BACKGROUND:

Mechanical ventilation (MV) is a clinically important measure for respiratory support in critically ill patients. Although moderate tidal volume MV does not cause lung injury, it can further exacerbate lung injury in pathological state such as sepsis. This pathological process is known as the 'two-hit' theory, whereby an initial lung injury (e.g., infection, trauma, or sepsis) triggers an inflammatory response that activates immune cells, presenting the lung tissue in a fragile state and rendering it more susceptible to subsequent injury. The second hit occurs when mechanical ventilation is applied to lung tissue in a fragile state, and it is noteworthy that this mechanical ventilation is harmless to healthy lung tissue, further aggravating pre-existing lung injury through unknown mechanisms. This interaction between initial injury and subsequent mechanical ventilation develops a malignant cycle significantly exacerbating lung injury and severely hampering patient prognosis. The two-hit theory is critical to understanding the complicated mechanisms of ventilator-associated lung injury and facilitates the subsequent development of targeted therapeutic strategies. METHODS AND

RESULTS:

CLP mice model was used to mimic clinical sepsis patients. After 12 hours the mice were mechanical ventilated for 2-6 hours. MV by itself didn't lead to HMGB1 release, but significantly strengthened HMGB1 in plasma and cytoplasm of lung tissue in septic mice. Plasma and lung tissue activation of cytokines and chemokines, MAPK signaling pathway, neutrophil recruitment, and ALI were progressively decreased in LysM HMGB1-/- (Hmgb1 deletion in myeloid cells) and iHMGB1-/- mice (inducible HMGB1-/- mouse strain where the Hmgb1 gene was globally deleted after tamoxifen treatment). Compared with C57BL/6 mice, although EC-HMGB1-/- (Hmgb1 deletion in endothelial cells) mice didn't have lower levels of inflammation, neutrophil recruitment and lung injury were reduced. Compared with LysM HMGB1-/- mice, EC-HMGB1-/- mice had higher levels of inflammation but significantly lower neutrophil recruitment and lung injury. Overall, iHMGB1-/- mice had the lowest levels of all the above indicators. The level of inflammation, neutrophil recruitment and the degree of lung injury were decreased in RAGE-/- mice, and even the above indices were further decreased in TLR4/RAGE-/- mice. Levels of inflammation and neutrophil recruitment were decreased in Caspase-11-/- and Caspase-1/11-/- mice, but no statistical difference between these two gene knockout mice.

CONCLUSIONS:

These data show for the first time that the Caspase-1/Caspase-11-HMGB1-TLR4/RAGE signaling pathway plays a key role in mice model of sepsis induced lung injury exacerbated by MV. Different species of HMGB1 knockout mice have different lung protective mechanisms in the 'two hits' model, and location is the key to function. Specifically, LysM HMGB1-/- mice due to the deletion of HMGB1 in myeloid cells resulted in a pulmonary protective mechanism that was associated with a downregulation of the inflammatory response. EC HMGB1-/- mice are deficient in HMGB1 owing to endothelial cells, resulting in a distinct pulmonary protective mechanism independent of the inflammatory response and more relevant to the improvement of alveolar-capillary permeability. iHMGB1-/- mice, which are systemically HMGB1-deficient, share both of these lung-protective mechanisms.