High mobility group box 1 (HMGB1) is an alarmin actively secreted by immune cells and passively released by necrotic nonimmune cells. HMGB1 has been implicated in both cardiac contractile dysfunction and the lethality associated with sepsis/endotoxemia. The aim of the current study was to assess whether viable cardiomyocytes could produce HMGB1 and whether HMGB1 can affect myocardial contractility. LPS was used as a model of sepsis/endotoxemia in mice and isolated cardiac myocytes. LPS increased myocardial expression of HMGB1 in vivo (immunohistochemistry) and production and secretion of HMGB1 by viable cardiac myocytes in vitro (Western). LPS increased the phosphorylation status of PI3Kgamma in cardiac myocytes, an effect not observed in TLR4(-/-) myocytes. Genetic (PI3Kgamma(-/-)) or pharmacologic (AS605240) blockade of PI3Kgamma ameliorated the LPS-induced 1) cardiomyocyte production and secretion of HMGB1 in vitro and 2) HMGB1 expression in the myocardium in vivo. The LPS-induced depression of myocardial contractility was prevented by the HMGB1 antagonist, A-box. Genetic (PI3Kgamma(-/-)) or pharmacologic (AS605240) blockade of PI3Kgamma ameliorated the LPS-induced decrease in myocardial contractility. No evidence of inflammatory infiltrate was noted in any of the in vivo studies. The findings of the current study indicate that 1) LPS can induce HMGB1 secretion by viable cardiac myocytes through a TLR4/PI3Kgamma signaling pathway, and 2) HMGB1 plays a role in the LPS-induced myocardial contractile dysfunction. The results of the current study also have broader implications (i.e., that viable parenchymal cells, such as cardiac myocytes, participate in the alarmin response).

Alveolar macrophages (AMφ) have been implicated in the polymorphonuclear leukocyte (PMN) recruitment to the lungs during sepsis. Using an in vivo murine model of sepsis (feces in the peritoneum), we show that peritonitis leads to increased activation of AMφ and PMN migration into pulmonary alveoli. To assess cellular mechanisms, an in vitro construct of the pulmonary vascular-interstitial interface (murine AMφ, pulmonary endothelial cells, and PMN) and a chimera approach were used. Using immunologic (Abs) and genetic blockade (CXCR2-deficient AMφ), we show that CXC chemokines in septic plasma are responsible for the activation of AMφ. The activated AMφ can promote PMN transendothelial migration, even against a concentration gradient of septic plasma, by generating platelet-activating factor and H2O2. Platelet-activating factor/H2O2 induce an oxidant stress in the adjacent endothelial cells, an event that appears to be a prerequisite for PMN transendothelial migration, since PMN migration is abrogated across Cu/Zn-superoxide dismutase overexpressing endothelial cells. Using gp91-deficient endothelial cells, we show that NADPH oxidase plays an important role in the AMφ-induced PMN transendothelial migration. Pharmacologic/small interfering RNA blockade of Src kinase inhibits AMφ-induced endothelial NADPH oxidase activation and PMN migration. Collectively, our findings indicate that the PMN transendothelial migration induced by septic AMφ is dependent on the generation of superoxide in endothelial cells via the Src kinase/NADPH oxidase signaling pathway.

Septic plasma can convert murine cardiac myocytes to a proinflammatory phenotype. These myocytes 1) have increased nuclear levels of nuclear factor-B (NF-B), 2) release CXC chemokines, and 3) promote polymorphonuclear neutrophil (PMN) transendothelial migration. The purpose of the present study was to evaluate the role of the mitogen-activated protein (MAP) kinases [p38 MAP kinase, extracellular signal-regulated kinase (ERK) 1/2, and c-Jun NH 2-terminal kinase (JNK)] as upstream intracellular signaling components involved in this phenomenon. Feces-induced peritonitis (FIP) was employed as a model of sepsis. In vitro, cardiac myocytes were treated with plasma (20%) obtained 6 h after either sham (saline) or FIP procedures. Myocyte supernatants were used for 1) detection of the CXC chemokines (enzyme-linked immunosorbent assay) and 2) assessment of their ability to promote PMN transendothelial migration. In vivo, myocardial PMN accumulation was assessed by measuring myeloperoxidase (MPO) activity and function (dF/dt and heart work). Treatment of cardiac myocytes with septic plasma activated p38 MAP kinase and ERK1/2, but not JNK. Blockade approaches (inhibitors or small-interference RNA) indicated that only p38 MAP kinase played a role in the conversion of the myocytes to a proinflammatory phenotype. Time course studies indicated that phosphorylation of p38 MAP kinase preceded the phosphorylation of NF-B p65. Inhibition of p38 MAP kinase (SB-202190) blocked both NF-B p65 phosphorylation and NF-B nuclear translocation. Confirmatory studies in vivo indicated that FIP resulted in an increase in myocardial MPO activity and dysfunction, events reversed by the inhibitor of p38 MAP kinase. Collectively, these data indicate that the cardiomyocyte p38 MAP kinase/NF-B signaling pathway plays an important role in the sepsis-induced conversion of myocytes to a proinflammatory phenotype.