Mitochondrial 2,4-dienoyl-CoA reductase (Decr) deficiency and impairment of thermogenesis in mouse brown adipose tissue

Abstract

A large number of studies have demonstrated significance of polyunsaturated fatty acids (PUFAs) for human health. However, many aspects on signals translating PUFA-sensing into body homeostasis have remained enigmatic. To shed light on PUFA physiology, we have generated a mouse line defective in mitochondrial dienoyl-CoA reductase (Decr), which is a key enzyme required for β-oxidation of PUFAs. Previously, we have shown that these mice, whose oxidation of saturated fatty acid is intact but break-down of unsaturated fatty acids is blunted, develop severe hypoglycemia during metabolic stresses and fatal hypothermia upon acute cold challenge. In the current work, indirect calorimetry and thermography suggested that cold intolerance of Decr^-/- mice is due to failure in maintaining appropriate heat production at least partly due to failure of brown adipose tissue (BAT) thermogenesis. Magnetic resonance imaging, electron microscopy, mass spectrometry and biochemical analysis showed attenuation in activation of lipolysis despite of functional NE-signaling and inappropriate expression of genes contributing to thermogenesis in iBAT when the Decr^-/- mice were exposed to cold. We hypothesize that the failure in turning on BAT thermogenesis occurs due to accumulation of unsaturated long-chain fatty acids or their metabolites in Decr^-/- mice BAT suppressing down-stream propagation of NE-signaling.

Figure 1

Degradation of polyunsaturated fatty acids in mitochondria.

Figure 2

Indirect calorimetry. (A) Basal metabolic rate (BMR) at +31 °C of Decr^−/− mice (KO) and wild type mice (WT) with or without prior 18 h-fasting indicated as milliliters of oxygen consumed per hour per gram of body weight. (B) Oxygen consumption (VO₂) and respiratory exchange ratio RER of WT and KO mice upon 6 h linear cooling exposure measured by indirect calorimetry. Upper graphs: oxygen consumption and respiratory exchange ratio of mice under fed state. Lower graphs: oxygen consumption and respiratory exchange ratio of mice after fasting of 18 hours. (C) Oxygen consumption and respiratory exchange ratio of WT and KO mice after NE-induction of BAT thermogenesis (1 mg/kg body weight, s.c.) at +31 °C. NE-injection was given after obtaining a 10-min stable basal value. The mean VO₂ from this period was set to 1 to calculate the relative increase in VO_2. After placing the mice back to the chambers, oxygen consumption was measured for 40 minutes until the experiment was terminated. Upper graphs: oxygen consumption and respiratory exchange ration of fed mice. Lower graphs: oxygen consumption and respiratory exchange ratio of mice with prior fasting of 23 hours. (D) Body temperature of WT and KO mice at the beginning and at the end of cold exposure of indirect calorimetry. Statistically significant differences between wild type and Decr^−/− mice are indicated by asterisks (***p < 0.001). Values are shown as group (n = 4) means ± SE.

Figure 3

Heat production by interscapular brown adipose tissue (iBAT). (A) Western blot analysis of iBAT and WAT tissue homogenates by using antibody against rat DECR showing presence or absence of the 33 kDa band corresponding to DECR in wild type (WT) and Decr^−/− (KO) mice, respectively. β-actin was used as a loading control. (B) Thermography of heat formation on the surface above iBAT was studied from fasted wild type (WT) and Decr^−/− mice (KO) pairwise (WT on the left, KO on the right) using infrared imaging. Mice were anesthetized and injected subcutaneously with NE (1 mg/kg body weight). Imaging of the heat formation on the surface above iBAT (circled area) was continued 30 min after the NE-injection. The general decreasing trend in both groups is due to the anesthesia. (C) Maximum surface temperatures were calculated using image analysis software. Values are expressed as means of ± SE of four mouse pairs. (D) iBAT mass index was analyzed from WT and KO mice from four treatment-groups: (i) normal feeding, no fasting, (ii) fasting (24 hours), (iii) cold exposure (3 hours) and iv) fasting (24 hours) including cold exposure (3 hours). Values are expressed as group (n = 9) means ± SE. Statistically significant differences are indicated by asterisks (*p < 0.05, **p < 0.01).

Figure 4

Electron microscopy analysis of iBAT, whole body magnetic resonance imaging and lipolysis in adipose tissues. (A) Representative EM images from iBAT of wild type (WT) and Decr^−/− (KO) mice on fed stage (images were taken with 13500 x magnification). (B) Representative EM images from iBAT of WT and KO mice on fed and fasted stage and after cold exposure (images were taken with 890 x magnification). (C) Phase analysis of lipid droplets vs. cytosolic area on EM images in fed and fasted state. The values are shown as group (n = 10 images) means of ± SE. Statistically significant differences between WT and KO mice are indicated by asterisks (**p < 0.01). (D) Relative body fat/water content in WT and KO mice before and after fasting as quantified using magnetic resonance imaging. Data are presented as group means ± SE. (n = 6). Ex vivo lipolysis rates in (E) WAT and (F) BAT under basal and isoproterenol-stimulated conditions (n =4–5).

Figure 5

MS analysis of iBAT and WAT fatty acids. Total triacylglycerol bound fatty acid content and composition in brown adipose tissue (A) and white adipose tissue (B) of wild type (WT) and Decr^−/− (KO) mice in fed and fasted stages were analyzed by mass spectrometry. WT fed mice versus Decr^−/− fed mice or fasted wt *p < 0.05, **p < 0.01, and ***p < 0.001; Decr^−/− fed mice versus fasted Decr^−/− mice ^#p < 0.05, ^##p < 0.01, and ^###p < 0^.001; and fasted wt versus fasted Decr^−/− mice ^$p < 0.05, ^$$p < 0.01, and ^$$$p < 0.001. AA = arachidonic acid.

Figure 6

Western blot analysis. Representative immunoblots of proteins involved in regulatory pathways of lipolysis and thermogenesis in iBAT of wild type (WT) and Decr^−/− (KO) mice. Test groups: (i) no fasting, no cold exposure, (ii) fasting (24 hours), no cold exposure, (iii) cold exposure (3 hours) and (iv) fasting (24 hours) including cold exposure (3 hours). (A) Factors involved in lipolysis cascade; ATGL, HSL, p-HSL, and UCP1 as the key factor in thermogenesis and kinases involved in intracellular signalling cascades; PKB, AMPK, p38 MAPK and transcription factor CREB. β-Actin served as a loading control. Relative protein quantities of p-HSL (B), UCP1 (C) and p-AMPK (D). Values are presented as group (n = 5) means ± SE from three technical repeats.

Figure 7

Quantitative RT-PCR analysis of thermogenic gene expression. (A) Expression of Ucp1 mRNA in iBAT of wild type (WT) and Decr^−/− (KO) mice. Treatment groups were (i) no fasting, no cold exposure, (ii) fasting (24 hours), no cold exposure, (iii) cold exposure (3 hours) and (iv) fasting (24 hours) including cold exposure (3 hours). Fold change of mRNA expression of fed WT mice was set as a reference and has a value of 1 on a linear scale. Mouse β-actin and GAPDH were used as endogenous controls to which sample values were normalized. Values are presented as group (n = 5) means ± SE from three individual measurements. Statistically significant differences between WT and KO mice are indicated by asterisks (***p < 0.001). (B) Gene expression of thermogenic genes Atgl, Cidea, Cox2, Dio2, Elovl3, Pgc-1α and Ucp1 in iBAT of fasted (24 h) and cold exposed (3 h) wild type (WT) and Decr^−/− (KO) mice. Fold change of mRNA expression of fasted and cold exposed WT mice was set as a reference and has a value of 1 on a linear scale. Mouse β-actin and GAPDH were used as endogenous controls to which sample values were normalized. Values are expressed as group (n = 5) means ± SE from three individual measurements. Statistically significant differences between WT and KO mice are indicated by asterisks (**p < 0.01, ***p < 0.001).

Figure 8

Fgf21 expression levels in liver and iBAT and FGF21 concentration in circulation. (A) Expression of Fgf21 mRNA in liver after 24 hours of fasting in wild type (WT) and Decr^−/− (KO) mice. Box whisker plot shows the data from four WT and five KO samples. The p-value is 0.0003. (B) Concentration of FGF21 in serum of wild type (WT) and Decr^−/− (KO) mice in fed and fasted stage (n = 5). Statistically significant differences are indicated by asterisk (*p < 0.05). (C) Fgf21 expression in iBAT of WT and KO mice under fed stage and after fasting. (D) Fgf21 expression in iBAT of WT and KO mice under fed stage and after fasting and cold exposure. Fed WT Fgf21expression was set as a reference and has a value of 1 on the linear scale. Statistically significant difference between WT and KO mice are indicated by asterisks (***p < 0.001). Values are expressed as group (n = 5) means ± SE. Mouse β-actin and GAPDH were used as endogenous controls to which sample values were normalized.