310
A. Kaczor and m. Pilarczyk
tissue can be accurately determined. In this context, a Raman spatially sensitive
technique—Raman imaging emerges a particularly promising tool of study.
Raman imaging is a non-invasive technique of studying components distribution
and concentration in a sample. the technique combines the structural specificity of
Raman spectroscopy with the high spatial resolution of microscopy (lateral resolution of 360 nm and depth resolution of ca. 1 µm @ 532 nm excitation) resulting in
possibility of measuring sample microstructure at the subcellular level. 3d images
of tissue and cellular components, illustrating high heterogeneity of the studied objects, have been already obtained with the application of confocal Raman microimaging [11].
due to the complexity and heterogenity of studied biochemical samples, data
analysis is usually supported by chemometrics. Chemometrics is mathematics and
statistics based discipline allowing for interpretation of complicated datasets, such
as the Raman spectral representation of a sample. Classical, particularly regressionbased chemometric methods concentrated on exact identification of the sample
composition. this approach, although valid for simple systems, does not work particularly well for biological samples of considerable heterogeneity and complex
spatial distribution. In such cases, data analysis aims rather at determination of
chemical markers that describe changes in the system [12]. depending on the studied system and research aim, both chemometric approaches were applied in carotenoid analysis of biological systems. For example, regression-based methods were
used to classify carotenoid-producing unicellular organisms of different type or in
modified environment (nitrogen-replete and nitrogen-starved conditions) [13–15].
In the “marker-aimed” approach, the application of K-means algorithm to analysis
of Raman images of cancerous versus non-cancerous tissues allowed for determination of molecular markers of the disease with a reduced carotenoid level being one
of them [6]. this potential of Raman spectroscopy to distinguish between cancerous
and non-cancerous tissues/cells is a hope for development of non-invasive Ramanbased systems for medical diagnostics (including biopsy-free diagnosis), although
routine clinical usage of Raman spectroscopy is still a song of the future.
Another potentially promising field of application of Raman spectroscopy is
monitoring of growth and morphogenesis of unicellular algae. unicellular algae,
some species of yeast, bacteria and fungi are natural de novo producers of carotenoids and, therefore, a potent biosource of these chemicals [16]. Additionally,
some algal strains are potential biofuel source due to massive lipid production (oil
content in the range of 20–50 % dry weight of biomass) [17]. therefore, it is not
surprising that already in 2004, the global market for microalgal biomass was estimated to be 5,000 t of dry matter per year [18]. Raman spectroscopy is an effective
tool to optimize conditions of carotenogenesis and lipid production by algal cells in
laboratory conditions and, undoubtedly, has a potential to became a routine control
system in industrial, large-scale algal cultures. As carotenogenesis processes in pigment producing unicellular organisms are not fully understood, there is a large field
of application of Raman spectroscopy to study these processes both qualitatively
and quantitatively, while Raman imaging can be used to spatially localize produced
pigments. Chemometric methods are often used to recognize between synthesized
A. Kaczor and m. Pilarczyk
tissue can be accurately determined. In this context, a Raman spatially sensitive
technique—Raman imaging emerges a particularly promising tool of study.
Raman imaging is a non-invasive technique of studying components distribution
and concentration in a sample. the technique combines the structural specificity of
Raman spectroscopy with the high spatial resolution of microscopy (lateral resolution of 360 nm and depth resolution of ca. 1 µm @ 532 nm excitation) resulting in
possibility of measuring sample microstructure at the subcellular level. 3d images
of tissue and cellular components, illustrating high heterogeneity of the studied objects, have been already obtained with the application of confocal Raman microimaging [11].
due to the complexity and heterogenity of studied biochemical samples, data
analysis is usually supported by chemometrics. Chemometrics is mathematics and
statistics based discipline allowing for interpretation of complicated datasets, such
as the Raman spectral representation of a sample. Classical, particularly regressionbased chemometric methods concentrated on exact identification of the sample
composition. this approach, although valid for simple systems, does not work particularly well for biological samples of considerable heterogeneity and complex
spatial distribution. In such cases, data analysis aims rather at determination of
chemical markers that describe changes in the system [12]. depending on the studied system and research aim, both chemometric approaches were applied in carotenoid analysis of biological systems. For example, regression-based methods were
used to classify carotenoid-producing unicellular organisms of different type or in
modified environment (nitrogen-replete and nitrogen-starved conditions) [13–15].
In the “marker-aimed” approach, the application of K-means algorithm to analysis
of Raman images of cancerous versus non-cancerous tissues allowed for determination of molecular markers of the disease with a reduced carotenoid level being one
of them [6]. this potential of Raman spectroscopy to distinguish between cancerous
and non-cancerous tissues/cells is a hope for development of non-invasive Ramanbased systems for medical diagnostics (including biopsy-free diagnosis), although
routine clinical usage of Raman spectroscopy is still a song of the future.
Another potentially promising field of application of Raman spectroscopy is
monitoring of growth and morphogenesis of unicellular algae. unicellular algae,
some species of yeast, bacteria and fungi are natural de novo producers of carotenoids and, therefore, a potent biosource of these chemicals [16]. Additionally,
some algal strains are potential biofuel source due to massive lipid production (oil
content in the range of 20–50 % dry weight of biomass) [17]. therefore, it is not
surprising that already in 2004, the global market for microalgal biomass was estimated to be 5,000 t of dry matter per year [18]. Raman spectroscopy is an effective
tool to optimize conditions of carotenogenesis and lipid production by algal cells in
laboratory conditions and, undoubtedly, has a potential to became a routine control
system in industrial, large-scale algal cultures. As carotenogenesis processes in pigment producing unicellular organisms are not fully understood, there is a large field
of application of Raman spectroscopy to study these processes both qualitatively
and quantitatively, while Raman imaging can be used to spatially localize produced
pigments. Chemometric methods are often used to recognize between synthesized