Association between the CDHR3 rs6967330 risk allele and chronic rhinosinusitis

Capsule summary

A retrospective, multi-center study of adults with and without chronic rhinosinusitis (CRS) identifies a significant association between rs6967330 in the viral receptor CDHR3 – known to be associated with wheezing and asthma in children – and the development of CRS.

To the Editor

Rhinoviruses (RV) are the most frequent virus isolated in chronic rhinosinusitis (CRS)¹ and asthma exacerbations.² CRS and asthma are both heterogeneous inflammatory disorders of the airway; however, their frequent co-occurrence suggests a unified airway endotype with a common pathophysiology. A recent genome-wide association study (GWAS) identified a missense single nucleotide polymorphism (SNP) in the Cadherin related family member 3 gene (CDHR3) that results in a cysteine to tyrosine substitution at position 529 (rs6967330). This rs6967330 risk allele (A) was found to be a significant additive risk factor for acute hospitalizations secondary to asthma in the first 6 years of life.³ Further mechanistic studies identified CDHR3 as the receptor for RV-C, an especially virulent subtype of RV highly correlated with severe airway disease.⁴ Given that sinonasal epithelia is the primary site for RV infection,⁵ we hypothesized that the combination of the rs6967330 genetic SNP with RV-C infections could be responsible for the development of CRS in adults.

We performed a retrospective case-control study of non-Hispanic white adults recruited from two different tertiary academic rhinology centers: the University of Arizona (UofA) and the University of Pennsylvania (UPenn) with diagnosed CRS (UofA n=76, UPenn n=267) and healthy controls (UofA n=52, UPenn n=337). Further details are available in the online methods and the participant’s characteristics are shown in Table E1. In the UPenn group, those with CRS were significantly older and more likely to be males compared to the controls. At both UofA and UPenn the prevalence of asthma was approximately 2-fold greater in the CRS compared to the control group.

DNA was extracted from buccal swabs or saliva samples collected from all study participants and rs6967330 was genotyped. In the control group for both populations, the allele frequency of the minor A allele was similar to that seen in the general non-Hispanic white population.³ There were no significant differences in subjective or objective measures of CRS including nasal polyp status in individuals with CRS related to CDHR3 genotype. We calculated the risk for CRS associated with rs6967330, stratified by location and adjusted for both sex and age in additive and dominant models. Additive models included genotype as an ordinal variable with three categories (0, 1, 2 for GG, AG, and AA genotypes) and dominant models included genotype as a binary variable (0, 1 for GG vs. AA+AG genotypes). Using an additive model, rs6967330 significantly increased the odds for CRS by 2.08 at the UofA (95% CI, 1.02–2.41, p=0.043) for each additional ‘A’ allele and remained significant when adjusting for sex and age. Using a dominant model, rs6967330 significantly increased the odds for CRS by 1.5 at UPenn (95% CI, 1.0–2.1, p=0.022) if an individual carried one or two copies of the minor A allele of rs6967330. Meta-analysis estimates of the UofA and UPenn populations demonstrated significantly increased risk for CRS in both additive and dominant models, which remained essentially unchanged after adjustment for age and sex. (Table 1) Excluding adults with co-occurring asthma did not significantly change this risk (mOR=1.55, 95% C, 1.11–2.17; p=0.011), suggesting that this association was not confounded by the presence of asthma (Figure 1).

Table 1.

Association between rs6967330 and CRS at UofA and UPenn

	Unadjusted					Adjusted for sex and age
All Participants	n	OR	95%CI	P	PHet*	n	adjOR	95%CI	P	PHet*
Additive Model
UofA	128	2.08	1.02, 4.21	0.043		126	2.14	1.04, 4.42	0.039
UPenn	604	1.31	0.97, 1.78	0.077		595	1.28	0.91, 1.80	0.162
Meta	732	1.41	1.07, 1.86	0.015	NS	721	1.40	1.03, 1.92	0.032	NS
Dominant Model
UofA	128	2.00	0.94, 4.23	0.071		126	2.06	0.95, 4.46	0.068
UPenn	604	1.50	1.06, 2.11	0.022		595	1.46	0.99, 2.17	0.057
Meta	732	1.57	1.15, 2.15	0.005	NS	721	1.57	1.11, 2.23	0.012	NS
Non-Asthmatic Participants
Additive Model
UofA	95	2.09	0.93, 4.72	0.076		93	2.05	0.86, 4.78	0.095
UPenn	442	1.48	1.02, 2.15	0.038		435	1.43	0.95, 2.15	0.087
Meta	537	1.57	1.12, 2.20	0.009	NS	528	1.53	1.06, 2.21	0.023	NS
Dominant Model
UofA	95	2.04	0.88, 4.74	0.095		93	2.01	0.84, 4.81	0.115
UPenn	442	1.85	1.22, 2.80	0.004		435	1.77	1.12, 2.81	0.014
Meta	537	1.88	1.30, 2.73	0.001	NS	528	1.82	1.22, 2.74	0.004	NS

^*Heterogeneity p-value based on chi-squared (NS indicates no heterogeneity between studies)

Figure 1.

Meta-analysis of rs6967330 from UofA and UPenn. The rs6967330 SNP is significantly associated with CRS in both additive and dominant models. This association remains significant in those without self-reported asthma (non-asthmatics). mOR was adjusted for sex and age.

RV infections are the most frequent cause of upper respiratory infections (URIs), otherwise known as the common cold. In most individuals, URIs resolve within 5–7 days. However, it is unknown why in adults, symptoms persist and can progress into CRS or trigger asthma exacerbations. The primary site of RV infection is the sinonasal epithelia. Until recently, the only known RV subtypes (A and B) bound to the low-density lipoprotein receptor (LDLR)⁵ and the intercellular adhesion molecule-1 (ICAM-1) receptors expressed on the apical surface of ciliated airway epithelia.⁶ In 2006, a third subtype, RV-C, was discovered using molecular detection techniques⁴, and found to be associated with more severe airway disease. RV-C isolates were identified with significantly higher frequency in the nasal washes of individuals with CRS,¹ and determined to be a trigger for acute wheezing lower respiratory illnesses (WLRI) and the subsequent development of childhood asthma.⁷

CDHR3 was first identified as a candidate gene for early childhood asthma with severe exacerbations in a large GWAS³ The CDHR3 protein is highly expressed in both upper and lower airway epithelia, and when plasmid DNA containing the mutation domain 5 (C₅₂₉Y) was transfected into 293T cells it resulted in a marked increase of CDHR3 trafficking to the cell membrane compared to the wild-type (G) major allele. Moreover, CDHR3 was identified as the putative receptor for RV-C, with the homozygous presence of the rs6967330 risk allele increasing RV-C binding by ten-fold in a transduced HeLA cell line stably expressing CDHR3-Y₅₂₉ compared to control HeLa cells.⁸

In summary, we have determined that carriers of rs6967330 are at significantly increased risk for CRS. The association between rs6967330 and airway disease was first reported in a study of acute wheezing lower respiratory illnesses (WLRI) in young children³. It is now well established that WLRI caused by rhinovirus in this age group are strongly associated with the subsequent development of asthma.⁷ Similarly, CRS usually starts with acute URI episodes, most often caused by rhinoviruses, and progresses into a disease characterized by quiescent periods followed by acute flare-ups. CDHR3 rs6967330 may be a common genetic risk factor between viral-associated childhood asthma and CRS.

Abbreviations

CRS

chronic rhinosinusitis

CDHR3

Cadherin related family member 3 gene

RV

Rhinovirus

URIs

upper respiratory infections

LDLR

low-density lipoprotein receptor

ICAM-1

intercellular adhesion molecule-1

WLRI

wheezing lower respiratory illnesses