The dose of intravenous normal saline at 20 mL/kg was selected based on our previous studies and aligned to clinical practice [29, 37]. In addition, all animals were provided with a subcutaneous reservoir of normal saline as a further precaution

against eliciting hydrodynamic differences. That this strategy was reasonably successful was indicated by our finding of a lack of significant difference among all experimental groups in two measures of dehydration: hematocrit and serum lactate. A limitation of our study was that we lacked the equipment to extend this observation to more discriminating measures, such as rodent blood pressure and vascular tone. We first compared the three resuscitation fluids in the simpler model of endotoxemia, using intraperitoneal LPS, a widely employed dose and route of administration ABC294640 order (e.g., [4, 17]). While no resuscitation fluid significantly influenced LPS-induced leukopenia or the number of adherent leukocytes in the sinusoids, AGP administration, but not that of saline or equimolar albumin in the form of HAS, clearly attenuated both leukocyte adhesion Decitabine in the PSV and blockage of sinusoids. AGP-treated mice also exhibited a reduction in average leukocyte adhesion in the sinusoids

that did not reach statistical significance. The incomplete concordance between sinusoidal blockage and sinusoidal leukocyte adherence is not surprising, given that blockage is likely an extreme example of sinusoidal narrowing, and our experimental approach did not permit measurement of overall sinusoidal flow or sinusoidal diameter. Reduced sinusoidal blood flow in sepsis and endotoxemia is derived from both leukocyte-, and platelet-mediated blockage of perfusion in the low shear environment of the sinusoids; perhaps, platelet effects, which we did not measure, predominated in this specific microvascular location. In addition, it is known that different mechanisms contribute to leukocyte adherence in the two hepatic vascular locations [30, 11]. Having demonstrated a superior

protective effect of AGP over HAS and saline in endotoxemia, we turned to ADAMTS5 the more complex but arguably more relevant CLP model, in which we focused on comparing AGP and saline. Administration of endotoxin replicates some of the clinical features of sepsis and septic shock and is consistent with the concept that it is the host response to bacteria, not the bacteria per se, that is most damaging, but only low levels of circulating endotoxin have been reported in clinical studies of septic patients [33]. The surgical CLP model provides a specific abdominal site for infection and exposes mice to a variety of bacterial danger signals [35]. Use of AGP as the resuscitation fluid in CLP demonstrated substantial overlap with the results in the endotoxemia model; its use led to better perfusion of the liver via its sinusoids, and to decreased adhesion to post-sinusoidal vessels.