Image analysis contributes to understand biological processes

SISTI MS; NISHIDA F; ZANUZZI CN; VON WERNICH M; GRANDINETTI JAB; BARBEITO CG; PORTIANSKY EL

Workshop; Imaging Techniques for Biotechnology and Biomedical Applications Workshop; 2016

The main work of our laboratory is oriented towards the animal nervous system, although we also contribute with other research areas and groups of our university and of the rest of the country. By using image analysis, we can determine among others, the intensity and distribution of immunofluorescent areas or the optical density of samples processed by immunohistochemical techniques. In this way, the presence and location of certain antigens can be achieved and thus, the possibility to evaluate different parameters, such as object counting and morphometric determinations, that may help to infer the pathogenic, physiological or developmental mechanisms occurring in different biological or pathological processes. Analyses are performed with images obtained from cell cultures, histological sections for light or electron microscopy, or from macroscopic samples observed under stereoscopy. Depending on the number of samples to analyze different algorithms to automate the process, both in capture and subsequent analysis, are generated. For analytical determinations, we use several commercial image analysis software (cellSens Dimension, v1.7 Olympus; ImagePro Plus - v6.3 MediaCybernetics), and free software (Fiji, NIH). We also use deconvolution software (AutoDeblur & AutoVisualize X, MediaCybernetics) to deconvolve stack images obtained either from the Z-axis automatized widefield microscope (Olympus BX53) or the confocal microscope (Olympus FV1000). With a motorized stage (Prior), we scan samples using the Multiple Image Alignment (MIA)/Stitching function. 3D reconstruction is generated using different functions of the abovementioned software and with others available on the Internet. Our laboratory is focused on the study of the nervous system aging and of neurodegenerative processes of the spinal cord caused by toxic agents. To identify morphological changes in different cell populations of the organ we use histological, histochemical, immunohistochemical and immunofluorescent techniques. Images are obtained with different microscopes and capture systems and lately analyzed using computerized image analysis. Our laboratory has been working on morphological and morphometric changes of the spinal cord establishing the normal parameters of this organ in rats (Portiansky et al 2004, 2006a, 2006b, Fontana et al 2009, Lozza et al 2009, Rodrigues de Amorim et al 2010, Nishida et al 2014a, 2014b, Portiansky 2015). Few years ago, we compared the morphometric characteristics of the entire cervical region (C1-C8) of the spinal cord of young (5 mo.) and aged (30 mo.) female rats (Portiansky et al., 2011). Gross anatomy of the organ showed a significant age-related increase in size of all those cervical segments that was not accompanied by an increase of the vertebral canal size. Morphometric analysis of cresyl violet stained sections also showed a significant increase in the area occupied by the gray matter with aging. The most interesting morphometric finding was that both the total area occupied by neurons and the number of neurons significantly increased with age. Some neurons of the aged rats also changed their perimeter size, mean feret diameter and roundness. Proliferation studies using bromodeoxyuridine showed positive labeling for neuron and glia in some cervical segments of both young and aged rats. In aged female rats, serum prolactin (PRL) was increased markedly in comparison with that of young rats. From these results, we hypothesized that in the cervical spinal cord of female rats both maturation of pre-existing neuroblasts and/or neurogenesis occurs during the entire life span, a process in which PRL may play a critical role.We are now working with an experimental model of neurodegeneration in rats injected with kainic acid (KA) into C5 segment of the cervical spinal cord, developed in our laboratory (Nishida et al., 2015). This toxin simulates the clinical signs, biochemical mechanisms and histological changes of different neurodegenerative entities of veterinary interest. Our results showed that neuronal cell count is significantly lowered at C4, C5 and C6 cervical segments of the injected animals. Using immunohistochemistry/immunofluorescence we found positive labeling for vimentin and neurofilament proteins in the perikarya of some neurons. Using colocalization analysis we also confirmed a high colocalization index between neuron specific enolase and vimentin or neurofilaments. We are now studying the response of the glial cells, their changes in number, distribution and morphology. For this purpose, we introduced some stereological methods adapted to the digital analyzers. These studies will allow a better understanding of the communication among neurons, microglia and astrocytes and their possible contribution to neuroregeneration or protection. One of the biggest obstacles in the study of biological systems is the presence of the lipid bilayer that creates a lipid/water interface, thus limiting the depth of the microscopic images obtained. We recently modified the CLARITY (Clear Lipid-Exchanged Acrylamide-hybridized Rigid Imaging / Immunostaining / In situ hybridization-compliant Tissue-Hydrogel) procedure that removes the lipid bilayer, preserving the native molecular information and structure of the organ and thus, facilitates the acquisition of high quality imaging from an entire section of a sample (Chung, et al., 2013). Our modification for processing the spinal cord simplifies the required equipment and adapts the technique for being used in a basic laboratory (Sisti, et al., 2015). Thus, we obtained a transparent spinal cord suitable for immunofluorescence analyzes and 3D image reconstruction. Nowadays, much of the biological data is reported as images. With the development of microscopic technologies, the volume and complexity of image data have increased to the point that it is no longer feasible to extract information without using computers (Portiansky, 2013). Image analysis involves the conversion of features and objects into image data of quantitative relevance. Image analysis replaces the subjective visual inspection into objective information and thus increases sensitivity, accuracy and reproducibility of the analyzed data.