Raman spectroscopy is the new black
Everyone that has been in a biology wet lab knows about fluorescence microscopy. In an easy case, you incubate a sample with a fluorescent dye like DAPI and Hoechst, and view it under a fluorescence microscope. In a trickier case where antibodies are involved (i.e. immunofluorescence microscopy), you have to block the sample (incubate it with proteins to reduce non-specific antibody binding, usually with serum from the same animal as your secondary antibody), incubate the sample with the fluorescent-labelled antibody, do multiple washes, and in the case where a secondary antibody is needed, repeat these steps.
Anybody who does immunofluorescence also knows it’s a pain in the butt. Antibodies are expensive—a 0.5 mL vial of Alexa Fluor 488, a Polyclonal Antibody, costs $652 CAD, and you have to get a different primary antibody for each sample you intend to stain. Autofluorescence, the phenomenon where background fluorescence mixes with the fluorescence dye’s signal, is a notable issue as well. Additionally, washing antibodies off takes time. If you are doing fluorescence microscopy at a single cell level, then the antibody washes present the additional challenge of cell loss–the cells just wash away as you are pipetting out the supernatant! Non-antibody based fluorescent dyes, on the other hand, are usually limited in their applications—there’s really only good dyes for nucleus stains, some actin fibres stains, endoplasmic reticulum stains, and mitochondrial stains. While you can use fluorescence microscopy on live animals, you cannot do a lot of fluorescence microscopy on a living human (you might not want random antibodies in your body or any exposure to fluorescence in general), limiting its use in intra-operation and other medical contexts.
I can go on and on about my frustrations about fluorescence microscopy, but what are the alternatives? What other tools can we use to visualize live cells in a subcellular resolution with good specificity?
Introducing Raman microscopy: a microscopy technique based on the Raman scattering of lights that enables subcellular resolution and live cell imaging in a label-free manner (no dyes!). Raman scattering refers to the phenomenon when a sample scatters light upon monochromatic light illumination, and a small component of the scattered light has altered frequencies due to interaction with the electronic subsystem of the sample. To put this in a biological context, the Raman spectrum from different biological molecules like lipids, proteins, and nucleic acids will all be different. If we take a series of pictures of a sample under different wavelengths where Raman scattering takes place, we would essentially create a hyperspectral image in which each pixel in the image represents a unique Raman spectrum. If we then unmix these macromolecules’ Raman spectrum from the Raman spectrum of a given sample, we can visualize the distribution of these macromolecules specifically and in sub-cellular resolution. In addition, there can be shifts in Raman signals that correspond tightly to a specific biological change: a decrease in the wavenumber at 939 cm-1, for example. This was observed to be associated with a reduction in the expression of collagen—a compound essential in multiple diseases. Understanding the Raman shifts with strong correlation with a biological process has enabled scientists to detect certain cell states that were previously undetectable or only detectable with long, costly assays. Combining Raman microscopy with machine learning techniques, such as deep learning, can reveal even more cellular mechanisms that are undiscoverable with current techniques.
Raman spectroscopy is the next big thing. It saves you time on sample preparation, money on antibodies and stains, and sanity from all the immunofluorescence BS. The next time your PI asks you to run a new immunofluorescence protocol, tell them (or grow the courage to tell them) about this new and beautiful thing called Raman microscopy.
P.S. They have a Raman microscope at AOMF, MaRs!