Brown Fat AKT2 Is a Cold-Induced Kinase that Stimulates ChREBP-Mediated De Novo Lipogenesis to Optimize Fuel Storage and Thermogenesis

Abstract

Brown adipose tissue (BAT) is a therapeutic target for metabolic diseases; thus, understanding its metabolic circuitry is clinically important. Many studies of BAT compare rodents mildly cold to those severely cold. Here, we compared BAT remodeling between thermoneutral and mild-cold-adapted mice, conditions more relevant to humans. Although BAT is renowned for catabolic β-oxidative capacity, we find paradoxically that the anabolic de novo lipogenesis (DNL) genes encoding ACLY, ACSS2, ACC, and FASN were among the most upregulated by mild cold and that, in humans, DNL correlates with Ucp1 expression. The regulation and function of adipocyte DNL and its association with thermogenesis are not understood. We provide evidence suggesting that AKT2 drives DNL in adipocytes by stimulating ChREBPβ transcriptional activity and that cold induces the AKT2-ChREBP pathway in BAT to optimize fuel storage and thermogenesis. These data provide insight into adipocyte DNL regulation and function and illustrate the metabolic flexibility of thermogenesis.

Keywords: Akt; SREBP; UCP1; insulin signalling; lipid metabolism; lipid synthesis; obesity; thermogenesis; white fat.

Figure 1. Profound BAT remodeling is induced by mild cold and strongly associates with increased anabolic de novo lipid synthesis

(A) Environmental temperatures and rectal, BAT and tail temperatures recorded at in random fed C57BL/6J mice (n=6). (B) Total body and fat depot weights of C57BL/6J mice adapted to the indicated temperatures (n=20). (C) Representative H&E images of fat depots from C57BL/6J mice adapted to the indicated temperatures. (D) qRT-PCR analysis of ucp1 mRNA levels from the fat depots of C57BL/6J mice adapted to the indicated temperatures (n=8). The y-axis for each is normalized to TN iBAT. (E) RNA-Seq based clustering of iBAT samples from C57BL/6J mice adapted to the indicated temperatures (n=4). (F) Pathway analysis of the top 100 most upregulated genes in RT iBAT versus TN (n=4). AdjP: Adjusted p-value. Category ratio of enrichment (R, Observed number of genes in the category versus expected) is shown in brackets. (G) The de novo lipogenesis pathway. Gene expression changes are indicated in the boxes to the right of each gene name. (H) qRT-PCR analysis on iBAT samples from C57BL/6J mice adapted to TN, RT and SC (n=8). (I) Western blots on iBAT samples from C57BL/6J mice. (J) D2O labeling showing the fraction of newly synthesized FA species over 3 days in the iBAT of C57BL/6J mice at each temperature (n=6). See also Figure S1.

Figure 2. Environmental temperature remodels the BAT lipid landscape

(A–C) Abundance and distribution of TG (triacylglyceride), DG (diacylglyceride) and PE (Phosphatidylethanolamines) species in iBAT of C57BL/6J mice at TN, RT and SC housing conditions (n=4). Each data point represents one species. (D–F) Heatmaps of the TG, DG and PE species in iBAT of C57BL/6J mice at TN, RT and SC housing conditions sorted by the fold change between TN and RT conditions. For each heatmap, the associated graphs represent relative abundance, total carbons and total desaturations in each species ordered according to the heatmap. Colors represent the fold-change compared to the average of RT levels, set to 1. See scale in each heat map. See also Figure S2.

Figure 3. Mild cold induces AKT2 in BAT to promote lipid storage and lipid remodeling

(A) BAT signaling pathways upregulated at RT compared to TN in C57BL/6J mice. AdjP: Adjusted p-value. Category ratio of enrichment (R, Observed number of genes in the category versus expected) is shown in brackets. (B) qRT-PCR analysis (n=8). (C) Corresponding Western blots. (D) Body mass and individual organ weights from Akt2^Ucp1CreER and Akt2 floxed littermate controls at 9 weeks old and 3 weeks post tamoxifen (n=11–13). (E) Corresponding Western blots. (F) Total body and fat depot mass (n=8 at TN n=7 at RT). (G) Glucose and insulin tolerance tests (n=5). (H) Representative H&E images of the indicated fat depots. (I) (left) RNA-Seq-based principal component (PC) analysis on sampled from iBAT (n=4). (right) Correlation scatter plot for each condition indicating the number of significantly affected genes between Akt2^Ucp1CreER and littermates. CPM: counts per million. (J) qRT-PCR analysis of iBAT samples (n=8 at TN n=7 at RT). (K) Corresponding Western blots. (L) Heatmaps of the TG species in the iBAT of Akt2^Ucp1CreER and littermate controls based on fold change between TN and RT conditions in controls (n=5). Associated graphs indicate relative abundance, total carbons and total desaturations in each species order according to the heatmap. Colors represent fold-change compared to the average of Control-RT levels, set to 1. See scale in each heat map. See also Figure S3.

Figure 4. AKT2 stimulates BAT DNL through ChREBP

(A) Gene expression levels of indicated genes in random fed C57BL/6J mice living at RT (n=4). iB, interscapular BAT, sB, subscapular BAT; cB, cervical BAT, asW, anterior subcutaneous WAT, psW, posterior subcutaneous WAT, rW, retroperitoneal WAT; pgW, pergonadal WAT, mW, mesenteric WAT; Tri, triceps; Quad, quadriceps; Gas; gastrocnemious; H, heart; Liv, liver; S Int, small intestine; Pan, pancreas; Spl, spleen; Thy, thymus; Kid, kidneys; Br, brain. ND: no detected. (B) qRT-PCR of iBAT samples from C57BL/6J mice adapted to TN, RT and SC (n=8). (C) qRT-PCR of iBAT samples from Akt2^Ucp1CreER and Akt2 floxed littermate controls (n=8 at TN n=7 at RT). (D) Western blot showing an insulin dose-response (1–100nM) in differentiated AKTiKO cells. (E) Western blot analysis of differentiated AKTiKO cells overexpressing plasmids containing ChREBP isoforms (pMSCV-Chα and pMSCV-Chβ) or the nuclear fragment of SREBP1c (pBABE-nBP1c). (F) qRT-PCR analysis of iBAT from Akt2^Ucp1CreER and controls with or without overexpressing human GLUT4 in adipose tissues (RT, n=7). (G) Western blot analysis of differentiated AKTiKO cells overexpressing recombinant GLUT1. (H) Glucose uptake in differentiated AKTiKO cells overexpressing GLUT1. Each genotype is shown relative to cells expressing empty vector (n=3). (I) qRT-PCR analysis using differentiated AKTiKO cells overexpressing GLUT1 (n=3). See also Figure S4.

Figure 5. BAT AKT2 is not required to maintain euthermia, but its loss enhances WAT browning

(A) qRT-PCR analysis of iBAT from Akt2^Ucp1CreER and floxed littermate controls (n=8 at TN n=7 at RT). (B) Rectal temperature in an acute cold challenge (4°C) starting from RT (n=7). (C) qRT-PCR analysis of iBAT after acute cold challenge (n=7). (D) Total body mass and fat depot weights after two weeks at 5°C (n=6–7). (E) Daily food intake per mouse during a two week adaptation to 5°C (n=2 cages). (F) Rectal temperature after two weeks at 5°C (n=6–7). (G) Thermal camera imaging after two weeks at 5°C (H) Representative H&E images after two weeks at 5°C (I) Western blot of iBAT, psWAT and pgWAT lysates from Akt2^Ucp1CreER and Akt2 floxed littermates after two weeks at 5°C. (J) Akt2 qRT-PCR analysis of fat tissues from Akt2^Ucp1CreER and Akt2 floxed littermates after two weeks at 5°C (n=6). (K) Corresponding ucp1 qRT-PCR analysis (n=6). See also Figure S5.

Figure 6. Short term high fat diet inhibits BAT DNL and the phosphorylation of AKT2, which also drives ChREBP-mediated DNL in WAT

(A) qRT-PCR analysis of iBAT from Akt2^Ucp1CreER and Akt2 floxed littermate controls at RT after two weeks of normal chow (NCD) or HFD feeding (n=3). (B) Corresponding Western blot analysis. (C) qRT-PCR analysis of posterior subcutaneous WAT from 8-week old Akt2^AdipoqCre and Akt2 floxed littermate control mice living at RT (n=8). (D) Corresponding Western blot for iBAT and pgWAT. See also Figure S6 and S7.

Figure 7. Chrebpβ and Fasn expression positively correlate with Ucp1 in human brown adipocytes

(A, C, E, G) Gene expression in human BAT with low (UCP1^low) or high (UCP1^high) levels of activation (n=9–10). (BAT, D, F, H) Correlation plots between indicated genes and ucp1 levels in human BAT. AU: arbitrary units.