Murine-recombinant leptin, rabbit anti-mouse CXCL1 antibody (catalog # 250-11), and control rabbit IgG (catalog # 500-P00) were purchased from Peprotech, Inc. (Rocky Hill, NJ, USA). Rapamycin was obtained from Sigma-Aldrich, Inc. (Saint Louis, MO, USA). Zileuton was obtained from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). U-75302 was purchased from Cayman Chemical (Ann Arbor, MI, USA). Osmium tetroxide was provided from Ted Pella, Inc. (Redding, CA, USA).

We used male mice of different strains: C57Bl/6, C3H/HeJ, C3H/He, 5-lipoxygenase (5-LO)-deficient (5-LO^−/−), CCL3-deficient (CCL3^−/−), tumor necrosis factor receptor 1 (TNFR1-deficient) (TNFR1^−/−), and PI3Kγ-deficient (PI3Kγ^−/−) mice and respective wild types (WTs) (5-LO^+/+, CCL3^+/+, TNFR1^+/+, and PI3Kγ^+/+), obtained as previously described (17–21). Mice were obtained from the FIOCRUZ breeding unit, as well as raised and maintained under the same housing conditions. All animal care and experimental protocols were conducted following the guidelines of the Brazilian Council for Care and Use of Experimentation Animals (CONCEA). The Oswaldo Cruz Institute Animal Welfare Committee (CEUA-IOC license number L-011/2015) approved all protocols used in this study.

The in vivo treatments were performed as previously described (11). Briefly, following the intraperitoneal (i.p.) administration of leptin (0.25, 0.5, 1, and 2 mg/kg, depending on the experiment) or vehicle (sterile, apirogenic saline), animals were euthanized at different time points (1, 6, or 24 h, as specified in each experiment). Alternatively, animals received three i.p. injections of rapamycin (12.5 μg/kg), or vehicle, 12 h before, 15 min before, and 12 h after the injection of leptin or saline, and the peritoneal lavage was harvested after 24 h. This treatment was established by us and was proved to be effective for the inhibition of leptin-induced lipid droplets in peritoneal macrophages (11). We also evaluated the effect of i.p. pretreatments, 15 min prior to leptin treatment, with the phospholipase A₂ inhibitor, Zileuton (60 μg/cavity), or the LTB₄ receptor BLT1-specific inhibitor, U-75302 (5 mg/kg). These drugs were administered according to data from previous works from our group and others (22–28). To block CXCL1, the antibody against CXCL1 (3 μg/animal) or the isotype control (diluted in sterile saline) was injected into the peritoneal cavity, 10 min before the leptin injection. After the time specified in each experiment, the peritoneal cells were harvested as follows. The peritoneal cavity was rinsed with HBSS (Hank’s balanced salted solution, 3 mL/cavity), and a volume of approximately 2.5 mL was recovered. Samples were diluted in Turk fluid (2% acetic acid) for total leukocyte counts using Neubauer chambers. Differential leukocyte counting was performed in cytospin smears stained by May–Grünwald–Giemsa, a classical staining for the differential identification of leukocytes (mononuclear cells, neutrophils, and eosinophils) (29). As a control of vascular integrity, we assured that leptin injection does not modify peritoneal lavage protein concentrations There is no difference in the total protein concentration between the samples at 6 h after injection of saline (0.754 ± 0.032 mg/mL) or leptin (0.729 ± 0.048 mg/mL), or at 24 h after injection of saline (0.776 ± 0.014 mg/mL) or leptin (0.749 ± 0.037 mg/mL).

The peritoneal cells including the macrophages were obtained from naïve C57Bl/6 mice, by peritoneal lavage with HBSS (5 mL). Cells were transferred to polypropylene tubes (1 × 10⁶ cells/mL) and then incubated with leptin (20 nM) for 4 h, under 37°C, 5% CO₂ atmosphere in RPMI 1604 medium. After incubation, tubes were centrifuged for supernatant collection. Cell viability was always >85% as determined by Trypan blue exclusion. For the in vitro incubation of neutrophils, murine bone marrow neutrophils were obtained as previously described (30). Briefly, the bone marrow from femurs and tibias was washed, and neutrophils were purified in a discontinuous Percoll gradient. Neutrophil viability was greater than 95% as assessed by the Trypan blue exclusion test, and purity was greater than 98% as analyzed by microscopy using Hemacolor staining (Merck, Darmstadt, Germany).

Peritoneal lavage or in vitro supernatants were analyzed for TNFα, CXCL1/KC, and LTB₄. TNFα and CXCL1 ELISA were performed with Duo Set kit, according to manufacturer’s protocol (R&D Systems, Minneapolis, MN, USA) and LTB₄ with enzyme immuno assay (EIA) kit (Cayman Chemical, Ann Arbor, MI, USA). For the investigation of the presence of intracellular LTB₄, we used the previously described Eicosacell protocol (31). Briefly, leukocytes were recovered from peritoneal cavities and immediately submitted to fixation and permeabilization with 0.5% 1-ethyl-3-(3-dimethylamino-propyl) carbodiimide (Sigma) in HBSS. After that, a common immunodetection protocol was performed using the following antibodies: the primary antibody anti-LTB4 (Cayman Chemical) or irrelevant IgG and the secondary antibody Alexa 488-labeled anti-rabbit IgG. Images were obtained using an Olympus BX51 fluorescence microscope and equipped with a Plan Apo ×100 objective and a DP72 camera (Olympus Optical, Japan) in conjunction with Cell^F Imaging Software (Olympus Life Science Europe, Germany).

Data were reported as the mean ± standard error of the mean (SEM). Data were statistically analyzed by the analysis of variance with Newman–Keuls–Student test, or Student’s t-test. Differences were considered to be significant when p ≤ 0.05 (see figure legends).

Leptin deficiency is associated with impaired cell-mediated immunity and increased susceptibility to infection (1), and neutrophil chemotaxis is a key component of inflammation and host response to infection. There are very few studies though addressing leptin-inflammatory effects in vivo, and the effects of leptin are confounded with the effects of other immunometabolism-modulating factors that are altered in diseases such as infection or obesity (32). To investigate the specific effect of leptin on neutrophil recruitment, here, neutrophil influx was evaluated at different time points after i.p. leptin (1 mg/kg) administration. Leptin-induced neutrophil accumulation in the peritoneal cavity of C57BL/6 mice was significant within 1 h, maximal within 6 h, and sustained for at least until 24 h (Figure 1A). This effect was observed at different leptin concentrations (0.5, 1, and 2 mg/kg) for 24 h (Figure 1B). It is known that after bacterial or LPS stimulation, activated neutrophils present enhanced lipid body numbers, which is considered characteristic of their activation (33–35). In our previous work, the enhancement of lipid droplets in leptin-stimulated macrophages was described (11). However, we did not observe the enhancement of lipid droplets above the basal levels in peritoneal neutrophils after leptin treatment (Figure S1 in Supplementary Material). The i.p. injection of leptin in C3H/HeJ LPS-resistant mice led to significant neutrophil accumulation in the peritoneal cavity within 24 h (Figure 1C). Leptin samples were confirmed negative for LPS contamination by LAL testing (<0.01 UI); therefore, our data indicate that LPS is not involved in the observed leptin response. Moreover, we confirmed the absence of a direct effect of leptin on neutrophils by in vitro neutrophil adhesion experiments (Figure S2 in Supplementary Material). Even with very high leptin concentrations (up to 200 nM), no adhesion was observed. In our preparation, we ensured >98% purity of neutrophils as previously described (30).

It had been previously demonstrated that leptin indirectly activates human neutrophils (10). This work showed that neutrophil CD11b expression can be enhanced in response to TNFα produced by leptin-stimulated monocytes in vitro. We confirmed that leptin induced TNFα production by the peritoneal cells in vitro (Figure 2A). Moreover, leptin injection into the peritoneal cavity induced in vivo TNFα production (Figure 2B). In order to investigate the role of TNFα receptor on leptin-induced neutrophil recruitment in vivo, TNFR1 knockout mice (TNFR1^−/−) or WT (TNFR1^+/+) animals (C57Bl/6) were used. As demonstrated in Figure 2C, leptin was unable to induce neutrophil accumulation in TNFR1^−/− mice, but interestingly, TNFR1 is not necessary for the effect of leptin in inducing lipid droplets in macrophages (Figure S3 in Supplementary Material). Leptin induced lipid body formation in peritoneal macrophages from TNFR1^−/− mice, similar to what was observed in WT (TNFR1^+/+ mice). Thus, leptin-induced neutrophil migration is dependent on TNFR1 signaling, but macrophage lipid body formation occurs independently of this TNFα receptor.

Several studies have demonstrated the importance of PI3K, specifically PI3Kγ, in leukocyte migration and activation in vivo (19, 36–39). To determine the role of PI3Kγ in the acute response induced by leptin, we examined neutrophil in WT mice or mice genetically deficient in PI3Kγ (PI3Kγ^−/−), 24 h after leptin i.p. administration. Leptin-induced accumulation of neutrophils in the peritoneal cavity was completely inhibited in PI3Kγ^−/− mice in comparison to that in WT control animals (Figure 3A). Thus, the PI3Kγ-signaling pathway is required for leptin-induced neutrophil recruitment.

We showed previously that mTOR was required for the effects of leptin on lipid droplets formation in macrophages both in vitro and in vivo (11). Others reported that the mTOR pathway was important for GM-CSF-dependent neutrophil activation in vitro (40). We investigated here the functional role of the mTOR pathway for the in vivo leptin-induced neutrophil activation and recruitment. We used the previously described treatment with rapamycin, a specific inhibitor and probe for mTOR activity (11, 41). The treatment of mice with rapamycin failed to modify leptin-induced neutrophil accumulation after 24 h (Figure 3B).

It had been shown that the lipid mediator LTB₄ directly exerts a chemoattractive effect on neutrophils in inflammatory conditions (42, 43). We performed several in vivo experiments in order to evaluate the role of LTB₄ in neutrophil migration stimulated by leptin. First, an EIA assay was performed to detect LTB₄ presence on peritoneal supernatants obtained after leptin i.p. injection. There was no difference in the concentration of LTB₄ between saline and 1-, 6-, or 24-h leptin-stimulated groups (Figure 4A). It could be that LTB₄ was produced but not released from the cell; therefore, we performed the Eicosacell assay. We did not observe intracellularly retained LTB₄ in the peritoneal cells submitted to the Eicosacell assay as performed before by our group (data not shown) (31). We also investigated the importance of the LTB₄ synthesis enzyme, 5-LO for neutrophil migration. Therefore, 5-LO^+/+ or 5-LO^−/− mice were injected with leptin in the peritoneal cavity (1 mg/kg), and neutrophil migration was investigated. As shown in Figure 4B, leptin induced neutrophil migration (24 h) in 5-LO knockout mice. The pharmacological inhibition of LTB₄ synthesis by Zileuton (60 µg/animal i.p.) and signaling on the receptor BTL1 by U-75302 (5 mg/kg i.p.) were tested. Figure 4C shows that peritoneal neutrophils migrate in response to leptin injection, despite pharmacological pretreatments. These results demonstrate that LTB₄ does not participate in leptin-induced neutrophil recruitment.

CXCL1 is known to mediate the neutrophil migration through the activation of CXCR2, a PI3Kγ-dependent receptor (44). Neutrophil migration can also be promoted by CCL3/MIP-1α signaling through CCR1, a PI3Kγ-dependent receptor (45). Because our result in Figure 3A showed that leptin-induced neutrophil recruitment is dependent on PI3Kγ, we investigated the chemokines CXCL1 and CCL3 as possible mediators for this effect. We observed that CXCL1 is secreted by the peritoneal cells stimulated by leptin both in vitro and in vivo (Figures 5A,B). Anti-CXCL1-neutralizing antibodies were used in vivo as a pretreatment before leptin injection. A partial but significant inhibition of neutrophil migration to peritoneal cavity was observed (Figure 5C). We investigated the participation of the chemokine CCL3 using the CCL3 knockout mice (CCL3^−/−). Mice were injected with leptin i.p. and, as seen in Figure 5D, neutrophil recruitment was not altered when compared to WT mice (CCL3^+/+). Taken together, our data show that leptin activates the peritoneal cells to induce neutrophil migration in a specific CXCL1-dependent manner.