The myocardial infarct-exacerbating effect of cell-free DNA is mediated by the high-mobility group box 1-receptor for advanced glycation end products-Toll-like receptor 9 pathway

Abstract

Introduction: Damage-associated molecular patterns, such as high-mobility group box 1 (HMGB1) and cell-free DNA (cfDNA), play critical roles in mediating ischemia-reperfusion injury (IRI). HMGB1 activates RAGE to exacerbate IRI, but the mechanism underlying cfDNA-induced myocardial IRI remains unknown. We hypothesized that the infarct-exacerbating effect of cfDNA is mediated by HMGB1 and receptor for advanced glycation end products (RAGE).

Methods: C57BL/6 wild type mice, RAGE knockout (KO), and Toll-like receptor 9 KO mice underwent 20- or 40-minute occlusions of the left coronary artery followed by up to 60 minutes of reperfusion. Cardiac coronary perfusate was acquired from ischemic hearts without reperfusion. Exogenous mitochondrial DNA was acquired from livers of normal C57BL/6 mice. Myocardial infarct size (IS) was reported as percent risk region, as measured by 2,3,5-triphenyltetrazolium chloride and Phthalo blue (Heucotech, Fairless Hill, Pa) staining. cfDNA levels were measured by Sytox Green assay (Thermo Fisher Scientific, Waltham, Mass) and/or spectrophotometer.

Results: Free HMGB1 and cfDNA levels were increased in the ischemic myocardium during prolonged ischemia and subsequently in the plasma during reperfusion. In C57BL/6 mice undergoing 40'/60' IRI, deoxyribonuclease I, or anti-HMGB1 monoclonal antibody reduced IS by approximately half to 29.0% ± 5.2% and 24.3% ± 3.5% (P < .05 vs control 54.3% ± 3.4%). However, combined treatment with deoxyribonuclease I + anti-HMGB1 monoclonal antibody did not further attenuate IS (29.3% ± 4.9%). In C57BL/6 mice undergoing 20'/60' IRI, injection of 40'/5' plasma upon reperfusion increased IS by more than 3-fold (to 19.9 ± 4.3; P < .05). This IS exacerbation was abolished by pretreating the plasma with deoxyribonuclease I or by depleting the HMGB1 by immunoprecipitation, or by splenectomy. The infarct-exacerbating effect also disappeared in RAGE KO mice and Toll-like receptor 9 KO mice. Injection of 40'/0' coronary perfusate upon reperfusion similarly increased IS. The levels of HMGB1 and cfDNA were significantly elevated in the 40'/0' coronary perfusate and 40'/reperfusion (min) plasma but not in those with 10' ischemia. In C57BL/6 mice without IRI, 40'/5' plasma significantly increased the interleukin-1β protein and messenger RNA expression in the spleen by 30 minutes after injection. Intravenous bolus injection of recombinant HMGB1 (0.1 μg/g) or mitochondrial DNA (0.5 μg/g) 5 minutes before reperfusion did not exacerbate IS (P = not significant vs control). However, combined administration of recombinant HMGB1 + mitochondrial DNA significantly increased IS (P < .05 vs individual treated groups) and this infarct-exacerbating effect disappeared in RAGE KO mice and splenectomized C57BL/6 mice. The accumulation of cfDNA in the spleen after combined recombinant HMGB1 + mitochondrial DNA treatment was significantly more elevated in C57BL/6 mice than in RAGE KO mice.

Conclusions: Both HMGB1 and cfDNA are released from the heart upon reperfusion after prolonged ischemia and both contribute importantly and interdependently to post-IRI by a common RAGE-Toll-like receptor 9-dependent mechanism. Depleting either of these 2 damage-associated molecular patterns suffices to significantly reduce IS by approximately 50%.

Keywords: HMGB1; RAGE; TLR9; cell-free DNA; ischemia/reperfusion; spleen.

Figure 1.

A. Role of HMGB1 or cfDNA in myocardial IRI. C57BL/6 mice underwent 40 min of ischemia and 60 min of reperfusion. Phatho Blue was used to stain the heart for delineation of infarct size (% of RR), ischemic risk region (RR, % of LV). Anti-HMGB1 mAb at a dose of 1 μg/g mouse weight or DNase I at a dose of 30 mU/g mouse weight administered at 5 min before the onset of reperfusion. B. Effect of cfDNA depletion on plasma HMGB1. C. Effect of protein depletion on plasma cfDNA.

Figure 2.

Levels of free HMGB1 and cfDNA in myocardial tissue during IRI. C57BL/6 mice underwent either sham thoracotomy or IR. Cytox green was injected i.v. bolus at a dose of 50 μl of 1 mM 1 minute before reperfusion. The heart was harvested after injection in sham mice, and 5 min, 15 min, or 30 min following reperfusion. The left ventricle was cut into non-ischemic posterior wall and ischemic anterior wall. A: Fluorescence of LV anterior and posterior walls imaged under IVIS Lumina Imaging System. B: Levels of cfDNA in the homogenates of LV anterior and posterior wall were measured using photometer. C: Cell-free DNA was extracted from LV posterior and anterior walls respectively in sham, 40’/5’ and 40’/15’ hearts. The cfDNA was quantitatively measured using Nanodrop 2000 Spectrophotometer. cfDNA: cell free DNA.

Figure 3.

Level of free HMGB1 and cfDNA in coronary perfusate and plasma during IRI Cardiac coronary perfusate was obtained by antegrade perfusing of the heart via catheter through ascending aorta. In the perfusate, the level of HMGB1 was measured by Western blot and the level of cfDNA was measured using Cytox green and photometry (Left column). The blood sample was obtained during IR by puncturing the right ventricle. The plasma level of HMGB1 was measured by Western blot and the level of cfDNA was measured using Cytox green and photometry (Right column).

Figure 4.

Role of exogenous cardiac coronary perfusate in exacerbating myocardial infarct size. Perfusate acquired from 10’/0’ or 40’/0’ donor hearts was administered i.v. bolus 5 min before reperfusion at a dose of 2 μl/g recipient mouse. The recipient mice underwent 20’/60’ IR. SPLX: splenectomy 5 min before occlusion of LCA.

Figure 5.

Role of exogenous plasma in exacerbating myocardial infarct size. A. Plasma acquired from 10’/5’ or 40’/5’ donor C57BL/6 mice was administered i.v. bolus 5 min before reperfusion at a dose of 2 μl/g recipient mouse. The recipient mice underwent 20’/60’ IR. Immunoprecipitation was used to deplete HMGB1 in the donor plasma; DNase I was used to deplete cfDNA in the donor plasma at a dose of 1 mU/μl donor plasma. On upper portion of the figure are representative TTC-blue staining of the middle left ventricular slice. On top of the figure, the ischemic area (in red), infarct area (in yellow) and non-ischemic area (in blue) are drawn respectively and included area can be calculated. SPLX: splenectomy 5 min before occlusion of LCA. B. Level of cfDNA in the 40’/5’ plasma with different treatments using Cytox green (Results of Proteinase K-treated plasma were also reported in Figure 1C).

Figure 6.

Splenic tissue level of IL-1β protein and mRNA. Spleen was harvested from sham, 10’/30’ and 40’/30’ mice and homogenized. IL-1β protein and mRNA were evaluated by ELISA and RTC-PCR respectively.

Figure 7.

Role of exogenous rHMGB1 and mtDNA in exacerbating myocardial infarct size. C57BL/6 mice and RAGE KO mice underwent 20’/60’ IR and were treated upon reperfusion with either mouse recombinant HMGB1 (rHMGB1) at a dose of 0.1 μg/g mouse weight and/or mouse mtDNA at a dose of 0.5 μg/g mouse weight. SPLX: splenectomy 5 min before occlusion of LCA. A: Combined treatments 5 min before reperfusion exacerbated infarct size in WT mice but not in RAGE KO or splenectomized WT mice. B: 30 min after treatment in mice without IRI, combined treatments increased splenic tissue DNA level WT mice, more significantly than combined treatment in RAGE KO mice or the individual treatment with either rHMGB1 or mtDNA in WT mice. C: 30 min after treatment in mice without IRI, the plasma level of cfDNA was similar among the five groups.

Figure 8.

Schematic mechanism of the free HMGB1 and cfDNA complex in acute myocardial I/R injury.