Fig 2: DNA-PKcs promotes endothelial barrier dysfunction, inflammation, and vasoconstriction in myocardial microvessels. (A to C) Western blot analysis of claudin-5 and VE-cadherin expression in DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice. (D and E) Immunohistochemistry was used to observe albumin leakage into myocardium in DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice. (F) FITC-dextran clearance assays were performed following 24-h LPS treatment (10 μg/ml) in CMECs isolated from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice. FITC-dextran permeation was measured to assess alterations in endothelial barrier function. (G) TER assays were performed in CMECs isolated from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice to evaluate changes in endothelial barrier integrity following LPS exposure. (H to J) Western blot analysis of VCAM1 and VCAM1 expression in cardiac microvessels from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice treated with LPS. (K and L) Immunofluorescence of Gr-1+ neutrophils in heart samples from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice treated with LPS. DAPI was used to stain nuclei and TnT was used to stain cardiomyocytes. (M to O). RT-qPCR was used to analyze the transcription of Mmp-9, IL-6, and Tnf-α in cardiac tissue from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice after LPS exposure. (P to R) Western blot analysis of p-eNOS and ET-1 expression in heart tissues from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice treated with LPS. (S) Endothelial-dependent and endothelial-independent relaxation responses were assessed in aortic rings by applying Ach (10−9–10−5 M) or SNP (10−10–10−6 M). Experiments were repeated at least 3 times. Data are shown as mean ± SEM (n = 6 mice or 3 independent cell isolations per group). *P < 0.05.

Fig 3: DNA-PKcs binds to cofilin2 by recognizing a TQ motif. (A and B) Western blot analysis of Cofilin1/2 in CMECs isolated from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice in the presence of LPS. (C and D) Co-IP assays were conducted to evaluate binding of DNA-PKcs to cofilin1/2 and F-actin using extracts from LPS-treated CMECs isolated from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice. (E) HCAECs were transfected with HA-DNA-PKCs before LPS treatment. Then, the interaction between DNA-PKcs and F-actin as well as the interplay between DNA-PKcs and cofilin1/2 were measured through Co-IP. (F) Immunoblots of DNA-PKcs and cofilin1 immunoprecipitates (Ips) in LPS-treated HCAECs. (G) Immunoblots of DNA-PKcs and cofilin2 Ips in LPS-treated HCAECs. (H) DNA-PKcs and ADF Ips from LPS-treated HCAECs were immunoblotted as indicated. (I) Docking analysis of the interaction between DNA-PKcs and cofilin2. (J and K) Putative hydrogen and hydrophobic bonds between DNA-PKcs and cofilin2 are indicated. (L) Amino acid sequences of cofilin2 in various species. (M) HCAECs were transfected with His-tagged cofilin2 constructs, including cofilin2 without Ser24 (His-cofilin2ΔS24), without Thr25 (His-cofilin2ΔT25), without Gln26 (His-cofilin2ΔQ26), or without Thr25 and Gln26 (His-cofilin2ΔT25Q26). After cells were exposed to LPS, His immunoprecipitates were collected and immunoblotted to determine their interaction with DNA-PKcs. (N and O) Immunofluorescence analysis of F-actin expression in HCAECs transfected with different mutant cofilin2 constructs. (P and Q) Endothelial barrier function was determined by FITC clearance assay and TER detection in HCAECs transfected with different cofilin2 constructs. (R) qPCR analysis of Icam1 expression in HCAECs transfected with different cofilin2 constructs. (S) ELISA assay of eNOS activity in HCAECs transfected with different cofilin2 constructs. Experiments were repeated at least 3 times and the data are shown as mean ± SEM (n = 6 mice per group). *P < 0.05.

Fig 4: DNA-PKcs phosphorylates cofilin2 at Thr25. (A) Cofilin2 phosphorylation was assessed in CMECs isolated from DNA-PKcsf/f/Tie2Cre and control DNA-PKcsf/f mice in the presence of LPS. (B) Cofilin2 phosphorylation was assessed in control CMECs and NU7441-treated CMECs after LPS exposure. (C) In vitro kinase assay results in HCAECs using recombinant mouse DNA-PKcs and recombinant mouse cofilin2 proteins in the presence or absence of NU7441. The levels of cofilin2Thr25 phosphorylation and DNA-PKcs were determined by western blotting. (D) HCAECs were transfected with phospho-mimetic (His-cofilin2T25D) or phospho-defective (His-cofilin2T25A) cofilin2 variants. HA-tagged immunoprecipitates were collected and immunoblotted to determine the interaction between HA-cofilin2 and F-actin. (E and F) Immunofluorescence imaging of F-actin in HCAECs transfected with His-cofilin2T25D and His-cofilin2T25A and treated with LPS. Scale bar, 45 μm. (G to I) Western blot analysis of claudin-5 and ICAM1 expression in HCAECs transfected with His-cofilin2T25D and His-cofilin2T25A and treated with LPS. (J and K) Endothelial barrier function was determined by FITC-dextran clearance assays and TER detection in HCAECs transfected with His-cofilin2T25D and His-cofilin2T25A and treated with LPS. Experiments were repeated at least 3 times. Data are shown as mean ± SEM (n = 6 mice or 3 independent cell isolations per group). *P < 0.05.

Fig 5: Repression of cofilin2Thr25 phosphorylation confers protection against endotoxemia-induced myocardial microvascular injury. WT, heterozygous cofilin2T25A/+, and homozygous cofilin2T25A/A mice were injected with LPS to model endotoxemic cardiomyopathy (n = 6 mice/group). (A and B) Western blot analysis of cofilin2 phosphorylation in CMECs isolated from mice. (C and D) Western blot analysis of ICAM1 in WT, heterozygous cofilin2T25A/+, and homozygous cofilin2T25A/A mice. (E and F) Immunofluorescence of VE-cadherin in myocardial microvessels. (G and H) Immunofluorescence of Gr-1+ neutrophils in WT, cofilin2T25A/+, and cofilin2T25A/A mice. DAPI was used to stain nuclei and TnT to stain cardiomyocytes. Scale bar, 65 μm. (I to K) Western blot analysis of p-eNOS and ET-1 expression in cardiac microvessels from WT, cofilin2T25A/+, and cofilin2T25A/A mice. (L to N) Western blots was used to evaluate cofilin2 phosphroyaltion in human circulating CD34+ ECs and EPCs. Experiments were repeated at least 3 times. Data are shown as mean ± SEM (n = 6 mice or 3 independent cell isolations per group). *P < 0.05. (O) LPS activates DNA-PKcs which ecognizes a TQ motif in cofilin2 and consequently induces cofilin2 phosphorylation at Thr25. Phosphorylated cofilin2 shows increased affinity for F-actin and promotes F-actin depolymerization, leading to disruption of the endothelial barrier integrity, microvascular inflammation, and defective eNOS-dependent vasodilation.