Fluorescence microscopy has transformed the life sciences. By attaching fluorescent dyes or proteins to cellular structures, researchers can image fine cellular morphology; track molecular localization, motion, and dynamics; and more. But fluorescence microscopy also presents significant obstacles. One of those is multiplexing.
In a multiplexing experiment, researchers select fluorophores that absorb and emit light in different parts of the visible spectrum in order to detect multiple molecules or events at once. But because those emission and excitation spectra are relatively broad, it is difficult in practice to image more than a handful of colors before the signals start to overlap. Now a new collection of molecular dyes promises to ease that limitation.
In a paper published 15 January in Nature Methods, Wei Min, a biophysical chemist at Columbia University in New York City, and colleagues describe a suite of molecular dyes that provide up to 20-plex multiplexing in a single experiment. They then used those dyes to stain 10 intracellular structures simultaneously, and to develop reagents that support highly multiplexed bioassays.
Called the ‘carbon rainbow’, or ‘Carbow’, the dyes exploit a technique called Raman scattering, a fundamentally different physical process from fluorescence.
In fluorescence, Min explains, molecules absorb energy of a specific wavelength, which ‘excites’ their electrons. As those electrons relax back to the ground state, they emit a photon of lower-energy (longer-wavelength) light — the fluorescent signal. In Raman, by contrast, the incoming light is scattered by interaction with the dye molecules, causing its wavelength to shift in characteristic ways. By measuring that color shift, researchers can determine what molecule the light interacted with.
As it turns out, Min says, these scattering spectra are much sharper than those used in fluorescence. Where a typical fluorophore might have an emission spectrum that is 50-nm wide, Raman spectra are about 100-times sharper, he says, leaving plenty of room for multiple easily resolved molecules. And Raman dyes are largely free of the photobleaching concerns that plague most fluorophores.
Min’s team constructed a collection of ‘polyynes’ — basically two benzene rings separated by a linker of alternating single and triple carbon-carbon bonds, which produce a strong Raman scattering signal. By alternating the length of that linker, doping it with heavy carbon isotopes, or adding different chemical groups to the benzene rings, the team ‘fine-tuned’ their properties to produce a palette of colors. Then, to make them amenable to cell biology, they conjugated those molecules to organelle-targeting groups or antibodies, and imaged them in both live and dead cells.
In a separate experiment, the team loaded microscopic polystyrene beads with different dye concentrations to create unique molecular barcodes, which can be used in highly multiplexed protein quantitation assays. Using 10 dyes at three concentration levels, for instance, supports some 59,048 resolvable barcodes, Min notes. (For comparison, Luminex Corp.’s FLEXMAP 3D technology, which is used for multiplexed genetic analyses, supports 500 unique codes.)
This isn’t the first time Min’s team has developed Raman probes for multiplexed microscopy. In a study in Nature published in April 2017, his team described a palette of 14 dyes based on a xanthene scaffold for a technique called MARS – Manhattan Raman Scattering. But as explained in the new study, Carbow dyes may be easier to use. “Carbow peaks are more evenly spaced and well resolved with substantially less cross-talk. For instance, the spectral separation of the closest Carbow peaks nearly doubles that in the MARS dyes,” the authors write.
So what kind of experiments might need that kind of multiplexing? According to Min, cell biologists could use it to answer questions that have been difficult or impossible to address using traditional fluorescent dyes. For instance, by attaching different dyes to different organelle-targeting chemicals and feeding them to live cells, it should be possible to record the dynamics of organelle motion and interaction in real time. And with some tweaks to the chemistry, he says, it should also be possible to boost the dyes’ brightness to enable single-molecule imaging.
Another possibility involves blending Carbow with commercial gold nanoparticles (which provide a strong boost to the Raman signal) for immunohistochemistry applications.
Carbow reagents are not yet commercially available, though Min says he is willing to share them with academic collaborators. But the microscopy hardware itself could prove to be a bigger hurdle.
Raman dyes cannot be imaged in a traditional fluorescence microscope; they require a Raman microscope, which has different types of lasers and detectors. Such instruments, available from companies such as Renishaw and HORIBA (or as a DIY project), are rarely found in life science laboratories, Min says, but are more popular in the physical and materials sciences. If you have access to one, and there’s a highly multiplexed experiment you’ve been dying to try, you might give Carbow-loading a try.
Jeffrey Perkel is Technology Editor, Nature
Image: Nature Methods