Applications of Raman and Surface-Enhanced Raman Scattering to the Analysis of Eukaryotic Samples - Emerging Raman Applications and Techniques in Biomedical and Pharmaceutical Fields - page 90

4.4 Conclusions
With Raman microspectroscopy, an in situ chemical characterization and clas-
sification of cellular samples can be achieved. As demonstrated, it is possible
to localize different spectral signals to specific morphological microstructures
in individual pollen grains. Maps of important molecular constituents in cryo-
sections of pollen grains could be constructed simultaneously and without
purification or staining. The data permit conclusions about the relative lev-
els and localization of aromatic amino acids along with aliphatic molecules,
nucleic acids, protein and carbohydrate storage in different structures of the
pollen. This combined analysis of different classes of molecular constituents
in the unaltered material differs from previous purification-based approaches.
Surface-enhanced Raman scattering (SERS) can significantly increase the sen-
sitivity of Raman spectroscopic approaches for pollen characterization and
allows a characterization of very small amounts of pollen cellular fraction,
which is water soluble. This is of particular interest for studies of pollen al-
lergens which are released by the pollen grains upon hydration, but also for
the identification of very small amounts of pollen material in aerosols.
By using a Raman spectroscopic approach, we could prove the working
hypothesis that the biochemical make-up of the pollen (as part of a recogni-
tion/mating system) is altered during formation of a new biological species
and that the species-specific chemical similarities and dissimilarities indeed
reflect in the Raman spectral fingerprint. Based on the chemical informa-
tion, unsupervised multivariate analysis consisting of hierarchical clustering
revealed in most cases chemical similarities between species that were indica-
tive of both taxonomic relationship and mating behaviour. As illustrated by
the data, the spectra of whole pollen can be used for their classification. Such
an approach is very promising, considering its potential to be employed for
automatic pollen detection and early warning.
Acknowledgements
We would like to thank the group of Dieter Naumann, Robert-Koch-Institut
Berlin, for providing the cryostat and Peter Lasch, RKI and CytoSpec, Inc.
for CytoSpec software. J.K. is grateful to Katrin Kneipp, Wellman Center for
Photomedicine, Harvard Medical School, Boston, for providing Raman setups
for SERS experiments in live cells. We gratefully acknowledge funding of this
research by Deutsche Forschungsgemeinschaft (Grants DFG KN 557/9-1 and
PA 716/9-1).
References
1. A.-K. Prahl et al . ,Zeitschriftfur Naturforschung 40c , 621 (1986)
2. C.J. Keijzer, New Phytol. 105 , 499 (1987)