Quantitative Analysis of the Whole-Body Metabolic Fate of Branched-Chain Amino Acids

Abstract

Elevations in branched-chain amino acids (BCAAs) associate with numerous systemic diseases, including cancer, diabetes, and heart failure. However, an integrated understanding of whole-body BCAA metabolism remains lacking. Here, we employ in vivo isotopic tracing to systemically quantify BCAA oxidation in healthy and insulin-resistant mice. We find that most tissues rapidly oxidize BCAAs into the tricarboxylic acid (TCA) cycle, with the greatest quantity occurring in muscle, brown fat, liver, kidneys, and heart. Notably, pancreas supplies 20% of its TCA carbons from BCAAs. Genetic and pharmacologic suppression of branched-chain alpha-ketoacid dehydrogenase kinase, a clinically targeted regulatory kinase, induces BCAA oxidation primarily in skeletal muscle of healthy mice. While insulin acutely increases BCAA oxidation in cardiac and skeletal muscle, chronically insulin-resistant mice show blunted BCAA oxidation in adipose tissues and liver, shifting BCAA oxidation toward muscle. Together, this work provides a quantitative framework for understanding systemic BCAA oxidation in health and insulin resistance.

Keywords: branched chain amino acids; insulin resistance; obesity; stable isotope tracing.

Figure 1.. In vivo isotopic tracing reveals rapid oxidation of BCAAs in multiple organs.

(A) Schematic of BCAA oxidation pathway. BCAT transaminates BCAAs to BCKAs, which are subsequently oxidized by BCKDH. By this reaction, BCKAs lose one carbon as CO₂ and are conjugated to Co-enzyme A (CoA). All downstream intermediates are trapped in the mitochondria by CoA and oxidized into the TCA cycle, except for 3-HIB, which can be released to the circulation. (B) Experimental scheme. Mice received an intravenous bolus injection of BCAAs with one labeled with ¹³C and the other two unlabeled. Blood was collected every 1 min, and tissues were harvested at 3 or 5 min. (C) Plasma levels of the indicated metabolites, normalized to time 0. (D) Valine oxidation is fast. Data show labeling (%) of the indicated metabolites in plasma. (E-H) Valine is oxidized by multiple organs. Data show labeling (%) of the indicated metabolites in plasma (shaded) and in the indicated tissues. N = 3-6. Data are means and error bars are ± SE. BAT and WAT, brown and white adipose tissue, respectively; Quad, quadriceps. See also Figure S1.

Figure 2.. Integrated quantitative analysis of whole-body BCAA disposal.

(A) Experimental scheme. Mice received a constant intravenous infusion of BCAAs (one labeled with ¹³C and the other two unlabeled). (B) Constant infusion achieves isotopic steady state. Data show labeling (%) of the indicated labeled BCAAs in plasma. (C) Perfect correlation between the content of each BCAA in protein and the rate of appearance (R_a) of each BCAA. (D) % of labeled carbons in malate in the indicated tissues, normalized to the % of label in plasma BCAAs. Labeled carbons refer to the sum of all labeled forms, with each form weighted by fraction of carbon atoms labeled. (E and F) Schematic model (E) and data (F) of tissue BCAA oxidation flux, calculated from the TCA labeling fraction from BCAAs and TCA turnover flux inferred from tissue oxygen consumption rate (VO₂). See Methods for details. (G) Lack of correlation between isoleucine oxidation flux and BCKDHA mRNA levels in each tissue. (H and I) Fractional synthesis rate (FSR) of proteins from BCAAs and calculated disposal flux into proteins in each tissue (I). (J) Lack of correlation between isoleucine oxidation and disposal to proteins in each tissue. (K and L) Schematic of tissue distribution of isoleucine oxidation (K) and disposal into protein (L) in healthy mice. N = 5, 8, and 5 for ¹³C-valine, ¹³C-leucine and ¹³C-isoleucine infusions, respectively. Data are means and error bars are ± SE. See also Figure S2.

Figure 3.. Transcriptional regulation of BCAA oxidation.

(A) Transgenic expression of PGC-1α in skeletal muscle induces BCAA catabolic genes. Heat map shows gene expression levels of BCAA oxidation enzymes in skeletal muscle in wild-type (WT) and PGC-1α muscle-specific transgenic mice (MCKα). N = 5. (B) Experimental scheme. Mice received an intravenous bolus injection of [U-¹³C]-valine with unlabeled leucine and isoleucine. Blood was collected at 1, 3, 5 min, and tissues were harvested at 5 min. (C) Plasma labeling (%) of the indicated metabolites. N = 3 and 4 for WT and MCKα, respectively. (D) BCAA oxidation is increased in skeletal muscle of MCKα mice. Data show labeled carbons in succinate in the indicated tissues normalized to WT. Data are means and error bars are ± SE. *p<0.05 by Student’s t-test. See also Figure S3.

Figure 4.. Whole body BCKDK deletion increases systemic BCAA oxidation and establishes new metabolic steady states.

(A) Schematic of BCKDK-mediated regulation of BCAA oxidation. (B-D) Plasma levels of the indicated metabolites after oral gavage with [U-¹³C]-isoleucine and unlabeled leucine and valine in control (WT, black) and BCKDK KO (pink) mice. TIC, total ion counts. Bar graph in D shows changes in area under curve (ΔAUC) from the base line (time 0). (E) BCKDK deletion increases systemic BCAA oxidation. Data show total labeled carbons in plasma succinate. (F and G) BCKDK deletion establishes new metabolic steady states. Data show plasma ratios of αKIV to valine (F) or 3-HIB to αKIV (G). (H) Metabolomics heatmap in plasma from WT and BCKDK KO mice. N = 6. Data are means and error bars are ± SE. *p<0.05 and ***p<0.001 by Student’s t-test. For the heatmap, FDR correction was performed with q<0.05. See also Figure S4.

Figure 5.. Acute pharmacological inhibition of BCKDK activates BCAA oxidation specifically in muscle.

(A) BT2 activates BCAA oxidation by inhibiting BCKDK activity. (B) Experimental scheme. Mice received a constant intravenous infusion of [U-¹³C]-isoleucine with unlabeled leucine and valine for 210 min. At 120min, mice received a bolus injection of vehicle (black) or BT2 (green) during the infusion. (C and D) BT2 acutely reduces plasma levels of isoleucine (C) and BCKAs (D). (E) Plasma labeling (%) of isoleucine. Inset shows calculated isoleucine appearance rate (R_a) (F) Western blot analyses of pBCKDHA and total BCKDHA (bottom) and their quantified ratios (top) in the indicated tissues in response to BT2. Soleus samples were not available for these analyses. (G) BT2 specifically activates isoleucine oxidation in skeletal muscle. Data show labeled carbons in the tissue TCA intermediates normalized to the plasma labeled isoleucine, expressed as the percent difference between vehicle and BT2-treated mice. N = 5 and 6 for saline and BT2 groups, respectively. Data are means and error bars are ± SE. *p<0.05 by Student’s t-test. See also Figure S5.

Figure 6.. Insulin acutely induces BCAA oxidation selectively in striated muscle.

(A) Experimental scheme. After hyperinsulinemic-euglycemic clamp for 90 min, mice received a primed-constant infusion of [U-¹³C]-isoleucine with unlabeled leucine and valine for additional 120 min. Blood was collected at 0, 90, 93, 130, 170, and 210 min, and tissues were harvested at 210 min. (B-D) Vehicle (black) or insulin (purple) was constantly infused to achieve hyperinsulinemic state (B). Glucose was also infused (C) to achieve euglycemic state (D). (E-G) Insulin reduces plasma BCAA levels (E) and slightly increases endogenous isoleucine release from proteolysis (F and G). (H) Insulin stimulates isoleucine oxidation specifically in striated muscle. Data show labeled carbons in the tissue malate normalized to the plasma labeled isoleucine, expressed as the percent difference between vehicle and insulin-treated mice. N = 9 and 10, respectively. Data are means and error bars are ± SE. *p<0.05 by Student’s t-test. See also Figure S6.

Figure 7.. Obese and diabetic mice redistribute whole-body BCAA oxidation.

(A) db/db mice have doubled body weight compared to the littermate control (WT). (B) Plasma labeling (%) of valine, after constant infusion of [U-¹³C]-valine with unlabeled leucine and isoleucine at the rate adjusted to total blood volume. (C) db/db mice show reduced capacity of valine disposal. Data show valine disposal rate (R_d) per g body weight (left) or total (right). (D) db/db mice show altered valine oxidation preference in multiple tissues. Data show labeled carbons in the tissue TCA intermediates normalized to the plasma labeled valine, expressed as the percent difference between control and db/db mice. N = 7 and 8 for control and db/db mice, respectively. (E) HFD increases body weight. (F) Plasma labeling (%) of isoleucine, after constant infusion of [U-¹³C]-isoleucine with unlabeled leucine and valine. (G) Mice fed HFD show normal capacity of valine disposal. Data show valine disposal rate (R_d) per g body weight (left) or total (right). (H) Mice fed HFD show reduced valine oxidation preference in adipose tissues. Data show labeled carbons in the tissue TCA intermediates normalized to the plasma labeled isoleucine, expressed as the percent differences between chow and HFD-fed mice. N = 5 and 6 for Chow and HFD groups, respectively. Data are means and error bars are ± SE. *p<0.05 by Student’s t-test. (I) Proposed model of casual relationship between altered tissue BCAA oxidation and elevated circulating BCAA levels in insulin resistance. In healthy conditions (left), BCAA oxidation is balanced among different organs. Genetic and environmental factors suppress BCAA oxidation in the liver and adipose tissues (right), causing increased circulating BCAA levels and overflow of BCAAs into skeletal muscle, resulting in lipotoxicity and insulin resistance. See text for details. See also Figure S7.