Platelet serotonin promotes the recruitment of neutrophils to sites of acute inflammation in mice

Abstract

The majority of peripheral serotonin is stored in platelets, which secrete it on activation. Serotonin releases Weibel-Palade bodies (WPBs) and we asked whether absence of platelet serotonin affects neutrophil recruitment in inflammatory responses. Tryptophan hydroxylase (Tph)1–deficient mice, lacking non-neuronal serotonin, showed mild leukocytosis compared with wild-type (WT), primarily driven by an elevated neutrophil count. Despite this, 50% fewer leukocytes rolled on unstimulated mesenteric venous endothelium of Tph1(-/-) mice. The velocity of rolling leukocytes was higher in Tph1(-/-) mice, indicating fewer selectin-mediated interactions with endothelium. Stimulation of endothelium with histamine, a secretagogue of WPBs, or injection of serotonin normalized the rolling in Tph1(-/-) mice. Diminished rolling in Tph1(-/-) mice resulted in reduced firm adhesion of leukocytes after lipopolysaccharide treatment. Blocking platelet serotonin uptake with fluoxetine in WT mice reduced serum serotonin by > 80% and similarly reduced leukocyte rolling and adhesion. Four hours after inflammatory stimulation, neutrophil extravasation into lung, peritoneum, and skin wounds was reduced in Tph1(-/-) mice, whereas in vitro neutrophil chemotaxis was independent of serotonin. Survival of lipopolysaccharide-induced endotoxic shock was improved in Tph1(-/-) mice. In conclusion, platelet serotonin promotes the recruitment of neutrophils in acute inflammation, supporting an important role for platelet serotonin in innate immunity.

Figure 1

Serotonin deficiency in Tph1^−/− mice is associated with mild leukocytosis. Whole blood counts of white blood cell (WBCs, A), red blood cells (RBCs, B), platelets (C), neutrophils (D), monocytes (E), and lymphocytes (F). N = 16-18. ns indicates not significant.

Figure 2

L-selectin shedding from neutrophils in Tph1^−/− mice. (A) Surface expression of P-selectin glycoprotein ligand (PSGL)–1 (A) and L-selectin (B) in flow cytometry. (C) Plasma soluble L-selectin was determined by ELISA. N = 7-10.

Figure 3

Leukocyte rolling on unstimulated endothelium is decreased and faster in the absence of platelet serotonin. (A) Rhodamine-stained leukocyte rolling in resting mesenteric veins of WT and Tph1^−/− mice after laparotomy without further stimulation (n = 8). (B) Resting leukocyte-rolling velocity in these mice. (C) In vitro uptake of labeled serotonin after incubation of isolated platelets with fluoxetine (n = 3). (D) Serum serotonin in Tph1^−/− and WT mice after a 3-week treatment with fluoxetine or vehicle in the drinking water (n = 11). Leukocyte rolling (E) and velocity (F) in WT mice after treatment with fluoxetine for 3 weeks (n = 9). Only negligible leukocyte adhesion was observed in these experiments. Bar represents 50 μm.

Figure 4

LPS-induced leukocyte adhesion is decreased in Tph1^−/− mice. (A) Leukocyte adhesion 4 hours after intraperitoneal injection of 20 mg/kg LPS in WT and Tph1^−/− mice. (B) Leukocyte adhesion in chronically fluoxetine-treated mice 4 hours after intraperitoneal injection of LPS. N = 10. Bar represents 50 μm.

Figure 5

Fewer neutrophils extravasate in acute peritonitis, lung inflammation, and skin wounds in the absence of non-neuronal serotonin. (A) Gr-1 and CD11b-positive neutrophils in abdominal lavages of WT and Tph1^−/− mice 4 hours after intraperitoneal injection of thioglycollate (n = 10) or zymosan (n = 4). (B) Neutrophil recovery in bronchoalveolar lavage 24 hours after inhalation with LPS (n = 6). (C) Tissue MPO concentration and hematoxylin-eosin staining in 24-hour-old skin wounds (n = 10). (D) Time to closure of skin wounds (n = 10).

Figure 6

In vitro neutrophil migration and platelet-neutrophil rosetting are not affected by serotonin. (A) Migration of isolated murine neutrophils toward serotonin or LTB₄. (B) LTB₄-induced migration of neutrophils that were isolated from WT and Tph1^−/− mice. (C) The capacity to form mixed aggregates was analyzed in isolated neutrophils and a 10-fold excess of thrombin-activated platelets after a 45-minute incubation period. Rosetting was defined as 2 or more platelets binding to 1 neutrophil. N = 4.

Figure 7

Survival of LPS endotoxic shock is improved in Tph1^−/− mice. (A) Kaplan-Meier survival analysis after intraperitoneal injection of E coli serotype 055:B5 LPS into male mice (n = 14). (B) Survival after intraperitoneally injection of E coli 0111:B4 LPS (n = 12). (C) Serum TNFα levels at the indicated time points after E coli 055:B5 LPS injection (n = 3). Heart rate (D) and heart rate variability (E) were measured noninvasively in awake mice at the depicted time points after E coli 055:B5 LPS challenge (n = 6).