Abnormal Cytokine Production By Mast Cell Cultures from Sickle Cell Anemia Patients in Response to Inflammatory Stimuli and to Co-Culture with Eosinophils

Mast cell function (MC) in pathologic states can be studied through their ability to secrete mediators in vitro depending on MC phenotype and the nature of the stimuli. Reports on MC mediators in sickle cell anemia (SCA) patients are scarce, but clinical signs of MC activation syndrome, such as increased plasma histamine in vaso‐occlusive crisis (VOC), and normal or slightly elevated serum tryptase have been reported. However, assessing the biological relevance of MC as a cytokine source is more challenging because it is unclear under which circumstances they secrete those products in vivo, or if the cytokines measured systemically stem from a different cell type. We aimed to investigate the profile of mediators involved in the inflammatory process produced by MC in SCA. Methods:The supernatant of 5-week old MC cultures (17 SCA, 8 HV) obtained from peripheral blood CD34+ cells from 29 SCA patients and 13 healthy volunteers (HV) was analyzed using a multiplex platform and colorimetric assays for endothelin-1 (ET-1) and substance P (SP) (10 SCA, 6 HV). A correlation matrix (Pearson correlations, R software, v. 3.6.1) was generated using laboratory and clinical data chosen based on their value as inflammatory or prognosis markers (hydroxyurea [HU] treatment, fetal hemoglobin [HbF], hemoglobin [Hb], vaso-occlusive crisis [VOC], percentage of peripheral blood neutrophils, eosinophils (Eos), basophils, erythroblasts, and reticulocytes), MC surface expression of CD117, CD48 and CD63, and the supernatant content of 11 cytokines. To investigate MC cytokine release, we tested the supernatants from Eos-MC co-cultures (3:1 ratio), and after stimulation with ET-1 (20 nM), SP (10 µM) and imatinib (20 µg/ml)(n=3 per treatment). Results: Out of 26 cytokines, we found elevated levels of the following in the supernatants of SCA-MC cultures (data represented as mean in pg/ml±SE): TNFα: SCA=88.7±18.4, HV=32.6±3.8; IFNγ: SCA=55.3±11.2, HV=15.7±1.8; MCP1: SCA=555.0±147.2, HV=145.3±35.2; RANTES: SCA=24.7±3.9, HV=10.7±1.8 (p<0.05). However, SCA-MC from patients treated with HU (n = 11) showed higher values of IL-1b, IL-4, IL-5, IL-9, IL-15, and FGF than HV (n=8) and HU-free patients (n=6) (p<0.05). Supernatants from SCA-MC had higher ET-1 production compared to HV-MC (SCA=16.3±1.2, HV=11.93±1.3, pg/ml, p=0.02) but SP production was similar (SCA=27.9±1.3; HV=31.49±0.7 pg/ml). ET-1 stimulation of MC cultures caused 2-fold increase in IL-1AR production on HV-MC, but failed to produce any effect on SCA-MC. Similarly, imatinib reduced FGF only in HV-MC samples (HV: 15.1±3.5, HV-HU: 4.0±1.6, pg/ml). No effect on cytokine production was observed with SP. Conversely, Eos-MC cocultures showed a 10- and 4-fold increase of IL-5 and IL-9, respectively, regardless of the origin of Eos (HV or SCA). SCA-MC/SCA-Eos co-cultures had elevated proinflammatory (IL-1b, IL-12, TNF-α) and angiogenic (FGF, VEGF) cytokines, RANTES, IL-7, IL-4, and IL1-RA compared with SCA-MC/HV-Eos and HV-MC/HV-Eos (p<0.05). Preliminary multiparametric analysis on data from SCA patients showed a strong negative correlation between HU therapy and VGEF production, and between HbF levels and CD63 expression (MC activation marker).We also found a positive correlation between history of VOC and eotaxin-1 produced by SCA-MC. Conclusions: We found that MC responses depend both on the origin of the cultured cell and the stimuli utilized. Despite differences between in vitro and in vivo MC populations, our data show that cultured SCA-MC have a sustained activated state and produce a repertoire of mediators that could contribute to a perivascular microenvironment in favor of leukocyte and endothelium activation. In terms of cytokine production, cultured SCA-MC were more sensitive to stimulation by SCA-Eos than by HV-Eos, which may be relevant to the pathophysiology of airway inflammation in SCA patients with asthma. Differences in cytokine production between SCA-MC cultures from patients treated or not with HU may reflect the variability in adherence to treatment, individual response to each compound, or epigenetic modifications during the MC differentiation process that affect the phenotype of the mature MC. These results support that mediators produced by MC can contribute to the chronic inflammatory state and may be implicated in exacerbated responses to eosinophil activation in SCA.