Grizzly bears (Ursus arctos horribilis) have evolved remarkable metabolic adaptations including enormous fat accumulation during the active season followed by fasting during hibernation. However, these fluctuations in body mass do not cause the same harmful effects associated with obesity in humans. To better understand these seasonal transitions, we performed insulin and glucose tolerance tests in captive grizzly bears, characterized the annual profiles of circulating adipokines, and tested the anorectic effects of centrally administered leptin at different times of the year. We also used bear gluteal adipocyte cultures to test insulin and beta-adrenergic sensitivity in vitro. Bears were insulin resistant during hibernation but were sensitive during the spring and fall active periods. Hibernating bears remained euglycemic, possibly due to hyperinsulinemia and hyperglucagonemia. Adipokine concentrations were relatively low throughout the active season but peaked in mid-October prior to hibernation when fat content was greatest. Serum glycerol was highest during hibernation, indicating ongoing lipolysis. Centrally administered leptin reduced food intake in October, but not in August, revealing seasonal variation in the brain's sensitivity to its anorectic effects. This was supported by strong phosphorylated signal transducer and activator of transcription 3 labeling within the hypothalamus of hibernating bears; labeling virtually disappeared in active bears. Adipocytes collected during hibernation were insulin resistant when cultured with hibernation serum but became sensitive when cultured with active season serum. Heat treatment of active serum blocked much of this action. Clarifying the cellular mechanisms responsible for the physiology of hibernating bears may inform new treatments for metabolic disorders.

The confluence of obesity and diabetes as a worldwide epidemic necessitates the discovery of new therapies. Success in this endeavor requires translatable preclinical studies, which traditionally employ rodent models. As an alternative approach, we explored hibernation where obesity is a natural adaptation to survive months of fasting. Here we report that grizzly bears exhibit seasonal tripartite insulin responsiveness such that obese animals augment insulin sensitivity but only weeks later enter hibernation-specific insulin resistance (IR) and subsequently reinitiate responsiveness upon awakening. Preparation for hibernation is characterized by adiposity coupled to increased insulin sensitivity via modified PTEN/AKT signaling specifically in adipose tissue, suggesting a state of "healthy" obesity analogous to humans with PTEN haploinsufficiency. Collectively, we show that bears reversibly cope with homeostatic perturbations considered detrimental to humans and describe a mechanism whereby IR functions not as a late-stage metabolic adaptation to obesity, but rather a gatekeeper of the fed-fasting transition.

Brown bears (Ursus arctos) exhibit hyperphagia each fall and can become obese in preparation for hibernation. They do this without displaying the physiological problems typically seen in obese humans, such as Type 2 diabetes and heart disease. The study of brown bear hibernation biology could therefore aid in the development of novel methods for combating metabolic diseases. To this end, we isolated mesenchymal stem cells from subcutaneous fat biopsies, and culture methods were developed to differentiate these into the adipogenic lineage. Biopsies were taken from 8 captive male (N = 6) and female (N = 2) brown bears, ages 2-12 years. Plastic adherent, fibroblast-like cells were proliferated and subsequently cryopreserved or differentiated. Differentiation conditions were optimized with respect to fetal bovine serum content and time spent in differentiation medium. Cultures were characterized through immunostaining, RT-qPCR, and Oil red O staining to quantify lipid accumulation. Adiponectin, leptin, and glycerol medium concentrations were also determined over the course of differentiation. The culturing protocol succeeded in generating hormone-sensitive lipase-expressing, lipid-producing white-type adipocytes (UCP1 negative). Serum concentration and time of exposure to differentiation medium were both positively related to lipid production. Cells cultured to low passage numbers retained similar lipid production and expression of lipid markers PLIN2 and FABP4. Ultimately, the protocols described here may be useful to biologists in the field investigating the health of wild bear populations and could potentially increase our understanding of metabolic disorders in humans.

Amgen requested the retraction as an outcome of an internal review where it was determined that one of the Amgen authors had manipulated specific experimental data presented in Figures 1 and 3. Because of data manipulation, this author is no longer employed by Amgen. The authors at Washington State University and University of Idaho are confident that the physiological data generated for this manuscript are accurate and representative of the true metabolic responses of these grizzly bears and are currently repeating the mechanistic portions of the study. Amgen deeply regrets this circumstance and extends their sincere apologies to the scientific community.