Optimally, we recommend using confocal scanning with 488 nm excitation light for the visualization of nonphotoconverted Dendra2. Short pixel dwelling time (the time the laser spends at every point) in the course of commonly applied medium speed scanning (400–800 Hz) by a 488-nm laser line excitation results in zero or negligible photoactivation. In a confocal mode, a low-intensity (0.1%–3% power) 488-nm laser should be used to excite the green fluorescent form (detection at 500–550 nm). A control image in the red channel (excitation by 543-nm laser, detection at 550–670 nm) shows a red signal, which should be close to zero level, before photoactivation. Ideally, you should obtain a clear cell image in the green channel and no signal above background in the red channel. If you see the same pattern for green and red fluorescence in the cell(s) of interest, this indicates that Dendra2 photoconversion has already occurred during the primary visualization. In this case, you should change the field of view using continuous scanning with 488-nm laser, to find Dendra2-expressing cells non-irradiated with a mercury lamp.Photoactivation of Dendra2
The nonactivated Dendra2 absorption spectrum possesses peaks at 386 and 490 nm corresponding to neutral and anionic GFP-like chromophore states (Figure 1B). Excitation at the shorter wavelength peak (e.g., 405-nm laser) leads to a very efficient conversion of Dendra2 into the red fluorescent state, similarly to other Kaede-like proteins. We recommend applying the 405-nm line from a mercury lamp or diode laser to activate Dendra2 in the UV-violet region (see Examples 1 and 2 below). Shorter wavelength light (e.g., 366 nm) can be used as well, but this irradiation is more harmful for cells.
Alternatively, Dendra2 can be activated by irradiation with blue light corresponding to its absorption peak at 490 nm (see Example 3). Compared to UV-violet, blue light activation has both advantages and disadvantages. Blue light is less harmful for living cells. Also, 488-nm lasers are more widely available in confocal microscopes. At the same time, Dendra2 activation with blue light is much less efficient than that with UV-violet light. Careful adjustment of several key parameters is necessary to obtain reliable photoconversion by a 488-nm laser.
This photoconversion appears to be a complex photochemical process, presumably requiring sequential absorption of two photons during some short time interval (15). Thus, the key features of the activating blue light are its intensity and continuity. Only continuous light of sufficient intensity can induce green to red conversion. At the same time, too strong irradiation may lead to bleaching of newly formed red fluorophores. In our experience, a mercury lamp (FITC filter set or similar) represents an efficient blue light source for Dendra2 photoconversion. In a confocal mode, two parameters of the 488-nm laser line are crucial: power and dwell time per pixel (the time that the laser spends at every point). An experimenter should vary these parameters and select optimal conditions for each experimental setup. The most efficient photoconversion occurs in a point bleach mode, when the laser beam stands still at a single pixel for some period of time (50–500 ms). For effective photoconversion in a scanning mode, we strongly recommend using a slow scan zoom mode to apply light onto a chosen cell region.Tracking Dendra2 After Activation
Activated Dendra2 possesses excitation-emission maxima at 553 and 573 nm, respectively. In a confocal mode, the red fluorescent signal can be acquired using 532-, 543-, 561-, or 568-nm excitation laser lines and detected at 550(570)–670 nm. For fast tracking (within the first several seconds after photoconversion), one should immediately start a time series of images to provide information on migration of the red Dendra2 from the region of activation. Since green fluorescence of Dendra2 decreases significantly upon its photoconversion to the red fluorescent form, one can follow migration of nonactivated green fluorescent Dendra2 into the region of activation (similarly to the photobleaching techniques). However, it is important, for such tracking, that a significant portion of the protein should be photoswitched from green to the red fluorescent form to allow monitoring of the green form redistribution.
Preferably, red and green signals from Dendra2 (as well as from any Kaede-like PAFP) should be collected separately, or in a sequential mode, for the clear separation of the two signals, because of the minor emission crosstalk of green fluorescence into the red channel. However, for fast dualcolor monitoring of both signals, simultaneous visualization mode can be used as well, using excitation with 488-nm and, for example, 561-nm laser lines.