Resonance raman probes for organelle-specific labeling in live cells

Raman spectroscopic imaging is among the most powerful tools available for the analysis of molecular organization of cells and tissues. Raman spectroscopy relies on inelastic scattering of incident monochromatic light, wherein the energy of photons is being changed, either by a Stokes or anti-Stokes process, upon interaction with molecules of the sample. Light scattering on different types of biomolecules generates corresponding bands in the Raman spectra, which enables to resolve certain amino acids, proteins, various classes of lipids and saccharides, as well as DNA and RNA. It is important to note that the intensity of Raman scattering is linearly dependent on the molecular concentrations at the site of spectral measurement, which uniquely allows for quantitative mapping of biomolecular distribution in situ1,2. This valuable feature of Raman spectroscopy has been realized using a Biomolecular Component Analysis (BCA), a powerful algorithm that identifies concentrations of different molecular groups which collectively contribute to the Raman spectrum of the sample3. In this regard, Raman spectroscopy has a breakthrough potential for the development of innovatory “omics” technologies (e.g. proteinomics, metabolomics, and lipidomics) at a single-organelle level. An ultimate goal of these research disciplines is a comprehensive characterization and monitoring of biochemical composition in specific cellular organelles, to unravel mechanisms of cellular regulation4,5. Up to date, most of the data on the molecular composition of subcellular structures have been obtained by mass-spectroscopy. Using this technique, several thousand diverse molecular species have been identified in various cellular compartments5,6,7,8. However, molecular profiling with mass-spectroscopy involves cell fractionation and extraction of biomolecules from various organelles, which inherently produces artifacts and is not compatible with live systems9. At the same time, conventional live cell imaging techniques, such as fluorescence microscopy, can identify only a few molecular species at a time, which significantly limits their efficiency for comprehensive molecular characterization of single cells and subcellular structures10,11,12,13. In comparison, Raman spectroscopy interrogates all molecules present in the sampling volume of the excitation beam, independently of any extrinsic labels, which is a major strength of this technique. Besides, the Raman signal intensity is not susceptible to photobleaching, thus enabling for long term monitoring of biological samples. Although Raman spectroscopy provides multiple benefits for molecular analysis of a cell, its capabilities as a single-organelle tool are rather limited. A major obstacle of this technique for subcellular analysis is that the location of specific organelles should be resolved to target acquisition of Raman spectra and analyze the content of a single organelle. Meanwhile, subcellular compartments, with a partial exception of mitochondria1,14, do not exhibit organelle-specific vibrational bands and, therefore, cannot be recognized by the Raman imaging of intrinsic cellular components. To circumvent this limitation, several research groups, including ours, have applied a bi-modal fluorescence/Raman approach, wherein conventional fluorescence reporters are used to target Raman spectral acquisition to specific cellular compartments3,15,16. However, it became apparent that conventional fluorescence reporters are not well suitable for Raman imaging, mainly due to a strong fluorescence background, masking the Raman signal. Moreover, fluorophores are quickly photobleached at high signal excitation power densities used in the Raman technique, thus making extended monitoring of organelles in live cells impossible17. In parallel to experimentations with fluorescence probes, a first generation of designated Raman reporters has been developed on the basis of deuterium, nitrile or alkyl containing molecular groups, which produce distinctive vibrational bands in the biologically silent region of Raman spectrum17,18. This advancement, for the first time, enabled detection of specific cellular structures in the Raman modality. However, the signal intensity of the aforementioned Raman tags does not exceed that of native cellular biomolecules, which implies severe limitations in the detection sensitivity. Since, exogenous molecular probes are typically applied at low (nanomolar to micromolar) concentrations to avoid cytotoxicity, the signal intensity from currently existing probes may not be sufficient for detection of labeled organelles in live cells. Plasmonic enhancement of Raman scattering on metal surfaces, known as Surface Enhanced Raman Spectroscopy (SERS)19, has been used to increase signal from Raman probes and improve the detection sensitivity. However, advancing the SERS technology for mapping of intracellular biomolecules is extremely challenging20,21. First, targeted molecules of interest greatly outnumber plasmonic nanoparticles that can be tolerated by cells without any adverse effect. Second, the size of most biomolecules is incomparably smaller than that of these nanoparticles. Therefore, any intracellular gradient in the distribution of biomolecules cannot be identified by SERS. Besides, bulky metal nanoparticles can disturb activities of biomolecules and produce mechanical damage and other artifacts. Finally, SERS is not applicable in fixed cells, as plasmonic nanoparticles do not penetrate into the cell, even after very extensive permeabilization of cellular membranes. All these factors limit the utility of the SERS technology for intracellular molecular probing. An optimistic strategy to achieve enhancement of Raman scattering utilizes the Resonance Raman (RR) phenomenon22. In this approach, the energy of Raman excitation is adjusted to overlap with an electronic transition of the molecule of interest, such as a Raman reporter, which results in a significant amplification of the light scattering process. This RR technique has been successively applied for biomedical tissue analysis and detection of cancer related abnormalities23,24. It has also been successfully applied for ultra-sensitive detection of nucleic acids and protein biomolecules. However RR enhancement of cellular biomolecules utilizes highly phototoxic UV light, which is a prohibitive limitation for most live cell studies. At the same time natural pigments, which absorb in visible light, have long been a target RR spectroscopic probing of live cells and tissues22. Besides, the advantages of RR and SERS can be combined into an approach known as Surface Enhanced Resonance Raman Spectroscopy (SERRS) when the RR enhancement for some analytes, or reporters, could be accomplished at the excitation wavelength which also excites the plasmonic particles and creates SERS. SERRS was reported to generate unsurpassed enhancement of Raman signal, and has been successfully used for cell-free assays25 . However as discussed above, the plasmonic particles have only a limited value for labeling of intracellular molecules. A new concept for Raman molecular probes, which produces unprecedentedly strong Raman signal through RR enhancement and provide capability for intracellular labeling, recently was developed by our group14. Our reporters utilized azobenzene (AZO) tags, which were modified to produce RR enhancement under excitation with visible light at 532 nm, which is far less cytotoxic than UV light used in conventional RR spectroscopy. We further synthesized an AZO-RR probe for organelle-specific labeling in live cells and demonstrated its exceptionally high photostability, enabling long term monitoring of the same organelle14. Current Raman microscopy involves fairly long acquisition time, up to several minutes per square micrometer of the sample as well as intense laser illumination (typically 10 mW or higher), and therefore the risk of photodamage to a living specimen has to be carefully considered. In this regards, it has been demonstrated that selecting the excitation wavelengths in the spectral region from the red to the infrared, significantly reduces phototoxicity and enables repeatable spectra acquisition, without compromising functions of the organelle26. Therefore, development of RR probes excitable by biologically safe wavelengths will be highly beneficial for Raman microscopy of live samples. In this paper, we introduce a novel RR molecular recognition probe, designed for identification of organelles or other cellular structures and demonstrate its application by Raman imaging using excitation in a biologically safe wavelength region. A RR reporter based on BlackBerry Quencher 650 (BBQ-650), was developed to produce RR enhancement under excitation in red spectral range, where cellular biomolecules practically do not absorb, thus minimizing phototoxicity. Amplification of Raman signal by the resonance mechanism drastically increases the detection threshold sensitivity as compared to that of conventional spontaneous Raman probes. Besides, this probe produces low fluorescence background that often limits the sensitivity of Raman technique. Using this novel RR reporter, we synthesized a probe for tracking lysosomes in live cells and demonstrate first Raman detection and RR imaging of this type of cellular organelles. An inherent advantage of our approach is that RR imaging can be combined with the mapping of unlabeled cellular macromolecules by spontaneous Raman technique.