Genome-Wide Association Study Reveals Novel Associations of Annexin A13 to Secretory and GAS2L2 with Mucous Otitis Media

doi:10.21203/rs.3.rs-4044059/v1

Download PDF

Research Article

Genome-Wide Association Study Reveals Novel Associations of Annexin A13 to Secretory and GAS2L2 with Mucous Otitis Media

https://doi.org/10.21203/rs.3.rs-4044059/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose: To evaluate the genetics of chronic non-suppurative otitis media (OM).

Methods: We performed a genome-wide association study of 429,599 individuals included in the FinnGen study using three different case definitions: any type of chronic non-suppurative OM (7,034 cases), mucous chronic OM (5,953 cases), and secretory chronic OM (1,689 cases). Individuals without otitis media were used as controls (417,745 controls). We used immunohistochemistry of the murine middle ear to evaluate the expression of annexin A13.

Results: Four loci were significantly associated (p<1.7x10-8) with non-suppurative OM. Three out of the four association signals included missense variants in genes having a plausible role in otitis media pathobiology. In subtype-specific analyses, one novel locus, near ANXA13, was associated with secretory OM. Three loci (near TNFRSF13B, GAS2L2, and TBX1) were associated with mucous OM. Immunohistochemistry of the murine middle ear samples revealed expression of annexin A13 at the apical pole of the Eustachian tube’s epithelium as well as variable intensity in the secretory cells of the glandular structure in the proximity of the Eustachian tube.

Conclusions: We demonstrate that secretory and mucous OM have distinct and shared genetic associations. The association of GAS2L2 implies a connection with ciliary epithelium function and the pathogenesis of dysfunctional mucosa in mucous OM. The abundant expression of annexin A13 in the Eustachian tube epithelium, along with its implicated role in apical transport for the binding and transfer of phospholipids, indicates the role of annexin A13 and phospholipids in Eustachian tube dysfunction.

Bioinformatics

Keywords: Finngen

otitis media

genome-wide association

annexin

GAS2L2

Eustachian tube

Acute otitis media (AOM) is characterized by acute infectious symptoms, otalgia, and suppurative middle ear effusion (MEE) [1-4,34]. Non-suppurative otitis media is an otitis media which involves transudation of fluid in the middle ear without pus formation. Acute nonsuppurative otitis media refers to the tubal pharynx, mouth, and cartilage segments, inflammatory mucosal hyperemia, swelling, and congestion after acute upper respiratory tract infection and may be accompanied by bacteria or viruses via the Eustachian tube, directly into the middle ear cavity, resulting in an inflammatory reaction in the middle ear mucosa [1-4]. Middle ear effusions persisting for more than three months are considered chronic and can be categorized as serous (or mucous based on the quality of the effusion. Chronic suppurative otitis media (CSOM) is a distinct entity from the non-suppurative mucous and secretory chronic otitis media with effusion (COME) and presents with infectious symptoms and purulent otorrhea and tympanic membrane perforation. The classification of middle ear fluid to either serous or mucous can, in some cases, be vague. However, the reason there is a separate diagnosis code for these two types of middle ear fluid is to precisely characterize the patients’ clinical findings [1,2,3].

Among young children with COME, effusion is most often mucoid [5], whereas in adults secretory COME is more common. As COME causes conductive hearing loss that may require surgical treatment, it significantly affects quality of life and increases health care costs. In childhood, there is usually a history of acute OM before the onset of COME, when secretion fails to resolve from the middle ear, even though the acute OM has been resolved. The etiology for adult-onset COME is variable and includes sinonasal and skull base lesions, chronic sinusitis, allergies, and Eustachian tube dysfunction. The precise mechanisms that contribute to COME are, however, poorly understood [3].

Genetics could potentially be used to gain a better understanding of OM pathophysiology. Based on the findings of twin studies, the heritability of OM (the fraction of phenotype variation that can be attributed to genetic variation between individuals) has been estimated to be between 40% and 70% [6-10]. In a further three studies, otitis media has been studied using genome-wide association study (GWAS) approaches [11-13]. To date, the largest study of childhood ear infections included 121,810 participants. In total, four genetic loci have been associated with risk for OM or similar outcomes [11-13].

At present, it is unclear whether secretory and mucous chronic otitis media are two phases of the same entity, or the underlying pathophysiology is distinct. Therefore, the aim of our study was to perform a separate genetic characterization of mucous and secretory OM to gain a further insight into the disease process and to elaborate on the shared and distinct genetic risk factors. We performed three genome-wide association studies (GWAS) with 417,745 individuals from the FinnGen study using different case definitions: any type of chronic non-suppurative OM (n cases=7,034), mucous chronic OM (ICD H65.3; n cases=5,953), and secretory OM (ICD H65.2; n cases=1,689). We revealed two distinct loci and two shared loci for the two subcategories of non-suppurative OM. Further, we evaluated the role of annexin A13 using immunohistochemistry in a murine model.

FinnGen study data release 10 and ethics

Launched in 2017, the FinnGen study (https://www.finngen.fi/en) is a public-private research project that will, over time, combine the genome and digital health care data of 500,000 Finns. The goal of the project is to provide a novel medically and therapeutically useful insight into human diseases. The FinnGen study is a pre-competitive collaboration between Finnish biobanks and their supporting organizations (universities and university hospitals) as well as international pharmaceutical industry partners and a Finnish biobank cooperative (FINBB). In the current study, we included data that were released in autumn 2022. (Release 10). For release ten, there was a total of 412,181 post-QC samples, 430,897 (pre-QC samples), 429,207 individuals with endpoints, and 429,861 individual with detailed longitudinal data. In the study, information about disease diagnoses and the procedures performed were collected from the Care Register for Health Care (THL) and the Causes of Death Register (Statistics Finland). As the population of northern and eastern Finland has expanded dramatically in isolation, following a series of bottlenecks, it harbors numerous deleterious alleles at a high frequency [42].

Ethics statement and materials & methods:

Patients and control subjects in the FinnGen study provided informed written consent for biobank research, based on the Finnish Biobank Act. Alternatively, separate research cohorts collected prior to the Finnish Biobank Act coming into force ( September 2013) and the start of the FinnGen study (August 2017), were collected based on study-specific consents and later transferred to the Finnish biobanks after approval by the Finnish Medicines Agency (Fimea), the National Supervisory Authority for Welfare and Health. Recruitment protocols followed the biobank protocols approved by Fimea. The Coordinating Ethics Committee of the Hospital District of Helsinki and Uusimaa (HUS) statement number for the FinnGen study is Nr HUS/990/2017.

The FinnGen study is approved by the Finnish Institute for Health and Welfare (permit numbers: THL/2031/6.02.00/2017, THL/1101/5.05.00/2017, THL/341/6.02.00/2018, THL/2222/6.02.00/2018, THL/283/6.02.00/2019, THL/1721/5.05.00/2019 and THL/1524/5.05.00/2020), Digital and population data service agency (permit numbers: VRK43431/2017-3, VRK/6909/2018-3, VRK/4415/2019-3), the Social Insurance Institution (permit numbers: KELA 58/522/2017, KELA 131/522/2018, KELA 70/522/2019, KELA 98/522/2019, KELA 134/522/2019, KELA 138/522/2019, KELA 2/522/2020, KELA 16/522/2020), Findata permit numbers THL/2364/14.02/2020, THL/4055/14.06.00/2020, THL/3433/14.06.00/2020, THL/4432/14.06/2020, THL/5189/14.06/2020, THL/5894/14.06.00/2020, THL/6619/14.06.00/2020, THL/209/14.06.00/2021, THL/688/14.06.00/2021, THL/1284/14.06.00/2021, THL/1965/14.06.00/2021, THL/5546/14.02.00/2020, THL/2658/14.06.00/2021, THL/4235/14.06.00/2021, Statistics Finland (permit numbers: TK-53-1041-17 and TK/143/07.03.00/2020 (earlier TK-53-90-20) TK/1735/07.03.00/2021, TK/3112/07.03.00/2021), and the Finnish Registry for Kidney Diseases permission/extract from the meeting minutes on 4^th July 2019.

The Biobank Access Decisions for FinnGen samples and the data utilized in FinnGen Data Freeze 10 include: THL Biobank BB2017_55, BB2017_111, BB2018_19, BB_2018_34, BB_2018_67, BB2018_71, BB2019_7, BB2019_8, BB2019_26, BB2020_1, BB2021_65, Finnish Red Cross Blood Service Biobank 7.12.2017, Helsinki Biobank HUS/359/2017, HUS/248/2020, HUS/150/2022 § 12, §13, §14, §15, §16, §17, §18, and §23, Auria Biobank AB17-5154 and amendment #1 (August 17 2020) and amendments BB_2021-0140, BB_2021-0156 (August 26 2021, Feb 2 2022), BB_2021-0169, BB_2021-0179, BB_2021-0161, AB20-5926 and amendment #1 (April 23 2020)and it´s modification (Sep 22 2021), Biobank Borealis of Northern Finland_2017_1013, 2021_5010, 2021_5018, 2021_5015, 2021_5023, 2021_5017, 2022_6001, Biobank of Eastern Finland 1186/2018 and amendments 22 § /2020, 53§/2021, 13§/2022, 14§/2022, 15§/2022, Finnish Clinical Biobank Tampere MH0004 and amendments (21.02.2020 & 06.10.2020), §8/2021, §9/2022, §10/2022, §12/2022, §20/2022, §21/2022, §22/2022, §23/2022, Central Finland Biobank 1-2017, and Terveystalo Biobank STB 2018001 and amendment 25^th Aug 2020, the Finnish Hematological Registry and Clinical Biobank decision 18^th June 2021, Arctic biobank P0844: ARC_2021_1001.

Phenotype descriptions and analysis

We began by performing a genome wide association study (GWAS) of all instances of chronic non-suppurative otitis media. The following codes were used as inclusion factors for the case cohort: at least one entry of the International Classification of Diseases, Tenth Revision (ICD-10) codes H65.2 or H65.3 and the International Classification of Diseases, Ninth Revision (ICD-9) codes 3812A and 3811A. Individuals with no record of chronic secretory otitis media (ICD-10 code H65.2 or ICD-9 code 3811A), chronic mucous otitis media (ICD-10 code H65.3 or ICD-9 code 3812A), nonspecific otitis codes (ICD-10 code H65.9, ICD-10 code H65.4, or ICD-9 code 3813X) or non-suppurative otitis (ICD-10 code H65.1 or ICD-9 code 3810A) were included in the control cohort. The GWAS of overall non-suppurative otitis media was performed with a case cohort of individuals of either secretory or mucous otitis media. The GWAS for chronic secretory otitis media was performed with a case cohort of individuals with chronic secretory otitis media (ICD-10 code H65.2). All cases with reported chronic secretory otitis media were included without any exclusion criteria. Finally, a case cohort of individuals with chronic mucous otitis media (ICD-10 code H65.3) was used for a GWAS of chronic mucous otitis media, without any exclusion criteria. For all three GWAS, the control cohort remained unchanged. Chronic non-suppurative otitis media can be secretory or mucus. In clinical practice, however, the borderline can be hazy in some instances, and there is known to be a case overlap. Recognizing a risk for bias, we found 608 cases among our analyzed data with both ICD-10 codes. After exclusion of the overlapping cases, 1,081 chronic secretory otitis media cases remained. The number of cases with only chronic mucous otitis media (excluding overlapped cases) was 5,345. However, we did not exclude the 608 overlapped cases from our primary analyses, as it was a considerable number of cases to remove from a case cohort of 1,689 individuals (ICD-10= H65.2 code). Therefore, according to the "less strict" definition, there were 1,689 participants in the case cohort for secretory otitis media and 5,953 participants in the case cohort for mucous otitis media. The overall number of cases of chronic secretory or mucous otitis media was 7,034, which included cases with either one or both diagnoses. The controls cohort included

417,745 cases. To correct for the population substructure, the outcome associations were tested using an additive model and the standard covariates: sex, age, and the first ten principal components from the genetic data. We performed a chi-square test of heterogeneity to explore whether the effect estimates differed significantly between the subtypes of OM.

Genotyping, imputation, and quality control

The FinnGen study uses a biobank sample consisting of 1) prospective samples (“legacy samples”) and 2) “new samples” that are collected upon request to obtain voluntary biobank consent and donate biobank samples (usually blood). Hospital Biobank samples and Terveystalo Biobank samples are typically collected during diagnostic sampling in hospital laboratories or on hospital wards. Blood Services Biobank approval and sample collection is also performed in connection with blood donation. Approval and sampling by the THL Biobank are typically done in connection with the collection of research samples. In contrast, legacy samples are older sample cohorts that were collected for a specific research project before the Finnish Biobank Act came into force (September 2013).
These older study cohorts were later transferred to a biobank according to the Finnish Biobank Act 13 §. FinnGen samples were genotyped with ThermoFisher, Illumina, and Affymetrix arrays.

Chip genotype data processing and QC Samples were genotyped with Illumina (Illumina Inc., San Diego, CA, USA) and Affymetrix arrays (ThermoFisher Scientific, Santa Clara, CA, USA). Chip genotyping data produced with earlier chip platforms and reference genome builds were lifted over to build version 38 (GRCh38/hg38) following the protocol described here (dx.doi.org/10.17504/protocols.io.xbhfij6). The ‘legacy samples’ were genotyped over the years using various generations of the Illumina and Affymetrix GWAS arrays.

Genotype calls were made with the GenCall and zCall algorithms for the Illumina data and the AxiomGT1 algorithm for the Affymetrix data. The ‘new samples’ were genotyped with the FinnGen ThermoFisher Axiom custom array at the ThermoFisher genotyping service in San Diego, CA, US. The current array (v2) contains 723, 376 probesets for 664,510 markers. In addition to the core GWAS markers (about 500,000), it contains about 116,000 coding variants enriched in Finland, >10,000 specific markers for the HLA/KIR region, almost 15,000 ClinVar variants, about 4,600 pharmacogenomic variants, and 57,000 selected markers that were of special interest for partners.

All GWAS data were imputed against a Finnish population-specific whole genome sequence (WGS) backbone. In sample-wise quality control steps, individuals with ambiguous sex, high genotype missingness (>5%), excess heterozygosity (+/-4SD), and non-Finnish ancestry were excluded. This resulted in the inclusion of samples from 412,181 individuals included in the association analysis. In variant-wise quality control steps, variants with high missingness (>2%), low Hardy-Weinberg equilibrium (HWE) P-value (<1x10^-6), and low number of minor alleles (MAC<3) were excluded. This resulted in the inclusion of 21,311,942 variants in the association analysis.

Genome-wide associations

To correct for the population substructure, the outcome associations were tested using an additive model and the standard covariates: sex, age, and the first ten principal components from the genetic data. We performed a chi-square test of heterogeneity to explore whether the effect estimates differed significantly between the subtypes of OM.

Characterization of the associated loci

The genomic regions within the +/- 1 Mb window around the primary variant are called the associated locus. Therefore, each connected locus consisted of at least one genome-wide significant variant (P < 5x10 ^-8) separated by at least a 1 Mb variable. According to the NHGRI-EBI catalog of human genome-wide association studies, the locus was considered novel. Candidate genes in each new locus were prioritized based on physical proximity to the index variant and the prior literature on the biological role of genes and clinical relevance. Variants with allele frequency (AF) of less than 1% were excluded.

By using bioinformatics, conservation of the residue (position) was analyzed using ConSurf with default settings. PolyPhen-2 was used to predict the pathogenicity of the variant [35]. The structural context of the target residue was analyzed using the Protein Data Bank and the AlphaFold database. The appearance of the mutation was assessed among somatic cancer mutations using ProteinPaint (see Table S1) .

Immunohistochemistry

The immunohistochemistry (IHC) protocol was modified as needed and several antibody dilutions were tested. Murine endogenous immunoglobulins were blocked with Rodent Block solution (Biocare Medical, Concord, CA, US). By using negative controls, we were able to test that the solution did not leak and that the mouse’s own antibodies were specifically blocked by the primary antibody. Thus, the technical part of the optimization of the protocol was successful.

We used immunohistochemistry of murine middle ear samples to evaluate the expression of annexin A13 in the Eustachian tube and the middle ear. Samples from murine colon were used as positive controls. According to the protein-atlas, there is an abundance of annexin A13 in the intestinal columnar epithelium. (https://www.proteinatlas.org/ENSG00000104537-ANXA13/tissue) Ethical permission for the use of mice for research purposes was obtained (permit number =ESAVI/23659/2018). Four adult mice (all of them were C57bl/6 wild type mice) were euthanized by CO2 inhalation according to guidelines for the euthanasia of rodents.

The rodents were then decapitated, and their brains removed using blunt dissection. Middle ear specimens and temporal bones were then dissected as previously reported (http://goodrich.med.harvard.edu/uploads/3/7/7/1/37718659/cochlearwholemountstainprotocol_120420_lw.pdf). After dissection, the tissue was fixed with 4% paraformaldehyde (PFA). The specimens were decalcified with 120 mM EDTA solution for 3 days at 4 ^°C and then embedded in paraffin. Sectioning was performed with Leica Microtome SM2010 using 4 µm-5 µm section thickness and 1-2 sections per slide. Paraffin sections were placed on Superfrost Plus microscope slides. Fully automated immunostaining was performed by a LabVision Autostainer instrument. Xylene steps were used for deparaffinization and ethanol series for the rehydration of the sections. Antigen retrieval was performed at 98 ^°C for 30 minutes (98 ^°C PT-Module) with 10 mM Tris-1mM EDTA Buffer (+0.05% Tween 20, pH 9.0). TBS-0.05%Tween washing buffer was used for the washing steps. Slides were then incubated in the LabVision Autostainer at room temperature for 30 minutes with an un-conjugated polyclonal primary antibody for annexin A13 (ABIN7006070) 1:600. Primary antibody specific to ANXA3 was bought from ELISA Kits & Proteins for Life Science Research (antibodies-online.com) (Catalog No. ABIN7006070). Endogenous peroxidase was blocked with 3% H₂0₂. Histofine SimpleStain Mouse Max Po Rabbit 414341F (Nichirei, Tokyo, Japan) was used as a detection method for primary antibody with diaminobenzidine (DAB) chromogen. As negative controls, we used both temporal bone and colon samples. Although the same protocol was followed, we excluded the antibody for annexin A13. Finally, the slides were counterstained with Hematoxylin, dehydrated, cleared, and mounted with xylene-based cover-slipping. Thereafter, the slides were scanned by Nanozoomer scanner (Hamamatsu photonics). The preparation of the slides was performed by an experienced laboratory technician. The sections were then examined with a light microscope by a pathologist and medical cell biologist.

In cases with a broad definition of OM (either mucous or secretory otitis media) in the GWAS, we detected an association in four loci at genome-wide significance (p<1.7x10^-8 and AF>1%) (Table 1). The chromosome 22 locus included a known leading variant rs1978060 near the TBX1 gene, which has previously been associated with ear infection [13,40]. In the proximity of that variant, there was a novel missense variant rs72646967 (N (Asn) > H(His) which, according to variant effect predictors, is tolerated, although with low confidence (CADD = 17.88) [35,44,45]. Based on a recent study, the rs72646967 variant has been linked to lung function. [28].

The chromosome 17 locus harbored a novel missense lead variant rs72553883 in TNFRSF13B that has not previously been associated with ear infection even though the gene TNFRSF13B has been linked to ear infection [13]. The second locus in chromosome 17 included a missense lead variant rs3744374 (A (Ala) > (Gly)) in the GAS2L2 gene. Based on a recent study, rs3744374 is protective for chronic rhinosinusitis [17]. According to the Ensembl database and the variant effect predictors Polyphen and SIFT [35,43], both rs72553883 (CADD=8.96) and rs3744374 (CADD=22) are predicted to benign and well tolerated [45,46]. In chromosome 8 locus, a novel leading variant, rs3779980, was near ANXA13. In addition to the lead variant, we observed a novel missense variant rs2294013 (R (Arg) > H (His)) in ANXA13, which had genome-wide significance. Bioinformatic analysis with PolyPhen-2 predicted this mutation to be rather well tolerated and the mutated residue is not highly conserved (Suppl. table S3).

In the GWAS of mucous otitis media, we revealed three associated loci of genome-wide significance (p<1.7x10^-8), (Table 1, Figure 1, Supplementary Figure 3(b-d).) The chromosome 17 locus harbored the same two missense variants that have already been reported above (Table 1) [13,14,17]. The chromosome 22 locus, near TBX1, included the same leading variant rs1978060-A/G (OR=1.15) that appears in the GWAS for the combined secretory and mucous otitis media phenotype (Figure 2, Forest plots). However, no association was seen for the locus on chromosome 8 observed in our first GWAS that included both secretory and mucous otitis media cases (Table 1). In contrast, in the GWAS of chronic secretory otitis media, we found one locus at genome-wide significance (p<5x10 ^-8, Table 1, Supplementary Figure 3a) on chromosome 8. All genome-wide significant variants in this locus were found within ANXA13. In addition to the lead variant, we also saw a missense variant rs2294013 (R (Arg) > H (His)) in ANXA13 at genome-wide significance. The missense rs2294013 variant is predictedto be benign based on the variant effect predictors PolyPhen and SIFT, with a CADD of 12.3 [35]. For all the referred variants, the GWAS analyses and Manhattan plots are available in Figure 1 and locus zooms for all the reported variants are available in Figure 5 in the Supplementary Information.

We further compared the effect sizes of index variants from any of the three GWAS studies to assess for heterogeneity based on phenotype definition (Table 3). The effect on chronic secretory otitis media was significantly higher in chromosome 8 locus, ANXA13 for both lead variants rs3779980 (OR=0.77), and missense variant rs2294013 (OR=1.28) than that of mucous otitis media and cross-phenotype meta-analysis without overlapping confidence intervals. In contrast, the effect on chronic mucous otitis media was higher for the GAS2L2 gene variant rs3744374 (OR=0.90) than that of chronic secretory otitis media without overlapping confidence intervals (Figure 2. Forest plots).

The critical value for Chi-Square for degrees of freedom (df) 1 and P=0.05 is 5.991. Thus, data for the ANXA13 (P=0.005) and GAS2L2 (P=0.05) variants revealed the variability across effect sizes exceeds what would be expected based on sampling error. However, for the TNFRSF13B (P=0.95) and TBX1 (P=0.9) variants, the variability across effect sizes did not exceed what would be expected based on sampling error.

In GWAS for both mucous and secretory otitis media (Figure 1-a) and for cases with only mucous otitis media (Figure 1-b), there was a strong locus in chromosome 6. This locus was in the human leukocyte antigen (HLA)/ (the major histocompatibility complex (MHC) area, and it was not studied further or analyzed at that time.

Phenome-wide association studies of the lead variants in the FinnGen study

We evaluated the co-associations of the lead variants in the FinnGen study. Among the variants associated with chronic mucous otitis media, the missense variant rs72553883 in TNFRSF13B was associated most strongly with chronic mucous otitis media in the primary GWAS. Furthermore, it was also strongly associated with multiple diseases, including chronic diseases of the tonsils and adenoids, nonsuppurative otitis media, eustachian salpingitis, and tonsillectomy (p<1.7x10^-8) (Supplementary Table S2). The missense variant rs3744374 near GAS2L2 was significantly associated with chronic sinusitis, diseases of the middle ear and mastoid, nonsuppurative otitis media, and other diseases of the upper respiratory tract (p<1.7x10^-8). The variant rs1978060 near TBX1 was significantly associated with suppurative and non-suppurative otitis media, nasal polyps, and diseases of the middle ear and mastoid (p<1.7x10^-8). The variant chromosome 8 locus and the reported variants near ANXA13 were associated with non-suppurative otitis media, even though the association is clearly stronger with the more specific phenotype of secretory otitis media.

Immunohistochemistry

To further assess the potential role of ANXA13 in otitis media pathophysiology, we conducted immunohistochemistry experiments that revealed expression of annexin A13 at the apical pole of the Eustachian tube’s epithelium as well as variable intensity in the secretory cells of the glandular structure in the proximity of the Eustachian tube (Figure 3 and Figure 4).

As previously reported in the literature [19], the expression of annexin A13 was intense in the intestinal epithelium. Furthermore, negative controls stay clearly negative.

Cross referencing with UKBB results

To further validate our findings, we investigated how the variants were associated with similar outcomes in publicly available UKBB data. The UKBB associations for the variants in Table 1 are shown in Supplementary Table S2. In short, the ANXA13 leading variant was associated with otitis media and Eustachian tube disorders. In particular, the phenotype otitis media and Eustachian disorder corresponds quite well with our chronic secretory otitis media phenotype. Additionally, the TBX1 leading variant was associated with the nasal polyps’ phenotype. For both the ANXA13 and TBX1 variants, allele frequency (AF) was the same in both the Finngen and the UKBB data.

This is the first description of a subtype GWAS for chronic non-suppurative otitis media. Here, we found four loci that were significantly associated (p<1.7x10^-8) with overall non-suppurative OM. Moreover, three out of the four association signals included missense variants in genes with a plausible role in otitis media pathobiology. In the subtype-specific analyses, one novel locus near ANXA13 was associated with the secretory subtype and three loci (near TNFRSF13B, GAS2L2, and TBX1) were associated with the mucous subtype. Interestingly, the association of ANXA13 is replicated in the UKBB data.

Immunohistochemistry revealed expression of annexin A13 at the apical pole of the Eustachian tube’s epithelium as well as at variable intensity in the secretory cells of the glandular structure in the proximity of the Eustachian tube. We discussed the potential mechanisms by which proteins encoded by the highlighted genes could lead to OM susceptibility.

There are two variants each for ANXA13 and TBX1; however, based on the odds ratios, their direction of effects is opposite to each other for both genes. (Figure 2) Regarding the ANXA13 locus, rs3779980 is the lead variant, whereas for the missense variant, rs2994013, r² was between 0.6 to 0.8, which means that these two variants are not in complete linkage disequilibrium (LD) but are highly correlated. Regarding the TBX1 variants, rs1978960 is the lead variant and for the missense variant, rs72646967, r² is between 0.2 to 0.4. This means that they are not in linkage disequilibrium.

Annexin-13 (ANXA13)

Annexin A13 is considered the original progenitor of the 11 other members of the vertebrate annexins. The specific function of this gene is not yet fully understood. However, it is a calcium-dependent, phospholipid-binding protein and an apical vesicle which may play a crucial role in the lipid raft-mediated delivery of apical proteins [20]. Earlier work has shown that ANXA13 may also play a role in epithelial cyst formation and ciliary function [48].

The Eustachian tube connects the middle ear to the nasopharynx and it is important that it is patent so that the middle ear ventilates. Previously, it was observed that similarly to the lung there is surfactant in Eustachian tube epithelium that consists partially of phospholipid proteins. Based on earlier research [21], it is known that the phospholipid layer is needed to make the surface of the epithelium of the Eustachian tube hydrophobic. This is crucial for opposing the strongly adhesive nature of proteins and easing the patency of the Eustachian tube.

Together with our mouse work that links the expression of ANXA13 to the Eustachian tube, the findings suggest for the first time that ANXA13 plays an important role in the function of the Eustachian tube and the middle ear [7,20,22,23].Further, according to the Human Protein Atlas [9],cytoplasmic expression of annexin-13 has been reported in the gastrointestinal tract, bile ducts, gallbladder, renal tubules, and fallopian tubes.

A query of our reported leading variants in the UKBB revealed the strong association of the variant rs3779980 with otitis media and Eustachian tube disorders [18].Interestingly, the allele frequency (AF) and odds ratio (OR) were similar to those in our GWAS study in patients with secretory otitis media (Figure 2. Forest plots). Even though TNFRSF13B and GAS2L2 have previously been linked to childhood ear infection, those specific missense variants (rs72553883 and rs3744374) were not significantly associated with any phenotype in the UKBB. Finally, the variant rs1978060 was associated with nasal polyps (P= 5,80x10^-7) in the UKBB. (Supplementary Table S2)

TNF Receptor Superfamily Member 13B (TNFRSF13B)

The protein encoded by this gene is a lymphocyte-specific member of the tumor necrosis factor (TNF) receptor superfamily. It interacts with calcium-modulator and cyclophilin ligand (CAML). The protein induces activation of the transcription factors NFAT, AP1, and NF-kappa-B and plays a crucial role in humoral immunity by interacting with a TNF ligand. TNFRSF13B mediates the calcineurin-dependent activation of NF-AT as well as the activation of NF-kappa-B and AP-1. Additionally, it engages in the stimulation of B- and T-cell function and the regulation of humoral immunity[14]. The missense variant rs72553883,which was significantly associated with mucous otitis media in our analyses, is reported in the literature in multiple individuals diagnosed with common variable immunodeficiency [24,25,26]. This variant is seen in clinically asymptomatic relatives as well as in the Finnish European population with an overall allele frequency of 2.4% (603/24990 alleles, including eight homozygotes) in the Genome Aggregation Database. This suggests that these variants may act as a susceptibility or a disease-associated variant with other genetic and/or environmental factors[27]. According to Clinvar[15,16] and Ensembl[43], the missense variant rs72553883 is mostly likely pathogenic. Other variants in TNFRSF13B (rs11649804, rs34557412) have previously been linked to tonsillectomy (rs34557412) and childhood ear infection [13,17,47]. In line with these findings, we saw co-associations with sinusitis and chronic diseases of the tonsils and adenoids in the FinnGen study [17]. Functional characterization of the rs72553883 variant in mice and human B-cells reveal disrupted NF-KB signaling and significantly impaired B-cell function [29,30].

GAS2L2 (Growth Arrest Specific 2 Like 2)

Growth Arrest Specific 2 Like 2 (GAS2L2) is a protein-coding gene which has previously been associated with Primary Ciliary Dyskinesia (PCD) by impairing cilia orientation and mucociliary clearance [31,32]. Gene ontology (GO) annotations related to this gene include regulation of microtubule dynamics and stability by interacting with microtubule plus-end tracking proteins, such as MAPRE1 [33]. GAS2L2 regulates ciliary orientation and performance in cells found in the airway [30]. The variant rs3744374 has been recently linked with chronic rhinosinusitis and other sinonasal and pharyngeal diseases [17], although based on previous studies it is predicted to be benign [35,36]. There is a mouse model (C57BL/6N-Gas2l2tm1a(KOMP)Wtsi/Wtsi)with abnormal auditory tube and middle ear [24].

T-Box Transcription Factor 1 (TBX1)

Loss-of-function of the TBX1 gene causes dysregulation of neural competence in the otocyst regions linked to the formation of either mechanosensory or structural sensory organ epithelia [36-38]. The TBX1 gene is also needed for inner ear morphogenesis [39]. The leading variant, rs1978060, that was associated with mucous otitis media in this FinnGen study, has also been associated with nasal polyps (p=9x 10^-09). In an earlier GWAS study in the Finnish population, the variant rs1978060 was reported to have a discordant association with chronic rhinosinusitis [17].

In a GWAS meta-analysis of Japan Biobank data with data from the UKBB and Finngen, rs1978060 has been associated with nasal polyps [11,18,40]. In previous studies, GWAS data from research participants of European ancestry have revealed the association of rs1978060 with childhood ear infection and myringotomy [13,18,47]. Childhood ear infections are strongly influenced by the shape of the Eustachian tube. Murine models knocked out for the TBX1 gene have not demonstrated any Eustachian tube abnormalities but have revealed abnormal development of the ear (outer, middle, or inner ear) and hearing (decreased hearing or even deafness) [41]. These findings show the crucial role TBX1 playsin ear development.

Limitations of the study

Because the population of northern and eastern Finland has expanded dramatically and in isolation following a series of bottlenecks, it harbors numerous deleterious alleles at relatively high frequency compared to other European countries. This increase in allele frequency aids the statistical power to detect certain disease alleles. In contrast, the population history has also removed some genetic variants or lowered their frequency substantially, making it much more difficult to find those variants using the Finnish population [44].

Although the number of cases in our study is substantially lower than is usual in complex traits GWAS, we were able to detect loci associating with the risk of chronic non-suppurative otitis media and its subtypes. The four lead variants in the discovered loci here explained 0.4% to 0.8% of the trait variation, which means around 0.1% to 0.2% per locus. This is larger than typically found for common complex traits.

We revealed distinct genetic associations of ANXA13 and GASL2L2 with secretory OM and mucous OM. In contrast, variants of TBX1 and TNFRSF13B were associated with both subtypes of non-suppurative otitis media. Immunohistochemistry of annexin A13 in the ear in a murine model confirmed abundant expression of annexin A13 in the glandular epithelium of the Eustachian tube, supporting the role of annexin A13 in the function of the Eustachian tube via regulation of phospholipid metabolism. Further studies are required to explore the potential pathophysiological role of annexin A13 in chronic secretory otitis media and a potential association of surfactant to annexin A13.

GWAS: genome-wide association study, OM: otitis media, ANXA13: annexin A13, ET: Eustachian tube, GAS2L2: Growth Arrest Specific 2 Like 2, TNFRSF13B: TNF Receptor Superfamily Member 13B, TBX1: T-Box Transcription Factor 1

We would like to acknowledge the participants and investigators of the FinnGen study. The FinnGen project was funded by two grants from Business Finland (HUS 4685/31/2016 and UH 4386/31/2016) and the following industry partners: AbbVie Inc., AstraZeneca UK Ltd, Biogen MA Inc., Bristol Myers Squibb (and Celgene Corporation & Celgene International II Sàrl), Genentech Inc., Merck Sharp & Dohme LCC, Pfizer Inc., GlaxoSmithKline Intellectual Property Development Ltd., Sanofi US Services Inc., Maze Therapeutics Inc., Janssen Biotech Inc, Novartis AG, and Boehringer Ingelheim International GmbH. The following biobanks are acknowledged for delivering biobank samples to the FinnGen project: Auria Biobank (www.auria.fi/biopankki), THL Biobank (www.thl.fi/biobank), Helsinki Biobank (www.helsinginbiopankki.fi), Biobank Borealis of Northern Finland (https://www.ppshp.fi/Tutkimus-ja-opetus/Biopankki/Pages/Biobank-Borealis-briefly-in-English.aspx), Finnish Clinical Biobank Tampere (www.tays.fi/en-US/Research_and_development/Finnish_Clinical_Biobank_Tampere), Biobank of Eastern Finland (www.ita-suomenbiopankki.fi/en), Central Finland Biobank (www.ksshp.fi/fi-FI/Potilaalle/Biopankki), Finnish Red Cross Blood Service Biobank (www.veripalvelu.fi/verenluovutus/biopankkitoiminta), Terveystalo Biobank (www.terveystalo.com/fi/Yritystietoa/Terveystalo-Biopankki/Biopankki/) and Arctic Biobank (https://www.oulu.fi/en/university/faculties-and-units/faculty-medicine/northern-finland-birth-cohorts-and-arctic-biobank). All Finnish Biobanks are members of the BBMRI.fi infrastructure (www.bbmri.fi). The Finnish Biobank Cooperative -FINBB (https://finbb.fi/) is the coordinator of BBMRI-ERIC operations in Finland. The Finnish biobank data can be accessed through the Fingenious® services (https://site.fingenious.fi/en/) managed by FINBB.

Conflict of Interest statement

The authors declare no conflicts of interest.

Williamson IG, Dunleavey J, Bain J, Robinson D. The natural history of otitis media with effusion--a three-year study of the incidence and prevalence of abnormal tympanograms in four South West Hampshire infant and first schools. J Laryngol Otol. 1994 Nov;108(11):930-4. doi: 10.1017/s0022215100128567. PMID: 7829943.
Midgley EJ, Dewey C, Pryce K, Maw AR. The frequency of otitis media with effusion in British pre-school children: a guide for treatment. ALSPAC Study Team. Clin Otolaryngol Allied Sci. 2000 Dec;25(6):485-91. doi: 10.1046/j.1365-2273.2000.00360.x. PMID: 11122285.
Rosenfeld RM, Shin JJ, Schwartz SR, Coggins R, Gagnon L, Hackell JM, Hoelting D, Hunter LL, Kummer AW, Payne SC, Poe DS, Veling M, Vila PM, Walsh SA, Corrigan MD. Clinical Practice Guideline: Otitis Media with Effusion (Update). Otolaryngol Head Neck Surg. 2016 Feb;154(1 Suppl):S1-S41. doi: 10.1177/0194599815623467. PMID: 26832942.
Rye MS, Blackwell JM, Jamieson SE. Genetic susceptibility to otitis media in childhood. Laryngoscope. 2012 Mar;122(3):665-75. doi: 10.1002/lary.22506. Epub 2012 Jan 17. PMID: 22252840.
Duah V, Huang Z, Val S, DeMason C, Poley M, Preciado D. Younger patients with COME are more likely to have mucoid middle ear fluid containing mucin MUC5B. Int J Pediatr Otorhinolaryngol. 2016 Nov;90:133-137. doi: 10.1016/j.ijporl.2016.09.009. Epub 2016 Sep 12. PMID: 27729120; PMCID: PMC5123680.
Visscher PM, Hill WG, Wray NR. Heritability in the genomics era--concepts and misconceptions. Nat Rev Genet. 2008 Apr;9(4):255-66. doi: 10.1038/nrg2322. Epub 2008 Mar 4. PMID: 18319743.
Casselbrant ML, Mandel EM, Fall PA, Rockette HE, Kurs-Lasky M, Bluestone CD, Ferrell RE. The heritability of otitis media: a twin and triplet study. JAMA. 1999 Dec 8;282(22):2125-30. doi: 10.1001/jama.282.22.2125. PMID: 10591333.
Casselbrant ML,et al.. The genetic component of middle ear disease in the first 5 years of life. Arch Otolaryngol Head Neck Surg. 2004 Mar;130(3):273-8. doi: 10.1001/archotol.130.3.273. PMID: 15023832.
Kvaerner KJ, et al.Distribution and heritability of recurrent ear infections. Ann Otol Rhinol Laryngol. 1997 Aug;106(8):624-32. doi: 10.1177/000348949710600802. PMID: 9270423.
Kvestad E et al. Otitis media: genetic factors and sex differences. Twin Res. 2004 Jun;7(3):239-44. doi: 10.1375/136905204774200514. PMID: 15193168.
Rye MS, et al. Genome-wide association study to identify the genetic determinants of otitis media susceptibility in childhood. PLoS One. 2012;7(10):e48215. doi: 10.1371/journal.pone.0048215. Epub 2012 Oct 25. PMID: 23133572; PMCID: PMC3485007.
Jiang L, et al.Genome-wide association analyses of common infections in a large practice-based biobank. BMC Genomics. 2022 Sep 27;23(1):672. doi: 10.1186/s12864-022-08888-9. PMID: 36167494; PMCID: PMC9512962
Tian C, et al.Genome-wide association and HLA region fine-mapping studies identify susceptibility loci for multiple common infections. Nat Commun. 2017 Sep 19;8(1):599. doi: 10.1038/s41467-017-00257-5. PMID: 28928442; PMCID: PMC5605711.
Vuckovic D et al. The Polygenic and Monogenic Basis of Blood Traits and Diseases. Cell. 2020 Sep 3;182(5):1214-1231.e11. doi: 10.1016/j.cell.2020.08.008. PMID: 32888494; PMCID: PMC7482360.
Landrum MJ, et a. ClinVar: improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 2018 Jan 4;46(D1):D1062-D1067. doi: 10.1093/nar/gkx1153. PMID: 29165669; PMCID: PMC5753237
Landrum MJ, et al. improving access to variant interpretations and supporting evidence. Nucleic Acids Res. 2018 Jan 4;46(D1):D1062-D1067. doi: 10.1093/nar/gkx1153. PMID: 29165669; PMCID: PMC5753237.
Saarentaus EC et al.Inflammatory and infectious upper respiratory diseases associate with 41 genomic loci and type 2 inflammation. Nat Commun. 2023 Jan 18;14(1):83. doi: 10.1038/s41467-022-33626-w. PMID: 36653354; PMCID: PMC9849224.
Zhou et al., ‘Efficiently Controlling for Case-Control Imbalance and Sample Relatedness in Large-Scale Genetic Association Studies.’
Thul PJ, Lindskog C. The human protein atlas: A spatial map of the human proteome. Protein Sci. 2018 Jan;27(1):233-244. doi: 10.1002/pro.3307. Epub 2017 Oct 10. PMID: 28940711; PMCID: PMC5734309.
Lizarbe MA, Barrasa JI, Olmo N, Gavilanes F, Turnay J. Annexin-phospholipid interactions. Functional implications. Int J Mol Sci. 2013 Jan 28;14(2):2652-83. doi: 10.3390/ijms14022652. PMID: 23358253; PMCID: PMC3588008.
Hills BA. Surface tension versus straight adhesion in the Eustachian tube. Otol Neurotol. 2002 Jul;23(4):620; author reply 621. doi: 10.1097/00129492-200207000-00040. PMID: 12170173.
Lecat S, Verkade P, Thiele C, Fiedler K, Simons K, Lafont F. Different properties of two isoforms of annexin XIII in MDCK cells. J Cell Sci. 2000 Jul;113 ( Pt 14):2607-18. doi: 10.1242/jcs.113.14.2607. PMID: 10862718.
Lafont F et al. Annexin XIIIb associates with lipid microdomains to function in apical delivery. J Cell Biol. 1998 Sep 21;142(6):1413-27. doi: 10.1083/jcb.142.6.1413. PMID: 9744874; PMCID: PMC2141766.
Castigli E, et al. TACI is mutant in common variable immunodeficiency and IgA deficiency. Nat Genet. 2005 Aug;37(8):829-34. doi: 10.1038/ng1601. Epub 2005 Jul 10. PMID: 16007086.
Salzer U et al .Mutations in TNFRSF13B encoding TACI are associated with common variable immunodeficiency in humans. Nat Genet. 2005 Aug;37(8):820-8. doi: 10.1038/ng1600. Epub 2005 Jul 10. PMID: 16007087.
Martinez-Gallo M, et al TACI mutations and impaired B-cell function in subjects with CVID and healthy heterozygotes. J Allergy Clin Immunol. 2013 Feb;131(2):468-76. doi: 10.1016/j.jaci.2012.10.029. Epub 2012 Dec 11. PMID: 23237420; PMCID: PMC3646641.
Dong X, Hoeltzle MV, Hagan JB, Park MA, Li JT, Abraham RS. Phenotypic and clinical heterogeneity associated with monoallelic TNFRSF13B-A181E mutations in common variable immunodeficiency. Hum Immunol. 2010 May;71(5):505-11. doi: 10.1016/j.humimm.2010.02.002. Epub 2010 Mar 12. PMID: 20156508.
Shrine N. et al. Multi-ancestry genome-wide association analyses improve resolution of genes and pathways influencing lung function and chronic obstructive pulmonary disease risk. Nat Genet. 2023 Mar;55(3):410-422. doi: 10.1038/s41588-023-01314-0. Epub 2023 Mar 13. Erratum in: Nat Genet. 2023 Oct;55(10):1778-1779. PMID: 36914875; PMCID: PMC10011137.
Lee JJ, Rauter I, Garibyan L, Ozcan E, Sannikova T, Dillon SR, Cruz AC, Siegel RM, Bram R, Jabara H, Geha RS. The murine equivalent of the A181E TACI mutation associated with common variable immunodeficiency severely impairs B-cell function. Blood. 2009 Sep 10;114(11):2254-62. doi: 10.1182/blood-2008-11-189720. Epub 2009 Jul 15. PMID: 19605846; PMCID: PMC2745845.
Fried AJ et al. Functional analysis of transmembrane activator and calcium-modulating cyclophilin ligand interactor (TACI) mutations associated with common variable immunodeficiency. J Allergy Clin Immunol. 2011 Jul;128(1):226-228.e1. doi: 10.1016/j.jaci.2011.01.048. Epub 2011 Mar 21.
Feng G, Xu Y, Saso S, Sasano H, Kondoh S, Itani H, Gotoh S, Nagao M, Ikejiri M, Tanabe M, Takeuchi K. A Novel Homozygous Variant in GAS2L2 in Two Sisters with Primary Ciliary Dyskinesia. Intern Med. 2022;61(18):2765-2769. doi: 10.2169/internalmedicine.8884-21. Epub 2022 Sep 15. PMID: 36104176; PMCID: PMC9556235.
Bustamante-Marin XM, et al. Lack of GAS2L2 Causes PCD by Impairing Cilia Orientation and Mucociliary Clearance. Am J Hum Genet. 2019 Feb 7;104(2):229-245. doi: 10.1016/j.ajhg.2018.12.009. Epub 2019 Jan 18. PMID: 30665704; PMCID: PMC6372263.
Stroud MJ, et a. GAS2-like proteins mediate communication between microtubules and actin through interactions with end-binding proteins. J Cell Sci. 2014 Jun 15;127(Pt 12):2672-82. doi: 10.1242/jcs.140558. Epub 2014 Apr 4. PMID: 24706950; PMCID: PMC4058111.¨
Schilder AG, Chonmaitree T, Cripps AW, Rosenfeld RM, Casselbrant ML, Haggard MP, Venekamp RP. Otitis media. Nat Rev Dis Primers. 2016 Sep 8;2(1):16063. doi: 10.1038/nrdp.2016.63.
Adzhubei I, Jordan DM, Sunyaev SR. Predicting functional effect of human missense mutations using PolyPhen-2. Curr Protoc Hum Genet. 2013 Jan;Chapter 7:Unit7.20. doi: 10.1002/0471142905.hg0720s76. PMID: 23315928; PMCID: PMC4480630.
Ng PC, Henikoff S. Predicting deleterious amino acid substitutions. Genome Res. 2001 May;11(5):863-74. doi: 10.1101/gr.176601. PMID: 11337480; PMCID: PMC311071.
Lim DJ. Functional morphology of the mucosa of the middle ear and Eustachian tube. Ann Otol Rhinol Laryngol. 1976 Mar-Apr;85(2 Suppl 25 Pt 2):36-43. doi: 10.1177/00034894760850S209.
Raft S, Nowotschin S, Liao J, Morrow BE. Suppression of neural fate and control of inner ear morphogenesis by Tbx1. Development. 2004 Apr;131(8):1801-12. doi: 10.1242/dev.01067. Epub 2004 Mar 17. PMID: 15084464.
Vitelli F, Viola A, Morishima M, Pramparo T, Baldini A, Lindsay E. TBX1 is required for inner ear morphogenesis. Hum Mol Genet. 2003 Aug 15;12(16):2041-8. doi: 10.1093/hmg/ddg216.
Sakaue S, et al.. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet. 2021 Oct;53(10):1415-1424. doi: 10.1038/s41588-021-00931-x. Epub 2021 Sep 30.
Tian C, Johnson KR. TBX1 is required for normal stria vascularis and semicircular canal development. Dev Biol. 2020 Jan 1;457(1):91-103. doi: 10.1016/j.ydbio.2019.09.013. Epub 2019 Sep 21. PMID: 31550482; PMCID: PMC7893642.
Kurki M.I., et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature 2023 Jan;613(7944):508-518. doi: 10.1038/s41586-022-05473-8. Epub 2023 Jan 18.
Cunningham et al. Ensembl 2022. Nucleic Acids Res. 2022, vol. 50(1):D988-D995; PubMed PMID: 34791404. doi:10.1093/nar/gkab1049
Locke AE et al. Exome sequencing of Finnish isolates enhances rare-variant association power. Nature. 2019 Aug;572(7769):323-328. doi: 10.1038/s41586-019-1457-z. Epub 2019 Jul 31. Erratum in: Nature. 2019 Nov;575(7783):E4. PMID: 31367044; PMCID: PMC6697530.
Kircher M, Witten DM, Jain P, O'Roak BJ, Cooper GM, Shendure J. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet. 2014 Mar;46(3):310-5. doi: 10.1038/ng.2892. Epub 2014 Feb 2. PMID: 24487276; PMCID: PMC3992975.
Schubach M, Maass T, Nazaretyan L, Röner S, Kircher M. CADD v1.7: using protein language models, regulatory CNNs and other nucleotide-level scores to improve genome-wide variant predictions. Nucleic Acids Res. 2024 Jan 5;52(D1):D1143-D1154. doi: 10.1093/nar/gkad989. PMID: 38183205; PMCID: PMC10767851.
Pickrell JK, Berisa T, Liu JZ, Ségurel L, Tung JY, Hinds DA. Detection, and interpretation of shared genetic influences on 42 human traits. Nat Genet. 2016 Jul;48(7):709-17. doi: 10.1038/ng.3570. Epub 2016 May 16. Erratum in: Nat Genet. 2016 Sep 28;48(10):1296. PMID: 27182965; PMCID: PMC5207801.
Torkko JM, Manninen A, Schuck S, Simons K. Depletion of apical transport proteins perturbs epithelial cyst formation and ciliogenesis. J Cell Sci. 2008 Apr 15;121(Pt 8):1193-203. doi: 10.1242/jcs.015495. Epub 2008 Mar 18. PMID: 18349078.

Table 1. Loci associated with non-suppurative otitis media and its subtypes from three GWAS. Nearest protein coding gene to the leading variant in the locus refers to the nearest protein-coding gene and the number of genes within 1 MB of the Functional Mapping and Annotation locus.

Chr, chromosome; Pos, position; EA, effect allele; NEA, non-effect allele; OR, odds ratio; CI, confidence interval EAF, effect allele frequency

Chr:Pos (hg38)	Rsid	AF	EA	NEA	Gene	Phenotype	OR	OR 95% CI	P-value
8:123697261 intron	rs3779980	0.530	G	A	ANXA13	Non-suppurative	0.90	0.867-0.929	6.24x10^-9
		0.530				Mucous chronic otitis media	0.93	0.898-1.060	1.13x10^-4
		0.530				Secretory chronic otitis media	0.77	0.724-1.174	1.45x10^-13
17:16940415 missense	rs72553883	0.024	G	T	TNFRSF13B	Non-suppurative	1.42	1.28-1.57	6.76x10^-11
		0.025				Mucous chronic otitis media	1.47	1.31-1.63	9.45x10^-12
		0.025				Secretory chronic otitis media	1.27	1.15-1.73	9.12x10^-05
17:35745536 missense	rs3744374	0.230	G	A	GAS2L2	Non-suppurative	0.90	0.87-0.93	6.01x10^-8
		0.230				Mucous chronic otitis media	0.86	0.82-0.90	5.7x10^-10
		0.230				Secretory chronic otitis media	1.00	0.92-1.09	9.80x10^-01
22:19762002 intron	rs1978060	0.640	A	G	TBX1	Non-suppurative	1.14	1.10-1.18	4.55x10^-13
		0.640				Mucous chronic otitis media	1.15	1.10-1.19	1.3x10^-11
		0.640				Secretory chronic otitis media	1.11	1.04-1.17	4 x10^-03

The authors declare no competing interests.

supplementaryInformationforannexinpaperABV.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Genome-Wide Association Study Reveals Novel Associations of Annexin A13 to Secretory and GAS2L2 with Mucous Otitis Media

Status:

Version 1

Abstract

Figures

Introduction

Methods

Genotyping, imputation, and quality control

Immunohistochemistry

Results

Discussion

Conclusions

Abbreviations

Statements and Declarations

References

Table

Additional Declarations

Supplementary Files

Status:

Version 1