Computational protocol: Effect of hypoxia on the retina and superior colliculus of neonatal pigs

Similar protocols

Protocol publication

[…] To analyse the cytoskeletal morphology of astrocytes in the retina, images were taken of whole mount retinas using the same microscope and criteria as those used for RGC quantification. The astrocytes in the pig retina are organized mainly like those in the human retina, running parallel to the RGC axons. A semi-automatic method to measure astrocyte organization was used to measure the morphological changes in astrocytes. The aim was to quantify the differences in the astrocyte network between the control and hypoxic retinas based on the local orientation of the astrocytic processes. An ad hoc programme was developed in Matlab R2010b (MathWorks, Inc) to quantify the morphological changes to astrocytes. This programme estimates the local orientation of the GFAP-positive lines in an image, taking advantage of the SURF extraction algorithm [] and its capacity to define local pixel orientation based on their neighbouring relationships. Following local characterization, a histogram of this orientation is extracted that allows the frequency of each direction to be measured (1-degree bins). This involves measuring the randomness of the histogram using Shannon’s Entropy equation [], a well-known function to measure the predictability of a variable. The higher the entropy the less predictable and consequently, the more disorganized the system. With this information, the degree of disorganization of control astrocytes and hypoxic astrocytes can be compared. The average entropy from the histogram of the images was calculated and a statistical analysis was performed to evaluate our hypothesis that the entropy is greater in hypoxic retinal astrocytes.To analyse the astrocytes in the superior colliculus, a Zeiss Axiocam MRM fluorescent microscope (Zeiss, Jena, Germany) and the Zeiss Zen software was used, obtaining 20 images (20X) at the same rostro-caudal level of the superior colliculus of the 4 control and 4 hypoxic animals. Since the astrocyte organization in the superior colliculus does not follow a well-established parallel pattern, the method used to analyse the astrocytes in this structure differed to that used in the retina. The morphology of the astrocyte cytoskeleton was analyzed using ImageJ (version 1.49), the image processing program developed at the National Institutes of Health (NIH). Using the “threshold color” tool, the region formed by colored pixels (labeled with the antibody against GFAP) was selected, and this area of the cytoskeleton of selected astrocytes was measured. Given the entire image area, we could calculate the proportion of the area occupied by astrocyte´s cytoskeleton in the control and hypoxic superior colliculus. [...] RGC density, the number of active caspase-3 positive cells and the morphological data from the astrocyte’s cytoskeleton were described as the mean and standard error of mean, and these parameters were compared between control and hypoxic tissues. Statistical analyses were carried out using IBM SPSS Statistics software v. 21.0 and the homogeneity of the variances was assayed with Levene´s test (p < 0.05). To assess whether there were significant differences in the number of RGCs, the number of active caspase-3 positive cells and in astrocyte morphology between control and hypoxia conditions, a Student T-test was used. In addition, a Mann-Witney U test was used to verify the differences between the control and hypoxic groups. For both tests, the minimum value of significance was defined as p<0.05. […]

Pipeline specifications

Software tools ImageJ, SPSS
Applications Miscellaneous, Microscopic phenotype analysis
Organisms Sus scrofa
Chemicals Oxygen