Dark states of monomeric red fluorescent proteins studied
Care must be taken when analyzing fluorescence correlation spectroscopy of these proteins.
You don’t have to be afraid of the dark, but you should be aware of it, according to researchers from Katholieke Universiteit Leuven in Belgium. In a study, they found that the monomeric red fluorescent proteins mRFP1, mCherry and mStrawberry all exhibit a flickering behavior because they enter and exit light-enhanced dark states. Some of the flickering is on a timescale close to that of diffusion, which can make it hard to make sense of experimental measurements.
Graduate student Jelle Hendrix, a research team member, has had some firsthand experience with this problem, having run into difficulties when trying to fit data to a fluorescence correlation spectroscopy model. Results show what appears to be a fast diffusion component in the microsecond timescale, but the models cannot be adjusted to fit the data.
Red fluorescent proteins are useful in biology because their emission is spectrally separated from that of green fluorescence.
Previous studies had shown that nonmonomeric red fluorescent proteins flicker, an event known to be dependent on light intensity. The research group decided, therefore, to study monomeric red fluorescent proteins in a controlled in vitro environment to better understand the effect of environmental and excitation conditions on the proteins’ dark states.
Such information could be of wide research applicability. The red emission of monomeric red fluorescent proteins has advantages in live-cell studies. The longer wavelength results in less scattering and in lower background fluorescence. In addition, because the emission is spectrally separated from that of GFP, the two can be used together in dual-color applications and in Förster resonance energy transfer measurements.
The researchers investigated three monomeric red fluorescent proteins and measured absorption spectra using a Shimadzu spectrophotometer. They collected excitation and emission spectra using a Photon Technologies International fluorometer. For excitation spectra, they set the excitation monochromator bandwidth to 2 nm and the emission monochromator to 640 to 660 nm. For the emission spectra, they set the excitation monochromator to 530 to 550 nm and the emission monochromator bandwidth to 2 nm.
Researchers recently studied the dark states of monomeric red fluorescent protein. An illustration of the molecule entering and exiting a dark state is shown.
They performed fluorescence correlation spectroscopy on the red fluorescent proteins in solution using a ConfoCor 2 system made by Carl Zeiss of Jena, Germany. This used a 543-nm helium-neon laser to excite the fluorescent protein; the resulting fluorescence was filtered by a long-pass filter that removed any component shorter than 560 nm and then detected with an avalanche photodiode. Hendrix noted that this setup worked well for single-color fluorescence correlation spectroscopy but that it would have trouble in dual-color mode, largely because of spectral crosstalk. For that, the group currently is investigating a technological change that would help.
“Our photonics equipment would be greatly improved by using pulsed interleaved excitation, the alternating excitation of green and red protein, thus separating the emission on the nanosecond timescale and completely avoiding cross-talk while still allowing fluorescence correlation spectroscopy measurements in the microsecond-to-second timescale,” he said.
Besides bulk measurements, the researchers also collected single-molecule data. They generated a 543-nm excitation pulse using a GWU Lasertechnik optical parametric oscillator pumped by a Spectra-Physics Ti:sapphire laser. They focused this onto the sample using an Olympus microscope and other optics. The same optics sent the fluorescence to a PerkinElmer avalanche photodiode, with a time-correlated single-photon-counting card made by Becker & Hickl of Berlin collecting time-resolved data.
Using this setup, they first showed that mRFP1, mCherry and mStrawberry all underwent dark-state formation. They examined the intensity dependence, ramping the excitation light from 2 to 122 kW/cm2. The flickering that signaled the transition from a fluorescent to a dark state proved to be intensity-dependent, with the data indicating at least two different dark states in each of the three fluorescent proteins. One dark state appeared to be always present, whereas the other seemed to exist only when excitation light of enough intensity was present.
They also examined the role pH plays in the dark states, finding that increasing the pH from 7 to 12 resulted in the protein’s spending more and more time in a dark state. Darkening also occurred below 5 pH, but the researchers attributed this to quenching brought about by the acidic environment.
The researchers were puzzled by the pH dependence in a basic environment, said Hendrix. However, by studying the structure of the proteins, they found that the higher pH induced a disruption of an H-bond to the chromophore. This favored conformational changes and made for more frequent dark-state formation.
Single-molecule spectroscopy also revealed reversible transitions from bright to dark states over three timescales. Some of these relaxation times are close to diffusion timescales. The result is that the dark states are rendered invisible to normal fluorescence correlation spectroscopy modeling. However, they still can influence concentration or diffusion parameters, thus throwing off experimental results. The work was reported online Jan. 30, 2008, in Biophysical Journal.
To illustrate potential problems, the investigators expressed a fusion protein of an mRFP and a GFP in live human HeLaP4 cells and then performed fluorescence correlation spectroscopy measurements on the green and red channels simultaneously. The extracted diffusion time was vastly different for the two fluorophores, although they actually were part of the same protein. In cases of very slow diffusion, such as might occur when large protein complexes are formed, the flickering might be less of a problem.
To summarize the findings, Hendrix said, “Semiquantitative fluorescence correlation spectroscopy and fluorescence cross-correlation spectroscopy measurements on mRFPs are indeed possible, but much care is advised when interpreting the experimental data.”
The researchers are using the results to justify future efforts. “We are also investigating the rational mutagenesis of mCherry to develop an mRFP that offers less dark-state formation, which is, of course, our ultimate goal,” he said.
MORE FROM PHOTONICS MEDIA