Background
Siglec-F
(Sialic acid-binding
Ig-superfamily
lectin-F) belongs to the CD33-related Siglec (CD33rSiglec) family which are a subclass of Siglecs defined by their mutual sequence similarity and clustered gene localization (chromosome 7 in mouse; chromosome 19q in humans) [
1]. Eosinophils express a restricted profile of Siglecs [
2‐
5]. Of the eight mouse Siglecs and fourteen human Siglecs that have been identified, eosinophils are reported to highly express significant levels of Siglec-F in mice [
2‐
5] and its functionally convergent ortholog Siglec-8 in human eosinophils [
6‐
8]. Most of the CD33rSiglecs are expressed on cells involved in innate immunity, such as monocytes, granulocytes, macrophages and natural killer cells [
1]. Siglec-F is a transmembrane receptor comprising a ligand binding V-set domain, three C-2 domains, a transmembrane domain, and a cytoplasmic ITIM motif (immunoreceptor tyrosine-based inhibitory motif), which is known to be involved in inhibitory signaling pathways in the immune system [
9,
10]. Support for inhibitory signaling by the cytoplasmic domain of CD33rSiglecs have come from studies which have demonstrated that antibody cross-linking of several CD33rSiglecs results in inhibition of cellular-activation signals, arrest of proliferation, or induction of apoptosis [
11‐
13].
Siglec-F is highly expressed on mouse eosinophils [
5] and levels of Siglec-F are up-regulated on peripheral blood eosinophils following acute OVA challenge in wild type (WT) mice [
5]. We have generated Siglec-F deficient mice and demonstrated that these mice have similar baseline levels of peripheral blood eosinophils as do WT mice [
5]. However, following acute OVA challenge Siglec-F deficient mice have significantly increased numbers of eosinophils in the bone marrow, blood, and lung compared to WT mice [
5]. These studies in Siglec-F deficient mice suggest that Siglec-F plays an inhibitory role in acute eosinophilic inflammation. Studies with an anti-Siglec-F Ab have demonstrated that it reduces levels of eosinophilic inflammation and induces eosinophil apoptosis when administered in mouse models of gastro-intestinal eosinophilic inflammation [
14], lung eosinophilic inflammation [
15], or a mouse model of the hypereosinophilic syndrome [
16]. Although studies have examined the role of Siglec-F utilizing Siglec-F deficient mice in acute antigen challenge models of asthma [
5], studies have not utilized Siglec-F deficient mice to examine whether Siglec-F plays a role in chronic antigen induced airway remodeling which is the focus of this study. As eosinophils may contribute to airway remodeling [
7,
17], we examined whether Siglec-F deficient mice would have increased levels of airway remodeling, and deposition of extracellular matrix proteins in the airway in vivo. In addition, as in previous studies we have demonstrated that WT mice challenged with allergen have increased levels of expression of Siglec-F ligands in the airway epithelium and peribronchial cells [
3,
5], we examined whether the absence of Siglec-F receptors in Siglec-F deficient mice would modulate levels of Siglec-F ligands expressed in the airway of Siglec-F deficient compared to WT mice.
Discussion
In this study we have utilized Siglec-F deficient mice to demonstrate an important role for Siglec-F in modulating levels of airway remodeling (peribronchial fibrosis, thickness of the smooth muscle layer), mucus expression, deposition of extracellular matrix proteins such as fibronectin, as well as an important role for Siglec-F in regulating levels of chronic allergen induced peribronchial Siglec-F ligands. This study confirms the importance of Siglec-F in airway remodeling and extends results obtained using an anti-Siglec-F Ab in WT mice [
15], to demonstrate an important role for Siglec-F in mucus expression, deposition of extracellular matrix proteins such as fibronectin, and regulating levels of chronic allergen induced peribronchial Siglec-F ligands which were not demonstrated in previous studies using an anti-Siglec-F Ab in WT mice [
15]. In addition, we demonstrated that Th2 cytokines such as IL-4 or IL-13 induce equivalent upregulation of Siglec-F ligand expression by airway epithelium in vivo. In contrast, TNF-α another cytokine expressed in the remodeled airway does not significantly regulate airway epithelial Siglec-F ligand expression, even though it induced a significant BAL neutrophil response. In addition to airway epithelium we have previously demonstrated that eosinophils also express Siglec-F ligands [
5]. All three cytokines (IL-4, IL-13, and TNF-α) increased the number of peribronchial cells expressing Siglec-F ligands and this was proportional to the induced lung eosinophil response which was strongest with IL-4 and IL-13 and significantly weaker with TNF-α. As activating Siglec-F receptors with anti-Siglec-F antibodies in vitro induces apoptosis [
5,
14‐
16], the up-regulated expression of Siglec-F ligands by airway epithelium in response to Th2 cytokines may be a mechanism by which airway epithelium downregulates eosinophilic inflammation when eosinophils come in contact with the airway epithelium. In vitro studies have demonstrated that a synthetic Siglec-8 ligand induces human eosinophil apoptosis [
24] underscoring the potential therapeutic utility of using Siglec ligands to limit eosinophilic inflammation. Chronic OVA challenge induced a similar significant increase in levels of lung IL-5 and IL-13 in Siglec-F deficient and WT mice. However there was a slight but statistically significant increase in levels of lung eotaxin-1 and RANTES in chronic OVA challenged Siglec-F deficient compared to WT mice which may reflect the increased numbers of peribronchial inflammatory cells in chronic OVA challenged Siglec-F deficient capable of expressing these chemokines. This study also demonstrated that the important eosinophil product LTC4, a mediator not investigated in previous studies using an anti-Siglec-F Ab in WT mice [
15], was unlikely to be contributing to enhanced remodeling in Siglec-F deficient mice. Siglec-F deficient mice challenged with chronic allergen had significantly enhanced peribronchial fibrosis, as well as increased mucus expression, and an increase in the thickness of the smooth muscle layer. Although Siglec-F deficient mice had increased thickness of the smooth muscle layer, the trend for increased airway responsiveness was not statistically significant.
The mechanism by which Siglec-F deficient mice have enhanced airway remodeling is not due to a direct effect of Siglec-F on fibroblasts, epithelial cells, or smooth muscle as these cells do not express Siglec-F. In the absence of an allergen stimulus to induce airway inflammation, non-OVA challenged Siglec-F deficient mice do not have evidence of airway remodeling, again supporting the concept that it is the exaggerated inflammatory response in Siglec-F deficient mice, rather than structural cells in the airway, that are the primary initiator of enhanced airway remodeling in Siglec-F deficient mice. Siglec-F is most highly expressed on eosinophils [
5], but can also be detected on macrophages and activated CD4+ T cells as previously demonstrated [
5]. As Siglec-F deficiency results in increased numbers of eosinophils in the allergen challenged lung, one potential explanation for the enhanced airway remodeling in Siglec-F deficient mice is the increased numbers of eosinophils in the lung expressing pro-fibrotic growth factors that may contribute to remodeling including TGF-β1 [
7,
22]. In support of this hypothesis are the increased numbers of TGF-β1+ cells we have identified in Siglec-F deficient mice. The importance of eosinophils and TGF-β1 to airway remodeling is suggested from murine studies in which airway remodeling is significantly reduced in mice treated with an anti-TGF-β1 Ab [
25], as well as in Smad 3 deficient mice [
22] which have impaired TGF-β signaling. In addition studies in IL-5 deficient mice [
7] as well as in human subjects with asthma treated with anti-IL-5 [
17] demonstrate reduced numbers of eosinophils, reduced TGF-β1+ eosinophils, and reduced airway remodeling. Although several studies show an important role for eosinophils, TGF-β1, and Smad signaling in airway remodeling [
22,
25], there are studies which have demonstrated that an anti-TGF-β1 Ab does not reduce remodeling in mice [
26], and that an anti-TGF-β1 Ab increase airway hyperreactivity in mice [
27]. Increased numbers of peribronchial cells expressing TGF-β1 in chronic OVA challenged Siglec-F decient mice may also contribute to increased smooth muscle thickness as TGF-β1 induces airway smooth muscle hypertrophy [
28]. As LTC4 is expressed by eosinophils [
19] and has pro-remodeling potential in asthma [
20], we examined whether Siglec-F deficient mice had increased levels of lung LTC4 in their remodeled airways. As there was no significant difference in levels of LTC4 in the remodeled airways of WT and Siglec-F deficient mice, LTC4 is unlikely to explain the differences in levels of airway remodeling in WT and Siglec-F deficient mice. The mechanism of increased mucus expression in chronic OVA challenged Siglec-F decient mice at present unknown. One possible explanation is that the increased numbers of eosinophils in the airway in Siglec-F deficient mice express increased levels of mediators that can influence mucus secretion including granule mediators (ECP, MBP)[
29,
30], lipid mediators (LTC4)[
31], cytokines (IL-13)[
31], or other at present unknown mediators. Alternatively eosinophils may release a mediator that influences a second cell type to subsequently influence mucus expression. Our initial experiments to address this question demonstrate no differences in levels of IL-13 or LTC4 between Siglec-F deficient and WT mice.
In this study we also made the novel observation that Siglec-F deficient mice had increased levels of peribronchial extracellular matrix remodeling as indicated by increased fibronectin deposition. Fibronectin is a large extracellular matrix glycoprotein molecule consisting of two similar subunits of 220-250 kDa [
32] that has been detected in increased amounts in the remodeled airway in human asthma [
33], as well as in fatal asthma [
34]. There are two forms of fibronectin, plasma fibronectin (dimeric and soluble) and cellular or extracellular matrix fibronectin (multimeric and insoluble)[
32]. Extracellular matrix deposition of fibronectin may enhance airway remodeling in asthma by contributing to the formation of collagen fibrils [
35], mediate the migration of fibroblasts [
36], and increase the proliferation of smooth muscle cells [
37]. In addition, fibronectin may enhance eosinophilc inflammation in the airway through its ability to increase CC chemokine expression by airway smooth muscle cells [
38].
In this study we demonstrated that Siglec-F plays an important role in mediating several key features of airway remodeling but did not a play a significant role in mediating AHR. Although mathematical models of asthma predict that the increased airway wall thickening in remodeled airways would result in disproportionately severe airway narrowing and responsiveness [
39], studies in human asthmatics have demonstrated that airway wall remodeling and thickening in asthma is associated with reduced rather than increased airway reactivity to MCh [
40]. One potential explanation suggested [
41] for the discrepancy in results between the mathematical modeling studies and the computerized tomography scan studies in asthma is that the mathematical modeling studies did not fully take into account the potential effect of airway wall thickening on the mechanical properties of the airway, e.g., stiffness of the airway [
41]. Our studies with Siglec-F-deficient mice underscore the fact that a gene such as Siglec-F, which plays a significant role in mediating several important aspects of airway remodeling, may not play an essential role in mediating AHR. Our studies using an anti-Siglec-F Ab in a model of chronic OVA allergen induced airway remodeling also demonstrated inhibition of airway remodeling but no effect on AHR measurements [
15], consistent with our observations in this study using Siglec-F deficient mice.
In addition to investigating the effect of Siglec-F on airway remodeling we examined whether the presence versus absence of Siglec-F receptors influenced levels of Siglec-F ligands being expressed. We have previously demonstrated that levels of Siglec-F are upregulated on blood eosinophils following OVA challenge [
5]. In this study we examined whether removal of the Siglec-F receptor in Siglec-F deficient mice influenced levels of Siglec-F ligand expression. Our study demonstrates that the number of Siglec-F ligand+ peribronchial cells was significantly increased in OVA challenged Siglec-F deficient compared to OVA challenged WT mice. This increase in level of peribronchial Siglec-F ligands in Siglec-F deficient mice is most likely accounted for by the increased numbers of inflammatory cells expressing Siglec-F ligands (i.e. eosinophils, macrophages) recruited to the airway following OVA challenge, though we cannot rule out a contribution from upregulation of Siglec-F ligand expression by inflammatory cells recruited to the lung. In contrast, constitutive and OVA induced levels of expression of the Siglec-F ligand in airway epithelium are similar in Siglec-F deficient and WT mice.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JYC, DJS, AP, PS, TD, and MM made significant contributions to design, acquisition of data, as well as analysis and interpretation of data. SD made significant contribution to acquisition of data. AV and DHB made significant contributions to conception, design, analysis and interpretation of data, and drafting of manuscript.