Multi-excitation hyperspectral autofluorescence imaging for the exploration of biological samples

Maize Internode Autofluorescence at the Macroscopic Scale: Image Representation and Principal Component Analysis of a Series of Large Multispectral Images

A quantitative histology of maize stems is needed to study the role of tissue and of their chemical composition in plant development and in their end-use quality. In the present work, a new methodology is proposed to show and quantify the spatial variability of tissue composition in plant organs and to statistically compare different samples accounting for biological variability. Multispectral UV/visible autofluorescence imaging was used to acquire a macroscale image series based on the fluorescence of phenolic compounds in the cell wall. A series of 40 multispectral large images of a whole internode section taken from four maize inbred lines were compared. The series consisted of more than 1 billion pixels and 11 autofluorescence channels. Principal Component Analysis was adapted and named large PCA and score image montages at different scales were built. Large PCA score distributions were proposed as quantitative features to compare the inbred lines. Variations in the tissue fluorescence were clearly displayed in the score images. General intensity variations were identified. Rind vascular bundles were differentiated from other tissues due to their lignin fluorescence after visible excitation, while variations within the pith parenchyma were shown via UV fluorescence. They depended on the inbred line, as revealed by the first four large PCA score distributions. Autofluorescence macroscopy combined with an adapted analysis of a series of large images is promising for the investigation of the spatial heterogeneity of tissue composition between and within organ sections. The method is easy to implement and can be easily extended to other multi-hyperspectral imaging techniques. The score distributions enable a global comparison of the images and an analysis of the inbred lines' effect. The interpretation of the tissue autofluorescence needs to be further investigated by using complementary spatially resolved techniques.

Coherent anti-Stokes Raman scattering cell imaging and segmentation with unsupervised data analysis

Coherent Raman imaging has been extensively applied to live-cell imaging in the last 2 decades, allowing to probe the intracellular lipid, protein, nucleic acid, and water content with a high-acquisition rate and sensitivity. In this context, multiplex coherent anti-Stokes Raman scattering (MCARS) microspectroscopy using sub-nanosecond laser pulses is now recognized as a mature and straightforward technology for label-free bioimaging, offering the high spectral resolution of conventional Raman spectroscopy with reduced acquisition time. Here, we introduce the combination of the MCARS imaging technique with unsupervised data analysis based on multivariate curve resolution (MCR). The MCR process is implemented under the classical signal non-negativity constraint and, even more originally, under a new spatial constraint based on cell segmentation. We thus introduce a new methodology for hyperspectral cell imaging and segmentation, based on a simple, unsupervised workflow without any spectrum-to-spectrum phase retrieval computation. We first assess the robustness of our approach by considering cells of different types, namely, from the human HEK293 and murine C2C12 lines. To evaluate its applicability over a broader range, we then study HEK293 cells in different physiological states and experimental situations. Specifically, we compare an interphasic cell with a mitotic (prophase) one. We also present a comparison between a fixed cell and a living cell, in order to visualize the potential changes induced by the fixation protocol in cellular architecture. Next, with the aim of assessing more precisely the sensitivity of our approach, we study HEK293 living cells overexpressing tropomyosin-related kinase B (TrkB), a cancer-related membrane receptor, depending on the presence of its ligand, brain-derived neurotrophic factor (BDNF). Finally, the segmentation capability of the approach is evaluated in the case of a single cell and also by considering cell clusters of various sizes.

The distribution and retained amount of benzo[a]pyrene at the micro-zones of mangrove leaf cuticles: Results from a novel analytical method

Plant leaf cuticles play a critical role in the accumulation and transport of atmospheric polycyclic aromatic hydrocarbons (PAHs). The relationship between the distribution and retained amount of PAHs on the leaf cuticles and the leaves micro-zone structures is still unclear. In this study, a confocal microscopic fluorescence spectral analysis (CMFSA) system with a spatial resolution of 200 nm was established as a direct and noninvasive means to determine the microscopic distribution and quantify the retained amount of benzo[a]pyrene (B[a]P) at Aegiceras corniculatum (Ac), Kandelia obovata (Ko) and Avicennia marina (Am) leaf cuticle micro-zones (0.096 mm²). The linear ranges for the established method were 10-1900 ng spot^-1 for Ac, 15-1700 ng spot^-1 for Ko and 30-1800 ng spot^-1 for Am, and the detection limits were 0.06 ng spot^-1 for Ac, 0.06 ng spot^-1 for Ko and 0.07 ng spot^-1 for Am. Notably, B[a]P formed clusters and unevenly distributed at the leaf cuticles. On the adaxial cuticles, B[a]P was mainly accumulated unevenly along the epidermis cell wall, and it was also distinctively distributed in the secretory cells around salt glands for Ac and Am. On the abaxial leaf cuticles, B[a]P was concentrated in the salt glands and stomata apart from being unevenly distributed in the epidermis cell wall. Moreover, the amount of B[a]P retained presented a negative correlation with the polarity of leaf cuticles, which resulted in the amount of B[a]P retained on the adaxial leaf cuticles being significantly higher than that on abaxial leaf cuticles. Our results provide a potential in situ method for investigating the distribution and retained amount of PAHs at plant leaf surface micro-zones, which would contribute to further studying and understanding the mechanism of migration and transformation of PAHs by plant leaves from a microscopic perspective.

The endosperm cavity of wheat grains contains a highly hydrated gel of arabinoxylan

Cereal grains provide a substantial part of the calories for humans and animals. The main quality determinants of grains are polysaccharides (mainly starch but also dietary fibers such as arabinoxylans, mixed-linkage glucans) and proteins synthesized and accumulated during grain development in a specialized storage tissue: the endosperm. In this study, the composition of a structure localized at the interface of the vascular tissues of the maternal plant and the seed endosperm was investigated. This structure is contained in the endosperm cavity where water and nutrients are transferred to support grain filling. While studying the wheat grain development, the cavity content was found to autofluoresce under UV light excitation. Combining multispectral analysis, Fourier-Transform infrared spectroscopy, immunolabeling and laser-dissection coupled with wet chemistry, we identified in the cavity arabinoxylans and hydroxycinnamic acids. The cavity content forms a "gel" in the developing grain, which persists in dry mature grain and during subsequent imbibition. Microscopic magnetic resonance imaging revealed that the gel is highly hydrated. Our results suggest that arabinoxylans are synthesized by the nucellar epidermis, released in the cavity where they form a highly hydrated gel which might contribute to regulate grain hydration.

Diagnosis of cholangiocarcinoma from microscopic hyperspectral pathological dataset by deep convolution neural networks

This paper focuses on automatic Cholangiocarcinoma (CC) diagnosis from microscopic hyperspectral (HSI) pathological dataset with deep learning method. The first benchmark based on the microscopic hyperspectral pathological images is set up. Particularly, 880 scenes of multidimensional hyperspectral Cholangiocarcinoma images are collected and manually labeled each pixel as either tumor or non-tumor for supervised learning. Moreover, each scene from the slide is given a binary label indicating whether it is from a patient or a normal person. Different from traditional RGB images, the HSI acquires pixels in multiple spectral intervals, which is added as an extension on the channel dimension of 3-channel RGB image. This work aims at fully exploiting the spatial-spectral HSI data through a deep Convolution Neural Network (CNN). The whole scene is first divided into several patches. Then they are fed into CNN for the tumor/non-tumor binary prediction and the tumor area regression. The further diagnosis on the scene is made by random forest based on the features from patch prediction. Experiments show that HSI provides a more accurate result than RGB image. Moreover, a spectral interval convolution and normalization scheme are proposed for further mining the spectral information in HSI, which demonstrates the effectiveness of the spatial-spectral data for CC diagnosis.

Darkfield and Fluorescence Macrovision of a Series of Large Images to Assess Anatomical and Chemical Tissue Variability in Whole Cross-Sections of Maize Stems

The proportion and composition of plant tissues in maize stems vary with genotype and agroclimatic factors and may impact the final biomass use. In this manuscript, we propose a quantitative histology approach without any section labelling to estimate the proportion of different tissues in maize stem sections as well as their chemical characteristics. Macroscopic imaging was chosen to observe the entire section of a stem. Darkfield illumination was retained to visualise the whole stem cellular structure. Multispectral autofluorescence images were acquired to detect cell wall phenolic compounds after UV and visible excitations. Image analysis was implemented to extract morphological features and autofluorescence pseudospectra. By assimilating the internode to a cylinder, the relative proportions of tissues in the internode were estimated from their relative areas in the sections. The approach was applied to study a series of 14 maize inbred lines. Considerable variability was revealed among the 14 inbred lines for both anatomical and chemical traits. The most discriminant morphological descriptors were the relative amount of rind and parenchyma tissues together with the density and size of the individual bundles, the area of stem and the parenchyma cell diameter. The rind, as the most lignified tissue, showed strong visible-induced fluorescence which was line-dependant. The relative amount of para-coumaric acid was associated with the UV-induced fluorescence intensity in the rind and in the parenchyma near the rind, while ferulic acid amount was significantly correlated mainly with the parenchyma near the rind. The correlation between lignin and the tissue pseudospectra showed that a global higher amount of lignin resulted in a higher level of lignin fluorescence whatever the tissues. We demonstrated here the potential of darkfield and autofluorescence imaging coupled with image analysis to quantify histology of maize stem and highlight variability between different lines.

Simultaneous quantification of multiple endogenous biothiols in cancer cells based on a multi-signal fluorescent probe

The feature information was extracted from the 3D spectrum by the TM method, and the simultaneous quantification of intracellular thiols was realized.