*4.5. Histopathology*

From the femoral sections of each group, sections thinly cut with a microtome from the femoral neck maximum cut surface of the forehead surface were used as hematoxylin and eosin (H&E)-stained specimens, and the presence/absence of osteonecrosis development was studied using a light microscope. Osteonecrosis was judged to be present when necrosis of medullary haematopoietic cells or fat cells or empty lacunae or condensed nuclei in osteocytes were noted [1]. Regarding the osteonecrosis development rate, all of the groups were compared with osteonecrosis considered to be present even if limited to a unilateral leg.

#### *4.6. Immunohistochemical Study*

To study the effect of taurine in a glucocorticoid-induced rabbit osteonecrosis model in vitro, an immunohistochemical study was performed using TFAM which has a mitochondria protective effect and ATP5A which shows ATP production. In all groups, femoral sections were prepared, and sections obtained from the femoral proximal medial diaphysis of each group were deparaffinized with xylene and ethanol, after which to activate antigen, protease K (Dako Cytomation, Santa Clara, CA, USA) was dribbled on tissue sections that were incubated at 37 ◦C for 40 min. Using 0.3% H2O2, endogenous peroxidase was eliminated, blocking was performed using mouse or goat normal serum, and the primary antibody was made to react. As the primary antibody, anti-TFAM rabbit polyclonal antibody (Invitrogen, Waltham, MA, USA) and anti-ATP5A goat polyclonal antibody (Lsbio, Seattle, MA, USA) were used at 0.1 μg/mL and 5.0 μg/mL, respectively. The reaction time was overnight at 4 ◦C in a cool dark room. After reacting the primary antibody, the secondary antibody (biotin), was reacted with an enzymatic agent (streptavidin), and after 5-minue immersion in DAB to allow for color development, the nuclei were stained and studied under a light microscope.

To quantify relative stainability, 3 fields were chosen at random from tissue adjacent to areas of osteonecrosis in the femoral proximal medial diaphysis, and the proportion of the positive cell count relative to that of the total cell number was calculated, and compared between each group as the percentage of positive cells (%PC).
