Hemoadsorption reprograms inflammation in experimental gram-negative septic peritonitis: insights from in vivo and in silico studies

Abstract

Improper compartmentalization of the inflammatory response leads to systemic inflammation in sepsis. Hemoadsorption (HA) is an emerging approach to modulate sepsis-induced inflammation. We sought to define the effects of HA on inflammatory compartmentalization in Escherichia coli-induced fibrin peritonitis in rats.

Hypothesis: HA both reprograms and recompartmentalizes inflammation in sepsis. Sprague Dawley male rats were subjected to E. coli peritonitis and, after 24 h, were randomized to HA or sham treatment (sepsis alone). Venous blood samples collected at 0, 1, 3 and 6 h (that is, 24-30 h of total experimental sepsis), and peritoneal samples collected at 0 and 6 h, were assayed for 14 cytokines along with NO(2)(-/)NO(3)(-). Bacterial counts were assessed in the peritoneal fluid at 0 and 6 h. Plasma tumor necrosis factor (TNF)-α, interleukin (IL)-6, CXCL-1, and CCL2 were significantly reduced in HA versus sham. Principal component analysis (PCA) suggested that inflammation in sham was driven by IL-6 and TNF-α, whereas HA-associated inflammation was driven primarily by TNF-α, CXCL-1, IL-10 and CCL2. Whereas -peritoneal bacterial counts, plasma aspartate transaminase levels and peritoneal IL-5, IL-6, IL-18, interferon (IFN)-γ and NO(2)(-)/NO(3)(-) were significantly lower, both CXCL-1 and CCL2 as well as the peritoneal-to-plasma ratios of TNF-α, CXCL-1 and CCL2 were significantly higher in HA versus sham, suggesting that HA-induced inflammatory recompartmentalization leads to the different inflammatory drivers discerned in part by PCA. In conclusion, this study demonstrates the utility of combined in vivo/in silico methods and suggests that HA exerts differential effects on mediator gradients between local and systemic compartments that ultimately benefit the host.

Figures

Figure 1

Effects of HA on circulating levels of inflammatory mediators. Postcannulation baseline values (that is, 24 h after implantation of an E. coli–impregnated fibrin clot and hence established sepsis) for all cytokines were not significantly different between the two experimental groups. Circulating TNF-α (A), IL-6 (B), CCL2 (C) and CXCL-1 (D) were significantly lower in the HA group versus the sham group (*p < 0.05 versus sham group; †p <0.05 versus HA at 0 h). Circulating NO2−/NO3− after 6 h of treatment was significantly lower (F) (*p < 0.001 versus sham; †p < 0.05 versus HA at 0 h). Note: Some error bars are not visible, since they are overlapped by the black symbol.

Figure 2

Rats were subjected to sepsis by the implantation of a fibrin clot impregnated with E. coli as descibed in the Materials and Methods. PCA suggests different drivers of inflammation in HA versus sham. Principal component analysis suggested that the circulating inflammatory response in the sham group was primarily driven by IL-6 and TNF-α (A), whereas the response to HA was primarily driven by TNF-α, CXCL-1, IL-10 and CCL2 (B).

Figure 3

Effects of HA on peritoneal inflammatory mediators and bacteria. Rats were subjected to sepsis by the implantation of a fibrin clot impregnated with E. coli followed by sham or HA treatment, and peritoneal fluid was obtained as descibed in the Materials and Methods. Peritoneal IL-5 (A), IL-6 (B), IL-18 (C), IFN-γ (D) and NO2−/NO3− (E) were significantly lower (n = 6 rats; *p < 0.05 versus sham group; †p < 0.05 versus baseline). CCL2 (F) and CXCL-1 (G) concentrations were significantly higher (n = 6 rats; *p < 0.05 versus sham group; †p < 0.05 versus baseline).

Figure 4

Ratios of local (peritoneal) mediators to systemic concentrations at 6 h for all the mediators that were measured in both sites. Rats were subjected to sepsis by the implantation of a fibrin clot impregnated with E. coli followed by sham or HA treatment, and plasma as well as peritoneal fluid was obtained as described in the Materials and Methods. Animals in the HA group had significantly higher localized TNF-α (*p = 0.0010 versus sham) (A), IL-10 (*p = 0.040 versus sham) (B), CXCL-1 (*p = 0.001 versus sham) (C), CCL2 (*p = 0.001 versus sham) (D) and IL-6 (*p = 0.029 versus sham) (E). The sham group had significantly higher ratios of IL-18 (*p = 0.001 versus HA) (F), IL-1β (*p = 0.003 versus HA) (G), IL-5 (*p = 0.001 versus HA) (H) and GM-CSF (*p = 0.005 versus HA) (I). There was no significant difference in the peritoneal/plasma ratios of IL-1α and IFN-γ.

Figure 5

Peritoneal bacterial count in HA and sham groups at 6 h versus the baseline at 0 h. Rats were subjected to sepsis by the implantation of a fibrin clot impregnated with E. coli followed by sham or HA treatment, and peritoneal fluid was obtained as described in the Materials and Methods. Peritoneal bacterial counts were significantly lower (n = 8 rats; *P < 0.001 versus sham) in the HA group (7 × 108 ± 3.0 × 107 CFUs/mL) when compared with the sham group (9.4 × 108 ± 5.3 × 107 CFUs/mL).