So, I sat with the biophysicist who built the microscope that I intend to use. We discussed my proposed experiment and the different capabilities of that particular microscope.
Immediately we discovered that I will not be able to use LSSmKat1. If you recall, the emission peak of LSSmKate1 is at 624nm. However, the dichromatic mirrors set in that microscope for the red channels are from ~500 to 620 and from 650 to 750. So the 624 falls just in the waveband that I cannot detect.
So I remain with LSSmKate2, with em. Peak at 605.
Ok, now, the available lasers in this microscope are 407, 436, 488, 561 and 640nm.
GFP is excited at 488. That the peak and that’s good. We will use the 525/30 filter (i.e. from 510 to 540).
mCherry is excited at 587, so using the 561 laser we can get a fair signal (~60%). We will use the 593/40 filter (i.e. from 573 to 613).
This means that in both cases I collect only part of the emitted photons (the ones within the filter range, see fig 1).
spectra of GFP & mCherry, with filters. (image: BD spectra viewer)
LSSmKate2 can be excited with the 436 or 488 laser. The 488 will give a ~8% signal for mCherry. But the 436 will not. Although LSSmKate2 is 3-fold less bright than mCherry, and is less photostable, to object using this marker is only for time zero, when the GFP and mCherry signal should be low. However, this needs to be experimentally tested. It could be that the mCherry signal could be mathematically subtracted later.
The problem is that I would not be able to excite simultaneously at 488 and 561 to get simultaneous data of GFP and mCherry. This is because I would also have LSSmKate2 excited. However, for my purposes, I do not have to take the images simultaneously. The microscope can take images with the different lasers at 50-100ms time intervals. Since I expect to detect changes in the seconds to minutes range, a 0.1sec difference between the green and red channel is acceptable.
There is another option and that is to use a far-red protein, such as TagRFP657.
| FP | Max ex. | Max em. | QY | EC(M-1 cm-1) | Brightness | Relative brightness | Photostability |
| TagRFP657 | 611 | 657 | 0.10 | 34,000 | 3400 | 0.1 | 55 sec |
| LSSmOrange | 437 | 572 | 0.45 | 52,000 | 21400 | 0.82 | 10 sec |
The advantage is that we can use the 650-750 dichromatic, and use excitation at 640, which still gives a fair excitation, and it will not excite mCherry. The down side is that the 561 laser will excite TagRFP657 too. However, the TagRFP657 emission at the 593/40 filter is fairly low and with a less than maximum excitation to begin with, and a lower brightness and photostability (compared to mCherry), this should not be a big problem (see figure 2) . But again – should be tested experimentally.
figure 2: TagRFP657 spectra (From: Morozova et al. Biophys J. 2010 99(2):L13-5.)
So here is another lesson: when designing a multi-color experiment, you need to take into account both the FP properties, and the capabilities of your microscope (lasers, filters, dichromatics, speed etc…).
Update 22.6.12:
After further discussion, another LSS FP was suggested, LSSmOrange. I added its characteristics to the table above. It is brighter than TagRFP657, and the Stokes shift of LSSmOrange is better than LSSmKate2 in relation to mCherry (bluer excitation, which means NO excitation of mCherry.) So I will test all three, and hopefully have an answer which is better within 2-3 weeks.
greenfluorescentblog 