Raman spectroscopy provides detailed information about the molecular structure of carotenoids. Advances in detector sensitivity and acquisition speed have driven the expansion of Raman spectroscopy from a bulk analytical tool to a powerful method for mapping carotenoid abundance in cells and tissues. In many applications, the technique is compatible with living organisms, providing highly specific molecular structure information in intact cells and tissues with subcellular spatial resolution. This leads to spatial-temporal-chemical resolution critical to understanding the complex processes in the life cycle of carotenoids and other biomolecules.
- vibrational spectroscopy
- multivariate analysis
- confocal Raman microscopy
- Raman imaging
- photosynthetic organisms
- resonance Raman scattering
Carotenoids are tetraterpenoids and have various functions in plants and algae. They extend the range of wavelengths for photosynthesis by harnessing solar radiation where chlorophyll pigments do not appreciably absorb and serve as structural elements within the photosynthetic apparatus . They also function to dissipate excess solar radiation and prevent the formation of harmful singlet oxygen species . Additionally, in specific photosynthetic eukaryotes, carotenoids accumulate in plastid organelles called chromoplasts and give many fruits, flowers, and roots their bright colors . For these reasons, assessing the presence and distribution of carotenoids at the subcellular level provides a keyhole through which to understand a variety of cellular processes. The use and utility of Raman imaging is addressed here toward this end.
The Raman effect involves an incoming photon being scattered inelastically by a polyatomic molecule. This phenomenon requires that energy is exchanged between the photon and molecule. Since energy levels in the molecule are discrete, the difference in energy between the incoming and scattered photon corresponds to a molecular transition, most typically a vibration . Molecular vibrations are influenced by the composition of atoms that make up the molecule, the types of bonds that connect atoms, and molecular symmetry. Thus, Raman spectroscopy has been exploited as a highly sensitive analytical tool for the determination of the chemical identity of many molecules [5, 6]. The probability of observing the Raman effect is low, however, as the intensity of scattered radiation scales with the fourth power of incident light’s frequency. The Raman effect is oftentimes considerably weaker as compared to the intrinsic fluorescence in cells and tissues, complicating and even prohibiting the detection of the Raman scatter in biologically relevant matrices.
However, if the frequency of incident light matches the energy of an electronic transition, an enhancement of the Raman signal may be observed. This is known as resonance Raman scattering (RRS), which can enhance the signal by several orders of magnitude over a “normal,” non‐resonance measurement. In fact, the resonance effect can increase the scattering cross section to exceed any off‐resonance Raman scatter and, importantly, even rival intrinsic fluorescence from the species. RRS thus has profound analytical potential and has been comprehensively reviewed elsewhere for the general life science field  and specifically for applications in photosynthesis . Here, its role in the assessment of carotenoids is examined explicitly.
The optical properties of carotenoids make them especially suitable for RRS. Carotenoids are π‐electron‐conjugated carbon‐chain molecules consisting of alternating C─C single bonds and C═C double bonds. Individual carotenoids can be distinguished by the number of conjugated carbon double bonds, the number of attached methyl side groups, and the number and type of end groups. These properties result in many of the highly abundant carotenoid molecules (e.g., β‐carotene, zeaxanthin, lycopene, and lutein) having distinct, yet broad (100 nm) absorption bands in the visible region of the spectrum. The absorption shifts to longer wavelengths as the effective conjugation length of the carotenoid increases. Fortuitously, these visible absorption bands overlap with the common laser wavelengths for Raman excitation. Thus, when excited under these conditions, carotenoids exhibit a very strong RRS response (enhancement factor of about five orders of magnitude relative to non-resonant Raman spectroscopy) and little to no fluorescence emission. Additionally, variations in absorption of the different carotenoids can be exploited by shifting the excitation wavelength (a technique also known as “tuning”) to preferentially excite different carotenoid molecules. Owing to these amenities, RRS therefore enables the detection of carotenoids, even in complex biological systems, such as living photosynthetic cells and tissues.
The majority of carotenoids have linear structures resulting in a limited number of Raman‐observable vibrations that are easily categorized into a distinct vibrational signature. There are three major Raman modes typically leveraged in the analysis of carotenoids [9, 10]. The
In recent years, advances in detector sensitivity and acquisition speed have driven the expansion of Raman spectroscopy from a bulk analytical tool to a powerful method for mapping molecular vibrations in cells and tissues by the addition of a spatial dimension. Spatial resolution available with Raman microscopy is dependent on the wavelength of the excitation light and in theory is the same as other optical microscopies, such as fluorescence (e.g., ∼250 nm lateral and ∼500 nm axial, for blue/green excitation). The addition of spatial resolution can be accomplished in several different ways: confocal point‐scanning Raman microscopy [12, 13], wide‐field Raman imaging , or Raman line scanning [15, 16]. The relative advantages and disadvantages of each of these approaches have been reviewed elsewhere and will not be covered here . In addition to the imaging methodology, spectral information can be obtained from either a single or a small number of bands through the use of discrete bandpass filters or tunable filters or in a hyperspectral fashion, where the entire Raman spectrum is dispersed via a prism or grating allowing for the simultaneous acquisition of hundreds of spectral bands. While there is typically a speed advantage associated with the filter‐based acquisition approach, recent commercial systems utilizing electron‐multiplied charge‐coupled device (EM‐CCD) detectors and cutting‐edge research systems  are able to approach confocal fluorescence microscopy speeds and further improvements are expected.
Given the complexity associated with biological matrices such as live cells and tissues, hyperspectral Raman approaches are often advantageous because they can provide detailed spectral signatures for subsequent analysis of weak or overlapping features and background noise reduction using chemometric algorithms. These approaches are collectively referred in the literature as Raman spectroscopic imaging, or hyperspectral Raman microscopy. Additionally, confocal point scanning or pseudo‐confocal line scanning can have advantages due to the inherent rejection of out‐of‐focus signal provided by the pinhole or entrance slit, respectively. A confocal, point‐scanning, hyperspectral Raman microscope was utilized for the work presented in this chapter as it adequately rejects the intense pigment signal from outside the focal plane, while providing reasonable acquisition speeds for live cells.
Recent literature shows that Raman spectroscopic imaging is emerging as a key technology for single‐cell analysis, including
While Raman spectroscopic imaging has potential for assessing carotenoid distributions in single cells and tissues for many applications in biomedicine and photosynthesis, there are some noteworthy limitations. First, not all carotenoids are enhanced to the same degree and some will not be enhanced at all. This differential enhancement can be advantageous, but it also can pose limitations for certain carotenoids. To some degree, the choice of laser excitation can target additional carotenoids. However, performing multiple Raman microscopy scans at two or more excitation wavelengths is prohibitive for live‐cell dynamics and may require fixed samples depending on the acquisition times. Second, even with high spectral resolution instruments, multiple carotenoid species with highly overlapping, similar peaks can be difficult to identify. Additionally, even with resonance enhancement, Raman peaks may still be weak at
In addition to the hardware approaches listed above, the first two limitations are actively being addressed with the development of robust multivariate analysis tools [29–31]. For example, multivariate curve resolution (MCR) has been developed by several research groups for the analysis of hyperspectral confocal fluorescence and Raman image data sets [29, 32–34] finding success in complex multicomponent biological samples. It is therefore used in the analysis presented in this chapter. Lastly, while recent advances in fluorescence microscopy have extended the spatial resolution beyond the diffraction limit, the spatial resolution of Raman spectroscopic imaging is still in many cases orders of magnitude larger than the biological processes being investigated. Near‐field approaches can provide higher spatial resolution, but have not found wide‐scale success with living cells and tissues .
This chapter presents three separate applications of confocal Raman microscopy to assess carotenoid localization and relative abundance within living cells. The green algae
2. Materials and methods
2.1. Cell culture and sample prep
Etiolated maize protoplasts were isolated from
2.2. Confocal Raman microscopy
Raman images were acquired with a WiTec Alpha300R system equipped with a WiTec UHTS spectrometer utilizing a 600‐l/mm grating and an Andor back‐illuminated electron‐multiplying charge‐coupled device (EMCCD). Light was incident at 532 nm and focused using a 50×/0.55 NA objective (
2.3. Image data analysis
All spectral image analysis was performed in Matlab 2012 or 2015 (Mathworks) equipped with the statistics and machine learning, signal processing, image processing, and curve fitting toolboxes leveraging in‐house written software, functions, and scripts. Hyperspectral confocal Raman images were preprocessed to remove cosmic spikes . When applicable, images from the same sample were compiled into a composite image data set. The use of composite images, rather than analyzing every image independently, increases the number of pixels for the MCR analysis and thus serves to improve spectral signature identification by adding additional variance . The spectral region was trimmed to exclude the excitation laser line but still include about 5–10 pixels of the spectrum where the signal was blocked by the Rayleigh filter. This technique assists in the analysis by providing a “zero‐signal region” for assessing baseline contributions and is discussed in detail by Jones et al. . In most cases, an image mask was created that excluded pixels outside the area of the cells from the analysis as they contain predominantly background signal.
Principal components analysis (PCA) was performed on the composite image and the Scree plot was inspected to determine the number of independent components present in the image data set (measured as being before the bend in the elbow of the Scree plot ). Multivariate curve resolution was then performed using a constrained alternating least‐squares algorithm and employing robust constraints for equality (offset/baseline) and nonnegativity (all true spectral components) to develop a spectral model that described the spectral variance within the data set. A PCA analysis of the residuals was used to confirm the appropriateness of the spectral model for describing the data and identify any unmodeled spectral signatures. The multivariate curve resolution algorithm [39–43] and specific approaches for success with biological images have been described in detail elsewhere [29, 38]. The details of the spectral models developed are presented during the results and discussion for each application in this chapter. The MCR‐identified spectra were then used in a classical least‐squares (CLS) analysis to predict the concentrations of each spectral component in each image pixel. Lastly, the resulting concentration maps were exported as 16‐bit grayscale tiffs such that they could be subject to traditional image analysis. Color image overlays and simple image cropping and scaling for visualization purposes was performed in Fiji .
3. Results and discussion
3.1. Resolving different carotenoids in living cells:
3.2. Carotenoids in high‐chlorophyll backgrounds: maize protoplasts
Similar to algae, plant carotenoids have roles in light harvesting and protection against light and heat stress. Metabolic engineering of carotenoids in plants has the potential to create varieties exhibiting increased adaptation to climate change as well as additional nutritional value. To develop plant cultivars with these enhanced properties, it is critical to understand the detailed spatial‐temporal arrangement of both carotenoids and chlorophyll
Confocal Raman spectroscopic imaging is capable of detecting carotenoids without the use of any exogenous labels and does so by detecting the resonance Raman signal of carotenoids. This signal can be observed also within the background of other pigments, such as the chlorophyll precursor, protochlorophyllide, which is found in cells of etiolated tissue  such as plant protoplasts from etiolated maize (Figure 3). Plant protoplasts are plant cells that have their cell wall removed through enzyme treatment and thus are an excellent experimental system for plant biologists because they improve the ease of introducing new genes.
Figure 3 highlights the spatial location and levels of carotenoids and protochlorophyllide in plastids (immature chloroplasts) of etiolated maize protoplasts as observed by confocal Raman microscopy. Previous high‐performance liquid chromatography data have shown that the dominant carotenoid in maize‐etiolated protoplasts is lutein with a minor amount of violaxanthin (data not shown). The peak positions (
3.3. Approaching the limits of spatial resolution:
Cyanobacteria and closely related organisms are thought to be the evolutionary ancestors to chloroplasts found in plants. Cyanobacteria perform photosynthesis to convert light energy to chemical energy and play key roles in the ecology of the earth. Light is harvested by photosynthetic antennae, which are pigment‐protein complexes composed of pigments with varying absorption and emission properties in order to “funnel” the energy to the reaction center where it is converted to chemical energy . Carotenoids are integrated into the membrane‐associated photosynthetic antennae complexes as well as the photosystems of all cyanobacteria and fundamentally coupled to the processes of light harvesting and photosynthesis. Unfortunately, carotenoids in cyanobacteria are often quite difficult to localize
Localizing carotenoids in living cells and tissues is challenging due to the complex biological matrices of the living cells and intense sometimes colocated interfering fluorescence. Raman spectroscopic imaging coupled with multivariate curve resolution analysis provides the necessary spatial and chemical resolution to identify highly similar carotenoids even in the midst of highly overlapping, strong chlorophyll emission and complex backgrounds. Three examples were presented that highlight different advantages of the methodology for investigating photosynthetic organisms. Recent technological advances in detector speed and sensitivity will most likely catalyze future investigations of carotenoid biogenesis in single cells, including population dynamics and response to changing environmental conditions, by facilitating time‐course studies that were previously prohibitive given the long scan times required for Raman spectroscopic imaging. These capabilities are anticipated to impact a variety of research areas including carbon partitioning and utilization, microbial ecology, and crop analytics.
The authors are grateful to the following people for their assistance with the research presented in this chapter: Anthony McDonald for assistance with the collection of Raman spectroscopic image data; Meghan Dailey for culturing the
This research was primarily supported as part of the Photosynthetic Antenna Research Center (PARC), an Energy Frontier Research Center funded by the U.S. Department of Energy (DOE), Office of Science, Basic Energy Sciences (BES), under Award # DE‐SC 0001035 (Maize protoplasts imaging, Synechocystis imaging, writing), by Sandia National Laboratories’ Laboratory Directed Research and Development (LDRD) Program under Award # 141528 (
Frank HA, Cogdell RJ. Carotenoids in photosynthesis. Photochem Photobiol. 1996; 63:257–64. DOI: 10.1111/j.1751‐1097.1996.tb03022.x.
Demmig‐Adams B, Adams Iii WW. The role of xanthophyll cycle carotenoids in the protection of photosynthesis. Trends Plant Sci. 1996; 1:21–6. DOI: 10.1016/S1360‐1385(96)80019‐7.
Bartley GE, Scolnik PA. Plant carotenoids: pigments for photoprotection, visual attraction, and human health. The Plant Cell. 1995; 7:1027–38.
Lewis IR, Edwards HGM, editors. Handbook of Raman Spectroscopy. New York: Marcel Dekker, Inc.; 2001.
Vankeirsbilck T, Vercauteren A, Baeyens W, Van der Weken G, Verpoort F, Vergote G, et al. Applications of Raman spectroscopy in pharmaceutical analysis. TrAC Trends in Analytical Chemistry. 2002; 21:869–77. DOI: 10.1016/S0165‐9936(02)01208‐6.
Clegg IM, Everall NJ, King B, Melvin H, Norton C. On‐line analysis using Raman spectroscopy for process control during the manufacture of titanium dioxide. Appl Spectrosc. 2001; 55:1138–50. DOI: 10.1366/0003702011953388.
Efremov EV, Ariese F, Gooijer C. Achievements in resonance Raman spectroscopy: review of a technique with a distinct analytical chemistry potential. Anal Chim Acta. 2008; 606:119–34. DOI: 10.1016/j.aca.2007.11.006.
Robert B. Resonance Raman spectroscopy. Photosynth Res. 2009; 101:147–55. DOI: 10.1007/s11120‐009‐9440‐4.
Merlin JC. Resonance Raman spectroscopy of carotenoids and carotenoid‐containing systems. Pure Appl Chem. 1985; 57:785–92.
Rimai L, Heyde ME, Gill D. Vibrational spectra of some carotenoids and related linear polyenes. Raman spectroscopic study. J Am Chem Soc. 1973; 95:4493–501. DOI: 10.1021/ja00795a005.
Gill D, Kilponen RG, Rimai L. Resonance Raman scattering of laser radiation by vibrational modes of carotenoid pigment molecules in intact plant tissues. Nature. 1970; 227:743–4.
Klein K, Gigler Alexander M, Aschenbrenner T, Monetti R, Bunk W, Jamitzky F, et al. Label‐free live‐cell imaging with confocal Raman microscopy. Biophys J. 2012; 102:360–8. DOI: 10.1016/j.bpj.2011.12.027.
Dieing T, Hollricher O, Toporski J, editors. Confocal Raman Microscopy: Springer-Verlag Berlin Heidelberg Springer; 2011.
Morris HR, Hoyt CC, Miller P, Treado PJ. Liquid crystal tunable filter Raman chemical imaging. Appl Spectrosc. 1996; 50:805–11.
Timlin JA, Carden A, Morris MD. Chemical microstructure of cortical bone probed by Raman transects. Appl Spectrosc. 1999; 53:1429–135.
Christensen KA, Morris MD. Hyperspectral Raman microscopic imaging using Powell lens line illumination. Appl Spectrosc. 1998; 52:1145–7.
Schlucker S, Schaeberle MD, Huffman SW, Levin IW. Raman microspectroscopy: A comparison of point, line, and wide‐field imaging methodologies. Anal Chem. 2003; 75:4312–8.
Kong L, Navas‐Moreno M, Chan JW. Fast confocal Raman imaging using a 2‐D multifocal array for parallel hyperspectral detection. Anal Chem. 2016; 88:1281–5. DOI: 10.1021/acs.analchem.5b03707.
Wu H, Volponi JV, Oliver AE, Parikh AN, Simmons BA, Singh S. In vivolipidomics using single‐cell Raman spectroscopy. Proc Natl Acad Sci. 2011; 108:3809–14. DOI: 10.1073/pnas.1009043108.
Schuster KC, Urlaub E, Gapes JR. Single‐cell analysis of bacteria by Raman microscopy: spectral information on the chemical composition of cells and on the heterogeneity in a culture. J Microbiol Meth. 2000; 42:29–38.
Gierlinger N, Keplinger T, Harrington M. Imaging of plant cell walls by confocal Raman microscopy. Nat Protocols. 2012; 7:1694–708. DOI: 10.1038/nprot.2012.092.
Huang WE, Griffiths RI, Thompson IP, Bailey MJ, Whiteley AS. Raman microscopic analysis of single microbial cells. Anal Chem. 2004; 76:4452–8. DOI: 10.1021/ac049753k.
Huang WE, Stoecker K, Griffiths R, Newbold L, Daims H, Whiteley AS, et al. Raman‐FISH: combining stable‐isotope Raman spectroscopy and fluorescence in situ hybridization for the single cell analysis of identity and function. Environ Microbiol. 2007; 9:1878–89. DOI: 10.1111/j.1462‐2920.2007.01352.x.
Matthäus C, Krafft C, Dietzek B, Brehm BR, Lorkowski S, Popp J. Noninvasive imaging of intracellular lipid metabolism in macrophages by Raman microscopy in combination with stable isotopic labeling. Anal Chem. 2012; 84:8549–56. DOI: 10.1021/ac3012347.
Pudney PDA, Gambelli L, Gidley MJ. Confocal Raman microspectroscopic study of the molecular status of carotenoids in tomato fruits and foods. Appl Spectrosc. 2011; 65:127–34. DOI: 10.1366/10‐06121.
Zheng Y‐T, Toyofuku M, Nomura N, Shigeto S. Correlation of carotenoid accumulation with aggregation and biofilm development in Rhodococcus sp. SD‐74. Anal Chem. 2013; 85:7295–301. DOI: 10.1021/ac401188f.
Collins AM, Jones HDT, Han D, Hu Q, Beechem TE, Timlin JA. Carotenoid distribution in living cells of Haematococcus pluvialis(Chlorophyceae). PLoS One. 2011; 6:e24302.
Toomey MB, Collins AM, Frederiksen R, Cornwall MC, Timlin JA, Corbo JC. A complex carotenoid palette tunes avian colour vision. J Royal Soc Interf. 2015; 12. DOI: 10.1098/rsif.2015.0563.
Haaland DM, Jones HDT, Timlin JA. Experimental and Data Analytical Approaches to Automating Multivariate Curve Resolution in the Analysis of Hyperspectral Images. In: Ruckebusch C, editor. Resolving Spectral Mixtures. Data Handling in Science and Technology. 30. Amsterdam: Elsevier; 2016. pp. 381–406.
Zhang J, O’Connor A, Turner II JF. Cosine histogram analysis for spectral image data classification. Appl Spectrosc. 2004; 58:1318–24.
Widjaja E, Crane N, Chen T‐c, Morris MD, Ignelzi Jr MA, McCreadie BR. Band‐target entropy minimization (BTEM) applied to hyperspectral Raman image data. Appl Spectrosc. 2003; 57:1353–62.
Felten J, Hall H, Jaumot J, Tauler R, de Juan A, Gorzsás A. Vibrational spectroscopic image analysis of biological material using multivariate curve resolution–alternating least squares (MCR‐ALS). Nat Protocols. 2015; 10:217–40. DOI: 10.1038/nprot.2015.008.
de Juan A, Tauler R. Chemometrics applied to unravel multicomponent processes and mixtures: Revisiting latest trends in multivariate resolution. Anal Chim Acta. 2003; 500:195–210. DOI: 10.1016/S0003‐2670(03)00724‐4.
Schoonover JR, Marx R, Zhang SL. Multivariate curve resolution in the analysis of vibrational spectroscopy data files. Appl Spectrosc. 2003; 57:154A–70A.
Webster S, Batchelder DN, Smith DA. Submicron resolution measurement of stress in silicon by near‐field Raman spectroscopy. Appl Phys Lett. 1998; 72:1478–80. DOI: 10.1063/1.120598.
Shumskaya M, Bradbury LMT, Monaco RR, Wurtzel ET. Plastid localization of the key carotenoid enzyme phytoene synthase is altered by isozyme, allelic variation, and activity. The Plant Cell. 2012; 24:3725–41. DOI: 10.1105/tpc.112.104174.
Anthony SM, Timlin JA. Removing cosmic spikes using a hyperspectral upper‐bound spectrum method. Appl Spectrosc. 2016:0003702816668528. DOI: 10.1177/0003702816668528.
Jones HDT, Haaland DM, Sinclair MB, Melgaard DK, Collins AM, Timlin JA. Preprocessing strategies to improve MCR analyses of hyperspectral images. J Chemomet Intell Lab Syst. 2012; 117:149–58. DOI: 10.1016/j.chemolab.2012.01.011.
Van Benthem MH, Keenan MR, inventors; Sandia Corporation, assignee. Fast combinatorial algorithm for the solution of linearly constrained least squares problems. US2008.
Van Benthem MH, Keenan MR. Fast algorithm for the solution of large scale non‐negativity constrained least squares problems. J Chemom. 2004; 18:441–50.
Ohlhausen JA, Keenan MR, Kotula PG, Peebles DE. Multivariate statistical analysis of time‐of‐flight secondary ion mass spectrometry images using AXSIA. Appl Surf Sci. 2004; 231‐232:230–4.
Kotula PG, Keenan MR, Michael JR. Automated analysis of SEM X‐Ray spectral images: a powerful new microanalysis tool. Microsc Microanal. 2003; 9:1–17.
Van Benthem MH, Keenan MR, Haaland DM. Application of equality constraints on variables during alternating least squares procedures. J Chemom. 2002; 16:613–22.
Schindelin J, Arganda‐Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an open‐source platform for biological‐image analysis. Nat Meth. 2012; 9:676–82. DOI: 10.1038/nmeth.2019.
Robertson DS, Bachmann MD, Anderson IC. Role of carotenoids in protecting chlorophyll from photodestruction—II. Studies on the effect of four modifiers of the albino cl1 mutant of maize. Photochem Photobiol. 1966; 5:797–805. DOI: 10.1111/j.1751‐1097.1966.tb05775.x.
Ruban AV, Pascal A, Lee PJ, Robert B, Horton P. Molecular configuration of xanthophyll cycle carotenoids in photosystem II antenna complexes. J Biol Chem. 2002; 277:42937–42. DOI: 10.1074/jbc.M207823200.
Blankenship RE. Molecular Mechanisms of Photosynthesis: Wiley‐Blackwell; Oxford 2013. 336 p.
Vermaas WFJ, Timlin JA, Jones HDT, Sinclair MB, Nieman LT, Hamad S, et al. In vivohyperspectral confocal fluorescence imaging to determine pigment localization and distribution in cyanobacterial cells. Proc Natl Acad Sci. 2008; 105:4050–5.