Serpina3k lactylation protects from cardiac ischemia reperfusion injury

Abstract

Lactate produced during ischemia-reperfusion injury is known to promote lactylation of proteins, which play controversial roles. By analyzing the lactylomes and proteomes of mouse myocardium during ischemia-reperfusion injury using mass spectrometry, we show that both Serpina3k protein expression and its lactylation at lysine 351 are increased upon reperfusion. Both Serpina3k and its human homolog, SERPINA3, are abundantly expressed in cardiac fibroblasts, but not in cardiomyocytes. Biochemically, lactylation of Serpina3k enhances protein stability. Using Serpina3k knockout mice and mice overexpressing its lactylation-deficient mutant, we find that Serpina3k protects from cardiac injury in a lysine 351 lactylation-dependent manner. Mechanistically, ischemia-reperfusion-stimulated fibroblasts secrete Serpina3k/SERPINA3, and protect cardiomyocytes from reperfusion-induced apoptosis in a paracrine fashion, partially through the activation of cardioprotective reperfusion injury salvage kinase and survivor activating factor enhancement pathways. Our results demonstrate the pivotal role of protein lactylation in cardiac ischemia-reperfusion injury, which may hold therapeutic value.

Fig. 1. Global lactylome profiling of myocardial ischemia/reperfusion injury.

a Experimental design. Adult C57BL/6 mice were subjected to sham, myocardial infarction (MI) or ischemia/reperfusion injury (I/R) surgeries, respectively. Protein extracted from the infarct area was used for global proteome and lactylome profiling by 4D-LC-MS/MS. n = 3 biological replicates, each biological replicate was a pooled myocardial sample from 3 mice. b, c Density gradient diagram of the Log₂ ratios of Kla sites and proteins in MI and I/R groups of the lactylomes (b) and proteomes (c), respectively. Red and green represent MI and I/R groups, respectively. Red and green numbers (percentages) represent the numbers and percentages of up- or down-regulated Kla sites or proteins, respectively (cutoff |Log₂FC | >1). d Numbers of Kla sites and Kla proteins exhibiting significant Kla-level changes after MI and I/R (cutoff of |Log₂FC | >1.3). e Numbers of proteins or sites exhibiting significant expression changes after MI and I/R (cutoff of |Log₂FC | >1.3). f Heat map showing hierarchical clustering of differentially lactylated sites in Sham, MI and I/R mouse hearts. Three biological replicates are shown separately. g Gene ontology (GO) analyses of enriched biological processes in (f). Selected top categories are shown. Statistical significance was performed using the two-sided hypergeometric distribution test. h Heat map showing hierarchical clustering of differentially expressed proteins in Sham, MI and I/R mouse hearts. Three biological replicates are shown separately. i Gene ontology (GO) analyses of enriched biological processes in (h). Selected top categories are shown. Statistical significance was performed using the two-sided hypergeometric distribution test. j Protein-protein interaction network of lactylated proteins in I/R based on the STRING database. The size of the circles represents the number of interacting proteins. Red represents upregulation, green represents downregulation, and yellow represents both. k Venn diagram of differentially expressed or lactylated proteins in MI or I/R compared to Sham control.

Fig. 2. SA3K and SA3K-Kla is increased upon I/R in vivo and upon normoxia-hypoxia in vitro.

a Western blotting and quantification of SA3K protein in mouse heart tissue. n = 3 mice per group. b Immunostaining and quantification of SA3K expression in mouse heart sections. Scale bar: 50 μm and 8 μm. n = 6 mice per group. c, d Immunoprecipitation and western blotting (c) and quantification (d) of the lactylation level of SA3K in mouse heart tissue. n = 4 mice per group. e Western blotting of SA3K using an SA3K-K351la-specific antibody in mouse heart tissues. Quantification is shown on the right. n = 4 mice per group. f Immunostaining of SA3K and vimentin (VIM) in mouse I/R heart sections. Arrows indicate co-localization. Scale bar: 50 μm (left) and 8 μm (right). g qRT-PCR of SA3K mRNA in neonatal mouse CMs and FBs. n = 3 independent experiments. h, i Western blotting of SA3K-K351la and SA3K in neonatal mouse FBs (h) and CMs (i) cultured under normoxia (N), hypoxia (H), or normoxia-hypoxia (H-N) conditions, n = 4 (FB) and n = 3 (CM) independent experiments. j qRT-PCR of SA3 mRNA in adult human primary CMs and cardiac FBs. n = 3 independent experiments. k Western blotting and quantification of SA3 expression in adult human primary FBs and CMs cultured under normoxia (N), hypoxia (H), or normoxia-hypoxia (H-N) conditions, n = 3 independent experiments. l ELISA of SA3 secreted from human primary FBs, n = 4 independent experiments. m, n Immunoprecipitation and western blotting of the lactylation level of SA3 in adult human primary cardiac FBs (m) and CMs (n) cultured under different conditions. Quantifications are shown on the right. n = 4 independent experiments. Significance was calculated using one-way ANOVA followed by Tukey’s multiple comparisons test (a, c, e, h, k, l), unpaired, one-tailed t-test (g, j), or the Friedman test followed by Dunn’s multiple comparisons test (d, m). For (a, c, m, n), GAPDH loading control was run on a separate gel. All data are means ± SD. Source data are provided as a Source Data file.

Fig. 3. L-lactate upregulates SA3K level by enhancing lactylation-dependent protein stability.

a Western blotting of SA3K protein and SA3K-K351la levels in neonatal mouse FBs cultured in increasing concentrations of L-lactate for 24 h. Quantification of relative SA3K and SA3K-K351la expression level is shown on the right. n = 4 independent experiments. b Schematic of regulatory factors in glucose metabolism. Rotenone inhibits oxidative phosphorylation, while oxamate inhibits lactate production. c Intracellular L-lactate levels were measured in mouse neonatal FBs treated with rotenone or oxamate for 24 h. n = 3 independent experiments. d, e Western blotting of SA3K protein and SA3K-K351la levels in mouse neonatal FBs cultured in increasing concentrations of oxamate (d) or L-lactate (e) for 24 h. GAPDH loading control was run on a separate gel in (e). Quantifications of relative SA3K and SA3K-K351la expression level are shown on the right, respectively. n = 4 independent experiments. f Measurement of endogenous SA3K protein half-life. Neonatal FBs were cultured in control medium, 25 mM L-lactate, or 20 mM oxamate for 24 h, and then treated with 50 μg/ml cycloheximide (CHX) for the indicated times in the presence or absence of MG132 before cellular proteins were extracted for western blotting. Quantitation was done by densitometry and expressed as the signal ratio of SA3K/GAPDH. n = 3 independent experiments. g Measurement of exogenous SA3K protein half-life. Neonatal FBs were transduced with SA3K-WT- or SA3K-K351R-expressing lentivirus. Cells overexpressing SA3K-WT or SA3K-K351R were treated with 25 mM L-lactate for 24 h, and then treated with 50 μg/ml CHX for the indicated times in the presence or absence of MG132 before cellular proteins were extracted for western blotting. Quantitation was done by densitometry and expressed as the signal ratio of SA3K/GAPDH. n = 3 independent experiments. Half-life calculation and visualization were performed by R (4.0.3) and ggplot2 (3.3.2) package in (f, g). Significance was calculated using one-way ANOVA followed by Tukey’s multiple comparisons test (a, c, d, e). All data are means ± SD. Source data are provided as a Source Data file.

Fig. 4. SA3K is an important driver of cardiac protection in reperfusion injury.

a Diagram outlining the experimental design for evaluating SA3K function in I/R. Wild-type (WT) or SA3K knockout (SA3K-KO) mice were subjected to I/R surgery. His-tagged SA3K (His-SA3K) was injected into a group of SA3K-KO at reperfusion. Analyses of myocardial ischemia, apoptosis, cardiac function, and fibrosis were performed at indicated time points. b Representative photographs showing TTC-stained transverse sections of Evans blue-perfused hearts. c, d Area at risk (AAR) (c) and infarct size (d) were quantified 1 day after IRI, n = 9 mice in each group. e–g Serum levels of cardiac troponin T (cTnT) (e), cardiac troponin I (cTnI) (f), and creatine kinase MB (CK-MB) (g) were measured 1 day after IRI, n = 10 mice in each group. h Representative terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining images showing apoptotic cell death in heart sections 1 day after IRI. An ACTN2 antibody was used to label cardiomyocytes. Scale bar, 50 μm and 8 μm. Quantification of TUNEL-positive cell rates is shown on the right. n = 11 mice in each group, 6 sections per animal. i Representative echocardiograms at D28. j Quantification of left ventricular ejection fraction (EF), left ventricular fractional shortening (FS), left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), left ventricular internal diameter at end diastole (LVIDd), and left ventricular internal diameter at end systole (LVIDs) at D28 based on echocardiography. n = 10 mice in each group. k Masson’s trichrome staining of heart sections 28 days after reperfusion. Scale bar: 2 mm. Quantification of fibrosis is shown on the right. n = 9 mice in each group, 4 sections per animal. Significance was calculated using one-way ANOVA followed by Tukey’s multiple comparisons test (d–h, j, k). All data are means ± SD. Source data are provided as a Source Data file.

Fig. 5. SA3K lactylation at K351 is required for cardiac protection.

a Diagram outlining the experimental design for evaluating the role of K351la in I/R. Mice were injected with AAV9 (empty vector (EV), wild-type SA3K (SA3K-WT) and mutant SA3K (SA3K-K351R)) 4 weeks before I/R surgery. Analyses of myocardial ischemia, apoptosis, cardiac function, and fibrosis were performed at indicated time points. b Representative photographs showing TTC-stained transverse sections of Evans blue-perfused hearts. c, d Area at risk (AAR) (c) and infarct size (d) were quantified 1 day after IRI, n = 9 mice in each group. e–g Serum levels of cardiac troponin T (cTnT) (e), cardiac troponin I (cTnI) (f), and creatine kinase MB (CK-MB) (g) were measured 1 day after IRI, n = 10 mice in each group. h Representative terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining images showing apoptotic cell death in heart sections 1 day after IRI. An ACTN2 antibody was used to label cardiomyocytes. Scale bar, 50 μm and 8 μm. Quantification of TUNEL-positive cell rates is shown on the right. n = 10 mice in each group, 6 sections per animal. i Representative echocardiograms at D28. j Quantification of left ventricular ejection fraction (EF), left ventricular fractional shortening (FS), left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), left ventricular internal diameter at end diastole (LVIDd), and left ventricular internal diameter at end systole (LVIDs) at D28 based on echocardiography. n = 10 mice in each group. k Masson’s trichrome staining of heart sections 28 days after reperfusion. Scale bar: 2 mm. Quantification of fibrosis is shown on the right. n = 9 mice in each group, 4 sections per animal. Significance was calculated using one-way ANOVA followed by Tukey’s multiple comparisons test (d–h, j, k). All data are means ± SD. Source data are provided as a Source Data file.

Fig. 6. SA3K/SA3 secreted by FBs protect CMs from IRI-induced apoptosis through paracrine signaling.

a Experimental design. Cardiac FBs from neonatal mice or adult humans were cultured under normoxia (N) or hypoxia-normoxia (H-N) conditions. CMs were treated with conditioned media, and subjected to H-N. Nab, neutralizing antibody. b TUNEL staining of apoptotic cells (white arrow). Scale bar, 50 μm. c Quantification of (b). n = 3 independent experiments. d Western blotting of cleaved caspase 3 (C-CASP3) and BAX in neonatal mouse and adult human CMs. e Quantification of (d). n = 4 independent experiments. f, g Quantification of LDH in mouse (f) or human (g) CM supernatant. n = 4 (mouse) or 5 (human) independent experiments. h Quantification of cellular ATP concentration in mouse (left) or human (right) CMs. n = 4 (mouse) or 5 (human) independent experiments. i Experimental design. SA3K and SA3 were silenced by shRNA in neonatal mouse and adult human FBs, respectively. Corresponding CMs were treated with conditioned media, and subjected to H-N. j, k Quantification of TUNEL-positive cells (j), and representative images of TUNEL staining (k). White arrows indicate TUNEL-positive cells. Scale bar, 50 μm. n = 3 independent experiments. l Western blotting of C-CASP3 and BAX in neonatal mouse and adult human CMs. m, n Quantification of (l) (m, mouse; n, human). n = 4 independent experiments. o Experimental design. Neonatal mouse FBs were transduced with lentivirus expressing empty vector (EV), SA3K-WT or SA3K-K351R. Neonatal CMs were treated with conditioned media, and subjected to H-N. p, q TUNEL staining (p) and quantification (q) of apoptotic CMs. White arrows indicate TUNEL-positive cells. Scale bar, 50 μm. n = 3 independent experiments. r Western blotting of C-CASP3 and BAX in neonatal mouse CMs. Quantification is shown on the right. n = 3 independent experiments. Significance was calculated by repeated-measures one-way ANOVA followed by Tukey’s multiple comparisons (c, f–h, j, q), or by the Kruskal-Wallis test followed by Dunn’s multiple comparisons test (e, m, p, r). GAPDH loading control was run on a separate gel in (l, r). All data are means ± SD. Source data are provided as a Source Data file.

Fig. 7. SA3K/SA3 protect CMs by inhibiting the WNT pathway and activating cardioprotective RISK and SAFE pathways.

a–d Western blots (a, c) and quantification (b, d) of p-AKT, AKT, p-ERK, and ERK in neonatal mouse (a, b) and adult human CMs (c, d) treated with conditioned media from FBs with SA3K/SA3 overexpression or depletion. GAPDH was used as loading control. n = 4 (b) or 5 (d) independent experiments. e–h Western blots (e, g) and quantification (f, h) of p-JAK2, JAK2, p-STAT3 (Tyr705), and STAT3 in neonatal mouse (e, f) and adult human CMs (g, h) treated with conditioned media from FBs with SA3K/SA3 overexpression or depletion. GAPDH was used as loading control. n = 5 independent experiments. All overexpression data were analyzed using the two-tailed Mann-Whitney test, while all knockdown data were analyzed using the Kruskal-Wallis test followed by Dunn’s multiple comparisons. For (a, c, g), GAPDH loading control was run on a separate gel. Data are means ± SD. Source data are provided as a Source Data file.