Expression profiling and functional analysis of Toll-like receptors in primary healthy human nasal epithelial cells shows no correlation and a refractory LPS response

The mucosal barrier of the nose forms the first line of defence against air pollutants, airborne allergens, and (non-) pathogenic microorganisms. Epithelial cells are the outer lining of the mucosa of the nasal airway and play, besides their role as passive physical barrier, an important role in orchestrating innate and adaptive immune responses [1‐3]. Epithelium can trigger anti-microbial responses by recognizing pathogen-associated molecular patterns (PAMPs) through the sentinel action of pattern recognition receptors (PRRs) like Toll-like receptors (TLRs).

Toll-like receptors are evolutionarily conserved pattern recognition receptors of the innate immune system [4]. Until now, 13 mammalian TLRs have been characterized and for most of the TLRs, except TLR10, TLR12 and TLR13, specific ligands have been identified [5, 6]. The first group of receptors recognize bacterial products. TLR2 is activated by a variety of bacterial lipoproteins, peptidoglycans (PGN), and lipoteichoic acids (LTA), by forming a heterodimer with TLR1 or TLR6 [7, 8]. TLR4 appears to form homodimers and under participation of adaptor molecules like MD-2 and CD14 this TLR recognizes lipopolysaccharide (LPS) from the outer membrane of gram-negative bacteria. Originally thought to be a receptor only for LPS, TLR4 now emerges as a molecule responsible for signalling induced by a broad variety of molecules such as respiratory syncytial virus protein F [9], fungal components [10], or endogenous ligands like heat shock proteins, lung surfactant protein A, and beta-defensin [11‐13]. Lastly, TLR5 recognizes bacterial Flagellin. Viral compounds trigger endosome-associated receptors, such as TLR3 by double-stranded (ds) RNA or its synthetic analogon polyinosinic polycytidylic acid (Poly I:C) and viral single stranded (ss) RNA signals via TLR7 and TLR8. DNA-containing CpG motifs are recognized via TLR9 [14, 15].

Although TLRs can be involved in the initiation of adaptive immune responses through their presence on dendritic cells (DCs), they may also indirectly affect the adaptive immune response. Innate immune recognition via TLRs on barrier cells like epithelial cells has been shown to determine the functional properties of tissue-resident DCs, thereby instructing the outcome of antigen-specific immunity [16]. The overall complexity of the contribution of TLRs on immune responses can be influenced by several factors like their relative abundance, their individual expression pattern, or the timing of exposure. For example, stimulation of TLR2 and TLR4 signalling pathways has been shown to both drive [16, 17] and inhibit [18, 19] the development of Th2-mediated allergic inflammation in different experimental mouse models. Moreover, the impact of TLR4 stimulation on allergic inflammation is highly dependent upon the dose of the TLR4 agonist, with high LPS concentrations inducing Th1-responses and low concentrations inducing Th2-polarized inflammatory responses [20].

The LPS-induced pulmonary burden varies between different respiratory compartments and might therefore explain the functional differences between bronchial and alveolar epithelial cells with respect to LPS-dependent cytokine release. It is widely believed that alveolar epithelial cells are unresponsive to LPS due to low or absent expression of TLR4 and/or CD14 or MD-2 [21]. In contrast to this it has been reported that lung epithelial cells do express TLR 1–6, including adaptor molecules like CD14 and MD-2, with bronchial epithelial cells showing CD14-dependent activation of TLR4 and alveolar epithelial cells showing LPS-binding protein (LBP)-dependent inhibition of TLR4 signalling [22, 23], confirming the relevance of co-factors for a proper TLR signalling. Small airway epithelial cells have been shown to express mRNA for the TLRs 1–6 and can respond to various stimuli such as viruses or bacteria resulting in the release of different pro- and anti-inflammatory cytokines and chemokines [24‐26].

At present only limited data is available on the expression of TLRs by nasal epithelial cells. Claeys et al. showed constant expression of TLR2 and TLR4 in tissue biopsies from patients with nasal polyposis or chronic rhinosinusitis and in healthy individuals [27]. Isolated nasal epithelial cells from nasal polyps were shown to constitutively express mRNA of all 10 TLRs, with more pronounced expression of TLR 1–6 [28]. Primary nasal polyp epithelial cells express functional TLR3 and TLR4 and release high concentrations of proinflammatory chemokines and cytokines upon stimulation with dsRNA [29]. Until only one study investigated the expression and function of some but not all TLRs on primary nasal epithelial cells from non-allergic, non-diseased individuals specifically [30]. Given the role of TLRs expressed on epithelial cells within the induction of immune responses, the relatively limited knowledge on the expression of TLRs in nasal epithelium, and the protective effect of TLR polymorphism in childhood asthma [31], the aim of this study was to determine the expression and functionality of TLRs expressed in primary nasal epithelial cells from healthy donors.

Epithelial cells are uniquely positioned at the interface between inside and outside of the organism, which makes them perfect candidates for initiating and orchestrating local immune responses. In addition to establishing which TLR receptors are expressed in primary nasal epithelial cells from healthy individuals, our data furthermore suggest that nasal epithelium has developed a delicate response system towards microbial exposures. Firstly, despite the presence of TLR4 and its prime co-stimulatory molecules CD14 and MD-2, nasal epithelium from healthy individuals does not respond to LPS. As the nasal mucosa is constantly exposed to high concentrations of endotoxin, this unresponsiveness could provide a mechanism to dampen the inflammatory response in the nasal mucosa in order to avoid a chronic inflammatory response. Secondly, levels of TLR expression in individuals varies strongly, to the extent that some individuals not express TLRs that others do. Thirdly, not only are the expression levels different between individuals, but independently of these differences, also the response induced by a specific TLR ligand varies strongly between individuals. And finally, the mediator response varies between different TLRs, even when they are thought to act through a common pathway. This complex level of variation in TLR signaling suggests that different healthy individuals may see different environments despite identical exposure. Although it should be noted that our sample size of 10 healthy individuals is relatively small, so that it would be difficult to generalize our conclusions for the general population

The expression of TLRs within the lower respiratory tract has been investigated intensively [22, 24, 25, 35], while data on TLR expression and functionality in nasal epithelial cells from healthy individuals is limited. Our experiments showed that primary nasal epithelial cells from most healthy donors express mRNA for TLR 1–6 and TLR9 and mainly respond to the TLR3 ligand Poly(I:C) and to the TLR2 and TLR5 agonists. The expression of TLR2, 3, and 4 has been shown in nasal epithelial cells derived from nasal polyps, with poly(I:C) inducing the secretion of RANTES, IP-10, IL-8 and GM-CSF [29]. We were able to confirm this outcome and in addition show a consistent and statistically significant up-regulation of IL-2R, VEGF, MIP-1β in all individuals. Closer inspection of our data shows strong up-regulation of other mediators as well (e.g. Mip-1α, MCP1, IL-7). However, as the induced expression of these mediators varies so strongly between individuals this up-regulation does not reach statistical significance. These observations show that healthy individuals differ strongly in their response to external triggers, which will contribute to differences between the ability to fight off viral and bacterial infections. The absence of TLR7 and TLR10 mRNA expression differs from the previous observations of Renkonen [36] and Tengroth [30]. In both previous reports expression levels at the mRNA were low, so that differences in our detection technique (real time PCR) versus microarray [36] or differences in growth conditions [30, 36] may help to explain the observed expression differences for TLR7 and TLR10. Allowing for the specificity of TLR antibodies both TLR7 and TLR10 could be detected by immuno-histochemistry with moderate biological activity for the TLR7 agonist relative to TLR3 activation [30, 36].

The most striking discrepancy between TLR expression and responsiveness we observed for TLR4. Despite the presence of the receptor on the cell surface, the presence of key co-stimulatory molecules (CD14 and MD-2) and a seemingly intact downstream signaling cascade (the cells do response to other TLR stimulations), nasal epithelial cells do not respond to LPS. This unresponsiveness has also been observed in the epithelia of the gut where it was attributed to missing MD-2 expression [37] and in nasal epithelium from polyposis patients by Wang and co-workers [29]. Nasal polyposis epithelium showed a much weaker response to LPS than to polyIC stimulation, indicating that even in diseased tissue the nasal epithelial response to LPS is affected. In contrast, lung or renal epithelia are able to respond to LPS which suggests that the hypo-responsiveness could be an adaptation in epithelia that are exposed to high concentrations of LPS, whereas epithelia that are relative sterile do show a response to LPS.

The functional consequences of responding to TLR ligation are many. Expression of IP-10, MIP-1α, MIP-1β, IL-8 and G-CSF after TLR3 activation contribute to the recruitment and activation of neutrophils or macrophages. Furthermore, IL-8 and RANTES have been shown to be involved in the recruitment and survival of eosinophils [38]. These findings imply a role for TLR3 in the nasal immune response not only in Th1-mediated responses, but also in viral induced allergic exacerbations. This notion would also be in line with our recent observation that many aspects of TLR3 activation of nasal epithelial cells resemble that of activation by house dust mite allergen [39]. The inflammatory features of dsRNA mediated by TLR3 are also thought to contribute to the exacerbation of CRS and nasal polyps during viral infection [40]. TLR4 expression on lung epithelial cells has been shown to be required for DC activation in the lung and for priming of effector T helper response to HDM [16]. In the absence of TLR ligation, lung DCs are minimally active. Binding with the TLR4 ligand LPS leads to enhanced motility and sampling behavior. This response strictly depends on neighboring epithelial cells being triggered by TLR4. In addition, responses to allergens are substantially altered when epithelial cells cannot detect the endotoxin in the allergen, indicating that TLR4 signaling in epithelial cells is critical for the initiation of Th2 responses to inhaled allergens [41].

Expression of TLRs on structural cells like epithelial cells and the respective functionality in terms of released mediators are important factors in the orchestration of local immune responses. We investigate the expression of all TLR receptors in primary nasal epithelial cells of healthy individuals and show an absence of TLR7 and TLR8 together with huge individual differences in mRNA expression level for TLR1-6 and TLR9. Although mRNA expression of TLRs of often taken as a measure of their activity we show that this is should be done with caution. Specific TLR agonist-induced mediator release in nasal epithelial cells is very variable between different individuals and does not correlate with the expression levels of the respective TLRs, although we show this only for a relative small number of individuals. Most notably we show that despite the presence of TLR4, CD14, and MD2 in nasal epithelial cells, stimulation with LPS does not induce any mediator response. Supporting and strengthening previous observations in nasal polyposis patients that nasal epithelial cells seem to resemble gut epithelial cells where a yet unidentified mechanism prevents epithelia routinely exposed to bacterial flora from fortuitous activation. Our data suggest that we should probably consider individual expression and activation levels better as this would affect how individuals see their microbial environment.