Fig. 6: Characterization of the effect of CD103 on thymic architecture.
A Histochemistry sections show representative hematoxylin and eosin (H&E)-stainings of the thymus of WT and 2D2 mice. Results are representative of two independent experiments. B Image scanning and medulla versus cortex area quantification of H&E-stained thymic sections of WT, 2D2, and 2D2.CD103Tgβ7Tg mice using ImageJ Macro. Pie graphs shows the relative medullary and cortical areas for each mouse strains. The demarcation of the medullary area in 2D2.CD103Tgβ7Tg mice was too blurry to be correctly identified by the ImageJ software such that it is considered as “disorganized”. C Immunohistochemistry of WT and 2D2 thymic sections, stained with anti-K14 or -UAE-1 and anti-CD205 or -Ly51 to identify mTECs and cTECs, respectively. D Medulla and cortex area quantification of H&E-stained thymic sections of WT, 2D2, and 2D2.CD103Tgβ7Tg mice by machine-learning random forest pixel classifier segmentation using Arivis Pro (v. 4.2) software. Whole thymic sections of the indicated mice were reimaged and reanalyzed using machine-learning software that permitted improved identification and segmentation of the cortical and medullary areas as shown in the bottom row of the image figure, cortex (cyan), medulla (yellow) and red blood cells (red). Bar graphs show the medulla/cortex ratio of the indicated mice, based on the scanned image areas. Data are presented as individual data points with the bar graphs showing the mean values.
