Photoactivatable, Photoconvertible, and Photoswitchable Fluorescent Proteins

Photoactivatable proteins can be "switched on" from a low fluorescent state to a higher fluorescent state. This switch is happening during less than a second by applying a short light pulse in the violet/blue spectrum and allows detection of dynamic cellular processes. Directed photoactivation of molecules in a region of interest (ROI) can be used to monitor the movement of these activated proteins inside the cell. Whereas in other pulse-chase experimental setups like FRAP one cannot distinguish between proteins which are re-entering the ROI and newly synthesized molecules, photoactivation is one way to bypass this problem.

The first photoactivatable protein was a wtGFP variant and had been created by Lippincott-Schwarz and Patterson. It comprised a single point mutation (T203H), which lead to a very low absorbance in the area from 450 to 550 nm. Being photoactivated with the help of violet light, PA-GFP (photoactivatable green fluorescent protein) switches its absorption maximum from 400 to 504 nm (s. Figure 2). Therefore its fluorescence increases ca. 100-fold when excited with a wavelength of 488 nm, what is reflected in a sharp contrast between activated and non-activated proteins.

The underlying process during the activation of PA-GFP seems to be a light-induced decarboxylation of the glutamic acid side chain in residue 222. This loss of a carbon dioxide alters the chromophore configuration from a neutral to an anionic state [1].

Another candidate which should be mentioned here is Phamret (photoactivation-mediated resonance energy transfer), which is a special tandem dimeric fluorescent protein, photoactivated by FRET. Phamret is a fusion protein, consisting of one PA-GFP coupled covalently to its FRET partner ECFP. Exposure to 458 nm wavelength leads to emission of ECFP at 480 nm. Adjacent photoactivation of PA-GFP is performed with a 405 nm laser beam. After that, repeated exposure to 458 nm leads to FRET between the excited ECFP and PA-GFP, resulting in a green fluorescence. So the "inactive" and "active" form can be excited by the same laser wavelength (s. Figure 3). A negative characteristic could be the size of the protein which is consisting of two fluorescent proteins, possibly evoking sterical problems.

It is noteworthy that photoactivatable proteins often display a reduced brightness compared to EGFP and show a reduced photostability (s.Table 1).

The underlying process of photoconversion is again a light induced process. In this case a histidine residue inside the chromophore (His61-Tyr63-Gly64) is cleaved by irradiation and finally leads to a formation of a highly conjugated dual imidazole ring system. This process is connected to a fluorescence shift to red wavelengths [1].

A red to green convertible FP is coming from the coral Dendronephtytha sp. Its name and the red activatable attribute are mirrored in the expression Dendra [4]. The commercial development Dendra2 is the first monomeric red-to-green photoconvertible protein.

Another widely used optical highlighter is tdEosFP [5,7] and was primarily isolated from Labophyllia hemprichii, a stony coral. The tandem dimer can convert from green to red fluorescence after near-ultraviolet exposure and is gladly used for super-resolution microscopy, what is also true for the monomeric versions mEos and mEos2.

A photoconvertible protein with very similar characteristics to Keade was found in a third stony coral: Favia favus. After near-ultraviolet irradiation tetrameric KikGR changes its fluorescence from green to red, but both in a much brighter way than Kaede. Its commercial variant is called Kikume and can be photoconverted by multiphoton excitation with 750 nm wavelength. Like this it can be used in thick tissue specimens. Mutagenesis of KikGR lead to the monomeric form mKikGR. Together with mEos2 and Dendra2 it is an often used optical highlighter for super-resolution imaging.

Furthermore there is a cis-trans isomerization connected with the protonation of the chromophore. In the neutral state of the chromophore the Tyr66 side chain is in a trans conformation (s. Figure 6). In its anionic state Tyr66 has a cis conformation. Upon irradiation with a 405 nm laser pulse Dronpa is forced into its fluorescent cis conformation. A 488 nm laser pulse then switches the Dronpa conformation to its non-fluorescent trans state. This cycle can be repeated several hundred times.

Besides fluorescent proteins which are susceptible to manipulation by light, there are additional proteins switchable by further triggers (s. Figure 8). For example, researchers who are interested in the ion status of living cells can choose GCaMP as a Ca²⁺ indicator. GCaMP changes its fluorescent behavior depending on its Ca²⁺ binding. Other fluorescent proteins such as VSFP1 are used to detect membrane potentials, whereas the yellow fluorescent protein YFP-H148Q is sensitive to Cl^- ions. Furthermore there are FPs changing their fluorescent characteristics depending on the pH of their surroundings. Superecliptic pHluorin (SEP) and pHuji are two examples. With their help endo- and exocytosis events can be tracked as well as intracellular sorting events.

Fluorescent proteins which are changing their emission spectrum by time are called fluorescent timers. This behavior is not the result of pH change, ionic strength or protein concentration but happens independently from these biochemical parameters. In other words the age of the relevant protein can be estimated by its color. With these characteristics it is possible to measure time – and time depending events – inside a living cell.

The first fluorescent timer was described by the lab of Sergey A. Lukyanov in the year 2000. This DsRed mutant named FT (for fluorescent timer) alters its emission spectrum from red to green wavelengths during 18 hours. Due to the fact that FT was a tetramer Vladislav Verkhusha saw the necessity to create a monomeric one. Mutagenesis of mCherry lead to 3 more fluorescent timers with different transformation times (s. Table 1). These proteins are turning from blue to red in fast, medium or slow time scales but still too slow to measure cellular processes happening during minutes.

Nevertheless, with the help of fluorescent timers it is possible to watch the behavior of proteins in living cells, both in spatial and temporal resolution. For example differentiation and cellular polarization processes can be observed. Moreover protein transport events or viral assembly can be tracked or gene activity can be monitored. This list of applications surely could be extended and shows the potential use of Fluorescent Timers in the future.

Ex_max: Excitation maximum, Em_max: Emission maximum, QY: Quantum Yield, Brightness is calculated by the product of Quantum Yield and molar extinction coefficient divided by 1,000