Single Nucleotide Polymorphisms in Chemosensory Pathway Genes GNB3, TAS2R19, and TAS2R38 Are Associated with Chronic Rhinosinusitis

Chronic rhinosinusitis (CRS) is a persistent inflammatory state of the sinus mucosa with symptoms of nasal obstruction or discharge with or without facial pain and changes in smell present for more than 12 weeks [1]. Prevalence estimates range from 10 to 12% of the population in the USA and Europe [2], resulting in health care costs greater than USD 770 million in the USA alone [3]. CRS patients have poor quality of life scores, similar to other chronic diseases, including CHF and COPD [4]. Evidence suggests a genetic component to CRS [5]. In a large population study, there was a 2.4-fold increased risk of CRS in first-degree relatives of patients [6]; this number is higher for first-degree relatives of patients with nasal polyps. Furthermore, several well-characterized genetic syndromes are associated with CRS, including cystic fibrosis [7] and primary ciliary dyskinesia [8].

An emerging concept regarding the chronic nature of CRS relates to a potential deficit in sensing and clearing bacterial infections; receptors for bitter and sweet taste have been found in the airway and are thought to play a role in sensing bacteria and regulating innate immune responses [9]. Bitter taste is initiated in the oral cavity through members of the T2R family of G protein-coupled receptors (GPCRs) [10‒12]. However, T2R bitter receptors are not limited to taste bud cells; they are also found in solitary chemosensory cells and ciliated epithelial cells of the nasal and sinus cavities, functioning in these locales to “taste” secreted bacterial products, such as acyl-homoserine lactones [13]. Gram-negative bacteria, such as Pseudomonas aeruginosa,use acyl-homoserine lactones as quorum-sensing molecules [14]. Binding of these bacterial products to T2Rs is thought to modulate innate host immune responses, such as release of antimicrobial peptides and nitric oxide.

Genome-wide association studies (GWAS) have identified several genes potentially associated with CRS [15, 16]. Recent genetic studies have underscored the importance of taste receptor signaling in innate immunity of the upper and lower airways, and paranasal sinuses [13]; single nucleotide polymorphisms (SNPs) in taste receptor genes have since been associated with altered bacterial immune response and, thus, with CRS [17, 18]. There are ∼25 type 2 taste receptors (TAS2R “bitterant receptors”), coupled to G protein signaling, that are expressed in multiple tissues. In the ciliated mucosa of the sinuses, these TAS2Rs respond to chemoirritants and bacterially produced secretions [19]. Some of us also recently showed that an inhibitor of GPCR signaling, regulator of G-protein signaling-21 (RGS21), opposes TAS2R-mediated bitterant signaling in immortalized airway epithelial cells and, given its expression in sinus mucosa and airway epithelia, may be involved in mucociliary clearance [20, 21]. Here, investigating candidate genes involved in bitterant signaling from among the TAS2R genes and RGS21, we identified three SNPs more common in West Virginia CRS patients than in healthy controls of European descent, including an SNP within GNB3, a downstream component of bitterant signaling previously unassociated with CRS.

Seventy-four volunteers were seen at an academic tertiary referral center in West Virginia. The study was approved by the Institutional Review Board of West Virginia University (WVU) Office of Research Integrity and Compliance. All volunteers met EPOS criteria [22] for CRS, and informed consent was obtained under IRB 1410476782. Age, sex, allergy, asthma, migraine, smoking status, presence of polyps, and previous surgical history were recorded (Table 1).CT scans of the sinuses were available for 66 of the 74 patients. Lund-Mackay CT score was calculated based on the standard radiologic staging of CRS for these patients [23]. Sino-nasal Outcome Test-22 (SNOT22) scores were obtained from all patients; total SNOT22 and the five standard quality of life domain scores were calculated (Table 1) [24]. Patients with cystic fibrosis, primary ciliary dyskinesia, known immunodeficiency, and craniofacial abnormalities were excluded from the study population.

Volunteers provided two buccal swabs, which were each placed in a 15-mL polystyrene conical centrifuge tube and stored at –80°C until DNA extraction using QIAamp DNA Mini Kits (exactly as per the manufacturer’s protocol). Purity and yield were assessed by 260/280 nm absorbance on a QIAxpert microfluidic spectrophotometer.

Genotyping was performed using TaqMan primer-probe sets (Table 2) and Type-it Fast SNP Probe PCR Kits (Qiagen) exactly as per the manufacturer’s protocols. Reactions were run on a Qiagen Rotor-Gene Q in duplicate; no-DNA negative control reactions were also performed for each primer-probe set (representative data obtained are displayed in Fig. 1).

Allele frequencies were compared with 1,006 Europeans (1,000 Genomes Project) [25], given that West Virginia is demographically homogenous [26], i.e., 94% White European ancestry (http://censusviewer.com/state/WV). Given no information for the frequency of the GNB3 SNP rs5443 in 1,000 Genomes, its allele frequency was instead compared with 116 Europeans (CEU) from HapMap [27]. Statistical analyses of SNP frequencies were performed using Pearson’s χ². A p value <0.05 was used to infer that the allele frequency from CRS patients is significantly different from demographically matched, public genome data. Subgroup analyses were performed comparing the SNP minor allele frequency (MAF) in patients with (CRSwNP) or without (CRSsNP) nasal polyps, and separately based on Lund-Mackay CT score, using standard χ² statistical tests (Tables 3, 4).

Given our recent findings that a negative regulator of bitterant GPCR signaling, RGS21, is expressed in sinus mucosa and airway epithelia [20, 21], we interrogated the status within a CRS patient population of two SNPs within RGS21, as well as four SNPs within other bitterant signaling genes with established or unknown association with CRS (i.e., TAS2R19, TAS2R20, TAS2R38, and GNB3). All chosen SNPs had an MAF greater than 30% so they would be identifiable in this patient population. Buccal swabs from CRS patients (Table 1) were used for genomic DNA isolation and genotyping.

TAS2R19 (previously known as TAS2R48) is a GPCR with bitterant taste receptor activity located on chromosome 12. SNP rs10772420 within TAS2R19was previously identified as being associated with RSV infections in a GWAS [28] and also identified in a GWAS of Canadian patients with sinusitis [15]. The minor allele (A) causes a missense mutation (arginine 299 to cysteine) in the encoded protein and is associated with intense quinine perception [29]. The “minor” allele is actually more prominent in Europeans (MAF 0.505; Table 2). In the CRS patients in this study, this allele is even more common (MAF 0.601; p = 0.024); 23% of the CRS patients were homozygous (A/A) for the minor allele.

TAS2R38, a bitterant receptor key to phenylthiocarbamide perception [30], is expressed in sinonasal ciliated epithelium [31] and upper airways [32]. TAS2R38 is implicated in innate immunity and the response to Pseudomonas [31]. rs713598 is a common missense SNP in TAS2R38 [30], associated with quinine intensity [29] and other taste preferences [33‒35]. rs713598 is one of the most common SNPs in the TAS2R38 gene, causing a missense mutation to the encoded GPCR (alanine 49 to proline) [30]. In the CRS patients in this study, this allele is more common (MAF 0.568; p < 0.001) than in Europeans of the 1,000 Genomes Project (MAF 0.423).

RGS21 is a Gα GTPase-accelerating protein expressed in the lingual epithelium, lung, and gastrointestinal tissues [21]; ablation of Rgs21 in mice blunts bitterant signaling [36]. rs7528947 is a 3′ untranslated region variant of RGS21 with no known disease associations. The rs7528947 MAF in CRS patients is 0.547, similar to the European MAF (0.516; p = 0.461). No difference in rs7528947 MAF was observed when stratified between CRS patients with or without nasal polyps (Table 3); however, a significant difference (p = 0.017) was observed across the spectrum of Lund-Mackay CT staging scores (Table 4), with the minor allele appearing more frequently with higher Lund-Mackay scores.

A second RGS21-associated SNP, rs1175152, is located in the intergenic region between RGS18 and RGS21 and is also of unknown consequence. To date, there have been no known studies identifying this marker with any known phenotype or in association with any disease. Its MAF among Europeans of the 1,000 Genomes Project is 0.433, similar to the frequency observed in the CRS patients of this study (MAF 0.439; p = 0.875).

TAS2R20 (also known as TAS2R49) is a bitterant receptor expressed in multiple tissue types [37]. Its missense SNP rs12226920 (histidine 143 to glutamine) was previously associated with CRS, with a biallelic difference of 16% versus controls, suggesting a possible role for variations in this GPCR in CRS pathogenesis [15]. In the CRS patients of this study, the MAF (0.458) was higher than that predicted by 1,000 Genomes (European MAF 0.383); however, this difference did not meet our statistical significance threshold (p = 0.062).

GNB3 encodes a Gβ subunit of the G protein heterotrimer. Its SNP, rs5443, has been studied extensively in other contexts and is associated with numerous metabolic-related diseases, including obesity [38], coronary artery disease, and hypertension [39], as well as the increased efficacy of antidepressants in major depressive disorder [40]. All of these associations are with minor allele (T) carriers. The HapMap European MAF is 0.367, whereas the CRS patients of this study showed an increased MAF of 0.493 (p = 0.014).

While RGS21 is functionally linked to bitterant GPCR signaling [20, 21, 36], neither RGS21 SNP tested was found in this study to be associated with CRS (although the minor allele of RGS21 SNP rs7528947 was seen more frequently in CRS patients with higher Lund-Mackay CT staging scores). Conversely, GNB3 SNP rs5443, a gene variation in an established component of bitterant GPCR signaling [41] but without a previous association with CRS, was found to be more highly prevalent in West Virginia CRS patients, along with two TAS2R variants previously identified as being associated with CRS. Gβ subunits of G protein heterotrimers, such as that encoded by GNB3, are known to assemble with E3 ubiquitin ligase complexes responsible for degradation of GRK2, a GPCR kinase [42]. Recent work has shown that GNB3 SNP rs5443, identified more commonly in the CRS patients of this study, is associated with decreased GRK2 ubiquitination [43]. Further studies are therefore required to clarify the role of altered GNB3 function (and potentially GRK2 function) in bitterant signaling, innate immunity activation/regulation, and CRS.

We thank the American Academy of Otolaryngology – Head and Neck Surgery Foundation (AAO-HNSF) Annual Meeting in Chicago, IL, USA, September 10–13, 2017, for accepting an oral presentation by one of the authors (B.L.A.) of an interim report on this study.

The study protocol was approved by the Institutional Review Board of WVU Office of Research Integrity and Compliance. The subjects of this study gave their written informed consent as obtained under WVU IRB 1410476782.

The authors have no conflicts of interest to declare.

This research was supported in part by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number 5U54GM104942-03 to the West Virginia Clinical and Translational Science Institute (to B.L.A. and D.P.S.). Funding support was also provided by the WVU E.J. Van Liere Endowed Medicine Professorship (to D.P.S.). Neither of these sources participated in the preparation of data or the paper.

David P. Siderovski: conception and design, data analysis and interpretation, revising the manuscript, final approval, and agreement to be accountable. Phillip R. Purnell: conception and design, acquisition of data, data analysis and interpretation, drafting and revising the article, final approval, and agreement to be accountable. Benjamin L. Addicks: conception and design, acquisition of data, data analysis and interpretation, revising the manuscript, final approval, and agreement to be accountable. Habib G. Zalzal: conception and design, acquisition of data, revising the manuscript, final approval, and agreement to be accountable. Scott Shapiro: conception and design, acquisition of data, revising the manuscript, final approval, and agreement to be accountable. Sijin Wen: data analysis and interpretation, revising the manuscript, final approval, and agreement to be accountable. Hassan H. Ramadan: conception and design, revising the manuscript, final approval, and agreement to be accountable. Vincent Setola: conception and design, data interpretation, revising the manuscript, final approval, and agreement to be accountable.

Edited by: I. Lehmann, Berlin.