Glucocorticoids antagonize tumor necrosis factor-α-stimulated lipolysis and resistance to the antilipolytic effect of insulin in human adipocytes

Abstract

High concentrations of TNF within obese adipose tissue increase basal lipolysis and antagonize insulin signaling. Adipocytes of the obese are also exposed to elevated levels of glucocorticoids (GCs), which antagonize TNF actions in many cell types. We tested the hypothesis that TNF decreases sensitivity to the antilipolytic effect of insulin and that GCs antagonize this effect in differentiated human adipocytes. Lipolysis and expression levels of lipolytic proteins were measured after treating adipocytes with TNF, dexamethasone (DEX), or DEX + TNF for up to 48 h. TNF not only increased basal lipolysis, it caused resistance to the antilipolytic effects of insulin in human adipocytes. DEX alone did not significantly affect lipolysis. Cotreatment with DEX blocked TNF induction of basal lipolysis and insulin resistance by antagonizing TNF stimulation of PKA-mediated phosphorylation of hormone-sensitive lipase (HSL) at Ser⁵⁶³ and Ser⁶⁶⁰ and perilipin. TNF did not affect perilipin, HSL, or phosphodiesterase-3B mass but paradoxically suppressed adipose tissue triglyceride lipase expression, and this effect was blocked by DEX. The extent to which GCs can restrain the lipolytic actions of TNF may both diminish the potentially deleterious effects of excess lipolysis and contribute to fat accumulation in obesity.

Fig. 1.

Dexamethasone antagonized TNF-induced lipolysis during culture. A: differentiated sc human adipocytes were treated with TNF (1 or 10 ng/ml) for 3 and 24 h, n = 2. B: adipocytes were treated with different concentrations of dexamethasone (DEX, 0, 10, and 1,000 nM) in the absence or presence of TNF (10 ng/ml) overnight, n = 6. C: adipocytes were treated in conditions of control (no treatments, ctrl), Dex (D, 10 nM), TNF (3 ng/ml), and TNF+D up to 48 h. Media along with hormones were replenished daily, n = 5. D: after treatment of differentiated human adipocytes, cell lysates were harvested and used for immunoblotting for fatty acid-bind protein 4 (FABP4) and heat shock protein-90 (HSP90) (a loading control). E: sc adipose tissue fragments were cultured in control (0.7 nM insulin), DEX (10 nM), TNF 3 ng/ml, or TNF + DEX for 24 h. Glycerol accumulation in culture media was measured, n = 3. TNF effects: *P < 0.05, **P < 0.01, ***P < 0.001; DEX effects: #P < 0.05, ##P < 0.01, ###P < 0.001.

Fig. 2.

Dex antagonized TNF stimulation of basal lipolytic rates. A: after overnight treatment with TNF (1 or 10 ng/ml), lipolytic rates were measured in KRB + 4% BSA in basal and isoproterenol (Iso, 10⁻⁶ M)-stimulated conditions, n = 2. B: after overnight pretreatments with media alone (control), Dex (D, 10 nM), TNF (3 ng/ml), or TNF + D, acute lipolysis was measured in KRB + 4% BSA in basal and Iso (10⁻⁶ M)-stimulated conditions. TNF effects: *P < 0.05; DEX effects: #P < 0.05, n = 4.

Fig. 3.

Dex blocked TNF-induced impairment in insulin antilipolysis. A: after overnight treatment in control, DEX (10 nM), TNF (3 ng/ml), or TNF + DEX (10 nM), acute lipolysis was performed in KRB + 4% BSA. Antilipolytic effects of insulin were measured against 8-bromo-cAMP-stimulated lipolysis at 0, 30, 120, and 600 pM. B: data from A were recalculated as percent suppression. TNF effects: *P < 0.05, **P < 0.01; DEX effects: #P < 0.05; TNF + DEX vs. TNF, P = 0.06 at 30 pM, n = 4.

Fig. 4.

Dex blocked TNF stimulation of hormone-sensitive lipase (HSL) and perilipin phophorylation. A: after treatment of differentiated human adipocytes at indicated conditions, cell lysates were harvested and used for immunoblotting. Representative blots of 4 experiments are shown. Relative expression of phospho-HSL (Ser⁵⁶³) to total HSL (B) and adipose tissue triglyceride lipase (ATGL) to RNA POL II (C) is quantified. TNF effects: *P < 0.05; DEX effects: #P < 0.05, n = 4.

Fig. 5.

Time course effects of Dex and TNF on phosphodiesterase (PDE)3B, HSL, perilipin, and ATGL protein expression levels. After treatment of differentiated adipocytes for 3, 24, and 48 h with medium alone (Ctrl), Dex (D, 10 nM), TNF (3 ng/ml), or D + TNF, cell lysates were prepared and used for immunoblotting. A: representative blots from 4 independent experiments are shown. Relative expression of phospho-HSL (Ser⁵⁶³) to total HSL (B) and ATGL to RNA POL II (C) are quantitated. TNF effects: *P < 0.05, **P < 0.01; DEX effects: #P < 0.05, ##P < 0.01, n = 4.

Fig. 6.

Time course effects of Dex and TNF on HSL, perilipin, and ATGL mRNA expression levels. Differentiated human adipocytes were incubated as indicated. Total RNA was isolated, and mRNA expression levels were measured by RT-qPCR. TNF effects: *P < 0.05; DEX effects: #P < 0.05, ##P < 0.01, n = 4.