Differential Expression and Release of Activin A and Follistatin in Chronic Rhinosinusitis with and without Nasal Polyps

BACKGROUND: Chronic rhinosinusitis with (CRSwNP) and without nasal polyps (CRSsNP) should be regarded as distinct clinical entities based on differential inflammatory mediator and remodeling profiles. Activin A, a member of the TGF-β superfamily, plays an important role in inflammation and remodeling in the lower airways, although its expression and release in the upper airways remain undescribed. OBJECTIVE: To investigate the expression of activin A and its inhibitor follistatin in nasal tissue samples from CRSsNP and CRSwNP patients, and to monitor the spontaneous release of these molecules in a human mucosal model. METHODS: Protein levels were determined using ELISA for activin A, follistatin, TGF-β1 and indicator proteins (IL-5, ECP, IFNγ) in 13 CRSsNP, 23 CRSwNP, and 10 control samples. The spontaneous release rate and the release ratios of activin A, follistatin and TGF-β1 were determined in 9 CRSsNP and 7 CRSwNP tissue fragments cultured ex-vivo. The induction of activin A and TGF-β1 by one another was studied in 7 CRSsNP tissue fragments cultured ex-vivo. RESULTS: Significantly higher concentrations of activin A, follistatin, TGF-β1, and IFNγ were observed in CRSsNP compared with CRSwNP samples, whereas the concentrations of IL-5 and ECP were significantly lower. Follistatin was positively and linearly correlated with activin A in CRSsNP and CRSwNP. Activin A, follistatin and TGF-β1 were all spontaneously released by the samples, although the relative ratios released by tissue fragments from CRSsNP and CRSwNP samples were significantly different, with a higher follistatin/activin A-ratio and a follistatin/TGFß1-ratio (with less overall TGF-β1) in CRSwNP than in CRSsNP. Furthermore, TGF-β1 enhanced activin A secretion in CRSsNP tissue fragments cultured ex-vivo. CONCLUSION: The differences in tissue concentrations and spontaneous release rates for activin A and follistatin in different CRS samples support the hypothesis that CRSsNP and CRSwNP are two distinct disease entities with respect to remodeling patterns.