Lipopolysaccharide (LPS), 5-HT, fetal bovine serum (FBS), bovine serum albumin (BSA), and Tyrode’s solution were obtained from Sigma-Aldrich (MO, United States). ELISA kits for tumor necrosis factor alpha (TNF-α) and interleukin 6 (IL-6) were obtained from Qiaoyi (Shanghai, China). HBSS and RPMI 1640 medium were obtained from Life Technologies (CA, United States). APC-Cy7-labeled anti-LY6G antibody and FITC-labeled anti-CD41 antibody were obtained from BD (NJ, United States). APC-labeled anti-CD62p antibody (p-selectin) was obtained from BioLegend (San Diego, United States). Rabbit anti-mouse citrullinated histone H3 (CitH3) was obtained from Abcam (Cambridge, United Kingdom). Rabbit anti-mouse Ly6G, rabbit anti-mouse CD41, and related fluorescent secondary antibodies were obtained from Cell Signaling Technology (MA, United States).

Wild-type (WT) and Tph1 knockout (Tph1 ^−/− ) C57BL/6 male mice (6–8 weeks, body weight 20 ± 2 g) were purchased from the Model Animal Research Center of Nanjing University (Nanjing, Jiangsu, China) and raised in the Experimental Animal Center of Jiangsu University (Zhenjiang, Jiangsu, China). All mice were randomly divided into four groups, including the WT and Tph1 ^−/− sham groups that received sham surgery as well as the WT and Tph1 ^−/− cecal ligation and puncture (CLP) groups that underwent CLP surgery. During CLP surgery, the mice were anesthetized with sevoflurane inhalation, the abdomen was disinfected with alcohol, and a 1-cm midline abdominal incision was made to expose the cecum of the mice. The end 1/3 of the cecum was ligated with 3–0 nylon thread, the end cecum was punctured with a 21-gauge needle, and an appropriate amount of intestinal content spilled out from the perforation. After the treatment, the cecum was reset, and the abdominal incision of the mice was sutured. Sham surgery requires only the cecum to be properly turned over and exposed, and then, the cecum is reset and the abdominal incision is sutured layer by layer. For survival analysis, in the sham groups (WT and Tph1 ^−/− ), each had six mice, while in CLP groups (WT and Tph1 ^−/− ), each had 25 mice. For other experiments, there were four mice in each group. The experimental protocol on animal protection and welfare was approved by the Council on Animal Care and Use at Jiangsu University.

Neutrophils were isolated from WT mice bone marrow by using density gradient centrifugation as previously described (Boxio et al., 2004; Szatmary et al., 2017). WT mice were sacrificed. Bone marrow was extracted from the femurs and tibias of the mice. Marrow cells were harvested from the bone marrow with a 70-µm cell strainer. Then, the cells were pelleted in a centrifuge and resuspended in HBSS. Following erythrocyte lysis, the neutrophil-containing solution was placed onto a discontinuous Percoll gradient solution (Percoll solution diluted to 78%, 69%, and 52% in HBSS), and the gradient was centrifuged at 1,500g at room temperature for 30 min. Neutrophils were collected from a band between the 78% and 69% layers. Finally, the neutrophils were resuspended in RPMI 1640 with 1% heat-inactivated FBS.

Platelets were isolated from WT and Tph1 ^−/− mice peripheral blood as previously described (Liu et al., 2013). Mice peripheral blood was collected into a vacuum-anticoagulated tube containing trisodium citrate, and Tyrode’s solution was added. Platelet-rich plasma were isolated by centrifuging at 180×g at room temperature for 10 min, and platelets were isolated by centrifuging at 1,250×g at room temperature for 10 min. Then, platelets were washed and resuspended in Tyrode’s solution for at least 1 h at 37°C before use.

The purity (>97%) of cells was detected and adjusted by flow cytometry and an APC-Cy7-labeled anti-LY6G antibody (neutrophil) and FITC-labeled anti-CD41 antibody (platelet). The cell concentration was maintained at 1 × 10⁶/ ml at every experiment. The activity of platelets was detected by flow cytometry and an APC-labeled anti-CD62p antibody.

WT neutrophils were randomly divided into three groups, including the WT neutrophil group, WT platelet and WT neutrophil co-culture group, and Tph1 ^−/− platelet and WT neutrophil co-culture group. In each group, cells were randomly divided into four groups, with each containing 4 × 10⁵ cells, including a control group, which received no intervention; an LPS group, which received LPS (1 μg/ ml) stimulation; a 5-HT group, which received 5-HT (100 μM) intervention; and an LPS + 5-HT group, which received 5-HT (100 μM) intervention and LPS (1 μg/ ml) stimulation. Cells were incubated in a 5% CO₂ incubator at 37°C and 95% humidity. Twelve hours later, cells were harvested for subsequent experiments.

Immunofluorescence was used to detect the NET formation in the lung of mice and in vitro cultured cells. Lung tissues were sequentially fixed by 4% paraformaldehyde, embedded in paraffin, sectioned, and permeabilized with 0.05% Triton X-100. Cells were harvested, fixed, and permeabilized. Then, both lung and cell specimens were stained with primary antibodies (1:200), including citH3, Ly6G, and/or CD41, overnight at 4°C. Fluorescent secondary antibodies (1:200) were added and incubated at room temperature for 1 h in the dark. DAPI (1 μm/ ml) was incubated in the dark for 15 min. Finally, NET formation was detected by using an inverted phase-contrast microscope (×400 magnification). NETs expression was calculated based on CitH3 staining.

Lung tissue specimens (approximately 0.4 g) were harvested 12 h after surgery and fixed in 10% formalin. Then, the fixed specimens were sequentially embedded in paraffin, sectioned, and stained with hematoxylin/eosin (HE). A light microscope (×400 magnification) was applied to examine alveolar structure, cellular edema, and granulocyte infiltration in lung specimens.

Mice were sacrificed, and endotracheal intubation was performed 12 h after surgery. Bronchoalveolar lavage fluid (BALF) was collected from the endotracheal tube by lavage with PBS without Ca²⁺ and Mg²⁺. ELISA kits were used to detect the level of TNF-α and IL-6.

A total of 62 mice were randomly divided into four groups: WT sham group (n = 6), WT CLP group (n = 25), Tph1 ^−/− sham group (n = 6), and Tph1 ^−/− CLP group (n = 25). All mice were raised in the same environment, with no restrictions on their food or water intake. The mice were monitored every 6 h for 72 h.

Statistical analyses were performed with GraphPad Prism version 9.0 (United States). One-way analysis of variance was used for comparison between multiple groups, t-test was used for comparison between two groups, and Dunnett’s test was used for post-hoc analysis comparison. All data are presented as mean ± SD. Survival rate analysis was performed using the Kaplan–Meier method. A p value <0.05 was considered to be statistically significant.

Peripheral 5-HT, a monoamine neurotransmitter that is mainly synthesized by TPH1, can promote the inflammatory response by activating immune cells and increasing inflammatory cytokine release (Duerschmied et al., 2013; Li et al., 2016; Zhang et al., 2017). TPH1 is a rate-limiting enzyme for the synthesis of peripheral 5-HT (Zhang et al., 2020). In order to explore the role of peripheral 5-HT in sepsis, Tph1 ^−/− mice were constructed and CLP was performed to induce sepsis. As shown in Supplementary Figure S1, the 5-HT concentration was markedly reduced in platelets of Tph1 ^−/− mice, which confirmed the functional deficiency of TPH1.

The effect of peripheral 5-HT on the survival rate of septic mice was examined postoperatively. Results showed that peripheral 5-HT deficiency significantly increased the survival of septic mice (Figure 1A). Further study showed that in the Tph1 ^−/− CLP mice, alveolar structure injury was less severe and the extent of alveolar wall granulocyte infiltration was reduced (Figures 1B, C). Meanwhile, we found that Tph1 ^−/− CLP mice had a minor increase in BALF TNF-α and IL-6. The differences were statistically significant when compared to WT CLP mice (Figures 1D, E). These results suggested that peripheral 5-HT deficiency improves the survival rate of septic mice and alleviates lung inflammatory injury.

Uncontrolled NET formation is reported to be a major contributor to sepsis-induced ALI and acute respiratory distress syndrome (ARDS) (Jhelum et al., 2018; Wang et al., 2020; Yaqinuddin and Kashir, 2020). Lefrancais et al. (2018) showed that inhibiting NET formation could reduce lung injury and improve mice survival. Lung specimens were collected, and immunofluorescence studies were performed to determine whether peripheral 5-HT deficiency affected NET formation. Results showed that lung tissues from sham mice had no NET formation. However, lung tissues from WT CLP mice had significantly enhanced NET formation, as indicated by the staining of CitH3. Interestingly, we observed that in the lung tissues of Tph1 ^−/− CLP mice, NET formation was significantly reduced (Figure 2). This result indicates that peripheral 5-HT deficiency can reduce NET formation in the lung of septic mice.

Peripheral 5-HT is mainly stored in dense granules of platelets and released into the plasma during platelet activation (Mohammad-Zadeh et al., 2008). In order to investigate whether platelet-derived 5-HT is associated with NET formation, lung specimens were collected and co-stained with anti-CitH3 (labeled NETs) and anti-CD41 (labeled platelets). Results showed that the lungs from sham mice did not show NET formation, while the lungs from WT CLP mice showed a mass of the accumulation of platelets and NETs. Interestingly, the lungs from Tph1 ^−/− CLP mice showed a significant decrease of the accumulation of platelets and NETs (Figure 3). We conclude that peripheral 5-HT deficiency can reduce the accumulation of platelets and NETs in the lung of septic mice.

Although accumulating evidences reveal the role of platelets in NET formation, its mechanisms is unknown. To address this issue, neutrophils were isolated from WT mice and stimulated by LPS with or without exogenous 5-HT in vitro. NET generation was detected by staining with anti-CitH3 and DAPI. We noticed that LPS stimulation induced NET formation, while exogenous 5-HT intervention did not increase the LPS-induced NET formation in WT neutrophils cultured alone (Figures 4A, D). Recent studies showed that in bacterial sepsis, LPS induces non-classical activation of platelets (Clark et al., 2007; Carestia et al., 2016; Martinod and Deppermann, 2021). Consistent with the above studies, our mean fluorescence intensity (MFI) result showed that P-selectin (also called CD62p—a platelet activation marker) (van Velzen et al., 2012) was highly expressed in LPS-stimulated platelets. Furthermore, the MFI of P-selectin in WT platelets was higher than that in Tph1 ^−/− platelets (Supplementary Figure S2). To further investigate whether platelet autocrine 5-HT accounted for NET formation, LPS was used to activate platelets. Platelets were isolated from WT or Tph1 ^−/− mice and co-incubated with the WT neutrophils. Interestingly, results showed that WT platelets promoted LPS-induced NET formation (Figures 4B, E), while Tph1 ^−/− platelets did not show a promotive effect (Figures 4C, F). However, exogenous 5-HT intervention remedied the deficiency of Tph1 ^−/− platelets and promoted LPS-induced NET formation (Figures 4C, F). In light of the results shown above, we conclude that activated platelets promote NET formation through autocrine 5-HT signaling.