This study included children with allergic rhinitis (AR) recruited by family paediatricians of Messina. The nasal HMGB1 concentration was measured in the laboratory of Unit of Genetics and Paediatric Immunology of the Policlinico of Messina. All participating subjects were of Italian origin.

The details of the study were explained to the parents/guardians of the children who were participating in the study, and written informed consent was obtained from each parent/guardian. A physician performed the physical examination of each child. Demographic data and other clinical characteristics of the study subjects were collected using the questionnaires. AR was diagnosed according to the presence of typical nasal symptoms (nasal itching, sneezing, watery rhinorrhea, and nasal obstruction) and sensitization to pollen allergen, documented by a positive skin prick test (SPT). Inclusion criteria were: age of between 6 and 16 years, a diagnosis of AR, sensitization to Parietaria pollen, and the presence of nasal symptoms.

Exclusion criteria were: perennial AR, any respiratory tract infection within 1 month before the start of the study, other nasal disease (including structural abnormalities, nasal polyposis and clinically relevant septum deviation), asthma comorbidity (excluded by history and physical examination), and any treatment in the previous month capable of interfering with the results. The study was carried out from March 2013 to May 2013, the time during which the subjects were naturally exposed to pollens. The study was conducted on 170 AR subjects (87 males and 83 females; median age, 10.3±3.4 years). All children presented with: nasal obstruction, nasal itching, sneezing and watery rhinorrhea, an increased level of total serum IgE and positive results for the Parietaria pollen SPT.

In accordance with the study protocol, patients were divided into three groups; one group received 1 puff of saline (29 males and 28 females), one group received 1 puff of Budesonide (32 males and 25 females) and one group received 1 puff of Narivent (DMG, Rome, Italy) (30 males and 26 females) into each nostril 2 times a day for 1 week. HMGB1 levels in nasal fluids were evaluated before and after treatment.

All subjects underwent nasal lavage as described by Salpietro et al. [12]. Briefly, healthy controls and patients affected by different types of rhinitis or nasal polyps as well as AR patients before and after 1 week of treatment with saline, Narivent or Budesonide were placed in a sitting position with the head slightly tilted. Nasal lavage was performed with a pipette, by injecting 1.5 mL of saline solution (0.9% NaCl at 37℃) into the nostril of the patient, after closing the possible escape route represented by the other nostril with a patch. The recovered liquid was collected after a few seconds with the same pipette. Sampling was performed for both nostrils. The fluid was stored at -20℃ until HMGB1 measurement. The recovery volume was on average more than 90% after nasal lavage.

HMGB1 concentration was measured in the nasal fluid collected from each patient (as described above) by performing an enzyme-linked immunosorbent assay (ELISA) using an HMGB1 ELISA Kit II developed by Shino-Test Corporation (IBL International GmbH, Hamburg, Germany) following the manufacturer's instructions. Briefly, samples were diluted 1:5 with sample diluent and added to microtiter plates, which were then incubated for 24 hours at 37℃. After washing, 100 µL/well of antihuman HMGB1 peroxidase-conjugated monoclonal antibody was added and the plates were incubated at room temperature for 2 hours. After extensive washing, tetramethylbenzidine dihydrochloride was added to each well. The enzyme reaction was allowed to proceed for 30 minutes at room temperature. The chromogenic substrate reaction was stopped by adding the stop solution, and the absorbance of each well was determined on a microplate reader at 450 nm. The HMGB1 concentrations were calculated with reference to a standard curve, with a detection limit of 0.2 ng/mL.

Plasma samples were collected from patients treated with acute (1 hour after one puff per nostril) or chronic (after 1 week of treatment with 1 puff per nostril twice a day) Narivent administration or from healthy subjects 1 hour after consumption of a licorice candy (containing 20.5±3 mg of GLT). Samples were quantitated by means of liquid cromatography-mass spectometry. Briefly, specimens were extracted with methanol and injected in a high performance liquid cromatography apparatus consisting of a reversed-phase C18 (ODS) column with a mobile phase of methanol 1%, formic acid (75:25, vol/vol). Mass spectometry determination was performed using negative electrospray ionization (negative ESI) in the selected ion monitoring mode. GLT was monitored at the m/z 821.1 channel, glycyrrhetinic acid (GA) at the m/z 469.2 channel, and internal standard (linuron) at the m/z 248.9 channel. The calibration curve was linear over the range from 0.05 to 10 µg/mL with a correlation coefficient above 0.99.

Blood samples from healthy donors were stratified on a Ficoll layer. After centrifugation at 1,500 rpm for 30 minutes at room temperature, peripheral blood mononuclear cells were collected at the interface with Ficoll and used for preparation of monocytes and lymphocytes as previously described [5]. Cellular pellet containing red blood cells and granulocytes was used for preparation of eosinophils (isolation kit 130-092010) and neutrophils (CD16 microbead 130-045-701) by means of magnetic separation procedure using Miltenyi kits following the manufacturer's instructions. Cells were cultured as previously described [5]. Cell death was assessed by Propidium iodide labeling and cytofluorometric analysis as previously described [13].

We first investigated whether the levels of HMGB1 were altered in nasal fluids of patients affected by different forms of rhinitis. To this end, we collected nasal fluids from healthy subjects and patients affected by AR, non-AR with eosinophilia syndrome (NARES) or nasal polyps (12 subjects per group), and evaluated their HMGB1 content by means of a commercial ELISA kit (see MATERIALS AND METHODS SECTION for further detail). As shown in Fig. 1A, we found significant increases in the HMGB1 content in all the three groups of rhinitis patients. When HMGB1 levels were compared with those in patients pooled into a single group, a 3-fold increase in HMGB1 levels was observed (Fig. 1B). The present data showing that the level of proinflammatory cytokine HMGB1 is increased in nasal fluids of patients with different forms of rhinitis underscore the therapeutic potential of HMGB1-inactivating strategies in treatment of inflammatory disorders of the nasal mucosa. The nasal formulation Narivent is a GLT-containing medical device used to treat inflammation of the nasal mucosa [14]. We therefore compared the efficacy of topical treatments with saline, Narivent, or Budesonide in reducing the HMGB1 content in nasal fluids of patients with AR (see MATERIALS AND METHODS SECTION for patient details). As shown in Fig. 1C, we found that, at variance with saline, Narivent and Budesonide equally reduced the HMGB1 content in nasal fluids after 1 week of treatment. Reportedly, GLT is known to inhibit 11p-hydroxysteroid dehydrogenase 2 and to promote arterial hypertension via cortisol accumulation [15]. However, it is not known whether nasal formulations containing GLT can increase cortisol levels because of systemic absorption. To address this question, we first evaluated the plasma levels of GLT and of its principal metabolite GA in healthy subjects. We found that both acute (i.e., 30 minutes after two puffs/nostril, 4.99±0.35 mg GLT/puff) or prolonged administration (i.e., two puffs/nostril twice a day per one week) resulted in GLT and GA concentrations in plasma in the order of the method detection limit (0.2 ng/mL). Notably, ingestion of a candy containing a standard amount of GLT (20.88±3.8 mg) resulted in plasma levels (30 minutes after ingestion) that were 2 folds higher than those found in patients receiving GLT for one week (Fig. 1D). In keeping with the very low plasma concentrations of GLT and GA, in subjects receiving Narivent, plasma cortisol levels were not different from those in controls (Fig. 1E). Taken together, these findings indicate that nasal treatment with GLT leads to poor systemic absorption that is not sufficient to affect the plasma cortisol levels and blood pressure, at least in the time window analysed. Data also indicate that ingestion of a single licorice candy is enough to cause significantly higher plasma GLT/GA concentrations than those in patients treated with nasal GLT formulation for an entire week.

The identification of increased HMGB1 concentrations in nasal fluids of rhinitis patients clearly underscores the pathogenetic relevance of this proinflammatory agent in immune disorders of the nasal mucosa. It is well known that the latter are frequently associated with eosinophilic infiltrates [16,17]. Given the ability of HMGB1 to activate and promote survival of eosinophils [18], as well as the high levels of the protein in nasal fluids of patients with AR and NARES (Fig. 1A) (in which the eosinophilic infiltrate is well represented and is of pathogenetic relevance), we next planned to evaluate the ability of different leukocytes to release HMGB1. Interestingly, we found that among human macrophages and granulocytes isolated from peripheral blood of healthy subjects, eosinophils were able to release a higher amount of HMGB1 within 12 hours of incubation (Fig. 1C). Of note, monocytes, lymphocytes, and neutrophils released an amount of HMGB1 (Fig. 2A) that was almost half of that released by eosinophils. To further investigate the ability of the protein to affect half-life of leukocytes in culture, we took advantage of the ability of GLT to bind and inactivate HMGB1. We therefore exposed cultured leukocyte populations to GLT and evaluated their survival. GLT was used at a concentration of 1%, 5 folds lower than that in the anti-inflammatory nasal preparation Narivent. As shown in Fig. 2B, we found that GLT selectively killed eosinophils and had no effect on the other types of leukocytes.