Antidiabetic screening of commercial botanical products in 3T3-L1 adipocytes and db/db mice

Numerous botanicals are purported to improve glucose metabolism and diabetic risk factors with varying degrees of supportive evidence. We investigated 203 commercially available botanical products representing 90 unique botanical species for effects on lipogenic activity in differentiating 3T3-L1 adipocytes. Anti-inflammatory activity of 21 of these products was further assessed in tumor necrosis factor alpha (TNFalpha)-stimulated, mature 3T3-L1 adipocytes. From these results, rho-isoalpha acids, Acacia nilotica bark, fennel, and wasabi were tested in the db/db mouse model. Fifty-nine percent of the 90 unique botanicals increased adipogenesis as did the standard troglitazone relative to the solvent controls. Botanical species with the greatest percentage of positive products were Centella asiatica, Panax quinquefolius, and Phyllanthus amarus at 100%, Vitis vinifera at 80%, Humulus lupulus at 71%, Aloe barbadensis at 66%, and Momordica charantia, Phaseolus vulgaris, and Punica granatum at 60%. All 21 subset samples inhibited TNFalpha-stimulated free fatty acid release and attenuated TNFalpha inhibition of adiponectin secretion. Both rho-isoalpha acids and A. nilotica reduced nonfasting glucose in the db/db mouse model, whereas A. nilotica also decreased nonfasting insulin levels. A post hoc analysis of the screening results indicated that the positive predictive value of the lipogenesis assay alone was 72%, while adding the criterion of a positive response in the anti-inflammatory assays increased this figure to 82%. Moreover, this large-scale evaluation demonstrates that antidiabetic, in vitro efficacy of botanicals is more a function of manufacturing or quality control differences than the presence of marker compounds and further underscores the need to develop functional as well as analytical bases for standardization of dietary supplements.