Bezafibrate (α/β,δ/γ agonist) did not decrease alcohol consumption in our study. There are a few considerations that might account for this. β/δ activation might oppose or prevent action at α and γ subunits. Testing a PPAR agonist with specificity for β/δ will be necessary to help determine why bezafibrate was not effective. Furthermore, it should be noted that α/β/γ isotypes function as a triad, and β/δ regulates the expression level and activity of PPAR α and γ 2, making it difficult to determine the role of a single PPAR isotype in alcohol drinking behavior following bezafibrate treatment. Finally, it is important to consider the doses of the drugs used for our studies. We compared the drug dosages in two ways: 1) in comparison to the human dose calculated as mg/body surface area, a validated measure for human-mouse comparisons 46 and 2) as a ratio of dose to the half maximal effective concentration (EC50) required for subtype-specific action 30, 65 (Table S6). Fenofibrate and tesaglitazar were used at about 5–7 times the human dose whereas bezafibrate was equivalent to the human
