Bz 2

Endogenous peroxidation was blocked with proteinase K (DAKO, Tokyo, Japan) and 0.1% (v/v) hydrogen peroxide. Polyclonal anti-rabbit AQP4 (#AB3594; 1:500; Millipore, Temecula, CA, USA) and streptavidin-biotin-labeled secondary antibody (DAKO, Tokyo, Japan) were used for AQP4 detection. For color development, an avidin-biotin complex staining kit (Thermo Fisher Scientific, Waltham, MA, USA) and diaminobenzidine substrate (DAKO, Tokyo, Japan) were used according to the manufacturer’s instructions. The stained sections were counterstained with hematoxylin and visualized under a microscope (BZ-X710; KEYENCE, Osaka, Japan). The images were quantitatively analyzed with BZ-II (KEYENCE, Osaka, Japan). The average values of the selected regions were quantified using six slices per mouse. The AQP4 expression levels in three randomly selected hippocampal areas (97,200 µm²; 360 × 270 µm/each area) were computed as the percentage of brain regions presenting with AQP4 immunoreactivity. AQP4 polarization around vessels was determined by calculating the ratios of AQP4 expression in the perivascular and parenchymal domains [33 (link)].

Brain tissue sections were stained with cresyl violet solution for 90 min and scanned under a microscope (BZ-X710; KEYENCE, Osaka, Japan). The average values for the selected regions were quantified using three slices per mouse. The average numbers of Nissl-positive cells were determined for three randomly selected areas (97,200 µm²; 360 × 270 µm/each area) in the CA1, CA2, and CA3 regions of the hippocampus. The cells were counted with BZ-II (KEYENCE, Osaka, Japan).

Total lesion area: Using histological pictures stained by HE and MT stains, the area of granulation tissue developed between the abdominal muscle wall, including fibrosis and regenerative and migrated muscle cells, was morphometrically measured with a computer-assisted-software BZ-II (Keyence).
Fibrosis area: representative lesions in MT stain were taken using 10-time lens in the Bio-Zero (Keyence), and the blue staining collagenous fibers were highlighted, binarized, and measured (VH-analyzer: Keyence, Osaka, Japan), and the percentage of collagen distribution in the granulation tissue was calculated.
Myoblast number: Myogenin-stained granulation tissue was taken using a 10-time lens, and the positive spots in the area of granulation tissue were calculated.
Regenerated myocyte area: Myoglobin-stained granulation tissue was taken using a 10-time lens, and the positive area of regenerative muscles in the granulation area was calculated.
Positive cell number for p-Smad3: Cells with a positive nucleus for p-Smad3 were counted out of the cells in tissue photographed with a 20-time lens. Examined tissues were of the granulation tissue and the stumped muscle (Fig. 7B).

Visceral adipose tissue was fixed with buffered formaldehyde solution (10%) followed by standard protocol for paraffin-embedded sections and hematoxylin-eosin (HE) staining. Images of the HE-stained adipose tissue were incorporated with a BIOREVO microscope (BZ-9000; KEYENCE), and morphometric analyses were performed with image analysis software (BZ-II) equipped on the microscope. Liver samples were snap-frozen in OCT compound (Sakura Finetek) with liquid nitrogen, and the cryosections were stained with Oil-Red-O (ORO) (Sigma). Images of lipid droplets in hepatocytes stained red were also quantified by image analysis software (BZ-II).