Altered ERK‐mediated control of AgRP and metabolic rate during obesity

Abstract

Adaptation of resting metabolic rate (RMR) contributes to the maintenance of human obesity, yet RMR control and how this system changes during obesity remains poorly defined. Desensitization of the brain to leptin has been proposed to contribute. Leptin normally suppresses Agouti‐related peptide (AgRP) expression in the arcuate nucleus of the hypothalamus (ARC) to disinhibit (increase) melanocortin‐mediated RMR stimulation. Diet‐induced obesity (DIO) increases leptin, but paradoxically exaggerates Agrp expression and causes RMR adaptation. Thus, something connecting leptin to Agrp control becomes dysfunctional during DIO. Our previous data indicate that within the ARC, disruption in the extracellular signal‐regulated kinase (ERK) cascade occurs during DIO. ERK normally phosphorylates Krüppel‐like factor 4 (KLF4), inactivating it. As KLF4 binds to the AgRP promoter to increase transcription, its inactivation results in reduced Agrp expression. We hypothesize that this mechanism becomes dysfunctional within AgRP neurons during DIO, such that KLF4 binding to the AgRP promoter increases despite increased ERK activity.Mouse hypothalamic GT1‐7 cells (model of AgRP neurons) were treated with leptin (1 uM, 15 min or 1 hr) and pERK/ERK ratio was measured by Simple WES and Agrp mRNA expression via qPCR, respectively. Leptin was given to chow‐fed adult mice (2 ug/g ip 30 min) and ARC pERK/ERK ratio was measured, as well as RMR assessed by respirometry. To examine consequences of ERK or KLF4 disruption upon cardiometabolic phenotypes, mice harboring a conditional version of endogenous Mapk1 or Klf4 were crossed with mice expressing Cre via the Agrp locus.In GT1‐7 cells, leptin increased pERK/ERK ratio (1 uM, 15 min leptin n=3, +89% vs vehicle n=2, p<0.05) and suppressed Agrp expression (1 uM, 1 hr leptin n=2, ‐54.1% vs vehicle n=6, p<0.05). In chow‐fed adult mice, leptin injection caused an increase in the ARC pERK/ERK ratio (2 ug/g ip 30 min leptin n=11, +43.8% vs saline n=9, p<0.05). Additionally, leptin stimulated RMR in adult‐chow fed mice (1 ug/g ip leptin n=8, 16.5‐fold of vehicle n=8, p<0.05). DIO caused increases in endogenous plasma leptin and an increase in the ARC pERK/ERK ratio (High Fat Diet (HFD; n=5), +54.4% vs chow n=4, p<0.05), but ARC Agrp expression was not suppressed (HFD n=4, 1.62 fold of chow n=4, p<0.05). There was a possible exaggeration of weight gain in Agrp‐Cre+, Mapk1F/F mice compared to controls following HFD (KOs n=2, +16.8% vs controls n=3). In contrast, weight gain was possibly attenuated in Agrp‐Cre+, Klf4F/F mice compared to controls following HFD (KO n=1, ‐10.7% vs controls n=2). RMR was higher in chow‐fed Agrp‐Cre+, Klf4F/F mice compared to controls (KOs n=5 +8.3% vs controls n=7, p<0.05) and may be exaggerated following HFD (KO n=1 +13.0% vs controls n=3). Together, these findings suggest that ERK and KLF4 within AgRP neurons appear to contribute to the control of AgRP and thus RMR. These findings may identify ERK control of KLF4 and KLF4 as a potential therapeutic target to maintain/restore RMR control after prolonged obesity.