Lactate-carried Mitochondrial Energy Overflow

Abstract

We addressed the question of mitochondrial lactate metabolism using genetically-encoded sensors. The organelle was found to contain a dynamic lactate pool that leads to dose- and time-dependent protein lactylation. In neurons, mitochondrial lactate reported blood lactate levels with high fidelity. The exchange of lactate across the inner mitochondrial membrane was found to be mediated by a high affinity H⁺-coupled transport system involving the mitochondrial pyruvate carrier MPC. Assessment of electron transport chain activity and determination of lactate flux showed that mitochondria are tonic lactate producers, a phenomenon driven by energization and stimulated by hypoxia. We conclude that an overflow mechanism caps the redox level of mitochondria, while saving energy in the form of lactate.

Fig. 1.. A dynamic mitochondrial lactate pool leads to protein lactylation.

(A) HEK293 cells stably expressing mitoCanlonicSF (HEK040) were sequentially imaged at 5 mM glucose/0.05 mM pyruvate/0.5 mM lactate (standard buffer), 6 mM oxamate (nominal zero lactate) and high lactate (10 mM). Bar represents 20 μm. Summary of lactate levels in standard buffer (five experiments, 53 cells). (B)The same protocol was applied to: HepG2 (three experiments, 51 cells), MDA-MB-231 (three experiments, 26 cells), astrocytes (eleven experiments, 45 cells), neurons (eleven experiments, 19 cells) and Cos7 cells (six, 32 cells). For brain cells, the standard buffer contained 2 mM glucose/0.05 mM pyruvate/0.5 mM lactate.(C) Effect of glucose (5 mM) or lactate (10 mM) on HEK293 cells expressing mitoCanlonicSF or cytosolic CanlonicSF, measured simultaneously. Box plot shows data from three experiments, 18 cells (Mann-Whitney, paired t-test). Bar represents 20 μm. NS, not significant.(D) Lactate dose- and time-response of whole-cell HEK293 protein lactylation (Kla antisera). (E) Effect of lactate (25 mM/5 hours) on protein lactylation (Kla antisera) of HEK293 subcellular fractions. Preparation purity was assessed by detection of VDAC (mitochondria) or β-actin (cytosol). (F) Effect of lactate (25 mM/5 hours) on the lactylation level of HEK293 mitochondrial proteins, identified by nHPLC-MS/MS. Inset shows that 13% of precipitated proteins were mitochondrial. NS, not significant.

Fig. 2.. Mitochondrial lactate in neurons covaries with plasma lactate.

(A) Mice expressing mito-CanlonicSF in neurons were i.v. injected with lactate (1.5 mmol/kg bodyweight). (B) Colocalization of mito-CanlonicSF with the neuronal marker NeuN and mitochondrial marker HSP60. Bar represents 20 μm. (C) Blood plasma lactate measurements upon lactate injection (data from three experiments in two animals). (D) Mito-CanlonicSF response upon lactate injection (data from eight experiments in four animals). (E) Paired measurement of lactate in mitochondria and cytosol, with mito-CanlonicSF and Laconic, respectively (normalized to peak, data from three experiments in three animals).

Fig. 3.. Characterization of mitochondrial lactate transport.

Direct access of lactate to mitochondria was obtained by permeabilizing HEK293 cells expressing mito-CanlonicSF or mitoSypHer with digitonin (A). Except for the lactate dose response curve, mitochondria were kept energized with 0.2 mM glutamate plus 0.1mM malate. (B) Dose response of lactate uptake (mM), in cells expressing mito-CanlonicSF. A rectangular hyperbola was fitted to the initial uptake rates giving a K_M= 0.8 ± 0.3 (four experiments, 75 cells). (C) Cells expressing the pH sensor mitoSypHer. The nucleus is indicated (n). Bar represents 20 μm. Effect of 10 mM lactate on matrix pH. Bar graph summarizes data from fifteen experiments (76 cells). (D) Mitochondria perfused with 0.5 mM lactate were exposed to oxamate (6 mM), pyruvate (0.5 mM) and glucose (5 mM). Initial slopes of lactate depletion are shown (two experiments,13 cells, paired t-test). (E) Mitochondria perfused with 0.5 mM lactate were exposed to increasing concentrations of pyruvate (mM; two experiments, 19 cells). (F) Lactate uptake was measured in the absence and presence of UK5099 (0.5 μM). Bar graph summarizes data from similar experiments with a panel of inhibitors. Ctrl (DMSO 0.01%–0.05%; nine experiments, 57 cells); UK5099 (0.5 μM; four experiments, 23 cells); AR-C155858 (1 μM; five experiments, 38 cells); AZD3965 (10 μM; five experiments, 36 cells); pCMBS (50 μM; four experiments, 25 cells); Inhibitor cocktail contains UK5099, AR-C155858, pCMBS and 5 μM syrosingopine (three experiments, 18 cells). Paired t-test. (G) Cells genetically depleted of MPC2 (sgMPC2KO7) were sequentially exposed to lactate (10 mM) or pyruvate (10 mM). Bar graph shown the summary of three KO experiments (27 cells) and three control experiments (40 cells). Studentś t-test. The right panel shows the cell-to-cell correlation between the initial rate of lactate production from pyruvate, and the rate of lactate uptake (●, ctrl; ○, sgMPC2KO7). Data obtained with a second RNA guide are also shown (Δ, sgMPC2KO9, three experiments, 22 cells).

Fig. 4.. Weak mitochondrial energization by lactate.

HEK293 cells were permeabilized with digitonin. (A) Left panel, image of FAD-FMN autofluorescence. A nucleus is indicated (n). Right panel, DIC image. Bar represents 20 μm. The graph shows the effect of pyruvate (0.1 mM) and lactate (0.1 mM) on FAD-FMN autofluorescence in the absence or presence of glutamate (0.2mM) and malate (0.1 mM). Summary of five experiments (28 cells; paired t-test). (B) Proton pumping by respiratory complexes is fueled by NADH and FADH_2. (C) The effect of pyruvate (0.1 mM), glutamate (0.2 mM) and malate (0.1 mM), or buffer adjusted at pH 7.8 was measured with mitoSypHer. Calibrated signals obtained at pH 7.2 and 7.8 are indicated. Bar graph shows the changes after 90 seconds of substrate exposure (nine experiments, 47 cells). (D) Mitochondria were exposed to glutamate and malate in the absence or presence of ETC inhibitors: rotenone (1 μM) and antimycin (1 μM). Bar graph summarizes data for pyruvate (three experiment, 9 cells) and glutamate plus malate (five experiments, 20 cells), paired t-test. (E) Pyruvate-energized mitochondria were exposed to ADP (2 mM) and Pi (2 mM) (three experiments, 21 cells). (F) After a control pulse of glutamate and malate, mitochondria were exposed to increasing concentrations of lactate (mM). Bar graph shows the change in pH signal relative to baseline. Data from six experiments (28 cells), paired-t test. (G) Glutamate plus malate -energized mitochondria were exposed to lactate (mM). Bar graph shows the change in pH signal elicited by 0.1 mM lactate relative to baseline (five experiments, 28 cells, paired t-test). NS, not significant; *; p < 0.05. (H) Neurons expressing the pH sensor mitoSypHer. The nucleus is indicated (n). Bar represents 20 μm. After a control pulse of pyruvate, neurons were exposed to increasing concentrations of lactate (mM). (I)Bar graph summarized data from five experiments (21 cells, paired t-test).NS, not significant; *; p < 0.05. Bar graph shows the change in pH signal elicited by 0.1 mM lactate relative to baseline (five experiments, 28 cells, paired t-test). NS, not significant; *; p < 0.05.

Fig. 5.. Lactate-carried mitochondrial overflow.

(A) Net mitochondrial lactate flux in the presence of 0.5 mM lactate was determined in permeabilized cells expressing mito-CanlonicSF using an inhibitor-stop protocol (1 μM UK5099 and/or 1 μM AR-C155858). (B) Effect of lactate transport inhibitors on the steady-state of mitochondrial lactate in HEK293 cells in the absence (left panel) or presence (right panel) of 0.1 mM pyruvate. Perfusate contained 0.2 mM glutamate and 0.1 mM malate. Summary of data from five experiments without pyruvate (22 cells) and seven experiments with pyruvate (27 cells). Wilcoxon Signed Rank test; *, p < 0.05. (C) Effect of inhibitors on the steady-state of HEK293 mitochondrial lactate in the absence of glutamate/malate (low redox, left panel) or presence of glutamate/malate (high redox, right panel). Perfusate contained 10 mM pyruvate. Summary of data from three experiments at low redox (23 cells) and four experiments at high redox (24 cells). Paired t-test; *, p < 0.05. (D) Effect of transport inhibitors on permeabiized neurons expressing mito-CanlonicSF at high redox (glutamate/malate) plus 10 mM pyruvate. Bar represents 20 μm. Bar graph summarizes nine experiments (65 cells; paired t-test; *, p < 0.05). (E) Effect of hypoxia (N₂-gassing) on the accumulation of lactate induced by transport inhibitors. Bar graph summarizes data from four experiments (34 cells, paired t-test; *; p < 0.05). (F) NADH fuels the electron transport chain (ETC) and also drives lactate production at mitochondrial LDH (mLDH). Lactate is then discarded via MPC and MLC. Protein lactylation (Kla) is determined by lactate level.