Figure 1.
Endothelial cells import exogenous lactate.
(A and B) HUVECs were incubated during 6-h with 10 mM of 13C-methy-labeled sodium lactate (left panels) or 10 mM of 12C sodium lactate (right panels). 13C-lactate was detected using nuclear magnetic resonance (NMR) in cell lysates. (A) In 13C NMR spectra, a 13C-Lactate peak is detected at 19 ppm. (B) In 1H NMR spectra, the methyl doublets at 1.56 and 1.30 ppm arise from 13C-lactate whereas the doublet at 1.43 ppm arises from a coincidental overlap of endogenous 12C-Lactate and threonine. (C) HUVECs were incubated with 10 mM sodium L-lactate. L-Lactate was detected using an enzymatic assay in the lysate of cells collected at different time points. *p<0.05; n = 4. Error bars reflect mean ± SEM.
Figure 2.
Lactate activates HIF-1 in normoxic endothelial cells.
HUVECs were incubated during 24-h with 10 mM lactate or not. (A) The transcriptional activity of HIF-1 was determined using a dual luciferase reporter assay in HUVECs. *p = 0.0207; n = 4. (B) VEGF protein level was measured using ELISA in the supernatant of cells. ns, non-significant; n = 6. (C) bFGF protein level was measured using ELISA in the supernatant of cells. ***p<0.0001; n = 5. (D) VEGFR2 protein expression was detected by western blot in cell lysates. **p = 0.0084; n = 4–5. (A–D) Error bars reflect mean ± SEM.
Figure 3.
Lactate induces an increase in HIF-1α protein expression in normoxic endothelial cells.
HIF-1α protein expression was detected by western blot in normoxic ECs. (A) HUVECs were incubated during 24-h with 10 mM lactate or not. **p = 0.0014; n = 8. (B) as in (A) but with BAECs. ***p<0.0001; n = 16–17. (C) BAECs were incubated during 24-h with the indicated concentrations of lactate. **p<0.01; n = 4. (D) BAECs were incubated during the indicated times with 10 mM of lactate. *p<0.05; n = 5–12. (A–D) Error bars reflect mean ± SEM.
Figure 4.
Lactate triggers HIF-1 induction in normoxic endothelial cells.
Lactate was used at 10 mM for 24-h. (A) HIF-1α mRNA was detected using qRT-PCR in HUVECs incubated with lactate. ns, non-significant; n = 4. (B and C). HIF-1α protein expression was detected by western blot in BAECs incubated with lactate and/or (B) 5 µg/ml Actinomycin D (**p<0.01, ***p<0.005, ns, non-significant compared to control, n = 3–4); (C) increasing concentrations of 2-oxoglutarate. *p<0.05, **p<0.01; n = 5–6. (D) ODD-driven luciferase activity was assessed in HUVECs incubated with 150 µM CoCl2 or lactate. **p<0.01; n = 5. (E) Intracellular pyruvate was detected using an enzymatic assay in HUVECs incubated with lactate. **p<0.01; n = 4. (F–H) HIF-1α protein expression was detected by western blot. (F) BAECs were incubated with lactate, oxamate (10 mM), and/or acetate (10 mM). **p<0.01, ***p<0.005, ns, non-significant compared to control cells without lactate; n = 4–16. (G) HUVECs transfected with the indicated siRNA were incubated with lactate. *p = 0.0170, **p = 0.0019 compared to control; ##p = 0.0056 compared to lactate alone; n = 3–9. (H) BAECs were incubated with pyruvate (10 mM), lactate, or not. **p<0.01 compared to control; n = 3–19. (A–H) Error bars reflect mean ± SEM.
Figure 5.
Targeting MCT1 inhibits lactate-induced HIF-1 activation in normoxic endothelial cells.
(A) Relative mRNA expression was determined in confluent HUVECs using quantitative RT-PCR. n = 3. Error bars reflect mean ± SEM. (B) Immunostaining showing the membrane expression of MCT1 in HUVECs (upper panel) and negative control (lower panel, omission of primary antibody). Bar = 40 µm. (C) HUVECs transfected with the indicated siRNA were incubated during 24-h with 10 mM lactate or not. HIF-1α protein expression was detected by western blot. *p<0.05, ***p<0.005 compared to control; ###p<0.005 compared lactate alone; n = 6–14. (D and E) HIF-1α and VEGFR2 protein expression was detected by western blot. (D) HUVECs were incubated during 24-h with 10 mM lactate and/or 5 mM α-cyano-4-hydroxycinnamate (CHC). *p<0.05, **p<0.01, ***p<0.005 compared to control; ###p<0.005 compared to lactate alone; n = 3–12. (E) BAECs were incubated during 24-h with 10 mM lactate, 5 mM Nω-nitro-L-arginine methyl ester (L-NAME), 5 mM α-cyano-4-hydroxycinnamate (CHC), and/or 5 mM N-acetyl-L-cysteine (NAC). *p<0.01, ***p<0.005 compared to control; ###p<0.005 compared to lactate alone; n = 3–14. (C–E) Error bars reflect mean ± SEM.
Figure 6.
Targeting MCT1 inhibits lactate-induced endothelial cell migration and vascular sprouting.
(A) Confluent BAECs on coverslips were placed into cell-free dishes in the presence of 10 mM lactate or not. Pictures show cell migration over time. Bar = 50 µm. (B) Quantification of EC migration was performed in a transwell assay in which calcein-labeled BAECs migrated during 24-h through a Matrigel-coated membrane towards serum-free medium containing 10 ng/ml VEGF, 10 mM lactate, or lactate and 5 mM CHC. Pictures show the membranes where pores occupied by endothelial cells (detected using a caveolin-1 antibody) are labeled with a red bullet. Bar = 50 µm. The graph shows calcein fluorescence in the bottom well. **p<0.01, ***p<0.005 compared to control; ###p<0.005 compared to lactate; n = 3. (C) Consecutive rings of a mouse aorta were cultured during 10 days in collagen gels containing 10 mM lactate and/or 2 mM CHC. Pictures show endothelial sproutings (arrowhead point at typical linear structures) and fibroblast seeding (scattered cells). Bar = 500 µm. The graph shows the number of sprouts per ring. ***p<0.005 compared to control; ###p<0.005 compared to lactate; n = 6–11. (B and C) Error bars reflect mean ± SEM.
Figure 7.
MCT1 inhibition blocks lactate-induced endothelial tube formation.
HUVECs were seeded on growth factor-reduced Matrigel containing 10 mM lactate or not and monitored for their ability to generate vascular tubes. (A) Pictures of the upper panel show endothelial network formation over time (bar = 500 µm), which was quantified in the lower graph. n = 4. Error bars reflect mean ± SEM and are sometimes smaller than symbols. (B) The tube formation assay was performed in the presence of 10 µM SU5402 (targeting VEGF and bFGF receptors), 10 nM echinomycin (targeting the transcriptional activity of HIF-1), or 2 mM of CHC (targeting MCT1). Pictures were captured after 8-h of culture in the Matrigel matrix. Bar = 200 µm.
Figure 8.
MCT1 inhibition blocks tumor angiogenesis.
(A) Two Matrigel plugs were implanted subcutaneously in the same mouse, one containing 10 mM lactate (left flank) and the other one an equal volume of saline (right flank). Plugs were microdissected 7 days after implantation. Pictures show CD31 immunostainings (bar = 100 µm; inset: a vascular structure carrying red blood cells), which were quantified in the graph. ***p = 0.0007; n = 4. Error bars reflect mean ± SEM. (B and C) On Day 0, mice were randomly assigned to treatment groups consisting of daily i.p. injections of CHC (1 mmol/Kg) or vehicle. (B) Mice carrying LLc tumors (4±1 mm) were treated. Pictures show CD31 immunostainings of size-matched tumors (bar = 20 µm), which were quantified in the graph. **p = 0.0051; n = 4. Error bars reflect mean ± SEM. (C) Intravital microscopy pictures were captured over time in treated mice carrying tumors generated with MCT1-negative TLT tumor cells. Bar = 1 mm.
Figure 9.
Model illustrating the anti-angiogenic activity of MCT1 inhibitors targeting lactate-induced HIF-1 activation in endothelial cells.
HIF-1 activity primarily depends on the stability of HIF-1α. When HIF prolylhydroxylase (PHD) is active as it is the case for normoxic endothelial cells (ECs) in oxidative tissues (left), HIF-1α is hydroxylated and addressed to the proteasome for degradation. PHD requires 2-oxoglutarate as a substrate and is competitively inhibited by pyruvate. Normoxic ECs in glycolytic tissues such as tumors (right) take up exogenous lactate originating from distant hypoxic cells, a process under the control of monocarboxylate transporter 1 (MCT1). ECs do not primarily use lactate as a fuel for oxidative phosphorylation (OXPHOS). Lactate oxidation by LDH-B rather increases the pool of intracellular pyruvate, now available to compete with 2-oxoglutarate from PHD, resulting in HIF-1α protein stabilization, HIF-1 activation, and angiogenesis. Because MCT1 is the main transporter for lactate uptake by ECs, targeting MCT1 inhibits lactate-induced angiogenesis in tumors. Abbreviations: asc., ascorbate; dehydroasc., dehydroascorbate; GLUT, glucose transporter.