules are, in fact, somewhat akin to fidgety children – they are constantly changing their shapes, wiggling around just a bit.

For a long time, such wiggling took place out of sight of the prying eyes of scientists, because by far the best available method to look at molecular structures was X-ray crystallography. That method has led to profound scientific insights, starting with the discovery of the structure of DNA by Francis Crick, James Watson and crystallographer Rosalind Franklin almost 60 years ago. But a crystal, by its nature, locks molecules into an unmoving lattice.

As modern imaging technologies such as nuclear magnetic resonance, or NMR, have come into more widespread use, the ever-changing shapes of proteins and nucleic acids have literally come into the view of structural biologists. One emerging realization is that proteins and nucleic acids also go in and out of what scientists have termed "excited states" – spontaneous, often localized and reversible shape changes that are not set off by interactions with another molecule, such as when a receptor binds its ligand or a chaperone folds a protein.

Such excited states are less frequent than the lowest-energy state. But in proteins, Hashim Al-Hashimi told BioWorld Today, they "play critical roles" in areas such as catalysis, binding and signaling.

And they may not be as rare as they look in an NMR sample. Al-Hashimi's group works with nucleic acids, and "in our NMR sample, [excited states] exist at a frequency of around 1 percent," he said.

But the energy difference between the lowest-energy state and the higher-energy excited states is comparatively slight. "The energy difference is one the order of what is contained in a few hydrogen bonds," Al-Hashimi said. And so, in the much more complex and interactive environment of a living cell, "you might end up with a much higher proportion of molecules in the excited state."

Al-Hashimi is a professor of chemistry and biophysics at the University of Michigan. He is a co-founder of Nymirum Inc., a start-up whose goal is to "build on the idea of the dynamic landscape of RNA . . . and use it to do virtual screening." Drug screening methods that are used for protein targets, he added, do not work as well for RNA, which is one reason that, although the known RNA world has increased vastly over the past decade or two, medical targeting of that world has lagged behind.

In the Oct. 7, 2012, advance online edition of Nature, senior author Al-Hashimi and his colleagues described how to work out the structure of RNA excited states, and presented several examples of such excited states – all of which, he pointed out, make potential drug targets.

In proteins, the structure of proteins in excited states can be determined from magnetic resonance imaging data. But for the less-studied nucleic acids, that has not been possible, because part of what it takes to get such structures are large databases – which exist for proteins but not, at this point, for nucleic acids.

The findings now published by Al-Hashimi and his team add to the evidence that such improved screening could open up currently undruggable targets.

The researchers used two very different approaches to solving such structures: computational biology and intuition.

NMR measures the position of hydrogen atoms in a molecule. Getting the actual structure from those measurements is done through structure prediction computer programs. And those programs deliver several possible structures – the one that is in best agreement with the experimental data, as well as a few runners-up.

Those runners-up "are typically ignored by the community," since the top prediction is, in fact, usually the correct one. But Al-Hashimi and his team realized that such second- and third-best predictions may, in fact, be shapes that RNA briefly goes into.

At the same time, NMR scientists will develop an instinct for what sorts of structures make sense for a given dataset. "It has to be energetically favorable – it can't be a wacko structure," he said.

In their work, the team looked for structures where bioinformatics and intuition converged, and identified putative excited states for three separate RNAs. They tested whether such excited states led to changes in function by testing RNAs with mutations that would lock or block the excited state.

The team showed that one of the RNA molecules they looked at, HIV's transactivation response element, or TAR, is inactive when it is in its excited state. That implies that any molecule that binds with TAR to lock it into such a state might interfere with HIV's function.

And excited TAR makes an interesting drug target for another reason. DNA and RNA bases are normally negatively charged. But in the excited state, the TAR RNA has positively charged bases.

"From a medicinal perspective, this is great, because you could finally have a negatively charged small molecule" that would bind to RNA, Al-Hashimi said. In fact, such a negatively charged small molecule would be quite specific to the excited state, since it would be repelled by the negatively charged bases of the regular state. In RNAs that feature positively charged bases in their excited state, "this might be a way to solve the specificity problem of targeting RNA."