Final Report Summary - NET-INJURY-IN-SEPSIS (Endothelial injury during sepsis: do NET-attached proteases participate in this process?)
Sepsis is a life-threatening condition because this systemic inflammation leads to multi-organ failure with severe injury to the liver and the lung. Especially Staphylococcus aureus-induced sepsis is becoming an alerting issue. In fact, S.aureus bacteraemia is one of the most common serious bacterial infections worldwide and is more strongly associated with death than when caused by any other bacterial pathogen. Moreover, methicillin-resistant S. aureus (MRSA) poses a serious threat as this originally hospital-associated S. aureus lineage has spread also into the community causing disease in otherwise healthy people.
One of main cell types participating in the process of inflammation, be it local or systemic, are neutrophils (see our review in Nat Rev Immunol. 2013; 13: 159-175). These cells kill pathogens in three ways: (i) intracellularly after phagocytosis of bacteria, (ii) extracellularly upon release of antimicrobial proteins and proteases (iii) extracellularly by a release of neutrophil extracellular traps (NETs). The function of NETs is to trap and kill bacteria but their formation is also accompanied by collateral injury to the lung and the liver endothelium if NETs persist in vasculature. Formation of NETs requires activity of peptidyl arginine deiminase 4 (PAD4) and neutrophil elastase (NE). NETs are composed of DNA which serves as a backbone decorated with histones and proteins from neutrophil granules, including proteases – NE and matrix metalloproteinase-9 (MMP-9). These enzymes were previously shown to primarily facilitate leukocyte migration and cytokine/chemokine activation. However, NE and MMP-9 also degrade numerous other extracellular and intracellular substrates. Thus the question arose: do MMP-9 and/or NE attached to NETs cause injury to bystander cells during sepsis? What is the mechanism of this process?
In order to answer the above questions, we utilized a technique of intravital (in vivo) microscopy which allows to observe processes occurring in real time inside of live mice. With this technique we focused on the liver as a target organ and we studied NET formation and its impact on the course of MRSA-induced sepsis. We observed that MRSA inoculation in mice leads to profound liver damage that is accompanied by NET formation. We confirmed that indeed the two proteases, NE and MMP-9, are attached to NETs. However, while NE is proteolytically active, MMP-9 is enzymatically silent. Importantly, genetic deficiency in NE or its pharmacological blockage prevented the liver damage. Similar prevention of collateral damage to the liver was observed in PAD4-deficient mice. Furthermore, we showed that it is NE attached to NETs that causes bystander cell damage when NETs persist in the liver vasculature for hours.
DNase which dissolves the backbone DNA of NETs is routinely used in experimental studies to remove NETs, and is being tested in some pre-clinical trials. We investigated how DNase treatment will impact removal of NETs by means of intravital microscopy. When injected intravenously and imaged in real time, DNase immediately removed extracellular DNA but failed to remove NE and histones. Even when DNase was left to act for several hours, it only partially removed other NET components. In subsequent studies we revealed that, once released into the vasculature, NE and histones of NETs attach to glycoproteins, such as Von Willebrand factor, lining the surface of endothelium. For this reason, removal of NETs by DNase is not sufficient to prevent the organ damage. These novel observations are presently “in press” in Nature Communications (2015).
Results of the studies undertaken during this fellowship might have important implications for treatment of sepsis and other inflammatory disorders in which NETs are formed. Now, it is clear that specific elements of NETs cause damage to the host organs, and that, in a relevant animal model of bacterial sepsis, the prevention of NET formation by inhibition of NE and/or PAD4 is much more effective than application of DNase. We have also revealed important novel mechanisms operating once NETs are released into the vasculature in regard to their binding to proteins lining the endothelial cells.
One of main cell types participating in the process of inflammation, be it local or systemic, are neutrophils (see our review in Nat Rev Immunol. 2013; 13: 159-175). These cells kill pathogens in three ways: (i) intracellularly after phagocytosis of bacteria, (ii) extracellularly upon release of antimicrobial proteins and proteases (iii) extracellularly by a release of neutrophil extracellular traps (NETs). The function of NETs is to trap and kill bacteria but their formation is also accompanied by collateral injury to the lung and the liver endothelium if NETs persist in vasculature. Formation of NETs requires activity of peptidyl arginine deiminase 4 (PAD4) and neutrophil elastase (NE). NETs are composed of DNA which serves as a backbone decorated with histones and proteins from neutrophil granules, including proteases – NE and matrix metalloproteinase-9 (MMP-9). These enzymes were previously shown to primarily facilitate leukocyte migration and cytokine/chemokine activation. However, NE and MMP-9 also degrade numerous other extracellular and intracellular substrates. Thus the question arose: do MMP-9 and/or NE attached to NETs cause injury to bystander cells during sepsis? What is the mechanism of this process?
In order to answer the above questions, we utilized a technique of intravital (in vivo) microscopy which allows to observe processes occurring in real time inside of live mice. With this technique we focused on the liver as a target organ and we studied NET formation and its impact on the course of MRSA-induced sepsis. We observed that MRSA inoculation in mice leads to profound liver damage that is accompanied by NET formation. We confirmed that indeed the two proteases, NE and MMP-9, are attached to NETs. However, while NE is proteolytically active, MMP-9 is enzymatically silent. Importantly, genetic deficiency in NE or its pharmacological blockage prevented the liver damage. Similar prevention of collateral damage to the liver was observed in PAD4-deficient mice. Furthermore, we showed that it is NE attached to NETs that causes bystander cell damage when NETs persist in the liver vasculature for hours.
DNase which dissolves the backbone DNA of NETs is routinely used in experimental studies to remove NETs, and is being tested in some pre-clinical trials. We investigated how DNase treatment will impact removal of NETs by means of intravital microscopy. When injected intravenously and imaged in real time, DNase immediately removed extracellular DNA but failed to remove NE and histones. Even when DNase was left to act for several hours, it only partially removed other NET components. In subsequent studies we revealed that, once released into the vasculature, NE and histones of NETs attach to glycoproteins, such as Von Willebrand factor, lining the surface of endothelium. For this reason, removal of NETs by DNase is not sufficient to prevent the organ damage. These novel observations are presently “in press” in Nature Communications (2015).
Results of the studies undertaken during this fellowship might have important implications for treatment of sepsis and other inflammatory disorders in which NETs are formed. Now, it is clear that specific elements of NETs cause damage to the host organs, and that, in a relevant animal model of bacterial sepsis, the prevention of NET formation by inhibition of NE and/or PAD4 is much more effective than application of DNase. We have also revealed important novel mechanisms operating once NETs are released into the vasculature in regard to their binding to proteins lining the endothelial cells.