Correcting optical aberrations in fluorescence microscopy

X-ray crystallography is still the only way to obtain the full structure of a protein, but crystallysing membrane proteins is notoriously fiddly.

Membrane proteins have a habit of resisting crystallisation attempts, so for many years the best and easiest technique for probing biological molecules in the one to 10 nanometre range has been fluorescence resonance energy transfer (FRET).

FRET measures the transfer of energy from one fluorescent probe called the donor to another probe, the acceptor, and thus the distance between them.

It is often used to investigate protein-protein interactions, but its efficiency depends on the orientation between the two probes and knowing how the probes are oriented relative to each other can be difficult.

A new experimental technique developed by Dr Pascal Vallotton, head of the biotech imaging group at CSIRO Mathematical and Information Sciences, may prove a cheaper and easier way to measure distances with nanometer accuracy in both two and three dimensions, using conventional fluorescence instrumentation.

Called Differential Aberration Correction (DAC) microscopy, the technique is aimed at bridging the gap between the resolution achieved by FRET in the 1-10 nm range and that of conventional diffraction limited microscopy beyond 300 nm.

“Our method is completely different from FRET,” he says. “It does not rely on the physical effect – the transfer of energy from one fluorophore to the other. It simply uses fluorescence filters to look first at one fluorophore and then the other one, thus avoiding the overlap of their image.

“By using the right image processing algorithms, it is possible to extend the useful range for making these distance measurements from 0-10 nm to 0-1 micron, in principle.”

---PB--- Methodology

In the DAC method, two single fluorescent interrogation probes emitting two different wavelengths are imaged. The two resulting images do not overlap exactly even if the two probes are co-localised because of chromatic aberrations, so the precise position of each probes has to be corrected using a mathematical algorithm.

To achieve this, reference probes consisting of the same dyes as the interrogation probes are used to estimate an aberration deformation field. This estimate is then applied to correct the position of the interrogation probes, and thus measure the distance between them accurately.

“The method is ‘differential’ because the probes of interest are affected in exactly the same manner as the reference probes,” Vallotton says. “It relies on a computational approach to accurately correct optical aberrations over the whole field of view.”

The method uses fairly standard instrumentation, with a few added extras. Vallotton uses as an analogy the difficulties encountered with the Hubble telescope. “Just before going into space and repairing the Hubble telescope, which was myopic, scientists developed very similar algorithms to cure it” he says.

“We are using a similar type of approach except that in this case, our microscope has reached its fundamental limits: they can only be improved computationally.”

The idea is that DAC will act as a cheaper and simpler complement to X-ray crystallography, especially when it comes to membrane proteins. It would be far easier – and better – to look at proteins in their natural, uncrystallised state, and this is what he hopes the technique will eventually achieve.

“To be fair, we are not quite there yet: We are using fluorescent microspheres that measure about 100 nm in diameter for our proof of principles,” he says.

“And X-ray crystallography gives you the full structure of the protein, with the arrangement of every atom. Our technique only gives you one distance between two positions in that protein, but it works in solution and – in principle – even in vivo.

“Eventually, we would like everybody to be able to do these measurements in the lab.”