Making Brighter Protein Predictions

Supercomputer simulations shorten development time of rigid fluorescent molecules used to clarify protein structure and dynamics

Computer models that describe how the vibrations of fluorescent probes influence light absorption and emission could lead to more accurate protein measurements. © Pobytov/DigitalVision Vectors/Getty

Most methods for the structural characterization of biomolecules, such as X-ray crystallography or electron microscopy, require static or crystallized samples. Attaching fluorescent molecules to protein surfaces, however, enables direct imaging of dynamic biomolecular interactions using light, which could be improved, say A*STAR researchers, with predictive modeling of fluorescence lifetimes1.

Fluorescence normally involves single molecules that spontaneously absorb light and then re-emit it as a different color. But under the right conditions, an absorbed photon can hop from a donor molecule to a nearby acceptor compound that also fluoresces. Researchers have recently exploited the strong distance dependence of this effect to produce ‘spectroscopic rulers’ that measure the nanoscale dynamics between donor and acceptor probes attached to different parts of a protein backbone.

A key challenge is to make spectroscopic rulers with acceptable accuracy. Conventional fluorophores have large, flexible structures that press against proteins in multiple ways, making it tricky to gage the ruler’s length. So to seek alternatives, Tsz Sian Chwee and co-workers from the A*STAR Institute of High Performance Computing investigated whether they could calculate the fluorescence of stiff and small molecules known as syn-bimanes, and then use such theories for probe design.

Typical quantum chemistry approaches, however, have trouble computing properties when a molecule absorbs a photon and enters an excited state. Chwee and his team hoped to overcome these inaccuracies using time-dependent density functional theory that treats the problem of excited electrons with an ‘exchange–correlation’ algorithm derived partly from experiments.

“Time-dependent density functional theory is used by the scientific community to study phenomenon such as absorption and emission, but the full potential of this approach hasn’t been harnessed yet,” says Chwee.

Using fluorescence lifetimes as a test parameter, the researchers compared how different exchange–correlation theories simulated syn-bimanes in realistic, solvent-filled situations. These trials revealed that models incorporating vibronic interactions — the synchronized coupling of molecular vibrations to electronic excitations — provided the most accurate predictions of fluorescent lifetimes. They discovered several exchange–correlation functions that are capable of handling these equations at minimal computational cost.

“Vibronic aspects have largely been overlooked, even though they play decisive roles in the photophysics of fluorescent molecules,” notes Chwee. “While we carried out our calculations on supercomputers, the computational resources are modest enough they could have been completed on a modern workstation in a couple of weeks.”

Chwee anticipates that rapid analysis using density functional theories might be better at spotting rare fluorescent probe candidates with strong absorption and tunable emission properties.

The A*STAR-affiliated researchers contributing to this research are from the Institute of High Performance Computing. For more information about the team’s research, please visit theMaterials Science & Engineering Department webpage.

Source : A*STAR Research