Lactate promotes endothelial-to-mesenchymal transition via Snail1 lactylation after myocardial infarction

Abstract

High levels of lactate are positively associated with the prognosis and mortality in patients with heart attack. Endothelial-to-mesenchymal transition (EndoMT) plays an important role in cardiac fibrosis. Here, we report that lactate exerts a previously unknown function that increases cardiac fibrosis and exacerbates cardiac dysfunction by promoting EndoMT following myocardial infarction (MI). Treatment of endothelial cells with lactate disrupts endothelial cell function and induces mesenchymal-like function following hypoxia by activating the TGF-β/Smad2 pathway. Mechanistically, lactate induces an association between CBP/p300 and Snail1, leading to lactylation of Snail1, a TGF-β transcription factor, through lactate transporter monocarboxylate transporter (MCT)-dependent signaling. Inhibiting Snail1 diminishes lactate-induced EndoMT and TGF-β/Smad2 activation after hypoxia/MI. The MCT inhibitor CHC mitigates lactate-induced EndoMT and Snail1 lactylation. Silence of MCT1 compromises lactate-promoted cardiac dysfunction and EndoMT after MI. We conclude that lactate acts as an important molecule that up-regulates cardiac EndoMT after MI via induction of Snail1 lactylation.

Fig. 1.. Increased lactate levels result in worsened cardiac dysfunction and increased cardiac fibrosis after MI.

Intraperitoneal injection of 2-DG or vehicle was administrated to wild-type mice daily since 1 day before MI or sham surgery. Three days after surgery, serum (A) and cardiac (B) lactate levels were measured by commercially available kit (n = 4 to 7 per group). (C to F) Left ventricular (LV) ejection fraction (EF%), fractional shortening (FS%), LV end-diastolic volume (LVEDV), and LV end-systolic volume (LVESV) were tested 7 days after MI or sham surgical operation (n = 4 to 6 per group). (G) To examine cardiac fibrosis, cardiac sections were stained with trichrome stain (Masson) kit (n = 4 per group). In separate experiments, mice were subjected to MI or sham surgery followed by supplemental lactate administration (intraperitoneal injection) every 7 days for 28 days. (H) Serum lactate levels were examined at 2, 4, 6, 8, 12, 24, 72, 120, and 168 hours after lactate injection (n = 3 per group). (I to L) To evaluate cardiac function, EF%, FS%, LVEDV, and LVESV were tested 7 and 28 days after MI or sham surgical operation (n = 6 to 10 per group). (M) Cardiac sections were stained with trichrome stain (Masson) kit to test cardiac fibrosis (n = 4 per group). Comparisons of data between groups were made using two-way analysis of variance (ANOVA) followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups. ns, not significant.

Fig. 2.. 2-DG attenuates MI-induced EndoMT.

2-DG or vehicle was administrated to wild-type mice via intraperitoneal injection daily since 1 day before MI or sham surgery, and heart tissues were harvested. (A) The expression of endothelial cell marker CD31, VE-cadherin, mesenchymal marker Collagen1a1, α-SMA, FSP1, and TGF-β in the myocardium was measured by Western blot (n = 4 per group). (B) Expression analysis by qRT-PCR of endothelial marker Cdh5 and Kdr and mesenchymal marker Acta2, Col1a1, Fn1, and S100a4 mRNA from the myocardium of MI or sham mice with 2-DG or vehicle administration (n = 4 to 6 per group). Endothelial cell–specific GFP-labeled (TIE2GFP) mice were subjected to MI or sham surgery. (C) Representative immunofluorescent staining images of GFP-labeled endothelial cell (green) and fibroblast marker FSP1 (red) in the heart tissues of TIE2GFP mice (n = 4 per group). Scale bar, 50 μm. (D) GFP-positive endothelial cells from heart tissues of TIE2GFP mice were also positive for endothelial cell marker anti-CD31 and anti-CD144 antibodies. (E) Representative flow density plot and quantitative analysis for gp38-positive endothelial cell frequency in all GFP-positive endothelial cells from heart tissues of TIE2GFP mice. n = 4 per group. Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 3.. Supplemental lactate promotes EndoMT following MI.

Mice were subjected to MI or sham surgery followed by supplemental lactate administration (intraperitoneal injection, 0.5 g/kg body weight). (A) The expression of endothelial cell markers CD31 and VE-cadherin and mesenchymal markers FSP1, α-SMA, and Collagen1a1 in the myocardium was measured by Western blot (n = 4 per group). (B to F) Expression analysis by qRT-PCR of endothelial marker Cdh5 and Kdr and mesenchymal marker Col1a1, Fn1, and S100a4 mRNA from the myocardium of MI or sham mice with lactate or vehicle administration (n = 4 to 6 per group). TIE2GFP mice were subjected to MI or sham surgery followed by lactate or vehicle administration. (G) Representative immunofluorescent staining images of GFP-labeled endothelial cell (green) and fibroblast marker FSP1 (red) in the heart tissues of TIE2GFP mice. Scale bar, 50 μm. (H) Representative flow density plot and quantitative analysis for gp38-positive endothelial cell frequency in all GFP-positive endothelial cells from heart tissues of TIE2GFP mice. n = 4 to 5 per group. Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 4.. Lactate induces endothelial cell migration and decreases VE-cadherin expression after hypoxia.

HUVECs were treated with lactate (5 or 10 mM) followed by normoxic or hypoxic challenge. (A) Endothelial cell migration was measured by wound-healing (or scratch) assay. Original magnification, ×20. (B) Endothelial cell proliferation was measured by EdU staining. Expression analysis by qRT-PCR (C) and immunofluorescent staining (D) of VE-cadherin (green) and nuclei (DAPI, blue). Scale bars, 50 μm. n = 3 per group. Comparisons of data between groups were made using one-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 5.. Lactate promotes EndoMT in endothelial cells after hypoxia.

HUVECs were treated with lactate (10 mM) followed by normoxic or hypoxic challenge. (A) Morphology of endothelial cells. Original magnification, ×20. n = 3 per group. (B) Immunofluorescent staining of CD31 (green), FSP1 (red), and nuclei (DAPI, blue). Scale bar, 10 μm. n = 3 per group. (C) The expression of endothelial marker VE-cadherin and mesenchymal marker α-SMA in endothelial cells was measured by Western blot (n = 3 per group). (D) Endothelial cell angiogenesis was examined by Matrigel assay (n = 3 per group). Scale bar, 200 μm. (E) Collagen gel contraction assay was performed to test fibroblast function (n = 4 per group). Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01 compared with indicated groups.

Fig. 6.. Lactate stimulates TGF-β/Smad2 signaling after MI/hypoxia.

(A to C) Mice were subjected to MI or sham surgery followed by supplemental lactate administration. The mRNA expression of Tgfb1, Smad2, and Smad3 was examined by qRT-PCR (n = 4 to 6 per group). (D and E) In separate experiments, mice were subjected to MI or sham surgery with 2-DG or vehicle administration. The mRNA expression of Tgfb1 and Smad2 was examined by qRT-PCR (n = 3 to 4 per group). (F to H) HUVECs were treated with lactate (10 mM) followed by normoxic or hypoxic challenge. The mRNA levels of TGFB1 and SMAD2 were examined by qRT-PCR. The expression of phospho (p)–SMAD2, phospho-SMAD3, and TGF-β was measured by Western blot (n = 3 to 4 per group). Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 7.. Inhibition of intracellular lactate attenuates hypoxia-induced EndoMT.

Endothelial cells were treated with the lactate transporter MCT inhibitor α-cyano-4-hydroxycinnamate (CHC) before lactate administration. (A) Intracellular lactate levels were measured by commercially available kit. (B) Endothelial cell migration was measured by wound-healing assay. Original magnification, ×20. (C) Immunofluorescent staining of VE-cadherin (green) and nuclei (DAPI, blue). Scale bar, 50 μm. (D) Endothelial cell angiogenesis was examined by Matrigel assay. Scale bar, 100 μm. (E to L) The mRNA levels of PECAM1, CDH5, KDR, ACTA2, FN1, COL1A1, TGFB1, and SMAD2 were examined by qRT-PCR. n = 3 to 4 per group. Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 8.. Lactate promotes Snail1 nuclear translocation and lactylation following hypoxia.

HUVECs were treated with lactate (10 mM) followed by normoxic or hypoxic challenge. The expression of cytoplasmic and nuclear Snail1 expression was measured by Western blot (A) and immunofluorescent staining (B). Scale bar, 50 μm. (C) ChIP assay was performed with anti-Snail1 antibody followed by qRT-PCR using primers specific for TGF-β. (D) Immunoprecipitation (IP) was performed to examine acetylation and lactylation of Snail1, as well as the interaction between Snail1 and CBP/p300. (E and F) Endothelial cells were treated with the MCT inhibitor CHC before lactate administration. Snail1 acetylation and lactylation, and the interaction between Snail1 and CBP/p300 were measured by immunoprecipitation. n = 3 to 4 per group. Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 9.. Silencing of Snail1 attenuates EndoMT and TGF-β/Smad2 activation after hypoxia.

HUVECs were transfected with siRNA specific for Snail1 (siSNAI1). Scrambled siRNA served as control (siNC). Twenty-four hours after transfection, cells were treated with lactate (10 mM) followed by hypoxic challenge. The expression of endothelial markers and mesenchymal markers was measured by Western blot (A) and qRT-PCR (B to G). The activation of TGF-β/Smad2 was also measured by Western blot (H) and qRT-PCR (I and J). n = 3 to 4 per group. Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.

Fig. 10.. Silencing of Snail1 attenuates Snail1 lactylation and EndoMT after MI.

Mice were subjected to MI or sham surgery followed by supplemental lactate or 2-DG administration. (A and B) Mouse hearts were collected and immunoprecipitation was performed to examine acetylation and lactylation of Snail1, as well as the interaction between Snail1 and CBP/p300. Mice were administered siRNA specific for Snail1 before MI or sham surgery. Western blot (C) and immunofluorescent staining (D) were performed to measure the expression of Snail1 expression in the myocardium. Scale bar, 200 μm. (E to H) Seven days after surgery, cardiac function (EF%, FS%, LVEDV, and LVESV) was measured by echo. (I to M) The mRNA levels of Cdh5, Kdr, Acta2, Fn1, and S100a4 were examined by qRT-PCR. (N) Scheme of lactate-induced EndoMT after MI. Comparisons of data between groups were made using two-way ANOVA followed by Tukey’s procedure or t test. *P < 0.05, **P < 0.01, ***P < 0.001 compared with indicated groups.