Carnitine octanoyltransferase is important for the assimilation of exogenous acetyl-L-carnitine into acetyl-CoA in mammalian cells

Abstract

In eukaryotes, carnitine is best known for its ability to shuttle esterified fatty acids across mitochondrial membranes for β-oxidation. It also returns to the cytoplasm, in the form of acetyl-L-carnitine (LAC), some of the resulting acetyl groups for posttranslational protein modification and lipid biosynthesis. While dietary LAC supplementation has been clinically investigated, its effects on cellular metabolism are not well understood. To explain how exogenous LAC influences mammalian cell metabolism, we synthesized isotope-labeled forms of LAC and its analogs. In cultures of glucose-limited U87MG glioma cells, exogenous LAC contributed more robustly to intracellular acetyl-CoA pools than did β-hydroxybutyrate, the predominant circulating ketone body in mammals. The fact that most LAC-derived acetyl-CoA is cytosolic is evident from strong labeling of fatty acids in U87MG cells by exogenous ¹³C₂-acetyl-L-carnitine. We found that the addition of d₃-acetyl-L-carnitine increases the supply of acetyl-CoA for cytosolic posttranslational modifications due to its strong kinetic isotope effect on acetyl-CoA carboxylase, the first committed step in fatty acid biosynthesis. Surprisingly, whereas cytosolic carnitine acetyltransferase is believed to catalyze acetyl group transfer from LAC to coenzyme A, CRAT^-/- U87MG cells were unimpaired in their ability to assimilate exogenous LAC into acetyl-CoA. We identified carnitine octanoyltransferase as the key enzyme in this process, implicating a role for peroxisomes in efficient LAC utilization. Our work has opened the door to further biochemical investigations of a new pathway for supplying acetyl-CoA to certain glucose-starved cells.

Keywords: acetyl coenzyme A (acetyl-CoA); acetylcarnitine; cell metabolism; energy metabolism; enzyme turnover; metabolic regulation; mitochondrial metabolism; peroxisome.

Figure 1

Carnitine acetyltransferase reaction.

Figure 2

LAC contributes to cytosolic acetyl-CoA pools in a concentration, time, and glucose-dependent manner in U87MG glioma cells.A, proposed utilization of exogenous acetyl-L-carnitine (LAC) in central metabolism. B, enrichment of M + 2 acetyl-CoA cultured with 1 mM unlabeled LAC and subsequently with 1 mM ¹³C₂ LAC for specific time periods. C, fractional enrichment of m + 2 acetyl-CoA in U87 cells cultured for 6 h with specified concentrations of ¹³C₂ LAC. D, fractional enrichment of m + 2 acetyl-CoA in U87 cells cultured with 1 mM ¹³C₂ LAC at specified glucose concentrations and time points. Statistical significance was calculated by two-way ANOVA. Each point represents a biological replicate and error bars represent standard error of the mean. ∗p ≤ 0.05; ∗∗∗∗p ≤ 0.0001. BHB, β-hydroxybutyrate; CRAT, carnitine acetyltransferase.

Figure 3

Differential partitioning and utilization of exogenous LAC in U87MG glioma cells.A, fractional enrichment of whole-cell m + 2, m + 3 acetyl-CoA in cells cultured without glucose for 6 h in the presence of 1 mM of ¹³C₄ BHB or ¹³C₂ LAC or d₃-LAC, with or without 4 mM Glutamine (Gln). Statistical significance between BHB and ¹³C₂ LAC labeling was calculated by multiple t-tests. B, fractional enrichment of palmitic acid (sum m + 2 m + 4) in U87MG cells cultured with 0.5 mM glucose for 48 h with 1 mM of specified metabolite. Statistical significance was calculated by one-way ANOVA. C, cellular oxygen consumption rates from glycolysis and mitochondrial respiration upon treatment with BHB or LAC measured using Seahorse analyzer under glucose starvation. D–F, fractional enrichment of tricarboxylic acid intermediates in cells cultured with 0.5 mM ¹³C₄ BHB or 1 mM ¹³C₂ LAC in the absence of glucose for 6h. Statistical significance was calculated by unpaired t tests. G, immunoblot comparing expression of FLAG-tagged SLC22A5 in U87 after transient transfection for overexpression. H, fractional enrichment of m + 2 acetyl-CoA in U87 cells overexpressing SLC22A5 cultured without glucose with 1 mM ¹³C₂ LAC for 6 h. Statistical significance was calculated by unpaired t test. Each point represents a biological replicate and error bars represent standard error of the mean. ∗∗∗∗p ≤ 0.0001. BHB, β-hydroxybutyrate; LAC, acetyl-L-carnitine.

Figure 4

Lack of CRAT requirement for LAC conversion to acetyl-CoA.A, immunoblot comparing CRAT expression in U87 cells edited with CRISPR/Cas9 to target CRAT. B, fractional enrichment of m + 2 acetyl-CoA in wildtype and CRAT-edited cells cultured with 1 mM ¹³C₂ LAC, no glucose. Statistical significance was calculated by unpaired t test. C, fractional enrichment of m + 3 acetyl-CoA in wildtype and CRAT-edited cells cultured in standard glucose conditions and 1 mM d₃-LAC. Statistical significance was calculated by unpaired t test. Each point represents a biological replicate and error bars represent standard error of the mean. CRAT, carnitine acetyltransferase; LAC, acetyl-L-carnitine.

Figure 5

CROT contributes to acetyl-CoA generation from acetyl-carnitine.A, cellular distribution of carnitine acyltransferases in facilitating fatty acid metabolism. B, immunoblot comparing expression of CROT/CPT2 in which Cas9 editing was used to target a safe-targeting region or CROT/CPT2. C, fractional enrichment of M + 2 acetyl-CoA in control and CROT/CPT2-knockout cells cultured in the absence of glucose and the addition of 1 mM ¹³C₄ BHB or ¹³C₂ LAC for 6 h. Statistical significance was calculated by two-way ANOVA. D, fractional enrichment of palmitic acid (sum m + 2 m + 4) in control and CRAT/CROT knockout cells cultured with 1 mM ¹³C₂ LAC and specified glucose concentrations for 24 h. Statistical significance was calculated by one-way ANOVA. E, immunoblot comparing CROT expression in wildtype, CROT-null, and CROT cDNA-rescued cells. ΔTHL indicates truncation of the C-terminal peroxisomal targeting sequence. F, fractional enrichment of m + 2 acetyl CoA after cDNA rescue in U87MG cells cultured with ¹³C₂ LAC without glucose for 6 h. Statistical significance was calculated by one-way ANOVA. Each point represents a biological replicate and error bars represent standard error of the mean. ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001; ∗∗∗∗p ≤ 0.0001. CPT, carnitine palmitoyltransferase; CRAT, carnitine acetyltransferase; CROT, carnitine octanoyltransferase; LAC, acetyl-L-carnitine.

Figure 6

Acetyl-carnitine utilization is cell-type specific. A, fractional enrichment of whole-cell m + 2 acetyl-CoA in cells cultured with no glucose for 6 h in the presence of 1 mM of ¹³C₄ BHB or ¹³C₂ LAC. Statistical significance was calculated by multiple t tests. B, palmitic acid (sum m + 2 m + 4) enrichment in C2C12 myoblasts cultured for 48 h in glucose-containing medium with 1 mM of the appropriate metabolite. Statistical significance was calculated by unpaired t test. C, acetyl-CoA m + 2 enrichment in C2C12 myoblasts, overexpressing CROT or wildtype, upon culture with 1 mM ¹³C₂ LAC for 30 min in media without glucose. Statistical significance was calculated by unpaired t test. Each point represents a biological replicate and error bars represent standard error of the mean. ∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗∗p ≤ 0.0001. BHB, β-hydroxybutyrate; CROT, carnitine octanoyltransferase; LAC, acetyl-L-carnitine.