Abstract
Purpose: Circadian rhythms are important in adipose physiology as adipokine plasma concentration, and approximately 20% of the murine adipose transcriptome, undergo diurnal variation. Additionally, overweight and diabetic mice exhibit suppressed amplitude clock and adipokine mRNA rhythms. However, the cellular basis of adipose rhythms is unclear, and clock gene rhythms have not been compared between lean versus overweight or diabetic human subjects. I tested the hypotheses that an endogenous adipocyte oscillator drives rhythmic synthesis and secretion of adipokines, and that rhythms of gene expression are suppressed in overweight or type 2 diabetic patients. Methods: Murine 3T3-L1 pre-adipocyte and adipocyte cultures were characterised. These cells were then ‘pulsed’ with 50% horse serum for 2 hours, and sampled every 4-hours over a 48-hour period. Additionally, lean, overweight, and type 2 diabetic subjects had 4 subcutaneous adipose biopsies taken, one every 6 hours. mRNA expression of clock genes and other genes key to adipose physiology were analysed by quantitative real-time PCR. Secretion of the adipokines leptin and adiponectin were also measured in culture medium from 3T3-L1 adipocytes. Results: Following a serum pulse of characterised 3T3-L1 cultures, circadian rhythms of clock genes Per2, NR1D1 (Rev-erba) and Dbp, but not Perl, Cryl and Arntl (Bmall), were observed in both pre- and postdifferentiated adipocytes. Ppara, Ppary, and Srebfl (Srebpl) also failed to exhibit temporal mRNA changes. Moreover, in the absence of Lep (Leptin) or AdipoQ (Adiponectin) mRNA rhythms in adipocytes, leptin accumulated in the culture medium in a circadian manner. In human adipose samples, PERI-3, CRY2, BMAL1, REV-ERBa and DBP showed diurnal oscillations, but CRY1 did not. In contrast, transription factors NRIP1 (RIP 140) and PPARGC1A (PGCla) exhibited temporal variation of expression, but this was not rhythmic over the timecourse. Interestingly, BMAL1 expression was lower in the overweight group, and PER2 showed lower mRNA levels in medicated versus non-medicated type 2 diabetic patients. Conclusion: 3T3-L1 pre- and post differentiated adipocytes possess an endogenous oscillator, which may control adipocyte leptin secretion, but only drives detectable rhythms of some genes. Additionally, there is only minimal effect of obesity and type 2 diabetes on circadian rhythms in human subcutaneous adipose tissue, although drug regimens may affect clock gene rhythms in type 2 diabetic patients. Together, these data increase understanding of circadian rhythms in adipose biology, and suggest avenues for future research.