key: cord-0276502-y3jka64k
authors: Tian, Chao; Hromatka, Bethann S.; Kiefer, Amy K; Eriksson, Nicholas; Tung, Joyce Y; Hinds, David A.
title: Genome-wide association and HLA region fine-mapping studies identify susceptibility loci for multiple common infections
date: 2016-09-09
journal: bioRxiv
DOI: 10.1101/073056
sha: bdba7f61e7a11d188a8de890f601ea57e7d0e45e
doc_id: 276502
cord_uid: y3jka64k

We performed 23 genome-wide association studies for common infections, including chickenpox, shingles, cold sores, mononucleosis, mumps, hepatitis B, plantar warts, positive tuberculosis test results, strep throat, scarlet fever, pneumonia, bacterial meningitis, yeast infections, urinary tract infections, tonsillectomy, childhood ear infections, myringotomy, measles, hepatitis A, rheumatic fever, common colds, rubella and chronic sinus infection, in more than 200,000 individuals of European ancestry. For the first time, genome-wide significant associations (P < 5 × 10−8) were identified for many common infections. The associations were mapped to genes with key roles in acquired and innate immunity (HLA, IFNA21, FUT2, ST3GAL4, ABO, IFNL4, LCE3E, DSG1, LTBR, MTMR3, TNFRSF13B, TNFSF13B, NFKB1, CD40) and in regulation of embryonic developmental process (TBX1, FGF, FOXA1 and FOXN1). Several missense mutations were also identified (in LCE5A, DSG1, FUT2, TBX1, CDHR3, PLG, TNFRSF13B, FOXA1, SH2B3, ST5 and FOXN1). Missense mutations in FUT2 and TBX1 were implicated in multiple infections. We applied fine-mapping analysis to dissect associations in the human leukocyte antigen region, which suggested important roles of specific amino acid polymorphisms in the antigen-binding clefts. Our findings provide an important step toward dissecting the host genetic architecture of response to common infections.

Infectious diseases, the second leading cause of death worldwide, represent persistent challenges to human health due to increasing resistance to established treatments, lack of life-saving vaccines and medications in developing countries, and increasing distribution 1 .

Studies have linked susceptibility to infectious agents to cancers, autoimmune diseases, and drug hypersensitivity. Human papillomaviruses are associated with multiple cancers 2 ; rubella and mumps infection have been linked to development of type 1 diabetes (T1D) in children 3 ; and risk of developing multiple sclerosis (MS) was greater among patients with herpes zoster (shingles) than in matched controls 4 . Reactivation of chronic persistent human herpesviruses has been linked to drug-induced hypersensitivity 5 . Thus, infectious diseases have a profound impact on our health, both directly as well as through connections with other diseases. Nevertheless, genetic studies of common infectious diseases lag somewhat behind those of other major complex diseases. Few genome-wide association studies (GWAS) have been undertaken for common infectious diseases of lower mortality or for infectious diseases for which vaccines are available 6 .

In this study, we conducted 23 imputed GWASs with more than 200,000 European research participants who were genotyped on the 23andMe platform and were asked to report on their history of infections (Table 1 and Supplementary note). Classical human leukocyte antigen (HLA) loci have been prototypical candidate genetic susceptibility to multiple infectious diseases 7 . We further imputed and tested the HLA classical alleles and amino acid polymorphisms to dissect independent HLA signals for multiple common infections that have significant HLA associations in our GWASs.

In total, 59 genome-wide significant (GWS) regions (P < 5 x 10 -8 ) were discovered ( Table   2 , see also Supplementary Figure 1 and Supplementary Figure 2 for GWAS Manhattan plots and regional plots). The HLA region is significantly associated with 13 of the infections that we studied and we dissected independent signals in each association ( Figure 1 and Table 3 , see also Supplementary Figure 3 for HLA regional plots). In the later sections, we describe the result for each phenotype in detail.

The pairwise LD score correlation on GWAS statistics (Figure 2, Supplementary Table 2) showed significant positive genetic correlations between strep throat and tonsillectomy (rg = 0.85, se = 0.09, P = 1.89 x 10 -21 ), between childhood ear infection and myringotomy (rg = 0.88, se = 0.14, P = 1.27 x 10 -9 ), and between chickenpox and shingles (rg = 0.56, se = 0.16, P = 4 x 10 -4 ), which are pairs of related phenotypes. A highly significant genetic correlation was found between tonsillectomy and childhood ear infections (rg = 0.42, se = 0.05, P=8.19

x 10 -15 ). Tonsillitis and ear infections are sometimes comorbid symptoms. A previous twin study suggested a substantial overlap in genetic factors influencing variations in liability to ear infection and tonsillitis 8 . In our GWASs, we identified the missense mutation (N397H) in TBX1 as a significant susceptibility locus for both. Significant positive correlations were also observed for pneumonia with chronic sinus infections (rg = 0.72, se = 0.19, P = 1.0 x 10 -4 ), childhood ear infections (rg = 0.39, se = 0.07, P = 1.47 x 10 -8 ), colds last year (rg = 0.57, se = 0.12, P = 2.66 x 10 -6 ) and strep throats (rg = 0.69, se = 0.13, P=1.0 x 10 -7 ). These are all upper respiratory infections. The same viruses that cause colds and sore throat (if these infect the throat, sinuses and upper respiratory tract) can also cause pneumonia (if these reach the lungs). We also saw a high genetic correlation between urinary tract infection (UTI) and yeast infections (rg = 0.68, se = 0.08, P = 4.30 x 10 -16 ). There was a phenotypic correlation (Pearson r = 0.41, P < 2.2 x 10 -16 ) among women who reported both UTI and yeast infections in our cohort. These two infections can have similar symptoms and both cause discomfort in vaginal area, but different infectious agents cause them. The genetic correlation from LD score regression removes the correlation in GWAS summary statistics due to the correlation in phenotypes and thus suggests a shared host genetic susceptibility to UTI and yeast infections in females.

Chickenpox, characterized by red, itchy bumps on the skin, is a highly contagious disease caused by primary infection with varicella-zoster virus (VZV). After the initial chickenpox infection, the virus remains dormant in the nervous system, though in approximately 20% of people it reactivates and manifests as shingles (see below). We identified independent associations in HLA-A and HLA-B in the class I region. On conditional analysis within HLA-A, the amino acid polymorphism HLA-A-Gly107 (P = 3.77 x 10 -10 ) accounted for the HLA allele association (HLA-A*02:01, P = 1.08 x 10 -9 , conditional p-value is 0.90) in this interval.

Although HLA-A-Gly107 is not located in the peptide-binding cleft, it is in high LD with HLA-A-Asp74, HLA-A-Gly62 and HLA-A-V95 (r 2 > 0. 95) , which are all in the peptide-binding cleft. In HLA-B, only rs9266089 (P = 1.00 x 10 -10 ) met the threshold for GWS.

Shingles, also known as herpes zoster, is characterized by a painful band-shaped rash 9 and is caused by reactivation of the VZV (see above) . We identified multiple independent HLA signals for shingles. Within HLA-A, the amino acid polymorphism HLA-A-Arg97 (P = 2.75 x 10 -22 ), located in the peptide-binding groove, accounted for most of the SNP effect (rs2523815, P = 1.83 x 10 -22 , conditional p-value is 0.04) and the HLA allele effect (HLA-A*02:01, P = 8.91 x 10 -20 , conditional p-value is 0.06). HLA-A*02:01 is implicated in both our chickenpox and shingles GWASs. The amino acid associations are different, but HLA-A-Arg97 is in high LD with HLA-A-Gly107 (r 2 = 0.73). Within HLA-B, rs2523591 (P = 1.74 x 10 -27 ) had a stronger association than HLA classical variants. There was a significant residual effect for rs2523591 after conditioning on HLA allele (conditional p-value is 1.71 x 10 -9 ) or amino acid (conditional p-value is 3.13 x 10 -5 ) associations in HLA-B region. The SNP rs2523591 is in LD with rs9266089 (r 2 = 0.24, D' = 0.98), which is a GWS signal in chickenpox GWAS, suggesting potential shared genetic factors for chickenpox and shingles.

After adjusting for the signals in HLA-A and HLA-B regions, we observed independent secondary signals in the class III region (rs41316748, P=1.44 x 10 -12 ) and in the class II region index by DRB1-PheSerHis13 (P= 5.9 x 10 -10 ), which is in the peptide-binding cleft of HLA-DRB1.

We identified a variant upstream of IFNA21 (rs7047299, P = 1.67 x 10 -8 ) as a GWS association with shingles. None of the variants within 500kb and in moderate LD (r 2 > 0. 6) with the index SNP rs7047299 were coding, nor were they reported as expression quantitative trait loci (eQTL). IFNA21, a member of the alpha interferon gene cluster at band 9p21, encoded type I interferon and is mainly involved in innate immune response against viral infection. It have been shown to be involved in the pathogenesis of rubella 10 , and may also influence susceptibility to asthma and atopy 11 .

Over half of the US population suffers from cold sores (herpes simplex labialis), which are most commonly caused by the herpes simplex virus type I (HSV-1) 12 . Once someone has been infected, the virus usually cannot be eliminated and lies dormant in nerve cells where it may reactivate years later. Although most people are infected with HSV at some point in their lives, it is not clear why only some people suffer from cold sore outbreaks and reactivation 13 . We identified two independent GWS SNP associations in the HLA class I region. One is 2kb upstream of POU5F1 indexed by rs885950 (P = 7.47 x 10 -13 ) and the other is upstream of HCP5 indexed by rs4360170 (P = 3.41 x 10 -9 ). We also found a GWS association with HLA-B-ThrGly45 in the peptide-binding cleft of the HLA-B protein (P = 4.91 x 10 -12 ). Upon conditioning on HLA-B-ThrGly45, the two SNPs had significant residual effects (conditional p-values are 6.28 x 10 -9 and 1.32 x 10 -5 ). However, the effect of HLA-B-ThrGly45 was largely removed (conditional p-value is 0.001) after conditioning on the two SNPs.

Over 90% of the world's adult population is chronically infected with Epstein-Barr virus (EBV) . Primary infection at childhood is usually asymptomatic or with only mild symptoms.

Primary infection later in life is often accompanied by infectious mononucleosis (IM) 14 , commonly referred to as "mono" and characterized by fever, tonsillitis, and fatigue. A longitudinal study of IM in a cohort of 2,823,583 Danish children found evidence of familial aggregation of IM that warrants GWAS on IM disease etiology 14 . We identified the HLA class I region upstream of HCP5 as a GWS association (indexed by rs2596465, P = 2.48 x 10 -9 ). No GWS HLA allele or amino acid variant was identified.

Mumps is an illness caused by the mumps virus. There is no disease modifying treatment for mumps infection. It is a rare disease in developed countries since the introduction of routine vaccination; however, outbreaks can still occur and mumps remains a significant health threat in developing countries 15 . We identified four independent GWS associations with mumps.

Variation in HLA-A, in the class I region, was significantly associated with mumps. The amino acid polymorphism HLA-A-Gln43 (P = 2.51 x 10 -17 ) accounted for most of the SNP association (rs114193679, P = 2.23 x 10 -17 , conditional p-value is 0.19) and the HLA allele association (HLA-A*02:05, P = 1.73 x 10 -17 , conditional p-value is 0.26) in the HLA-A region.

We identified FUT2 (indexed by rs516316, P = 9.63 x 10 -72 ) and ST3GAL4 (indexed by rs3862630, P = 1.21 x 10 -8 ) as highly significant associations with mumps. Both belong to the glycosphingolipid (GSL) biosynthesis pathway, which was the most significant MAGENTA-analyzed pathway (Table 4 : P = 1 x 10 -4 and FDR = 0.003) identified using mumps GWAS data. This pathway also includes the ABO gene (indexed by rs8176643, P = 5.81 x 10 -5 ). Glycosphingolipids (GSLs), found in the cell membranes of bacteria all the way to humans, play important biological roles in membrane structure, host-pathogen interactions, cell-cell recognition, and modulation of membrane protein function. Genetic variation in the human FUT2 gene determines whether ABO blood group antigens are secreted into body fluids 16 . Detailed functional annotations of the genes identified in our mumps GWAS are provided in Supplementary table 3. The high-risk allele rs516316 (C) is in complete LD with rs601338 (A), a nonsense variant in FUT2 that encodes the "nonsecretor" (se) allele, which has been reported to provide resistance to Norovirus 17 and susceptibility to Crohn's disease 18, 19 and T1D 20 . The surface glycoprotein of mumps virus, hemagglutinin-neuraminidase, attaches to sialic acid receptors and promotes fusion and viral entry into host cells 21 . ABO antigens can modulate the interaction between pathogens and cell surface sialic acid receptors 22 ; this presents a plausible mechanism by which secretor status could modify susceptibility to mumps infections. ST3GAL4 encodes a member of the glycosyltransferase 29 family, a group of enzymes involved in protein glycosylation.

We also identified a significant association between mumps and a variant on chromosome 14 (rs11160318, P = 4.56 x 10 -12 ) that is located in the intergenic region upstream of BDKRB2. The role of BDKRB2 in mumps susceptibility is not clear, but the BDKRB2 product has high affinity for intact kinins, which mediate a wide spectrum of biological effects including inflammation, pain, vasodilation, and smooth muscle contraction and relaxation 23 .

GWASs have been carried out to identify genetic loci involved in HBV susceptibility and immune response to hepatitis B vaccine, and the HLA-DRB1-HLA-DQB1 region has been consistently shown to be GWS 24-26 . In our GWAS of HBV infection we identified a GWS association with rs9268652 (P = 3.14 x 10 -9 , in HLA-DRA), consistent with the association of the HLA class II region with HBV susceptibility. Fine mapping analysis failed to identify GWS associations with HLA classical variants, but DQB1*06:02 was moderately associated with hepatitis B (P = 5.47 x 10 -6 ). HLA-DQB1*06:02 is protective against HBV in our data, which is consistent with previous studies that the DRB1*15:01-DQB1*06:02 haplotype is associated with protection from type 1 autoimmune hepatitis and development of hepatocellular carcinoma 27 . Although chronic infection with the HBV has been linked to the development of hepatocellular carcinoma for more than 30 years 28 , the mechanism by which HBV infection leads to hepatocellular carcinoma is unclear.

Infection with hepatitis A virus (HAV) can be asymptomatic or result in an acute illness characterized by flu-like symptoms and jaundice. Serious complications are rare and the hepatitis A vaccine is effective for prevention 29 . No large cohort study has been done for HAV infection. Our GWAS showed a suggestive association with IFNL4, indexed by rs66531907 (P = 5.7 x 10 -8 , OR = 1.23, 1kb upstream of IFNL4). IFNL4 is located upstream of IFNL3 (also known as IL28B). The high-risk allele rs66531907 (C) in our hepatitis A GWAS is in almost complete LD with rs8099917 (T) (r 2 = 0.992), which has been associated with progression to chronic hepatitis C infection (HCV) and poor response to HCV therapy in multiple studies 30 .

The human papillomavirus (HPV) causes plantar warts, which occur on the soles of the feet. Some strains of HPV are also associated with certain forms of cancer 2 . We identified two GWS associations with plantar wart. Our first association was with the HLA-DRB1 in class II region. The amino acid polymorphism DRB1_PheSerTyr13 (P = 1.84 x 10-39), in the peptide binding clefts of HLA-DRB1, was highly significant. Conditioning on it, the effects of HLA-DRB1*15:01 (P=4.48e-14, conditional p-value is 0.36), HLA-DRB1*04:01 (P = 4.88 x 10 -18 , conditional p-value is 0.001), and rs9272050 (P = 9.66 x 10 -31 , conditional p-value is 0.06) were largely removed, but a small residual effect remained for rs28752534 (P = 1.50

x 10 -30 , conditional p-value is 2.15 x 10 -8 ).

We also identified a GWS association with rs6692209 (P = 5.25 x 10 -9 ) near LCE3E in the Epidermal Differentiation Complex (EDC) on 1q21. The EDC comprises a remarkable density of gene families that determine differentiation of the human epidermis 31 . The highrisk allele rs6692209 (T) is in high LD (r 2 = 0.86) with rs2105117 (A) (P = 2.98 x 10 -8 in our GWAS), a missense variant in LCE5A. Variants in late cornified envelope (LCE) genes have been implicated in GWASs of atopic dermatitis (rs3126085-A, P = 6 x10 -12 , r 2 = 0.047 with rs6692209) 32 and psoriasis (rs4085613-A, P = 7 x 10 -30 , r 2 = 0.002 with rs6692209) 33 , although the index SNPs identified in those GWASs were in low LD with our identified region. The skin is the primary interface with the external environment and provides a critical first barrier of the innate immune defense to infections 34 . While studies have linked the epidermal differentiation genes to various skin diseases, the role of the LCE proteins in epidermal biology has not been extensively studied. One study suggested that LCE proteins of groups 1, 2, 5 and 6 are involved in normal skin barrier function, while LCE3 genes encode proteins that are involved in barrier repair after injury or inflammation 35 .

Tuberculosis is a common and, in many cases, lethal infectious disease caused by various strains of mycobacteria, usually Mycobacterium tuberculosis. A chest x-ray and a sample of sputum are needed for a diagnosis of active TB disease 36 . The phenotype used in our GWAS was defined by asking customers whether they had had a positive response (red bumps) to a tuberculosis (TB) skin test. A positive TB test indicates that a person has at some time been exposed to and infected with TB bacteria. A negative TB test indicates that latent TB infection or TB disease is unlikely. In our GWAS of positive TB tests, the cases have either active TB disease, latent tuberculosis infections, or were vaccinated against TB, while controls may or may not have been exposed to TB bacteria and have never been infected.

We identified multiple independent HLA associations. In the most significant HLA-DRA-DQA1 region, conditioning on HLA-DRB1-Leu67 (P = 1.14 x 10 -29 ) and HLA-DQA1-His129 

Strep throat (also called streptococcal tonsillitis or streptococcal pharyngitis) is characterized by fever, headache, severe sore throat, swollen tonsils and lymph nodes, and is caused by group A streptococcus (GAS) bacteria 37 . It is routinely treated with antibiotics.

The symptoms of untreated strep throat typically resolve within a few days, but can sometimes cause complications such as rheumatic fever and kidney inflammation. We found a HLA-B class I region (index SNP rs1055821, P = 7.69 x 10 -11 ) as a GWS association with strep throat. HLA-B*57:01 was almost GWS (P = 5.06 x 10 -8 ). HLA-DRB1_CysTyr30 in class II region also showed a significant effect (P = 2.14 x 10 -9 ). Conditioning on rs1055821, the effect of HLA-B*57:01 was removed (conditional p-value is 0.84) and the effect of HLA-DRB1_CysTyr30 was substantially decreased (conditional p-value is 0.005).

We also found an GWS association with a small insertion, rs35395352, in 1p36.23 (P = Interestingly, the strep throat risk allele rs11121129-A is protective against psoriasis.

Studies found that exacerbation of chronic psoriasis can be associated with streptococcal throat infections, and the T-cells generated by palatine tonsils can recognize skin keratin determinants in patients' blood 44 .

Scarlet fever (also known as scarlatina) is caused by group A streptococcus (GAS) and is characterized by a bright red rash that covers most of the body, sore throat, and fever. It was a major cause of death before the discovery of effective antibiotics. We found a GWS association between scarlet fever and the HLA class II region. The amino acid polymorphism HLA-DQB1-Gly45 (P = 6.35 x 10 -14 ), located in the peptide-binding cleft of HLA-DQB1, accounted for the SNP association (rs36205178, P = 9.49 x 10 -14 , conditional pvalues is 0.83) and the HLA allele association (HLA-DQB1*03:01, P = 2.98 x 10 -14 , conditional p-values is 0.14) in this region. Although scarlet fever develops in some people with strep throat, we did not observe any cross disease effect between 'scarlet fever' and 'strep throat' in our analysis.

Pneumonia is characterized by inflammation of the alveoli in the lungs and can be caused by a variety of organisms, including bacteria, viruses, and fungi. The most common cause of bacterial pneumonia in adults -isolated in nearly 50% of community-acquired cases --is Streptococcus pneumonia 45 . We found a GWAS association between pneumonia and a HLA class I region indexed by rs3131623 (P = 1.99 x 10 -15 ). Conditioning on rs3131623, the residual effects of HLA-B-SerTrpAsn97 (P = 2.37 x 10 -12 , conditional p-value is 9.33 x10 -4 ) and HLA-DRB1-LS11 (P = 4.10 x 10 -11 , conditional p-value is 3.89 x 10 -6 ), both in the peptide-binding cleft, remained significant. There was also a residual effect left for rs3131623 (conditional p-value is 6.94 x 10 -5 ) after conditioning on the two amino acid associations.

Bacterial meningitis is a severe infection causing substantial neurological morbidity and mortality worldwide. Streptococcus pneumoniae is the leading cause of bacterial meningitis and is associated with a 30% mortality rate 46 . We identified a GWS association with CA10 (carbonic anhydrase X) indexed by rs1392935 (P = 1.19 x 10 -8 , intronic of CA10).

None of the variants within 500kb and in moderate LD (r 2 > 0.6) with rs1392935 were coding, nor were they reported as an eQTL. However, rs1392935 falls within an enhancer that is defined in the fetal brain. The protein encoded by CA10 is an acatalytic member of the alpha-carbonic anhydrase subgroup and it is thought to play a role in the central nervous system, especially in brain development 47 .

A yeast infection, also called candidiasis, is usually a localized fungal infection of the skin or mucosal membranes, typically affecting the oral cavity, esophagus, gastrointestinal tract, urinary bladder, toenails, or genitalia 48 . Most yeast infections are caused by species of Candida, often Candida albicans. We found GWS associations between yeast infections and variants in DSG1 (rs200520431, P = 1.87 x 10 -16 ). The index SNP rs200520431 is intronic and in high LD (r 2 > 0.8) with multiple missense mutations (rs8091003, rs8091117, rs16961689, rs61730306, rs34302455) in DSG1. The DSG1 gene product is a calciumbinding transmembrane glycoprotein component of desmosomes in vertebrate epithelial cells. It connects the cell surface to the keratin cytoskeleton and plays a key role in maintaining epidermal integrity and barrier function 49 . This glycoprotein has been identified as the autoantigen of the autoimmune skin blistering disease pemphigus foliaceus 50 and homozygous mutations in DSG1 have been showed to result in severe dermatitis, multiple allergies, and metabolic wasting syndrome 49 . A variant downstream of PRKCH (indexed by rs2251260, P = 3.46 x 10 -10 ) was also significantly associated with yeast infections. None of the variants within 500kb and in moderate LD (r 2 > 0.6) with the index SNP rs2251260 were coding, nor were they reported as an eQTL. However, rs2251260 falls within strong enhancers defined in hepatocellular carcinoma. PRKCH is a member of a family of serine-and threonine-specific protein kinase and is predominantly expressed in epithelial tissues. The PRKCH protein kinase can regulate keratinocyte differentiation 51 . We also found a significant association with a variant in the 14q32.2 gene desert (rs7161578-T, p = 4.04 x 10 -8 ). The index SNP is in high LD with many enhancer sequences defined in epidermal keratinocytes. A variant in the same region (rs7152623-A, p = 3 x 10 -15 , r 2 = (+) 0.98 with rs7161578-T) was implicated in aortic stiffness 52 .

About 80%-85% of urinary tract infections (UTIs) are caused by Escherichia coli (E. coli), and 5-10% are caused by Staphylococcus saprophyticus 53 . Very rarely, viruses or fungi may cause UTIs 54 . UTIs occur more commonly in women than men 55 and in our dataset, significantly more female than male customers reported having two or more UTIs.

We found an association between UTIs and variants in PSCA (indexed by rs2976388, P = 3.27 x 10 -10 ). The index SNP rs2976388 falls within strong enhancers defined in epidermal keratinocytes. It is also in high LD (r2 = 0.95) with rs2294008 (P = 1.01 x 10 -9 ), which is in the 5'-UTR promoter sequence of PSCA and can cause changes in transcriptional repressor CTCF motif. The PSCA gene variants conferred risk of UTI only in female when we further tested the effect in female and male cohort separately (P < 1e-11 in female, and P >0.1 in male). Prior GWASs have found associations between the PSCA gene and duodenal ulcer (rs2294008-C, p = 2e-33) and bladder cancer (rs2294008-T, p = 4 x 10 -11 ) 56,57 . PSCA encodes a glycosylphosphatidylinositol-anchored cell membrane glycoprotein of unknown function. This glycoprotein was initially identified as a prostate-specific cell-surface marker 58 and is overexpressed in a large proportion of prostate cancers, but also detected in bladder and pancreatic cancers 57 .

We also identified a variant intronic of FRMD5 (rs146906133, P = 2.02 x 10 -8 ) that is in LD with rs138763871 (r 2 = 0.8, p = 1.75 x 10 -6 in our GWAS), a missense variant in STRC (R1521W). The role of FRMD5 or STRC in UTIs is unclear.

Tonsils are considered the first line of defense against respiratory infections. Previous twin studies have suggested a substantial genetic predisposition for recurrent tonsillitis 59 .

Tonsillectomy is a surgery to remove the tonsils typically due to repeated occurrence of We detected 34 additional GWS associations (Table 2 and Supplementary Table 4 Although tonsillectomy has been performed for over 100 years, possible immunological effects of this procedure remain controversial. Many reports have demonstrated that in certain patients tonsillectomy is an effective therapy for psoriasis 44 and rheumatoid arthritis (RA) 62 ; the rationale for this effect is unknown. The 'Intestinal immune network for IgA production' pathway was also identified as the most significant pathway in a recent GWAS of IgAN 63 . In that study, they also linked the intestinal mucosal inflammatory disorders and inflammatory bowel disease (IBD) with risk of IgAN. Our tonsillectomy GWAS showed significant overlap of the identified loci with those autoimmune and inflammatory disorders 64 . We found five risk loci that are also associated with RA (HLA, TNFIP3, CD40, SPRED2 and SH2B3); four risk loci that are also associated with IBD (HLA, MTMR3, TNFAIP3 and CD40) and ulcerative colitis (HLA, GNA12, MAPK3 and NFKB1); three risk loci that are also associated with psoriasis (HLA, SLC12AB and 1p36.23); and two risk loci (HLA and MTMR3) that are shared with IgAN. We observed both concordant and opposing effects compared to these immune-mediated diseases. Our results may help to elucidate the connection between tonsillectomy and these diseases and may provide insight into clinical markers that could be used as indicators of tonsillectomy as a therapy for these diseases.

Although ear infections can have a bacterial origin, they are often secondary to an upper respiratory infection such as a cold or sore throat, which can be caused by viruses or bacteria 65 . Children are more likely than adults to have ear infections because their Eustachian tubes are less effective at draining fluid and their immune systems are not fully developed. In our GWAS, we identified 14 regions that were significantly associated with childhood ear infection. Signals in the HLA region mapped to HLA-DRB1 in class II region.

Our strongest association within HLA-DRB1 was with HLA-DRB1-Gln96 in the peptidebinding groove. After conditioning on HLA-DRB1-Gln96 (P = 2.11 x 10 -12 ), the effects of rs4329147 (P = 9.55 x 10 -12 , conditional p value is 0.27) and DQB1*06:02 (P = 5.79 x 10 -10 , conditional p value is 0.85) were eliminated.

We found 13 additional GWS signals ( Table 2 and Supplementary Table 5 ). Although MAGENTA pathway analysis did not identify significant pathways ( 68, 69 . We also identified other genes involved in developmental processes, including MKX (rs2808290, P = 5.09 x 10 -16 ), FGF3 (rs72931768, P = 2.63 x 10 -9 ) and AUTS2 (rs35213789, P = 3.75 x 10 -9 ). The MKX product is an IRX family-related homeobox transcription factor and has been shown to play a critical role in tendon differentiation by regulating type I collagen production in tendon cells 70 . FGF3 is a member of the fibroblast growth factor family, which is involved in a variety of biological processes including embryonic development, cell growth, morphogenesis, tissue repair and invasion. Study of similar genes in mouse and chicken suggested the role in inner ear formation 71 .. AUTS2 has been implicated in neurodevelopment and is a candidate gene for autism spectrum disorders and developmental delay 72, 73 . Finally, we identified missense mutations in CDHR3 (rs114947103, P = 5.40 x 10 -9 ) and PLG (rs73015965, P = 3.78 x 10 -8 ). The index SNP rs114947103 is in almost complete LD with rs6967330 (r2=0.97), a missense variant in CDHR3, that was recently identified in a GWAS as a susceptibility locus for asthma 74 . The biological role of CDHR3 is not clear, but it is highly expressed in airway epithelium and belongs to the cadherin family that is involved in cell adhesion and epithelial polarity 74 .

Mutations in the PLG gene could cause congenital plasminogen deficiency, which results in inflamed growths on the mucous membranes. Studies in mice have showed that PLG plays an essential role in protecting against the spontaneous development of chronic otitis media (middle ear infection) and have also suggested the possibility of using PLG for clinical therapy of certain types of otitis media 75 .

Myringotomy is a surgical procedure in which a tiny incision is created in the eardrum to allow the release of middle-ear fluid. Before the invention of antibiotics, myringotomy without tube placement was a common treatment for severe or frequent acute otitis media 76 . The same variant in TBX1 that was associated with tonsillectomy and childhood ear infection (see earlier) was also found to be significantly associated with myringotomy (rs1978060, P = 3.34 x 10 -10 ).

We identified the greatest number of independent associations for tonsillectomy (n=35) and childhood ear infection (n=14), two relatively nonspecific phenotypes. For many of the infections, we only identified HLA associations. These findings may be due to our relatively larger sample sizes for the two traits and may also be due to the fact that tonsillectomy and childhood ear infections are heterogeneous phenotypes, influenced by a variety of pathogens and also by anatomical abnormalities. People undergo tonsillectomy for different reasons, including recurrent tonsillitis, obstructive sleep apnea, and nasal airway obstruction 77 . The causes of childhood ear infections are also diverse, involving different types of bacteria and viruses, and ear developmental defects 78 . Variants in the TBX1 gene, which is essential for inner ear development 68, 69 , were associated with both tonsillectomy and inner ear infections. Many genes that are involved in embryotic development were also identified in our tonsillectomy and childhood ear infection GWASs. The vulnerability to tonsillectomy and childhood ear infections may also be largely inherited. There were earlier studies calculated that heredity account for over 70% of susceptibility to recurrent ear infections in children 79, 80 . Previous twin studies had suggested a substantial genetic predisposition for recurrent tonsillitis 59 and sleep-disordered breathing 81 . According to the LD score regression on our GWAS data, tonsillectomy and childhood ear infections have relatively higher heritability on the observed scale than most of the other infections we studies. We did not calculate the liability scale heritability due to the lack of population prevalence information.

In the HLA region, we found that viral diseases -e.g., chicken pox, shingles, cold sores, mononucleosis (all caused by herpesvirus) as well as mumps (caused by mumps virus)were mainly associated with variation in class I molecules (Figure 1 ). The bacterial diseases -specifically having a positive TB test (caused by Mycobacterium tuberculosis), scarlet fever (caused by GAS), and childhood ear infection (mostly caused by bacteria)were mainly associated with variation in class II molecules. Tonsillectomy and pneumonia, caused by either bacteria or viruses, were associated with both class I and class II molecules. These observations are consistent with previous knowledge about antigen presentation; viruses mostly replicate within nucleated cells in the cytosol and produce endogenous antigens that are presented by class I MHC molecules, while bacteria grow extracellularly and are taken up by endosomal compartments where they are processed for presentation by class II MHC molecules 82 . These two intracellular pathways of protein processing may not be completely separate. Activation of CD8+, HLA class I-restricted Tcells by exogenous antigens has been reported 83 and HLA class II-restricted CTLs that recognize endogenously synthesized antigen, such as HBV envelope antigens, have also been described 84, 85 .

Some diseases, however, do not follow the common principles of antigen presentation. 88 . Since most sore throats are caused by viruses 89 , another possibility is that some of the reported strep throat cases actually had a misdiagnosed or misreported viral infection. A final possibility is that HLA class I molecules can bind bacterial peptides derived from exogenous proteins that are internalized by endocytosis or phagocytosis, a process called cross-presentation 85, 90 .

We failed to detect any GWS association for certain infectious diseases, including rubella, measles, chronic sinus infection, the common cold, or rheumatic fever. It is known that HLA affects responses to measles and rubella vaccinations 91, 92 and may thus also be associated with susceptibility. One limitation of our study that might affect findings with respect to measles and rubella is our lack of data on vaccination status. Since there is now a common vaccination for measles and rubella, misclassifying vaccinated people as controls would theoretically reduce our power to detect associations. This limitation might also affect the power of our shingles GWAS. The controls used for the shingles GWAS were not filtered for a positive history of chickenpox, though having chickenpox is a prerequisite for shingles.

Another reason might be the heterogeneity of the phenotype. Colds are caused by several different viruses (mostly rhinoviruses and coronaviruses), and larger sample sizes may be required to identify alleles only associated with specific pathogens.

Although additional studies will be required to validate our associations, our findings are an important step towards dissecting the host genetic contribution to variation in response to infections. Research insights into infectious diseases will help to derive new diagnostic approaches and perhaps new therapies and preventions. This study may also help us to understand some of the autoimmune disorders that are associated with various infection triggers. Many of the identified associations were also found in autoimmune diseases. One postulated connection between infectious diseases and immunological disorders is that the immune cells that are activated in response to a pathogen epitope are also cross-reactive to self and lead to direct damage and further activation of other arms of the immune system 93 .

We conducted GWASs of 23 infectious diseases. All participants were drawn from the customer base of 23andMe, Inc., a personal genetics company. All individuals included in the analyses provided informed consent and answered surveys online according to our human subjects protocol, which was reviewed and approved by Ethical & Independent Review Services, a private institutional review board (http://www.eandireview.com).

All participants were of primarily European ancestry (>97% European ancestry) as determined through an analysis of local ancestry 94 . A maximum set of unrelated individuals was chosen using a segmental identity-by-descent (IBD) estimation algorithm 95 .

Individuals were defined as related if they shared more than 700 cM of IBD, including regions where the two individuals share either one or both genomic segments IBD. This level of relatedness (involving ~20% of the human genome) corresponds approximately to the minimal expected sharing between first cousins in an outbred population. Table 1 shows our discovery sample sizes, drawn from more than 650,000 genotyped customers who reported via web-based questionnaires whether they had been diagnosed with one or more infectious diseases. (For more detail on survey questions used to collect the phenotype data, see the Supplementary Notes). For the non-pseudoautosomal region of the X chromosome, males and females were phased together in segments, treating the males as already phased; the pseudoautosomal regions were phased separately. We then imputed males and females together using Minimac as with the autosomes, treating males as homozygous pseudo-diploids for the non-pseudoautosomal region.

For case control comparisons, we tested for association using logistic regression, assuming additive allelic effects. For quantitative traits, association tests were performed using linear regression. For tests using imputed data, we use the imputed dosages rather than best-guess genotypes. We included covariates for age, gender, and the top five principal components to account for residual population structure. The association test pvalue was computed using a likelihood ratio test, which in our experience is better behaved than a Wald test on the regression coefficient. Results for the X chromosome were computed similarly, with men coded as if they were homozygous diploid for the observed allele.

We computed LD scores as previously described 99 using the samples in the 1000 Genomes Project reference panel (phase 3 verion 5a) with a MAF cutoff of 5%. We calculated the pairwise genetic correlation (rg) on the GWAS summary statistics (including MHC region) using LD score regression 99 . The cluster map of genetic correlation, base on the lower bound of the 95% confident interval of calculated rg, was plotted using 'Nearest Point Algorithm' implemented in python "Searborn" package (https://stanford.edu/~mwaskom/software/seaborn/index.html). We found no significant negative correlations in our data, thus we set all negative values to zero.

To explore whether any of the significant SNPs identified might link to functional mutations or have potential regulatory functions, we used the online tools HaploReg (http://www.broadinstitute.org/mammals/haploreg/haploreg.php) to confirm the location of each SNP in relation to annotated protein-coding genes and/or non-coding regulatory elements. We queried only the variants that were within 500kb of and in moderate LD (r 2 >0.6) with the index SNP.

HLA imputation was performed with HIBAG 100 , an attribute bagging based statistical method that comes as a freely available R package and includes a pre-fit classifier. This classifier was trained from a large database of individuals with known HLA alleles and SNP variation within the HLA region. Over 98% of the tagging SNPs used in HIBAG were genotyped and passed quality control on 23andMe's platform. We imputed allelic dosage for HLA-A, B, C, DPB1, DQA1, DQB1, and DRB1 loci at four-digit resolution. We used the default settings of HIBAG and the run time for 100,000 samples was about 10 hours on our cluster.

Using an approach suggested by P. de Bakker 101 , we downloaded the files that map HLA alleles to amino acid sequences from https://www.broadinstitute.org/mpg/snp2hla/ and mapped our imputed HLA alleles at four-digit resolution to the corresponding amino acid sequences; in this way we translated the imputed HLA allelic dosages directly to amino acid dosages. We encoded all amino acid variants in the 23andMe European samples as biallelic We imputed 292 classical HLA alleles at four-digit resolution and 2395 bi-allelic amino acid polymorphisms. Similar to the SNP imputation, we measured imputation quality using r 2 , which is the ratio of the empirically observed variance of the allele dosage to the expected variance assuming Hardy-Weinberg equilibrium. The imputation quality (r 2 ) of the top associated HLA alleles and amino acids are in Supplementary Table 6 .

To test associations between imputed HLA allele/amino acid dosages and phenotypes, we performed logistic or linear regression using the same set of covariates used in the SNPbased GWAS for that phenotype. We applied a forward stepwise strategy, within each type of variant, to establish statistically independent signals in the HLA region. Within each variant type(e.g. SNP, HLA allele, and HLA amino acid), we first identified the most strongly associated signals (lowest p-value) for each disease and performed forward iterative conditional regression to identify other independent signals if the conditional p-value was < 5 x 10 -8 . All analyses controlled for sex and five principal components of genetic ancestry.

The p-values were calculated using a likelihood ratio test. The iterative conditional regression dissected HLA signal into independent HLA associations. Within each identified HLA locus (HLA-A, B, C, DPB1, DQA1, DQB1, and DRB1), We further carried out reciprocal analyses, which are the conditional analyses across variants types, to see if the association can be attributed to the amino acid polymorphism within each HLA locus.

In order to better understand how multiple genes in the same pathway may contribute to certain infections, we performed pathway analysis using Meta-Analysis Gene-set Enrichment of variaNT Associations (MAGENTA) 102 , which tests for enrichment of genetic associations in predefined biological processes or sets of functionally related genes, using GWAS results as input. We used gene sets of 1320 canonical pathways from the Molecular Signatures Database (MsigDB) compiled by domain experts 103 The strongest associated signals (red) and putative independent secondary signals (blue) are shown for each disease along with their location within the region. Independent secondary signals were defined as those with residual conditional association using the same significant threshold as the primary association signal mentioned in the main text. a. We identify independent associations for each category of variants: SNP, HLA (classical allele) and AA (HLA amino acid), separately using iterative conditional regression. For amino acid polymorphism, the label specifies the amino acid and the position. For example, 'A-Val95' means amino acid Val at position 95 of the HLA-A protein. We use this naming convention throughout the paper, which is composed from the name of the HLA protein (A), followed by the three-letter code for the amino acid (Val) and its location in the protein (95) . The amino acid polymorphism is highlighted as 'red' if the signal in that gene region can be attribute to it by reciprocal analysis. b. The alleles for SNP are represented as risk allele/non-risk allele; the allele for HLA classical allele is recorded as 'NA', which means it either has or does not has the tested allele; the alleles for HLA amino acid polymorphisms are all the possible amino acids (one-letter code) at that position. c. Strep throat, plantar warts and UTI frequency are quantitative traits, for which we reported the effect size instead of the odds ratio. d. The 'Cond. P-value' column contains conditional p-values that are after iterative conditional regression within each category of variants (SNP, imputed HLA alleles and imputed HLA amino acid polymorphisms). The 'P-value' column contains the unconditional association test p-values. The most significant pathway (gene sets) with a gene set enrichment FDR <0.01 are presented. The genes that contain genome-wide significant SNPs in the GWAS analysis are highlighted red. The most significant pathway (gene sets) with a gene set enrichment FDR <0.01 are presented. The genes that contain genome-wide significant SNPs in the GWAS analysis are highlighted red. The most significant pathway (gene sets) with a gene set enrichment FDR <0.01 are presented. The genes that contain genome-wide significant SNPs in the GWAS analysis are highlighted red.

Infectious diseases -a global challenge

The role of human papillomaviruses in oncogenesis. Recent Results Cancer Res

Viral trigger for type 1 diabetes: pros and cons

Increased risk of multiple sclerosis following herpes zoster: a nationwide, population-based study

Fever, rash, and systemic symptoms: understanding the role of virus and HLA in severe cutaneous drug allergy

HLA and infectious diseases

Genetics of susceptibitlity to human infectious disease

Recurrent otitis media and tonsillitis: common disease predisposition

Herpes zoster: epidemiology, natural history, and common complications

Microarray analyses of differentially expressed human genes and biological processes in ECV304 cells infected with rubella virus

Variation in the type I interferon gene cluster on 9p21 influences susceptibility to asthma and atopy

Trends in herpes simplex virus type 1 and type 2 seroprevalence in the United States

C21orf91 genotypes correlate with herpes simplex labialis (cold sore) frequency: description of a cold sore susceptibility gene

A genetic basis for infectious mononucleosis: evidence from a family study of hospitalized cases in Denmark

Harrison's principles of internal medicine

Sequence and expression of a candidate for the human Secretor blood group alpha(1,2)fucosyltransferase gene (FUT2). Homozygosity for an enzyme-inactivating nonsense mutation commonly correlates with the non-secretor phenotype

Antibody prevalence and titer to norovirus (genogroup II) correlate with secretor (FUT2) but not with ABO phenotype or Lewis (FUT3) genotype

Fucosyltransferase 2 (FUT2) non-secretor status is associated with Crohn's disease

Declining hepatitis A mortality in the United States during the era of hepatitis A vaccination

Genetic variation in IL28B is associated with chronic hepatitis C and treatment failure: a genome-wide association study

Genes encoding structural proteins of epidermal cornification and S100 calcium-binding proteins form a gene complex ('epidermal differentiation complex') on human chromosome 1q21

Genome-wide association study identifies two new susceptibility loci for atopic dermatitis in the Chinese Han population

Psoriasis genome-wide association study identifies susceptibility variants within LCE gene cluster at 1q21

Innate immunity mediated by epidermal keratinocytes promotes acquired immunity involving Langerhans cells and T cells in the skin

Psoriasis risk genes of the late cornified envelope-3 group are distinctly expressed compared with genes of other LCE groups

Learn the Signs and Symptoms of TB Disease

New understanding of the group A Streptococcus pathogenesis cycle

Serum and salivary IgA levels in normal subjects: comparison between tonsillectomized and nontonsillectomized subjects

Simultaneous, clonally identical T cell expansion in tonsil and synovium in a patient with rheumatoid arthritis and chronic tonsillitis

Genome-wide association study identifies susceptibility loci for IgA nephropathy

The NHGRI GWAS Catalog, a curated resource of SNP-trait associations

Importance of respiratory viruses in acute otitis media

Association between secretor status and respiratory viral illness

Tbx1 haploinsufficieny in the DiGeorge syndrome region causes aortic arch defects in mice

Suppression of neural fate and control of inner ear morphogenesis by Tbx1

TBX1 is required for inner ear morphogenesis

The Mohawk homeobox gene is a critical regulator of tendon differentiation

LAMM syndrome with middle ear dysplasia associated with compound heterozygosity for FGF3 mutations

The role of AUTS2 in neurodevelopment and human evolution

De novo intragenic deletion of the autism susceptibility candidate 2 (AUTS2) gene in a patient with developmental delay: a case report and literature review

A genome-wide association study identifies CDHR3 as a susceptibility locus for early childhood asthma with severe exacerbations

Spontaneous development of otitis media in plasminogen-deficient mice

To tube or not to tube: indications for myringotomy with tube placement

Clinical practice guideline: tonsillectomy in children

Ear Infections in Children

The heritability of otitis media: a twin and triplet study

Distribution and heritability of recurrent ear infections

Genetic and environmental influences in sleep-disordered breathing in older male twins

Pathways of antigen processing

Presentation of exogenous antigen with class I major histocompatibility complex molecules

Hepatitis B virus (HBV)-specific cytotoxic T-cell (CTL) response in humans: characterization of HLA class II-restricted CTLs that recognize endogenously synthesized HBV envelope antigens

Class II MHC molecules can use the endogenous pathway of antigen presentation

HPV -immune response to infection and vaccination

Immune responses to human papillomavirus

Immune dysfunction and bacterial coinfections following influenza

Cross-presentation: underlying mechanisms and role in immune surveillance

Genome-wide characterization of transcriptional patterns in high and low antibody responders to rubella vaccination

Consistency of HLA associations between two independent measles vaccine cohorts: a replication study

The role of infections in autoimmune disease

Ancestry Composition: A Novel, Efficient Pipeline for Ancestry Deconvolution

Cryptic distant relatives are common in both isolated and cosmopolitan genetic samples

A map of human genome variation from population-scale sequencing

Rapid and accurate haplotype phasing and missingdata inference for whole-genome association studies by use of localized haplotype clustering

A. minimac2: faster genotype imputation

LD Score regression distinguishes confounding from polygenicity in genome-wide association studies

HIBAG--HLA genotype imputation with attribute bagging

Imputing amino acid polymorphisms in human leukocyte antigens

Common inherited variation in mitochondrial genes is not enriched for associations with type 2 diabetes or related glycemic traits

Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles

We thank the customers of 23andMe for participating in this research and the employees of 23andMe for their contributions to this work.

All authors listed were/are employed by 23andMe, and own stock or stock options in 23andMe.