Elevated Toll-Like Receptor-Induced CXCL8 Secretion in Human Blood Basophils from Allergic Donors Is Independent of Toll-Like Receptor Expression Levels

Human blood basophils have recently gained interest in addition to their function as allergic effector cells. Previous work suggests the involvement of innate immune mechanisms in the development and exacerbation of allergic responses, which might be mediated by basophils. We assayed the expression levels of Toll-like receptor (TLR) 1, 2, 4 and 6 on purified basophils from birch pollen-, house dust mite-, and non-allergic individuals. Additionally, we compared cytokine and chemokine secretion upon TLR stimulation in these basophil donor groups. Expression of TLR4 on the basophils of the allergic donor groups was decreased and CXCL8 secretion was elevated upon stimulation of TLR1/2 and TLR2/6 compared to the non-allergic donors. Decreased TLR expression and elevated CXCL8 secretion may represent possible mechanisms for aggravation of allergic symptoms in case of parasitic infection.


Introduction
TLRs are central molecules of innate immune responses, as they interact with pathogen-associated molecular patterns derived from microbial invaders [1]. Recent publications suggest that TLRs may not only be involved in host defense against microbial infection, but may also contribute to the formation and exacerbation of allergic responses by influencing IgE-mediated pathways. Tulic et al. [2] compared TLR function and expression in allergic versus non-allergic children. They found excessive TLR function in allergic children at birth and a relative decrease with age thereafter, which they discussed to result in an upregulated T H 2 response. A link between TLR signaling and allergy was further strengthened by the studies of Kucuksezer et al. [3] who showed that stimulation of myeloid dendritic cells via TLR4 or TLR8 resulted in enhanced allergen-specific CD4 + T cell proliferation [3].
Human basophils are, although in very low abundance in blood, important key players in allergic reactions as they release first line defense mediators and T H 2 skewing cytokines upon allergen-mediated crosslinking of Fc epsilon receptor 1 (FcεRI) [4]. The presence of TLRs in basophils was investigated by Sabroe et al. [5] detecting mRNA of TLR2 and 4 in human basophils. This study was corroborated by Komyia et al. [6] who found TLR2, 4, 6, and in some donors also TLR1 on a transcriptional level. Bieneman et al. [7] confirmed reactivity of human blood basophils to ligands of TLR2 but not to TLR4, whereas Watanabe et al. [8] showed IgG1 and IgG4 production by B cells mediated upon basophil activation by TLR2 and 4 ligands from patients with autoimmune pancreatitis. However, an involvement of human blood basophils via their TLRs in inflammatory or immunomodulatory processes like allergy is still scarce. Gyimesi et al. [9] showed a slight upregulation of one activation marker CD63 and histamine release in basophils from atopic patients at high concentrations of TLR4 ligand. Suurmond et al. [10] observed an increased cytokine release from basophils upon simultaneous stimulation of the high-affinity IgE receptor and TLRs. It was shown that TLR4 or TLR9 stimulation combined with FcεRI activation by anti-IgE resulted in a synergistic increase of IL-4, CXCL8, IL-13 and RANTES/CCL5, all of which are thought to enhance IgE-mediated responses, such as allergy or parasitic infection.
To further investigate the potential role of TLRs in augmenting allergic symptoms mediated by basophils, we compared TLR expression levels between allergic and non-allergic individuals and analyzed their capacity to trigger mediator release upon TLR stimulation of the different donor groups. We analyzed the surface expression of TLR1, 2, 4 and 6 on purified human basophils from non-allergic, birch pollen (BP)-allergic, and house dust mite (HDM)-allergic donors. In addition, basophils were stimulated with TLR ligands Pam 3 CSK 4 (TLR1/2), Pam 2 CSK 4 (TLR2/6) and LPS (TLR4) in order to investigate cyto-/chemokine release (CXCL8, IL-6, IFN-γ, TNF-α) and basophil activation marker CD203c surface expression.

Basophil donors
70 ml EDTA-anticoagulated whole blood were collected from birch pollen-allergic, house dust mite-allergic, and non-allergic donors (n = 8, each; Table 1). Next to clinical history, allergies were confirmed in vivo by positive skin prick tests (Pangramin, ALK-Abelló, Horsholm, Denmark). ImmunoCAP™ (Thermo Scientific, Waltham, MA, USA) was used to quantify allergenspecific IgE. IgE reactivity 0.7 kU/l was considered positive (Table 1). This study and its experiments were performed in accordance with the local institutional guidelines strictly adhering to methods and experimental protocols approved by the Ethics Committee of Land Salzburg (No.: 415-E/1/1117-2009). Written informed consents were obtained from the blood donors.

Basophil purification
EDTA whole blood diluted 1:2 using DPBS (Sigma-Aldrich) was pipetted on top of 12 ml of a mixture (17:1) of Ficoll-Paque™ PLUS (GE Healthcare, Little Chalfont, UK) and Percoll 1 (Sigma-Aldrich, St. Louis, MO, USA). Upon 30 minutes centrifugation at 500 g the PBMCcontaining interphase was collected and washed. Basophils were purified by negative magnetic sorting using the Basophil Isolation Kit II (Miltenyi Biotec, Bergisch Gladbach, Germany). All buffers and solutions were without calcium and magnesium and pre-chilled at 4°C. Basophil purity was determined using flow cytometry by staining the purified cell fraction with anti- Competing Interests: Buehlmann Laboratories AG did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries (Michael Schneider) and/ or research materials (stimulation buffer for basophils). This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. TLR staining and flow cytometry The highly purified basophils of all donors were resuspended in stimulation buffer (IMDM, Sigma-Aldrich) supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin (all from Gibco 1 , Life Technologies, Carlsbad, CA, USA), and 2.5 ng/ml IL-3 (Sigma-Aldrich). 40.000 cells in 100 μl were used per reaction. Basophils were incubated for 30 minutes at 4°C with anti-TLR1-PE (clone GD2.F4), anti-TLR2-eFluor450 (clone T2.5), anti-TLR4-APC (clone HTA125), and anti-TLR6-Biotin (clone h-Per6) (20 minutes followed by 10 minutes Streptavidin-APC-eFluor780 treatment) or the respective isotype controls (eBioscience, San Diego, CA, USA and ImmunoTools, Friesoythe, Germany). Specificity of the anti-TLR antibodies has previously been demonstrated [11][12][13][14]. The cells were measured on a FACS Canto II (BD, Franklin Lakes, NJ, USA).
Basophil stimulation and cytokine/chemokine determination

Results
This work was performed using highly purified basophils (>90%) from allergic and non-allergic individuals (Fig 1). The gating strategy comparing TLR expression of allergic patients and non-allergic donors is depicted in Fig 2.

TLR4 expression is altered in allergic basophil donors
Expression of TLR1, TLR4 and to a lesser extent TLR2 was observed on up to 60% of basophils in all donor groups, whereas TLR6 expression was generally very low and could only be detected in 4 donors (Fig 3a and 3b). Comparative analysis of TLR1 and TLR2 revealed almost similar expression levels in non-allergic vs. allergic donor groups. TLR4 levels were significantly lower in HDM-allergic patients compared to non-allergic donors (Fig 3b).
CXCL8 secretion from allergic donor basophils is augmented upon TLR1/2, TLR2/6 stimulation From the cyto-/chemokines tested, we observed differences in release only for CXCL8. Stimulation with TLR1/2 ligand Pam 3 CSK 4 (Fig 4a) resulted in upregulation of CXCL8 secretion in basophils from BP-allergic donors (p<0.05) and TLR2/6 ligand Pam 2 CSK 4 stimulation (Fig  4b) resulted in upregulation of CXCL8 in BP-and HDM-allergic patients compared to the non-allergic donors (both p<0.05). Comparisons between non-stimulated samples of nonallergic and allergic donors were not significant. Stimulation with the TLR4 ligand LPS did not result in significant differences in CXCL8 secretion between the three groups, although single individuals displayed a strong CXCL8 response (Fig 4c).

Activation marker expression and cyto-/chemokine secretion
The basophil activation marker CD203c was significantly upregulated upon stimulation with the two positive controls (anti-FcεRI antibody and fMLF) but not upon stimulation with TLR ligands (Fig 4d). A release of IL-6, IFN-γ and TNF-α was neither detected from unstimulated nor stimulated basophils with the exception of 1 to 3 donors from each group, which showed low secretion levels.

Discussion
Innate immunity, especially TLR signaling, is increasingly considered crucial for influencing the development of allergic reactions [2,3]. Hence, basophils, important cells of innate  immunity and main elicitors of allergic symptoms, could be key cells in humans to link innate immune reactions to allergy [10,15,16]. In this study we investigated TLR expression of basophils and their stimulation-induced cyto-/chemokine secretion in seasonally affected BP-pollen-and perennially affected HDM-allergic patients compared to non-allergic donors. We observed enhanced CXCL8 secretion in BP-allergic patients despite of similar expression levels of TLR1 and 2 on basophils in allergic patients and non-allergic donors. In contrast, TLR4 was reduced in both allergic patient groups. This observation is consistent with a proposed role of TLR4 signaling in allergy development based on the findings of Hollingsworth et al. [16] who  reported severe allergic inflammation in TLR4-deficient (C3H/HeJ) mice upon prolonged allergen stimulation. The fact that CXCL8 levels upon TLR stimulation of BP-and HDM-allergic patient groups were higher compared to those of non-allergic individuals adds to Gilmartin et al. [17] who additionally found increased CXCL8 expression upon FcεRI crosslinking in basophils of allergic donors, which they discussed as crucial for the recruitment of neutrophils, eosinophils, and monocytes to areas of inflammation. CXCL8 might, therefore, aggravate latephase reactions by enhanced inflammatory cell recruitment to the sites of parasitic infection. According to our data, this could be mediated by the amplified CXCL8 secretion of basophils from the allergic patients upon TLR1/2 and TLR2/6 stimulation and, hence, boost the allergenmediated inflammation as shown in the study of Gilmartin et al. Basophils were highly purified, however, a minor contribution in CXCL8 secretion by contaminating cells cannot be excluded.
As stimulation by TLR ligands is well-known to primarily induce secretion of pro-inflammatory cyto-/chemokines in a number of different cell types, we were interested in a potential induction of IFN-γ, TNF-α, CXCL8, and IL-6. With the exception of CXCL8, basophils did not respond with secretion of pro-inflammatory cyto-/chemokines. Our study stands in agreement with data recently described by Jeon et al. [18] who found IL-6 and CXCL8 secretion in flagellin-stimulated KU815 cells but not IFN-γ, IL-1β or TNF-α. Next to the selected cyto-/chemokine panel we analyzed the expression of CD203c, a very sensitive basophil activation marker, on the cell surface [19]. However, no effects of the TLR ligands on its expression could be detected, which indicates that basophils do not degranulate upon TLR stimulation.
In conclusion, elevated CXCL8 secretion upon TLR stimulation despite comparable TLR1 and 2 expression may represent an additional mechanism of how tissue inflammation might be exacerbated in allergic individuals and amplify allergic responses, like suggested by Yi et al. [20]