Resistin production does not affect outcomes in a mouse model of acute surgical sepsis

Abstract

Introduction: Because of the strong correlation between the blood concentration of circulating resistin and the illness severity of septic patients, resistin has been proposed as a mediator of sepsis pathophysiology. In vitro data indicate that human resistin directly impairs neutrophil migration and intracellular bacterial killing, although the significance of these findings in vivo remain unclear.

Objective: The objectives of the present study were: (1) to validate the expression of human resistin in a clinically relevant, murine model of surgical sepsis, (2) to assess how sepsis-induced changes in resistin correlate with markers of infection and organ dysfunction, and (3) to investigate whether the expression of human resistin alters immune function or disease outcomes in vivo.

Methods: 107 male, C57BL/6 mice transgenic for the human resistin gene and its promoter elements (Retn+/-/-, or Retn+) were generated on a Retn-/- (mouse resistin knockout, or Rko) background. Outcomes were compared between age-matched transgenic and knockout mice. Acute sepsis was defined as the initial 24 h following cecal ligation and puncture (CLP). Physiologic and laboratory parameters correlating to the human Sequential Organ Failure Assessment (SOFA) Score were measured in mice, and innate immune cell number/function in the blood and peritoneal cavity were assessed.

Results: CLP significantly increased circulating levels of human resistin. The severity of sepsis-induced leukopenia was comparable between Retn+ and Rko mice. Resistin was associated with increased production of neutrophil reactive oxygen species, a decrease in circulating neutrophils at 6 h and an increase in peritoneal Ly6Chi monocytes at 6 h and 24 h post-sepsis. However, intraperitoneal bacterial growth, organ dysfunction and mouse survival did not differ with resistin production in septic mice.

Significance: Ex vivo resistin-induced impairment of neutrophil function do not appear to translate to increased sepsis severity or poorer outcomes in vivo following CLP.

Fig 1. Serum concentrations of human resistin in hRetn-/- mice (Rko) and in hRetn-/- mice transgenic for the human resistin gene (Retn+).

Concentrations were measured at 6 h and 24 h post-surgery. n = 3 for Rko with acute sepsis at 6 h and Retn+ with acute sepsis at 24 h; n = 4 for Rko acute sepsis and control at 24 h; n = 6 for Retn+ acute sepsis and control at 6 h; n = 6 for Retn+ control at 24 h; n = 7 for Rko control at 6 h.

Fig 2. Complete blood count and leukocyte differential analysis at 6 h and 24 h post-surgery, in resistin knockout C57BL/6 mice (Rko) and transgenic C57BL/6 mice on an Rko background (Retn+).

A—E represent mean (+/- SEM) serum concentrations of the named parameters following surgery. For A-C, n = 6 per time point per group. For D and E, n = 5–7 and results reflect cell staining followed by manual cell differential count. P values represent statistically significant results of ordinary 3-way ANOVA (time x resistin genotype x sepsis) within each cell type.

Fig 3. Flow cytometry assessment of cellular differential at 6 h and 24 h post-cecal ligation and puncture, in knockout C57BL/6 mice (Rko) and mice expressing human resistin (Retn+).

A, Blood neutrophils at 6 h; B, Blood monocytes and T lymphocytes at 6 h; C, Peritoneal neutrophils at 6 h; D, Peritoneal monocytes, macrophages and T lymphocytes at 6 h; E, Blood neutrophils at 24 h; F, Blood monocytes and T lymphocytes at 24 h; G, Peritoneal neutrophils at 24 h; H, Peritoneal monocytes, macrophages and T lymphocytes at 24 h. n = 4 for Rko at 6 h and 24 h, and for Retn+ at 24 h. n = 3 for Retn+ at 6 h.

Fig 4. Serum concentrations of human cytokines in hRetn-/- mice (Rko) and mice expressing human resistin (Retn+).

Concentrations were measured at 6 h and 24 h post-CLP. n = 4 for Retn+ at 6 h and 24 h, n = 2 for Rko at 6 h and n = 4 for Rko at 24 h. P values for statistically significant relationships on ordinary 3-way ANOVA (time versus sepsis effect versus genotype) are shown.

Fig 5. Outcomes for knockout mice (Rko) and those producing human resistin (Retn+) in acute sepsis induced by CLP.

A, Kaplan-Meier curve following acute sepsis (time = 0 h), with n = 14 for Retn+ of whom 11 underwent CLP and 3 underwent sham surgery; n = 18 for Rko of whom 13 underwent CLP and 5 underwent sham surgery. High apparent mortality in sham-operated Retn+ is attributed to one mortality in a group with n = 3, which we surmise to be an outlier. B, Peritoneal bacterial load in relation to expression of human resistin. There was no bacterial overgrowth in Rko as compared with Retn+ mice with sepsis. n = 4 for Rko and n = 3 for Retn at 6 h; n = 9 for Rko and Retn+ at 24 h. C, Change in fluorescence intensity between neutrophils stimulated with PMA and unstimulated cells following exposure to CellROX reagent at 24 h following CLP (n = 6 for Rko and Retn+ mice). D. Change in fluorescence intensity between neutrophils stimulated with LPS and unstimulated cells following exposure to CellROX reagent at 24 h following CLP (n = 6 for Rko and Retn+ mice).

Fig 6. Physiologic parameters at 0 h, 6 h and 24 h post-surgery in mice lacking resistin (Rko) and in mice producing human resistin (Retn+).

A, Rectal temperature, B, Mouse Severity of Sepsis Score; C, Breath Rate; D, Oxygen saturation by pulse oximetry. N = 10–25 per group. Box and violin plot line at mean value for each group.