Blinded evaluation of potential obesogens in murine 3T3-L1 preadipocytes

Blinded evaluation of potential obesogens in murine 3T3-L1 preadipocytes following selection from database results


Manushree Bharadwaj




Bharadwaj M, Chappell VA, Thayer KA, Allen DG, Polk WW, Rice J, Fenton SE


National Toxicology Program Laboratory, DNTP, NIEHS, Office of Health Assessment and Translation, DNTP, NIEHS, Integrated Laboratory Systems, Inc., Durham, NC, USA


Identifying environmental chemicals which play a role in triggering obesity is essential to propose new research directions and elucidate critical signaling pathways. Our objective was to screen numerous chemicals, selected as potential obesogens based on their receptor activity data in the Tox21 phase II high-throughput screening (HTS), for their adipogenic and lipogenic effects on murine 3T3-L1 preadipocytes. There were two rounds of chemical selection, the first was based strictly on Tox21 HTS data, the second selection criteria were made more stringent by harmonizing Round I data with animal study data in EPA’s Toxicity Reference (ToxRef) and the National Toxicological Program’s Chemical Effects in Biological Systems (CEBS) databases; both sets of chemicals were similarly tested in a blinded-to-treatment design. Low passage 3T3-L1 preadipocytes were exposed in 96-well plates to a range of concentrations (log dilutions from 10 µM to 1 pM) of the 50 selected positive and 17 negative (did not activate the receptors of interest) chemicals for eight days in quadruplicate. Cells treated with 0.05% DMSO and 1 µM rosiglitazone were used as the negative and positive controls, respectively. For each chemical, half the wells were induced to differentiate into adipocytes by treating them with a mixture containing 0.5 mM methyl isobutylxanthine, 1 µM dexamethasone, and 10 mg/L insulin. Cells were fixed in 4% paraformaldehyde on day 8 and stained with Hoechst and Nile red dyes to quantitate, by high-content microscopic imaging, the adipocyte count and extent of lipid droplet accumulation per adipocyte. Our data demonstrated that a positive response was predicted better than a negative response based on our selection criterion. Amongst the 17 Round I chemicals predicted as positive, 13 induced adipo-/lipogenesis in 3T3-L1 preadipocytes. However, in Round II, only 20 and 15, out of 33 predicted positives, significantly increased adipo-/lipogenesis, resulting in accuracy of prediction for positives decreasing from approximately 76% (Round I) to 60% (Round II). Those chemicals predicted to be negative gave more than a 50% positive response rate, suggesting that signaling pathways outside those selected in our criterion are needed for adipo-/lipogenic responses. These important findings have spurred investigation into alternative pathways involved in obesogenic responses and additional transcript-/lipidomic studies for several positive response chemical families.