316
differences between nitrogen-replete and starved cells, i. e. to elucidate chemical
information encoded in the Raman spectra. It was stressed out that the single cell
approach enables assessing population variability and the individual cellular answer
for environmental stimuli, for instance nutrient conditions, without losing the overall information about the cellular system [14].
the insightful in vivo study of carotenogenesis in Haematococcus Pluvialis living cells by means of confocal Raman imaging supported by multivariate analysis
was reported by Collins et al. [19] the application of multivariate Curve Resolution
(mCR) resulted in generating of the reconstructed cell images in which the relative concentration and location of four pure components (astaxanthin, β-carotene,
chlorophyll and cellular autofluorescence) were determined at every spatial pixel
(Fig. 11.3).
In flagellated motile cells, astaxanthin was practically absent, and only the weak
intensity signal of this pigment was observed in the palmelloid cells. In both these
morphotypes, β-carotene was found to be co-localized with chlorophyll in chloroplasts. upon further stress conditions and production of palmelloid cells, distribution of both carotenoids changed and both co-localized β-carotene and astaxanthin
were found outside the chloroplasts, supporting the hypothesis that upon stress
conditions β-carotene is transported via a chloroplast membrane and transformed
into astaxanthin. While β-carotene was found also in chloroplasts, astaxanthin is
localized solely outside chloroplasts, in the center of the palmelloid cells. Finally, a
massive production of astaxanthin was observed in aplanospores with the increased
concentration of the pigment in the periphery of cysts. therefore, the application
of mCR analysis allowed not only for discrimination of two components of a very
similar spectral profile (β-carotene and astaxanthin), but also for extracting of information about relatively weak spectral features (β-carotene bands are about 4-fold
less intense compared to astaxanthin ones) [19].
Abbas et al. [15] described the application of target orthogonal partial least
squares (t-oPLS) to the analysis of the subtle differences between carotenoid spectra of two species Dunaliella and Phaeodactylum. In t-oPLS, a known reference
spectrum (in this case a standard carotenoid spectrum) is designated as the target
and denoted as the single y column in an oPLS regression model, while the X matrix consists of the unfolded image spectra as variables in its columns. Finally, the
carotenoids correlation images from Dunaliella and Phaeodactylum images in the
model were reconstructed from the oPLS p loadings.
Both the original and reconstructed images showed distribution of carotenoids
in the studied species. In Dunaliella, pigments are dispersed over the cell, with the
high concentration of carotenoids in the eye, while in Phaeodactylum carotenoids
are localized mainly in the center of cells. the orthogonal scores of the t-oPLS
model enabled to find small differences between the reference and the studied
samples, including variations between different samples of the same species (most
probably due to the environmental influence) [15]. the work confirms the high
potential of multivariate analysis in the extraction of detailed and exact information
from the samples in situ in their complicated biological context.
A. Kaczor and m. Pilarczyk
differences between nitrogen-replete and starved cells, i. e. to elucidate chemical
information encoded in the Raman spectra. It was stressed out that the single cell
approach enables assessing population variability and the individual cellular answer
for environmental stimuli, for instance nutrient conditions, without losing the overall information about the cellular system [14].
the insightful in vivo study of carotenogenesis in Haematococcus Pluvialis living cells by means of confocal Raman imaging supported by multivariate analysis
was reported by Collins et al. [19] the application of multivariate Curve Resolution
(mCR) resulted in generating of the reconstructed cell images in which the relative concentration and location of four pure components (astaxanthin, β-carotene,
chlorophyll and cellular autofluorescence) were determined at every spatial pixel
(Fig. 11.3).
In flagellated motile cells, astaxanthin was practically absent, and only the weak
intensity signal of this pigment was observed in the palmelloid cells. In both these
morphotypes, β-carotene was found to be co-localized with chlorophyll in chloroplasts. upon further stress conditions and production of palmelloid cells, distribution of both carotenoids changed and both co-localized β-carotene and astaxanthin
were found outside the chloroplasts, supporting the hypothesis that upon stress
conditions β-carotene is transported via a chloroplast membrane and transformed
into astaxanthin. While β-carotene was found also in chloroplasts, astaxanthin is
localized solely outside chloroplasts, in the center of the palmelloid cells. Finally, a
massive production of astaxanthin was observed in aplanospores with the increased
concentration of the pigment in the periphery of cysts. therefore, the application
of mCR analysis allowed not only for discrimination of two components of a very
similar spectral profile (β-carotene and astaxanthin), but also for extracting of information about relatively weak spectral features (β-carotene bands are about 4-fold
less intense compared to astaxanthin ones) [19].
Abbas et al. [15] described the application of target orthogonal partial least
squares (t-oPLS) to the analysis of the subtle differences between carotenoid spectra of two species Dunaliella and Phaeodactylum. In t-oPLS, a known reference
spectrum (in this case a standard carotenoid spectrum) is designated as the target
and denoted as the single y column in an oPLS regression model, while the X matrix consists of the unfolded image spectra as variables in its columns. Finally, the
carotenoids correlation images from Dunaliella and Phaeodactylum images in the
model were reconstructed from the oPLS p loadings.
Both the original and reconstructed images showed distribution of carotenoids
in the studied species. In Dunaliella, pigments are dispersed over the cell, with the
high concentration of carotenoids in the eye, while in Phaeodactylum carotenoids
are localized mainly in the center of cells. the orthogonal scores of the t-oPLS
model enabled to find small differences between the reference and the studied
samples, including variations between different samples of the same species (most
probably due to the environmental influence) [15]. the work confirms the high
potential of multivariate analysis in the extraction of detailed and exact information
from the samples in situ in their complicated biological context.
A. Kaczor and m. Pilarczyk