henhaobio
Sign in

ScienceBiochemistryPeer-reviewed journal article

Mitochondrial control of fuel switching via carnitine biosynthesis

Christopher AugeriD, Hiroshi NishidaiD, Bo YuaniD, Guilherme Martins SilvaiD, Masanori FujimotoiD, Mark LiiD, Daisuke KatohiD, Dandan WangiD, Melia Granath-PaneloiD, Jihoon ShiniD, Rose WitteiD, Jin-Seon YookiD, Anthony R.P. VerkerkeiD, Alexander S. BanksiD, Sheng Tony HuiiD, Lijun SuniD, Shingo KajimuraiD

PDFDOIPreprintCodeData
In-depth full-text digest available · Jump to digest ↓

Abstract

Environmental adaptation often involves a shift in energy utilization toward mitochondrial fatty acid oxidation, which requires carnitine. Besides dietary sources of animal origin, carnitine biosynthesis from trimethyllysine (TML) is essential, particularly for those who consume plant-based diets; however, its molecular regulation and physiological role remain elusive. Here, we identify SLC25A45 as a mitochondrial TML carrier that controls carnitine biosynthesis and fuel switching. SLC25A45 deficiency decreased the carnitine pool and impaired mitochondrial fatty acid oxidation, shifting reliance to carbohydrate metabolism. Slc25a45-deficient mice were cold-intolerant and resistant to lipid mobilization by GLP1 receptor agonist (GLP-1RA), rendering them resistant to adipose tissue loss. Our study suggests that mitochondria serve as a regulatory checkpoint in fuel switching, with implications for metabolic adaptation and the efficacy of GLP-1RA-based anti-obesity therapy.

In-depth summary

Claude Opus 5.5 · xhigh effort · 5 Nov 2026 · from the full text of version 2 (20 Dec 2025) · CC BY 4.0

SLC25A45 is the inner mitochondrial membrane transporter for trimethyllysine required for de novo carnitine biosynthesis, enabling metabolic fuel switching to fatty acid oxidation during cold exposure and semaglutide treatment.

Mitochondrial long-chain fatty acid oxidation requires carnitine, which is synthesized de novo from trimethyllysine (TML) when dietary animal-derived carnitine is limited, but the carrier importing TML across the inner mitochondrial membrane and its physiological role remained poorly understood. Identifying SLC25A45 as the mitochondrial TML carrier reveals how endogenous carnitine synthesis governs systemic fuel choice, cold thermogenesis, and GLP-1 receptor agonist-induced fat loss.

Key results

  1. ShownSLC25A45 selectively transports trimethyllysine into mitochondria to support de novo L-carnitine biosynthesis.

    Motivated by a human GWAS correlation between SLC25A45 and serum TML (P = 1 x 1.9^-80, n = 6136), Slc25a45 knockout (KO) mouse mitochondria showed lower D9-TML uptake (p = 0.0006) and reduced HTML, gamma-butyrobetaine, and carnitine synthesis, while MBP-SLC25A45 reconstituted in E. coli transported D9-TML (P = 0.00013) but not gamma-butyrobetaine or choline. · INTRODUCTION; SLC25A45 is required for de novo carnitine…; Reconstitution of TML transport by SLC25A45; Figure legends

  2. ShownDocking and molecular dynamics simulations define a stable, specific TML binding pocket inside the SLC25A45 transmembrane cavity.

    Across three 300 ns simulations of AlphaFold structure AF-Q8N413-F1-v4, the SLC25A45-TML complex showed an average MMGBSA binding free energy of -17.10 +/- 3.75 kcal/mol, reduced backbone RMSD compared to apo protein, hydrogen bonding to R268 in ~90% of simulations, and destabilization upon D174E mutation. · Reconstitution of TML transport by SLC25A45; Structural Prediction of SLC25A45; Molecular dynamics studies; Figure legends

  3. ShownWhole-body Slc25a45 deletion causes systemic carnitine deficiency, renal stress, impaired long-chain fatty acid oxidation, and a shift toward carbohydrate catabolism.

    Slc25a45 KO mice had lower serum, muscle (p = 0.01), and heart carnitine, elevated serum creatinine, Cystatin C, blood urea nitrogen, and renal malondialdehyde, higher dark-phase RER, lower 12 h fasting glucose (p = 0.005), lower State 3 palmitoyl-CoA heart respiration (p = 0.018), and 1675 altered liver transcripts (1078 up, 597 down). · SLC25A45 loss results in systemic carnitine…; Figure legends

  4. ShownSLC25A45-mediated carnitine synthesis is induced by cold and required for brown adipose tissue thermogenesis and acute cold tolerance.

    When shifted from 30 C to 6 C, whole-body Slc25a45 KO mice dropped below 30 C rectal temperature within 5 hours with lower brown adipose tissue State 4 mitochondrial respiration (p < 0.05) and lower serum C12-C18 acylcarnitines; 5 days of 100 mg/kg L-carnitine supplementation rescued hypothermia. · Carnitine biosynthesis via SLC25A45 is required…; Mouse physiology; Figure legends

  5. ShownSlc25a45 deficiency blunts lipid mobilization, appetite suppression maintenance, and adipose loss induced by the GLP-1 receptor agonist semaglutide.

    During 5-day semaglutide treatment (30 nmol/kg), Slc25a45 KO mice maintained higher RER, failed to elevate serum long-chain acylcarnitines or reduce triglycerides/diglycerides, resumed higher nighttime food intake in Phase II, and resisted fat mass loss despite intact glucose lowering; L-carnitine supplementation restored adipose loss. · Carnitine biosynthesis is required for the…; Figure legends

Methods, data and reproducibility

Generated whole-body (CMV-Cre), liver-specific (Alb-Cre), and kidney-specific (Cdh16-Cre) Slc25a45 KO mice, CRISPR KO/rescue HEK293 cells, and MBP-hSLC25A45 E. coli C43(DE3). Combined D9-TML LC-MS flux tracing, plasma lipidomics, AlphaFold docking and 300 ns Desmond MD simulations, liver RNA-seq, Oroboros O2k respirometry, and Promethion calorimetry under fasting, 6 C cold challenge, and 30 nmol/kg semaglutide.

  • Mouse liver bulk RNA-seq and in vivo calorimetric, physiological, and serum/tissue metabolomic/lipidomic cohorts (whole-body, liver, and kidney Slc25a45 KO vs control) · mouse · n = n = 3 to 16 mice per group · PRJNA1345061
  • LC-MS metabolomics of D9-TML and D9-gamma-BB tracing in isolated mitochondria, mouse hepatocytes, inguinal WAT adipocytes, HEK293T cells, and E. coli C43(DE3) · mouse, human, E. coli · n = n = 3 to 6 biological replicates per group · ST003920
  • Published mouse kidney single-nuclei RNA-seq dataset reanalyzed for Slc25a45 and carnitine synthesis enzymes · mouse · GSE164274
  • AlphaFold predicted 3D protein structures of human and mouse SLC25A45 and human SLC25A48 · human, mouse · n = 3 structures · AF-Q8N413-F1-v4
Methods & tools
AlphaFold 2.0, Schrodinger Maestro (Suite 2024-4), GlideSP / GlideXP, AutodockVina, JAMDA, Prime MMGBSA, Desmond, El-Maven, LipidSearch (v5.0), STAR (v2.5.2b), edgeR, Seurat (v4.0.1)
Software & environment
R (v4.0.4), Seurat (v4.0.1), Trimmomatic (v0.36), STAR (v2.5.2b), Subread featureCounts (v1.5.2), edgeR, pheatmap, Schrodinger Maestro (Suite 2024-4), Pymol (3.0.2), ChimeraX (1.9), LipidSearch (v5.0), GraphPad Prism 8.
Assay conditions & compute
Molecular dynamics simulations (three 300 ns replicates per condition) were run on the O2 High Performance Compute Cluster at Harvard Medical School.
Reproducibility gaps
Custom R and Python scripts for bulk RNA-seq differential expression, PCA, and pathway enrichment are not linked in a public repository.
Limitations (4)
  • Slc25a45 knockout cells and mice still synthesize low levels of carnitine via alternative pathways, potentially mediated by upregulated Slc25a29 or conversion of downstream gamma-butyrobetaine. · stated by the authors
  • Neither liver-specific nor kidney-specific Slc25a45 deletion caused systemic carnitine deficiency, kidney injury, or cold intolerance because circulating carnitine is supplied across multiple organs. · stated by the authors
  • Structural analysis of TML binding to SLC25A45 relied on AlphaFold models, docking, and molecular dynamics simulations rather than an experimentally resolved 3D structure. · stated by the authors
  • Experiments were conducted in mice fed a standard chow diet with undetectable carnitine, whereas human omnivores obtain approximately 75% of their carnitine pool from diet. · stated by the authors

Read this if you study mitochondrial solute carriers, carnitine and fatty acid metabolism, thermogenesis, or metabolic determinants of GLP-1 receptor agonist efficacy.

Citations
12
FWCI
54.85
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISYook '22Oikawa '26Karavaeva '26Kingren '26Verkerke '24Prola '19
73% semantic matchPhysiology
J.-S. Yook, Z. H. Taxin, B. Yuan+6 more·bioRxiv·Dec 6, 2022