Celastrol Preparation and Administration

Celastrol (Sigma, Missouri, USA) stock solution of 10 mg/ml was prepared using ethanol 100% as solvent (vehicle). A recent study has tested the solubility of celastrol in different vehicles, and we have previously demonstrated that ethanol is one of the most efficient solvents for this compound (21). This celastrol stock solution was further dissolved in PEG400 (Sigma, St. Louis, USA) (1, 2.5, 5, 7.5, 12.5, and 25 μg/g in 1 ml) and administrated by oral gavage to AIA rats for 14 consecutive days (N = 5 rats/group). The sample size in each group was calculated using free sample size calculating G^*Power version 3.1.9.2 software [type of power analysis: a priori; α error probability: 0.05; power (1-β error probability): 0.95; effect size d: 2.59; actual power: 0.976]. This calculation was based on our own previous data (7, 8, 17).

Our study was approved by the institutional animal welfare body, licensed by the Portuguese competent authority, and complied with good ethical, scientific, legal, and economic reasons for using laboratory animals, including the 3R principle (replace, reduce, and refine). Focused on the “reduce” rule, we used the minimum number of animals, calculated by the free sample size calculating G^*Power software, in order to perform this study.

The need for daily administrations is supported by the study of Zhang et al., which showed that the half-life of pure celastrol is ~10 h (22). Based on this publication, we have also calculated the oral dose of 2.5 μg/g/day as the equivalent to the intraperitoneal dose of 1 μg/g/day that we had previously found to be effective in the treatment of arthritis in the same rat model (7, 8). The calculation was based on the fraction of the celastrol dose absorbed in the portal blood after oral administration, which was 17.6%. Intraperitoneal dose calculation took into consideration that the intraportal dose will have higher bioavailability. Therefore, the relationship between the area under the curve and the doses administered in the oral and the intraportal dose was used for determining the intraperitoneal dose. Treatment was initiated after 8 days of disease induction, at the acute clinical stage of arthritis progression (therapeutic intervention) (32). Healthy non-arthritic (N = 8) and arthritic untreated (N = 15) female age-matched Wistar rats were used as controls. The arthritic untreated group received an equal volume of vehicle through oral gavage. The vehicle proportion of ethanol and PEG400 used was in the same proportion as the one used in the celastrol-treated groups.

The rats were sacrificed after 22 days of disease progression by CO₂ narcosis, and blood, internal organs, as well as paw samples were collected. Studies using the AIA model are generally completed at this time point due to a plateau effect of inflammatory manifestations (7, 31).