Inhibition of leukocyte rolling by nitric oxide during sepsis leads to reduced migration of active microbicidal neutrophils

Abstract

We developed two models of sepsis with different degrees of severity, sublethal and lethal sepsis, induced by cecal ligation and puncture. Lethal sepsis induced by cecal ligation and puncture (L-CLP) resulted in failure of neutrophil migration to the infection site and high mortality. Treatment of septic animals with aminoguanidine (AG), a nitric oxide (NO) synthase inhibitor, precluded the failure of neutrophil migration and protected the animals from death. However, cytokine-induced NO synthase (iNOS)-deficient (iNOS(-/-)) mice subjected to L-CLP did not present neutrophil migration failure, but 100% lethality occurred. iNOS(-/-) mice subjected to sublethal sepsis induced by cecal ligation and puncture (SL-CLP) also suffered high mortality despite the occurrence of neutrophil migration. This apparent paradox could be explained by the lack of microbicidal activity in neutrophils of iNOS(-/-) mice present at the infection site due to their inability to produce NO. Notably, SL- and L-CLP iNOS(-/-) mice showed high bacterial numbers in exudates. The inhibition of neutrophil migration by NO is due to inhibition of a neutrophil/endothelium adhesion mechanism, since a reduction in leukocyte rolling, adhesion, and emigration was observed in L-CLP wild-type mice. These responses were prevented by AG treatment and were not observed in the iNOS(-/-) L-CLP group. There was no significant change in L-selectin expression in neutrophils from L-CLP mice. Thus, it seems that the decrease in leukocyte rolling is due to a defect in the expression of adhesion molecules on endothelial surfaces mediated by iNOS-derived NO. In conclusion, the results indicate that despite the importance of NO in neutrophil microbicidal activity, its generation in severe sepsis reduces neutrophil migration by inhibiting leukocyte rolling and their firm adhesion to the endothelium, in effect impairing the migration of leukocytes and consequently their fundamental role in host cell defense mechanisms.

FIG. 1.

Lack of NO₂⁻ production and absence of detection of iNOS mRNA in iNOS^−/− mice. (A) Concentration of NO₂⁻ in the supernatant of macrophages stimulated with RPMI medium and LPS (200 ng ml⁻¹) plus IFN-γ (200 IU ml⁻¹) and incubated for 12, 24, and 48 h. The results were expressed as means ± SEM of quadruplicates. ∗, P < 0.05 compared with RPMI group (control) (analysis of variance, followed by Bonferroni's test). (B) iNOS mRNA products of 754 bp were obtained after amplification using total RNA from peritoneal macrophages. Lane 1, nonstimulated macrophages from wild-type mice; lanes 2 and 3, macrophages from wild-type and iNOS^−/− mice, respectively, stimulated with LPS (200 ng ml⁻¹) plus IFN-γ (200 IU ml⁻¹) for 6 h. The same result was repeated two times, and 10 wild-type and iNOS^−/− animals were used to obtain peritoneal macrophages in each experiment.

FIG. 2.

Neutrophil migration into the peritoneal cavity in wild-type and iNOS^−/− mice subjected to SL- and L-CLP. Assessment of neutrophil migration into the peritoneal cavity was performed 4 h after surgery. Wild-type mice subjected to L-CLP were treated 30 min before the surgery with saline and 30 mg of AG kg⁻¹ subcutaneously. Results are expressed as mean numbers of neutrophils per cavity ± SEM. ∗, P < 0.05 compared with sham-operated animals; #, P < 0.05 compared with wild-type SL-CLP group; ∗∗, P < 0.05 compared with wild-type L-CLP group (analysis of variance, followed by Bonferroni's test).

FIG. 3.

Survival of wild-type and iNOS^−/− mice after SL- and L-CLP. (A) The survival rates of sham-operated (n = 25), wild-type (n = 15), and iNOS^−/− SL-CLP (n = 15) mice were determined daily up to 5 days after surgery. The iNOS^−/− SL-CLP group was significantly different from sham-operated and wild-type SL-CLP mice. P < 0.05, Mantel-Cox log rank test. (B) The survival rates of wild-type mice treated subcutaneously with saline (n = 30) and AG (30 mg kg⁻¹, 30 min prior) (n = 15) and iNOS^−/− mice (n = 15), both subjected to L-CLP surgery, were determined daily up to 5 days after surgery. The wild-type and iNOS^−/− L-CLP mice were significantly different from sham-operated and wild-type L-CLP mice treated with AG. P < 0.05, Mantel-Cox log rank test. Results are expressed as percent survival.

FIG. 4.

Bacterial counts in the peritoneal fluid of wild-type and iNOS^−/− mice subjected to SL- and L-CLP. Quantification of the amount of bacteria in the peritoneal cavity was performed 4 and 24 h after SL-CLP and 4 h after L-CLP surgery. The number of bacteria present in the peritoneal cavity is expressed as mean CFU per cavity. The numbers of animals in the different experimental groups are indicated below the bars. ND, value not determined because the animals died before this time. ∗, P < 0.05 compared with wild-type SL-CLP group/4 h; #, P < 0.05 compared with iNOS^−/− SL-CLP group/4 h.

FIG. 5.

TNF-α, IL-1β, and IL-10 levels in mice subjected to CLP. The cytokine levels in peritoneal exudates were determined at 4 h after surgery in sham-operated, wild-type, and iNOS^−/− CLP mice. Results are expressed as means ± SEM, and each group had 15 mice. ∗, P < 0.05 compared with sham-operated animals; #, P < 0.05 compared with SL-CLP animals (analysis by unpaired Student's t test).

FIG. 6.

Leukocyte rolling, adherence, and migration to mesentery in wild-type and iNOS^−/− mice subjected to SL- and L-CLP. Bars show the number of rolling, adherent, and migrating leukocytes in postcapillary venules of mesentery, using an in vivo intravital microscopy assay. One group of L-CLP mice received 30 mg of AG kg⁻¹ subcutaneously 30 min before surgery. Sham-operated animals served as controls. The parameters were evaluated 3 h after the surgery. The results are expressed as mean numbers of leukocytes ± SEM, and each group had 15 mice. ∗, P < 0.05 compared with sham-operated animals; #, P < 0.05 compared with wild SL-CLP group; ∗∗, P < 0.05 compared with wild L-CLP group (analysis of variance, followed by Bonferroni's test).

FIG. 7.

Flow cytometric analysis of CD62L (L-selectin) and CD18 (β₂-integrin) expression on neutrophils obtained from sham-operated and CLP mice. Three hours after the surgery, whole blood taken from wild-type sham-operated (solid thin line), SL-CLP (dashed line), and L-CLP (thick line) mice were incubated with FITC- and PE-conjugated monoclonal antibodies. These monoclonal antibodies were used to detect the phenotypes of CD62L and CD18, respectively, in neutrophil populations, discriminated in previous experiments. Also, L-CLP mice treated subcutaneously with saline and with 30 mg of AG kg⁻¹ 30 min before surgery (dotted line) and iNOS^−/− L-CLP mice (dotted/dashed line) were analyzed. The same result was repeated two times, and 10 animals in each group were used to obtain the purified peripheral neutrophils.