BGPT: Paper Review: Glucose feeds the TCA cycle via circulating lactate

Explore by Goal

Quick Explanation Copied

Core claim (skeptical read)

In fasted and fed mice, the authors infer that circulating lactate is a dominant carbon carrier into the TCA cycle, so much that glucose largely feeds TCA indirectly via lactate in most tissues (except brain), based on whole-body ¹³C tracer turnover/labeling and modeling.

Long Explanation

Paper: Glucose feeds the TCA cycle via circulating lactate

DOI: 10.1038/nature24057
Study type: Whole-body in vivo isotope tracing in mice + quantitative modeling + genetically engineered tumor models.

1) Visuals first — what the paper reports (from provided table)

The paper’s Table of circulatory turnover fluxes (fasted mice) includes high lactate turnover relative to glucose.

Figure A — Circulatory turnover fluxes (fasted mice, nmol·g⁻¹·min⁻¹)

2) One-sentence essence (with evidence boundary)

The authors argue that lactate is a primary circulating carbon source feeding tissue TCA cycle labeling, so that glucose’s contribution to TCA is mostly indirect via circulating lactate during fasting across tissues except brain—based on tracer-derived turnover fluxes and TCA labeling attribution, with additional tumor model experiments.

3) What exactly is measured vs inferred

Category	Specific outputs	Confidence boundary
Measured	Serum and tissue ¹³C labeling patterns of TCA intermediates and selected metabolites after infusions; circulatory metabolite enrichment kinetics used to estimate turnover flux.	High, within the tracer + MS measurement pipeline described. (Still, exact MS isotopologue correction details are only partially visible in the provided text.)
Modeled/attributed	Attribution of direct vs indirect carbon contributions to tissue TCA metabolism (e.g., glucose → lactate → TCA) using a quantitative accounting/linear algebra framework and whole-body flux balance consistent with tracer labeling.	Moderate: depends on modeling assumptions (steady-state, compartmentalization, completeness of included flux routes).
Biological mechanism	Proposed rationale: lactate supports whole-body sharing/decoupling; requires LDH activity and monocarboxylate transport and rapid exchange.	Low-to-moderate: consistent with tracer outcomes but not directly mechanistically proven in the provided excerpt.

4) Key results (skeptical interpretation)

4.1 Lactate circulatory turnover is highest

The paper reports lactate’s circulatory turnover flux (F_circ) exceeding glucose in fasting mice.

4.2 Glucose feeds tissue TCA mostly via circulating lactate (fasting), except brain

Using ¹³C-glucose vs ¹³C-lactate infusions, the authors quantify how much of TCA labeling after labeled glucose is explained by labeled circulating lactate. They conclude that in fasted mice, glucose’s contribution to tissue TCA intermediates is primarily indirect via lactate in all tissues except brain.

4.3 Tumors: lactate contribution to TCA exceeds glucose in the studied GEMMs (fasted)

In genetically engineered lung and pancreatic cancer mouse models, the paper reports that circulating lactate’s contribution to tumor TCA intermediates exceeds glucose, with glutamine contributing more than lactate in pancreatic cancer.

5) Critical appraisal: what could make this wrong?

Model identifiability / completeness: The “glucose feeds TCA via circulating lactate” conclusion is attributional. If other unmodeled glucose-derived routes (e.g., direct intracellular cycling, compartment-specific transport, or alternative intermediary labeling paths) contribute comparably, attribution could be biased toward lactate simply because lactate is both abundant and heavily recycled. The paper partially addresses exchange/labeling artifacts (arterial-vs-venous enrichment differences) but attribution still inherits assumptions of the flux framework.
Steady-state and equilibration: Turnover flux inference and labeling attribution rely on steady-state labeling and effective mixing across capillary beds. If steady-state is not achieved uniformly across tissues, fitted labeling kinetics could mis-estimate flux. The paper reports specific infusion durations and fitting to labeling kinetics, which supports steady-state assumptions, but steady-state validity can still be tissue-dependent.
Species/state generalization: Results are in C57BL/6 mice under fasting/fed and specific GEMM tumors. Lactate dynamics in other species and in chronic disease states may differ. The paper acknowledges tissue heterogeneity and that not all tumor regions may be equally perfused.
Brain exception interpretation: The brain is a stated exception. That exception is valuable because it suggests the mechanism is not a universal rule—but it also means the mechanistic story needs a more detailed substrate-transport/accounting explanation for the brain.

6) What information would disprove or sharply revise the conclusion?

Demonstrate (in the same tissue set and comparable metabolic states) that glucose-derived TCA labeling cannot be accounted for by lactate labeling once accounting models include additional plausible intermediates/compartments, i.e., that glucose has substantial direct TCA feeding in multiple tissues during fasting.
Show that manipulating lactate production/transport changes TCA labeling attribution in the opposite direction (e.g., TCA carbon from glucose remains high while lactate attribution collapses), indicating the “carrier” role was incidental rather than causal for carbon delivery.

7) Practical “use this” takeaways for a metabolism modeler

Use the tracer accounting framework: The paper’s logic separates abundance/turnover from contribution to TCA labeling by combining Fcirc estimates with enrichment-based accounting. That’s a template for building and testing metabolic network hypotheses where indirect intermediates dominate.

8) Go deeper on BGPT

Author reviews (BGPT)

Feedback:

Updated: May 01, 2026