Raman spectroscopy comes to flow cytometry

One thread in the history of flow cytometry is the quest to expand the number of parameters that can be measured for each particle. This expansion has involved both cytochemical probe developments and instrumental innovations. Numerous new fluorescent molecules to quantify cellular components have been developed over the years. The development of fluorescently labeled antibodies that have enabled the dissection of the immune system and signaling pathways is perhaps the most notable. The report by Watson et al. (1) in this issue of Cytometry is the first description of a flow cytometer with sufficient sensitivity and spectral resolution to analyze surface enhanced Raman spectra (SERS) emitted by multiple dye- labeled nanoparticls bound to microspheres. Raman scattered light is ubiquitous. Stokes Raman scattering (2), an inelastic scattering of light, occurs when a photon incident on a molecule excites the molecule into a vibrational, or other type of excited state, and is re-emitted with lower energy or longer wavelength. The difference in energy between the incident and scattered photons is a measure of the energy of the excited state. If the molecule is in an excited state when the incident photon is scattered, it can interact in such a manner as to de-excite the molecule, resulting in the scattered photon being more energetic than the incident photon (Anti-Stokes Raman). The more complex a molecule is, the more virbational modes it has, resulting in Raman spectra that are rich in information and can be quite complex with many sharp features. Although the process is weak, especially compared to fluorescence from fluorophors used in cytometry, the Raman scattered light can interfere with measurements of low levels of fluorescence. One can directly observe Raman scattered light by viewing the region of a flow cytometer that the laser beam passes through. It appears as a faint line of light that is red-shifted from the laser beam. The laser light scatters off of the water molecules in the flow cell and results in photons of two wavelengths. One comes from excitation of the stretching of the OH bonds and the other from a molecular vibration in which the angle between the two H atoms oscillates. Silver and gold nanoparticles that have roughened surfaces greatly enhance Raman scattering of photons by molecules on the surface of the particles. A recent article discusses the mechanisms, which are still a matter of debate, that cause surface enhanced Raman scattering (3). The degree of enhancement can be as much as 14 orders of magnitude over nonenhanced Raman scattering, bright enough to record the SERS emissions from a single molecule (3). Compared to fluorescence, the SERS emissions can be on the same order of brightness as a typical fluorophor. However, the SERS spectra are rich in features that reflect the vibrational modes of the molecules bound to the metal surface. Fluorescence spectra, in general, are quite broad and featureless. Fluorescence emission from the types of dyes used in flow cytometry typically emit over a 50–100 nm wide range. If one excludes UV excitation wavelengths, the typical photomultier used in flow cytometry for fluorescence detection has a useful range that is less than 300 nm wide. Even with full spectral analysis, the number of fluorophors that can be measured is limited. The advent of quantum dots was believed to provide a new type of fluorescent body that has a very narrow emission range. While that is true for individual Q-dots, populations of them have wider emission spectra due to the variability in size. Nevertheless, emission spectra of a population of Q-dots is, in general, narrower that most fluorescent dye emission spectra. There are numerous instances of the application of Raman spectroscopy in image cytometry. Taking advantage of the chemical information provided by Raman spectral analysis in a confocal microscope, the locations and interactions of specific molecules within cells have been studied (4-7). Under the right conditions, it is possible to identify and track individual chemical species in a complex background by Raman spectroscopy (8, 9). A more rapid approach to Raman image acquisition was demonstrated by de Grauw et al. (10) in the development of a line scan imaging microscope system used for the analysis of banding patterns in polytene chromosomes. Interestingly, recent work has shown that deconvolution of Raman spectra from unstained cells can yield data on basic biochemical composition (11). Instrumentally, the expansion of the number of measurable parameters has followed, and sometimes led, probe development. The addition of more spectral filters and wavelength splitting mirrors has enabled analysis of as many as six fluorophors plus light scatter excited by a single laser. Multiple excitation lasers multiply the number of labels that can be detected in a single cell or particle. Full spectral analysis with 1 nm resolution was reported recently (12), which will potentially enable addition of more probes excited by a single laser. The article by Watson et al. (1) is the first report of high-resolution spectral analysis by flow cytometry of SERS-complex labeled microspheres. Approximately 8,000 rough metal nanoparticles (∼67 nm diameter) with 1 of 4 dyes on their surface are bound to larger microspheres. Each type of microsphere then has a distinguishable SERS spectrum. The recorded single particle SERS spectra are analyzed by two techniques to distinguish the four dyes. In the first method, the spectra were binned in a manner to emphasize differences between the four types, which is equivalent to having “after the fact” programmable optical filters. Secondly, spectra recorded for all four types of microspheres were analyzed by principal component analysis demonstrating that complete separation could be achieved. This advance opens up several avenues for development. New cellular probes can be developed that have SERS reporter molecules with the possibility of analyzing cellular content of more molecular species within cells. New approaches to microsphere encoding are enabled that can increase the degree of multiplexing available for analyses. This new method will enable flow measurements of samples using SERS enhancement mechanisms that are currently being carried out with static systems. In addition, applications of flow cytometric high-resolution spectral analysis opens up new avenues of investigation in biological studies where additional cellular labels are of value and where more highly multiplexed bead-based analyses are needed. Currently, the analyses rates with high-resolution flow analyses are relatively slow. Future advances in camera readout and digital signal processor speeds will enable higher analysis rates and ultimately sorting based upon spectral striping. Support of the National Flow Cytometry Resource (NIH Grant RR-01315) and contributions by James Freyer are gratefully acknowledged.