Aperio imagescope software

The tissue of one-year-old female BALB/c mice intranasally infected with 10³ PFU of mouse-adapted SARS-CoV-2 (MA10) or mock-infected [16 (link)] were stored in 10% phosphate-buffered formalin fixative and processed for paraffin embedding. After paraffin embedding, paraffin sections (5 μm in thickness) were used for hematoxylin and eosin (H&E) stains to identify morphological changes in brains. Slides were scanned with a digital slide scanner (Zeiss Axio Scan; Zeiss, White Plains, NY, USA). Representative photomicrographs at 20× magnification were acquired from whole scanned coronal sections using the Aperio Image Scope software (version 12.3.2.8013, Leica, Buffalo Grove, IL, USA). The number of perivascular cuffing sites was counted from all the mice in each group (uninfected control n = 5; SARS-CoV-2 (MA10)-infected n = 5) and was quantified by a pathological score. Whole specimens were examined by using Aperio Image Scope software to establish a histopathological score in each case.

Three or four micron sections of human omentum were subject to immunohistochemistry using the aforementioned primary antibodies, as previously described^{23 (link)}. To quantify RNase 3 reactive cells in omentum, slides were scanned at 20x magnification (NanoZoomer, Hamamatsu, Japan) and analyzed with Aperio ImageScope software (Leica, Wetzlar, Germany). Large vessels were excluded. The Positive Pixel Count Algorithm was used to determine the fraction of RNase3+ cells among total cell nuclei, expressed as a percentage.

IHC staining was performed using Ventana Benchmark XT automated closed system following their recommended protocol. EZH2 and H3K27me3 antibodies were purchased from Cell Signaling Technology, and KRT14 was obtained from Abcam. Colon, Skin, and Appendix were used as internal positive control for KRT14, EZH2, and H3K27me3 staining, respectively. All stained slides were analyzed by Leica Aperio ImageScope software. The staining intensity of each section was scored as 0 (no staining), 1+ (weak staining), 2+ (moderate staining), or 3+ (strong staining). The tumor cell positive rate (0–100%) per slice was multiplied by the staining intensity to get an overall H-scores ranging from 0 to 300.

Heart and kidney samples fixed in 4% PFA were processed to paraffin and sectioned at 4 μm. Representative midline sections from each kidney and heart were stained with Periodic acid–Schiff (PAS) and Masson’s Trichrome by Monash Histology Platform. The Aperio Scanscope AT Turbo scanner (Leica Microsystems, NSW, Australia) was used to generate digital images of the sections (20 × magnification). All analyses were performed by a researcher blinded to treatment groups using Aperio ImageScope software (Leica, RRID:SCR_020993), Fiji ImageJ (RRID:SCR_002285) or QuPath version 0.2.3 (RRID:SCR_018257) [19 (link)].

Histologic evaluation was performed by an experienced cytotechnologist (R. Cora) with specific training and expertise in assessing TDLU parameters and who was blinded to all participant characteristics and genotyping. H&E-stained digital images were reviewed using the Aperio ImageScope software from Leica (version 12.3.3); no samples contained either preneoplastic or malignant cells, and samples without any obvious epithelial component were considered ineligible for review. The mean tissue area scored was 35.26 mm², and total counts of normal TDLUs were calculated per 100 mm² and included any TDLU with at least 2 acini associated with a discernable lumen. Ninety-one of our participants had TDLU counts independently determined for another KTB study by Mayo Clinic pathologist M. Sherman, MD; these results were made available to us after our counts were completed and the paired readings showed high correlation, r = 0.89.
Nine of the 200 participants were not included in the final analysis due to lack of detectable epithelium on their biopsy slide (3) or inconclusive rs2016347 genotyping (6), resulting in an analytical dataset of 191.