Combined Administration of Recombinant TGF-β1 and DMSO Decreases the in Vitro Inflammatory Properties of Neutrophils from Sickle Cell Anemia Individuals

Sickle cell anemia (SCA) is a chronic inflammatory disease, in which activated neutrophils play a role in initiating vaso-occlusion. Two populations of circulating neutrophils have been described, denominated as low-(LDN) and high-(HDN) density neutrophils. Circulating numbers of LDN (a less inflammatory subset) are normally minimal, but this population augments under inflammatory stress, such as that seen in cancer. Transforming growth factor beta-1 (TGF-β1) is a cytokine with anti-inflammatory properties that is elevated in SCA. In conditions such as Chron's disease, TGF-β1 has protective effects, mediated by its immuno-suppressive functions. In macrophages, it is thought to trigger the polarization from pro- (M1) to anti-inflammatory phenotypes (M2), which hypothetically occur in neutrophils too (N1 and N2). Moreover, dimethyl sulfoxide (DMSO) reportedly increases TGF-β receptors expression on epithelial cells. We aimed to characterize the subsets of circulating neutrophils in SCA patients and investigate the effects of TGF-β1 and DMSO on these cells. Neutrophils from healthy (CON) and SCA individuals, in steady state and without blood transfusion for 90 days, were isolated from peripheral blood by Ficoll-Paque density gradient centrifugation. HDN and LDN were obtained from the granulocyte and mononuclear layers, respectively, and stained with CD66b for neutrophil identification by flow cytometry. As no significant effect of hydroxyurea (HU) therapy on the data obtained was observed, patients' data were grouped together irrespective of HU use. Percentages of LDN were calculated based on the total of gated CD66b⁺ cells. SCA patients had higher levels of LDN than CON (3.2±0.9%, N=7 vs 1.3±0.3%, N=13; p=0.02). We next investigated the presence of CD66b⁺/CD206^- and CD66b⁺/CD206⁺ cells, to infer the presence of N1 and N2 phenotypes, respectively. N2 were more frequent in the LDN than in the HDN subset (CON: 68.1±3.3% vs 52.0±4.4%, N=9, p=0.01; SCA: 77.6±8.9% vs 44.1±5.0%, N=3, p=0.03). To determine whether TGF-β1 and DMSO could shift HDN to a LDN profile, cells were treated (2h) with TGF-β1 (50pM) and/or DMSO (1.5%). Treatments with DMSO alone or combined with TGF-β1 increased the percentage of CD206⁺ cells in CON (45.7±2.1% vs 61.9±7.6 and 53.6±2.6% respectively, N=6, p=0.04), as well as CD206⁺ expression on each cell (mean fluorescence intensity, MFI) (137.5±16.9 MFI vs 293.6±71.2 MFI and 210.1±23.9 MFI, respectively, p=0.03). In SCA, only the combined TGF-β1/DMSO treatment increased the MFI of CD206 in HDN (115.7±10.2 vs 255.8±29.7 MFI, N=4, p=0.03). We next investigated whether TGF-β1/DMSO could reduce the adhesion of HDN to fibronectin ligand (FN, 20μg/mL) using static adhesion assays (30 min, 37ºC). HDN from CON and SCA were treated with TGF-β1 and/or DMSO (90min) and stimulated with TNF-α (200ng/mL, 30min). Although TGF-β1 alone did not reduce the adhesion of HDN to FN (p>0.05), the addition of DMSO decreased TNF-α-induced adhesion in CON (16.5±1.8% to 11.3±1.5%, p=0.03, N=10) and SCA HDN (38.9±23.9% to 13.9±1.5%, p=0.04, N=3). Subsequently, HDN were stimulated (4h) with LPS (100ng/mL) and INF-γ (20ng/mL), to induce N2 polarization, in the presence/absence of TGF-β1 and DMSO. The combined treatment again reduced adhesion in both groups (CON: 11.1±1.5% to 4.2±1.2%, p<0.01, N=4; SCA: 47.4±12.3% vs 21.9±4.5%, p=0.04, N=4). To assess whether TGF-β1 and DMSO could affect the production of proinflammatory cytokines by HDN after stimulation with LPS/INF-γ, TNF-α and IL-1β levels in cell supernatants were measured by ELISA. TGF-β1 and DMSO, in combination, decreased both TNF-α and IL-1β release from CON (TNF-α: 39.7±8.7pg/mL to 8.7±0.9pg/mL, p<0.01, N=6; IL-1β: 66.9±10.5pg/mL to 16.5±5.1pg/mL, p=0.02, N=4) and SCA HDN (TNF-α: 174.0±55.5pg/mL to 21.8±6.6pg/mL, p=0.01, N=8; IL-1β: 103.6±25.6pg/mL to 43.1±15.5pg/mL, p=0.01, N=8). Our results demonstrate for the first time the presence of elevated numbers of LDN in SCA patients, indicating an increased basal response to inflammatory stress. However, this shift in the anti-inflammatory subset does not appear enough to control inflammatory responses in the disease, and the use of agents capable of inducing this polarization may be a promising approach. Moreover, the anti-inflammatory effects of TGF-β1 on HDN seem to be enhanced by DMSO and suggest this combination as an effective modulator of the inflammatory profile of neutrophils.