Integrator system software

After formalin fixation, dehydration, and paraffin embedding, 4 µm sections were stained for Hematoxylin Eosin (HE), PicroSirus red (PSR), and immunohistochemistry (IHC). IHC was performed on an automated Ventana Ultra system (Ventana Medical Systems, Roche Group, USA). The rat monoclonal (M3/38) antibody against MAC2 (Galectin-3, CL8942AP, Cedarlaine, 1:5000 dilution) or rabbit polyclonal antibody against COL1A1 (LS-C343921, LSBio, 1:1000 dilution) were used to detect macrophages or collagen deposits in liver sections, respectively. Antibody binding was visualized using DAB, hematoxylin was added as a nuclear counterstain.
Image analysis was performed on digital images of the HE, MAC2, and COL1A1 stained slides, using Visiopharm Integrator System software (v2019.2, Visiopharm). Lipidosis was calculated based on the unstained area (fat) on the HE stained slides and corrected to the total section area. For MAC2 and COL1A1 quantification, DAB positive area was quantified and normalized to the total section area. The areas were detected by threshold and machine learning.
Fibrosis score assessment was performed on PSR stained slides at 200x magnification, according to Kleiner et al.^{56 (link)}. The definitions of fibrosis stages are: 0= no fibrosis, 0.5-1= perisinusoidal or periportal, 2= perisinusoidal and portal/periportal, 3= bridging fibrosis, 4= cirrhosis.

Immunohistochemistry staining was carried out for the podocyte markers nephrin and WT1 (see Table 1 for antibody details).
A novel method was introduced to allow automated unbiased quantification of nephrin and WT1 staining in multiple, adjacent, immunostained, 2‐μm transverse kidney sections that were digitized using slide scanners (Zeiss Axioscan or Zeiss Mirax, Carl Zeiss AG, Munich, Germany), fitted with a 20× air objective yielding images with 4.545 pixels/μm. The resulting virtual slides were imported into Visiopharm Integrator System software (version 5.3, Visiopharm, Hørsholm, Denmark).
For analysis, the tissues from sequential sections were aligned and then resampled at 1.138 pixels/μm using systematic random sampling of the cortical region with coverage of 60%. Machine learning and Bayesian classification (Visiopharm, Visiomorph DP module), followed by morphological filtering, identified tissues and stained areas. Individual glomeruli were identified using nephrin staining and then used to generate regions of interest on the adjacent tissue section that had been stained for WT1 expression. For each glomerulus, the number of WT1‐positive nuclei and the fractional area of WT1 staining were calculated.

To assess cell proliferation in the dentate gyrus (DG) using stereological analysis, brains were embedded in optimum cutting temperature compound and snap-frozen. Serial coronal 20 μm sections were cut in a cryostat, extending over the entire length of the hippocampus. To detect Ki67, a nuclear protein expressed in all phases of the cell cycle except the resting phase G0, a mouse monoclonal anti-Ki67 (Novocastra, UK; 1:100 dilution) was used accordingly with standard procedures. The primary antibody was detected by the Ultravision Detection System (Lab Vision, CA, USA), and the reaction developed with 3,3′-diamino- benzidine substrate (Sigma Aldrich, MO, USA; DAB: 0.025 and 0.15% H₂O₂ in Tris-HCl 0.05 M, pH 7.2). Sections were then counterstained with hematoxylin.
Hippocampal cell proliferation was measured by counting the cells expressing Ki-67 in the subgranular zone (SGZ), considered as the 3-cell-body-wide zone at the border of the DG and normalized by the respective area (results are presented as number of Ki67⁺ cells per mm²). The use of the visiopharm integrator system software (Visiopharm, Denmark) allowed the delimitation, at low magnification (40x), of the areas of interest and the identification of the Ki67⁺ cells within the defined areas was performed at higher magnification (400x). Counts were performed by one researcher blind to the experimental conditions.

Follicle density was evaluated in the xenografts after 4 weeks. Only morphological ‘normal’ follicles with clear nucleus were counted, and follicles were defined and counted as previously described [26 (link)]. Primordial follicles were characterized by an oocyte surrounded by a single layer of flat granulosa cells. Primary follicles as oocytes surrounded by a layer of cuboidal granulosa cells, and secondary with two or more layers of cuboidal granulosa cells [46 (link)]. Morphologically, follicles were subcategorized as healthy follicles or atretic follicles [47 (link), 48 (link)]. For area measurement Olympus BH-2 microscope with Visiopharm Integrator System software (Visiopharm, Hoersholm, Denmark, version 4.6.1779) was used. Density was calculated by the total number of follicles and the volume of the graft based on the area measurements. Follicle morphology was subcategorized (healthy or atretic), percentages from total number of the subcategories was used to visualize the data (Fig. 6).