By Christoph Krafft // Micheal Schmitt // Jan Rüger // Fisseha-Bekele Legesse // Tobias Meyer // Jürgen Popp
The interest in growing and harvesting microalgae is constantly increasing mainly driven by production of biofuels. In parallel, extraction of other valuable co-products also gained attraction such as carotenoids that have found diverse applications in pharmaceutical, food and cosmetic industries. The xanthophyll fucoxanthin (Fx), for example, shows anti-oxidant, anti-inflammatory and anti-cancer properties. High performance liquid chromatography and mass spectrometry are common analytical methods to investigate the chemical content of microalgae which require, however, extensive sample preparation and bulk sample volumes. Raman-based methods offer analysis on the single cell and even subcellular level, in a non-destructive way and without sample preparation. Raman microspectroscopy was used to identify chemical fingerprints of lipids and pigments in microalgae, assess the growth phases of diatoms (1), and image the silica shell of diatoms (2). Disadvantages of Raman spectroscopy for microalgae research are relatively small cross sections of biomolecules and autofluorescent background of chlorophyll pigments upon visible laser excitation. These issues give relatively long acquisition times for Raman imaging and require several steps to process the images. Our recent publication solved these issues by using coherent anti-Stokes Raman scattering (CARS) microscopy (3).
In the CARS process, high energy picosecond pulses of pump frequency νpand Stokes frequency νsare tuned to coherently drive Raman active vibrations at their difference frequency (νp− νs). The inelastic scattering of a third pulse of probe frequency νpoff this vibration results in an anti-Stokes signal at frequency νas= 2νp− νs. CARS microscopy enables to acquire images at video frame rates using laser scanning microscopes and avoids the problem of one-photon fluorescence as the anti-Stokes signal is blue shifted from the excitation pulses. Another benefit of properly chosen laser pulse frequencies is that two-photon excited fluorescence (TPEF) images can simultaneously be registered. Previous CARS studies on algae have focused on examining the lipid metabolism as lipids provide intense signals due to abundant CH2stretching vibrations.
In (3) CARS microscopy, TPEF microscopy and Raman spectroscopy was employed to investigate the effect of different light cycles on pigment accumulation in two diatoms. CARS images focused on the prominent carotenoid and Fx band near 1528 cm-1and its morphological change as the conditions were varied. TPEF mainly probes chlorophyll in chloroplasts and its arrangement relative to the structure of the cell. The field of view using a 25×, 1.1 NA water immersion objective lens was 200×200 µm2at 1024×1024 pixel resolution. The 2 µs dwell time per pixel scales to a total acquisition time of 8 s. 10 Raman spectra with an exposure time of 0.5 s each were collected along a line within diatoms at 785 nm excitation with a 60×, 1.0 NA objective lens. They identified differences in the molecular fingerprints of the cells that correlate with the variations in light cycles (light:dark 20:4 h, 12:12 h, and 4:20 h).
Figure 1 depicts multimodal images of the diatoms Ditylum brightwellii and Stephanopyxis turris. The red channel represents the CARS modality while the TPEF modality is co-displayed in the green channel. As multiphoton effects are limited to a small confocal volume, depth profiles were constructed from images taken at 1 µm axial steps. Carotenoids were found to be localized within the chloroplasts of the diatoms. The chloroplasts containing the carotenoids in S. turris are arranged around the periphery, as such the boundaries of the cell are already demarked by the two modalities. Image analysis revealed that cells kept under longer dark cycles (4:20) contained chloroplasts that showed relatively stronger CARS signal intensity and were bigger than those kept under (20:4) dark cycles. This implies that the prolongation of the dark cycle resulted in accumulation of more pigments.
Figure 2 shows a mean Raman spectrum of all investigated diatom cells with labels of prominent bands due to chlorophyll a at 746, 988 and 1328 and due to carotenoid species at 1162 and 1528 cm-1. In a 100 times repeated procedure, PLS-LDA classification models were trained with random subsamples of half of the dataset (180 spectra) whereas the remaining 180 spectra were used for validation. The score plot of one randomly chosen iteration step visualizes classification performance. Two independent score layers are combined. In the first layer, colored hexagons represent a two-dimensional histogram of the training data in linear discriminant subspace. In the second layer, a scatter plot with circles squares and triangles shows the respective model projections of the test dataset. All predicted objects are correctly assigned to their appropriate class. Interpretation of the model coefficients (not shown) suggests that a control mechanism leads to relative change in the carotenoid pool. The ratio of light-protecting pigments diadinoxanthin and diatoxanthin Ddx/Dtx relative to the accessory pigment increases at higher light conditions.
In summary, we demonstrated that CARS, TPEF and spontaneous Raman microscopy are complementary techniques to investigate the influence of metabolic factors on the pigment accumulation of diatoms.
Funded by: DFG
(1) J. Rüger, N. Unger, I.W. Schie, E. Brunner, J. Popp. Assessment of growth phases of the diatom ditylum brightwellii by FT-IR and Raman spectroscopy. Algal Research (2016) 19, 246-252.
(2) M. Kammer, R. Hedrich, H. Ehrlich, J. Popp, E. Brunne, C. Krafft. Spatially resolved determination of the structure and composition of diatom cell walls by Raman and FTIR imaging. Anal. Bioanal. Chem. (2010) 398, 509-517.
(3) F.B. Legesse, J. Rüger, T. Meyer, C. Krafft, M. Schmitt, J. Popp. Investigation of microalgal carotenoid content using coherent anti-Stokes Raman scattering (CARS) microscopy and spontaneous Raman Spectroscopy. ChemPhysChem (2018) 19(9) 1048-1055.