Fluorescence imaging is used in a wide number of biological applications, but has resolution limited by the diffraction of light. Super-resolution imaging avoids the limitations of diffraction and allows imaging of single molecules in biological specimens on nanometer length scales. Fluorescence Photoactivation Localization Microscopy (FPALM) is a type of super-resolution fluorescence imaging based on localizing many sets of individual molecules whose visibility can be controlled with light. This is useful for biology because it can allow us to see how individual molecules move. We would like to do simultaneous imaging of three different types of molecules because we can learn about how these three different things interact with each other. The problem has been that imaging more than two types of molecules has been difficult because the emission spectra of fluorophores cover large portions of the visible emission spectrum.
In our experiments, the goal is to localize three different types of fluorescent molecules, or fluorophores, using FPALM imaging. FPALM excites labeled molecules using a laser. The molecule emits fluorescence when it relaxes from the excited state. The FPALM method captures the light emitted using a highly sensitive digital EMCCD camera. The fluorophores that will be localized are photoactivatable variants of three optimized red fluorescent proteins (RFPs) derived from corals, such as mOrange, mCherry and mKate. The optimized RFPs and enhanced variants of a green fluorescent protein (GFP) from a jellyfish Aequoria victoria, are excellent labels of intracellular structures because they are monomeric, have high pH stability, low cytotoxicity and brighter than wild-type fluorescent proteins. Another advantage of RFPs and GFPs is that they are genetically encoded meaning they can be attached genetically to any protein of interest, thus resulting in a so called fusion chimera of that protein with either an RFP or GFP label. Also we can image live cells using enhanced GFPs or RFPs because the excitation wavelength maximum is not in the ultraviolet wavelength range. We will use the photoactivatable variants of mOrange, mCherry and mKate, each having different excitation and emission maxima. The photoactivatable variants were developed using molecular mutagenesis. The maximum emission wavelength is the most probable wavelength of the fluorescence that is emitted from RFPs. The three variants used are PAmKate, which has the longest maximum emission wavelength, PAmCherry, which as the middle maximum emission wavelength, and PAmOrange, which has the shortest maximum emission wavelength. We can determine positions of the specific molecules based on an image of the fluorescence that is detected. To localize each photoactivatable protein, we separate the fluorescence emitted by the source by splitting it into two paths using a dichroic mirror. A dichroic mirror splits a beam of light into two paths based on wavelength. Our dichroic mirror is a long pass dichroic mirror, which means wavelengths higher than the mirror's edge will pass through the mirror and be imaged in one channel, and wavelengths lower than the edge will get reflected and imaged in a second channel. The dichroic mirror has an edge of around 600 nm in our setup. Then we use fluorescence emission filters to narrow the viewed range of wavelengths in each path. To identify the type of each imaged molecule, we count the number of photons detected from the molecule in each channel.
Using FPALM imaging, we were able to obtain clear images of PAmKate, PAmCherry and PAmOrange individually. We were able to measure the fluorescent ratio FR (equal to detected photons in the long-wavelength channel divided by the total detected photons) for a fusion chimera of a transferin receptor with PAmKate, which agreed with our calculated ratio within 0.01. We measured FR for a fusion construct of actin with PAmCherry to agree with our calculated ratio within about 0.06, and we measured FR for PAmOrange purified protein to agree with our calculated ratio within 0.02. This compares well with previous measurements of PAmCherry by two-color FPALM. In our experiments we have successfully imaged three distinct photoactivatable RFPs independently using two-channel FPALM imaging, suggesting that simultaneous imaging of the same three proteins will be possible. Successful simultaneous imaging of three fusion proteins in cells will open the door for many more possibilities in super-resolution fluorescence imaging in biomedical applications.
This work was funded by NIH K25-AI65459, NSF CHE 0722759, Maine Technology Institute MTAF 1106 and 2061, and University of Maine startup funds.