Autoregulation of H+/lactate efflux prevents monocarboxylate transport (MCT) inhibitors from reducing glycolytic lactic acid production

Abstract

Background: Pharmacological inhibition of membrane transporters is expected to reduce the flow of solutes, unless flux is restored (i.e., autoregulated) through a compensatory increase in the transmembrane driving force. Drugs acting on monocarboxylate transporters (MCTs) have been developed to disrupt glycolytic metabolism, but autoregulation would render such interventions ineffective. We evaluated whether small-molecule MCT inhibitors reduce cellular H⁺/lactate production.

Methods: Cellular assays measured the relationship between MCT activity (expressed as membrane H⁺/lactate permeability; P_HLac) and lactic acid production (inferred from H⁺ and lactate excretion; J_HLac) in a panel of pancreatic ductal adenocarcinoma (PDAC) cells spanning a range of glycolytic phenotype.

Results: MCT activity did not correlate with lactic acid production, indicating that it is not set by membrane permeability properties. MCT inhibitors did not proportionately reduce J_HLac because of a compensatory increase in the transmembrane [lactate] driving force. J_HLac was largely insensitive to [lactate], therefore its cytoplasmic build-up upon MCT inhibition does not hinder glycolytic production. Extracellular acidity, an MCT inhibitor, reduced J_HLac but this was via cytoplasmic acidification blocking glycolytic enzymes.

Conclusions: We provide mathematically verified evidence that pharmacological and physiological modulators of MCTs cannot proportionately reduce lactic acid production because of the stabilising effect of autoregulation on overall flux.

Fig. 1. In a system that auto-regulates efflux, pharmacological inhibition of a transporter does not necessarily reduce the carried flux.

a By definition, net metabolic production of lactic acid (represented by tap) is equal to H⁺/lactate efflux across the membrane, largely mediated by surface-expressed MCTs (represented by green vent). The level of H⁺/lactate in the cytoplasmic compartment is set by the production flux and permeability. A drug-evoked reduction in membrane permeability will initially reduce efflux. This results in a build-up of lactate/H⁺ in the cell, which increases the driving force for efflux. A new steady-state is attained, at which efflux is restored, despite the lower permeability. For this to take place, the build-up of H⁺/lactate must not block production, the ultimate driver of lactic acid release from cells. b Schematic of the relationship between glycolytic flux and MCT-dependent H⁺/lactate efflux, indicating the variables interrogated by assays. Flux (typically in units of mM per min) described the rate at which a solute is produced (e.g., by glycolysis) or transported (e.g., by MCTs). Permeability is a property of the cell membranes that measures how easily solutes can cross (e.g., aboard MCTs). Pharmacological drugs are often designed to influence flux by interrupting permeability, but this may not hold true if the system shows autoregulation of flux.

Fig. 2. Membrane H+/lactate permeability does not correlate with lactic acid production.

a Measuring MCT-dependent H⁺/lactate permeability (P_HLac) using rapid solution switcher alternately releasing 30 mM lactate-containing and lactate-free microstreams. Exemplar pHi time course (cSNARF1). P_HLac in six PDAC cell lines (n = 2500–3500 cells/N = 4–6 platings). b Lactic acid production rate inferred from the immediate effect on pHi of MCT blockade with α-cyano-4-hydroxycinnamate (CHC; 2 mM) (n = 500/N = 2). c Lactic acid production measured from medium acidification (cSNARF1-dextran) and end-point lactate by biochemical assay (n = 9/N = 3). Yellow traces are cell-free wells; grey lines indicate the initial rate. d Concordance between lactate and H⁺ production in media (Pearson’s R = 0.93; P = 0.008). Medium pH time courses converted to cumulative acid production, also showing endpoint [lactate] measured biochemically. e, f Lack of significant correlation between P_HLac and lactic acid production measured by the two assays (R² is the Pearson correlation coefficient).

Fig. 3. Pharmacological MCT inhibitors do not block lactic acid production.

a Confirmation of the inhibitory effect of SR13800 (SR), AR-C155858 (AR-C) and syrosingopine (Syro) on H⁺/lactate permeability. b Western blot showing expression of MCT1, MCT2 and MCT4. c Membrane permeability measured for H⁺/lactate in the presence and absence of MCT1/MCT2 inhibitor AR-C (1 µM), and compared to H⁺/acetate (n = 500–1000/N = 3). **P < 0.01, multiple comparisons t test. d Medium acidification (cSNARF1-dextran) in presence of AR-C (10 µM), SR (10 µM) or CHC (2 mM). Cumulative acid production and endpoint [lactate] are unaffected by MCT inhibitors (n = 3 repeats each; not significant; one-way ANOVA) The data are plotted as mean ± SD. e Western blot showing response of protein lactylation in media containing galactose instead of glucose (GAL) or media supplemented with lactate (LAC), and densitometric analysis. f Lactylation increases after treatment with 10 µM AR-C (n = 3) and 2 mM CHC (n = 3) and (g) 10 µM Syro (n = 4). *P < 0.05, repeated measures ANOVA). Densitometric analyses were normalised to controls (dashed lines).

Fig. 4. Lactic acid production is strongly influenced by pH but not [lactate].

a Fluorimetric assay for medium acidification (cSNARF1-dextran). Starting medium pH and [lactate] were varied (n = 12/N = 4). Two-way ANOVA: P < 0.001 for effect of pH, P > 0.05 for effect of lactate. b Acid production into media supplemented with weak acids of increasing pK_a (thick line: 5 mM of Ace-acetate, Pro-propionate, But-butyrate, Pho-phosphate, Imi-imidazole), compared to controls (thin line). Error bars not shown for clarity. Endpoint [lactate] measured by biochemical assay (n = 12/N = 4). Significant effect of buffer tested by t test: *P < 0.05; **P < 0.01. c Glycolytic production of lactic acid inferred from the extracellular acidification rate (ECAR; Seahorse) in response to reducing pH from 7.4 with HCl injection, and then restoring pH to 7.4 with NaOH injection (n = 8/N = 2). Exemplar time course shows ECAR normalised to initial rate (i.e., prior to first pH change). Grey lines show pHe-ECAR curve averaged for all six lines.

Fig. 5. MCT-dependent H⁺/lactate permeability is sensitive to extracellular pH.

a Protocol for mapping pHe-P_HLac relationship by altering pH of lactate-free superfusate (shaded). pHe-P_HLac relationship fitted with a non-cooperative Hill curve for six PDAC cell lines; grey lines show mean for all six lines (pK = 7.19, Hill coefficient 1.26). b Protocol for measuring P_HLac in PDAC cells after 48 h treatment with DMOG (1 mM) to induce HIF signalling. pHe-dependence fitted with a non-cooperative Hill curve for six PDAC lines. All cells could be described by the same Hill coefficient and apparent pK_a. c Six-state model fit to pHe-P_HLac data. d Estimates of deprotonation rate constant k_b and equilibrium constant K₂ and the copy-number of MCT proteins at the membrane.

Fig. 6. Low extracellular pH reduces lactic acid production through intracellular acidification.

a Response of pHi to sudden extracellular acidification to pHe 6.6 or 5.7, converted to flux (n = 1000–3000/N = 3). b Resting pHi measured after a >10 min equilibration period at test pH (n = 1000–3000/N = 3). c Incubation in low-chloride medium raises pHi at constant pHe (n = 12/N = 4) and accelerates glycolysis (n = 12/N = 4; effect of low chloride: *P < 0.05, **P < 0.01 by two-way ANOVA). d ECAR data from five PDAC lines is described by a unique function of pHi (pKa = 7.10, Hill coefficient 2.25). e pHi-sensitivity is sufficient to predict the time course of medium acidification and the effect of 5 mM imidazole.