Shinshu Med J, 63⑶ :157～165, 2015Gene-gene Interaction Between EPAS1 and EGLN1 in Patients with High-Altitude PulmonaryEdema , Masao OTA(cid:50892)(cid:50903) Toshimichi HORIUCHI(cid:50891)(cid:50903) , Yunden DROMA(cid:50891)(cid:50903) Nobumitsu KOBAYASHI(cid:50891)(cid:50903)and Masayuki HANAOKA(cid:50891)(cid:50903)(cid:50924)1) The First Department of Medicine, Shinshu University School of Medicine 2) Department of Legal Medicine, Shinshu University School of MedicineObjective:EndothelialPER-ARNT-SIM (PAS)domain protein 1(EPAS1,also known as HIF2α)and eglnine homolog 1 (EFLN1), also known as prolylprolyl hydroxylase domain protein 2 (PHD2), have key functions in the upstream ofthehypoxia-induciblefactor (HIF)pathway.In order to dissect thegeneticbiologyofsusceptibili-tyto high-altitude pulmonaryedema (HAPE)from the viewpoint oftheHIF pathway,weidentified theallelic discriminations ofthreesignificant tag single-nucleotidepolymorphisms (SNPs)in EPAS1 and threetag SNPs in EGLN1 in HAPE-susceptible (HAPE-s)Japanese subjects.Methods: Alleles were determined for the six SNPs (rs13419896, rs4953354, and rs4953388 in EPAS1 ;rs1435166, rs7542797, and rs2153364 in EGLN1) by the TaqM an(cid:51378) SNP Genotyping Assay in a group of 59 HAPE-s subjects and a controlgroup of67HAPE resistant (HAPE-r)subjects.In addition to thecase-control analysis,multi-dimensional reduction (M DR)methodologywas applied to a gene-gene interaction analysis to evaluate the association of HAPE-s with gene-gene interactions.Results:TheEGLN1 rs2153364(A/G)x EPAS1 rs13419896(G/A)interaction was significantlyassociated with HAPE-s in the pairwise model (P＝0.0049)based on the balanced accuracy of 63.23% in M DR.However,no significance was detected for the association with HAPE-s in the single gene model.Conclusion :The EPAS1-EGLN1 interaction appears to be associated with HAPE-s in the Japanese popula-tion despite the individual genes not being associated with HAPE-s. Shinshu Med J 63 : 157 ―165, 2015 (Received for publication January 29, 2015;accepted in revised form February 17, 2015)Key words:hypoxia-inducible factor, gene-gene interaction, high-altitude pulmonary edema(cid:50303) Introduction Mountains cover more than one-fifth of the Earth’s surfaceand arepopular tourist destinations.Large numbers ofpeopletravelto high altitudes for various reasons,including trekking,pilgrimage,and tourism. In Japan, approximately 38-65 million people climb mountains (＞2,000 m above sea level)for recreation each year according to the Japan (cid:50924)Corresponding author:Masayuki Hanaoka The First Department of Medicine, Shinshu University School of M edicine, 3-1-1 Asahi, Matsumoto, Nagano 390-8621, Japan E-mail:masayuki＠shinshu-u.ac.jpScience Research Society (http://soukaken.com/,in Japanese). Exposure to the oxygen-depleted envi-ronment triggers the onset of a range ofphysiologi-cal and biochemical reactions that enhance the efficacy of the body’ respiratory(cid:50891)(cid:50903), cardiovascular,and oxygen utilization systems(cid:50892)(cid:50903)through the proc-ess of acclimatization. Failure to acclimatize leads to high-altitude pulmonary edema (HAPE), a form ofaccelerated pulmonaryedema ofnon-cardiogenic origin that occurs in lowlanders who rapidlyascend above 2,500 m ;it has a considerablyhigh mortality if hypoxic exposure persists(cid:50891)(cid:50903)(cid:50873)(cid:5089 The inception and progression of HAPE are governed by both genetic and environmental factors(cid:50894)(cid:50903)(cid:50895)(cid:50903 3, 2015Horiuchi.Ota.Droma et al.Thehypoxia-induciblefactor (HIF)pathwaytrig-gers downstream cascades of multiple molecular events involving the pathophysiology of hypoxia(cid:50896)(cid:50903).More than a dozen candidate genes in the HIF pathway have been shown to contribute to the geneticadaptation to hypoxia in high-altitudepopu-lations across multiple studies(cid:50897)(cid:50903)(cid:50873)(cid:50891 including endothelial PER-ARNT-SIM (PAS)domain protein 1(EPAS1), also known as HIF2α, and egl nine homolog 1 (EGLN1), also known as prolylprolyl hy-droxylase domain protein 2 (PHD2). These two key genes function in the upstream of the HIF pathway in human high-altitude adaptation(cid:50891)(cid:50891)(cid:50903)(cid:50 the single-nucleotide polymorphisms(SNPs) of rs13419896 (G/A), rs4953354 (A/G), and rs4953388 (G/A)in EPAS1 were demonstrated to be significantly associated with hypoxia tolerance in indigenous high-altitude Tibetan(cid:50897)(cid:50903)and Sherpa(cid:50891)(cid:50896)(cid:50903)rs4953354, and rs4953388) and three tag SNPs in EGLN1 (rs1435166, rs7542797, and rs2153364) in HAPE-susceptible (HAPE-s)Japanesesubjects and compared the results to HAPE-resistant (HAPE-r)Japanese subjects in a case-control studyand gene-gene interaction analysis.(cid:50304) M ethods(cid:50771) Ethics statement The current study was approved by the Ethics Committee of Shinshu University School of Medi-cine (M atsumoto, Japan). The protocol was perfor-med in accordance with the principles outlined in the Declaration on Helsinki of the World Medical Association(cid:50891)(cid:50899)(cid:50903)and was approved by the Ethics Committee of Shinshu University School of Medi-cine (Permission numbers: 106). Written informed consent was obtained from each subject after pro-populations. Moreover, the SNPs of rs1435166 (A/ viding them with a full explanation of the study.G), rs7542797 (A/C), rs2153364 (A/G) in EGLN1 weredocumented to besignificantlyassociated with human high-altitude adaptation in the Indo-Aryan population(cid:50891)(cid:50894)(cid:50903).Obviou HIF to hypoxia adaptation in indigenous high-alti-tude populations maynot be applicable to the eluci- dation ofgeneticbiologyin HAPE,an acutedisease caused bysudden oxygen deprivation,despite high- (cid:50772) Subjects The case-control study included a case group with 59 HAPE-s subjects (52 males, 7 females,average age 34.2 years)and a control group with 67 HAPE-r subjects (58 males,9 females,average age 37.0 years).Allsubjects wereunrelated natives born in Japan and resided at low altitudes.The HAPE-s subjects were patients admitted to Shinshu Univer-altitude hypoxia being a common initial element in both disorders. Indigenous high-altitude popula-sityHospital because of HAPE that occurred while climbing in the Japan Alps at heights over 2,500 mtions, such as Tibetans and Sherpas, have success-fully lived and reproduced at high altitudes for hundreds of generations with hypoxia as a constant evolutionary pressure,bringing about natural selec-tion toward phenotypes (e.g., relatively low hemo-globin) that tend to offer beneficial adaptations to sustained hypoxia(cid:50891)(cid:50897)(cid:50903)(cid:50891 contrast,HAPE occurs in lowlanders rapidly exposed to a high altitude,resulting in a failure to acclimatize to high-altitude hypoxia,which is characterized byhypoxia-induced acute pulmonary vasoconstriction(cid:50892)(cid:50903)(cid:50873)( order to dissect the genetic biology of HAPEsusceptibility from the viewpoint of the HIF path-way, we identified the allelic discriminations of three significant tag SNPs in EPAS1 (rs13419896,between 1971and 2013.Thediagnosis ofHAPE was based on diagnostic criteria at the onset of the disorder(cid:50892)(cid:50900)(cid:50903).Allpatie within one week of hospitalization.Related clinical examinations and cardiovascular tests were con-ducted to exclude any pre-existing cardiopulmo-naryproblems.TheHAPE-rsubjectswereelit from the M ountaineering Association of Nagano Prefecture (Nagano,Japan)and the Alpine Club ofShinshu University(Matsumoto,Japan)and often climbed mountains higher than 3,000 m. We defined these subjects as HAPE-r due to their resis-tance to HAPE during exposure to high-altitude environments.During recruitment,no HAPE-r sub-jects reported a historyofmedicalproblems relatedShinshu Med J Vol.63EPAS1 and EGLN1 interaction in high-altitude pulmonary edema to altitudeor cardiopulmonarydisorders in a questi-are located in introns.onnaire containing the components of the Lake . Venous blood samples were colLouise Score(cid:50892)(cid:50891)(cid:50903)lected and frozen at(cid:62781)-70℃ for research purposes.(cid:50774) Selection ofThe EGLN1 gene contains four coding exons andmaps to chromosome 1q42-q43. Three SNPs (cid:50773) Selection of［rs1435166 (A/G), rs7542797 (A/C), and rs2153364The tag SNPs act as direct proxies for all otheruntyped SNPs because they are highly correlated with one another by means of pairwise tagging(cid:50892)(cid:50892)(cid:50903).Setting the threshold ofcorrection coefficient (r(cid:50892))at 1.0 resulted in a perfect correlation between the tag SNP and untyped SNPs, capturing the untyped SNPs perfectly and efficiently.The human EPAS1 gene contains 16 exons, withspans of at least 120 kb, and maps to chromosome 2p21-p16. The introns within the genomic region encode the NN-terminal basic helix-loop-helix (bHLH)/PAS proteins,which arecriticalregulators of gene expression networks of transcriptional responses to low oxygen tension mediated by the HIF signaling pathway(cid:50892)(cid:50893)(cid:50903). Three SNPs［rs13419896(G, ancestral allele/A, derived allele) according to the NCBI dbSNP database, rs4953354 (A/G), and rs4953388 (G/A)］in EPAS1 were selected as tag SNPs because they can capture (r(cid:50892)＝1.0) numerous untyped SNPs in EPAS1 from the genotype data in the International HapM ap Consortium 2005 (http:// hapmap.ncbi.nlm.nih.gov/)with the Japanese popu-(A/G)］were selected as tag SNPs due to an evalua-tion oftheir tagging efficiency(r(cid:50892)＝1.0)for available SNPs in EGLN1 from the genotype data in the InternationalHapM ap Consortium 2005(http://hap-map.ncbi.nlm.nih.gov/) with the JPT population(Table 2). In particular, EGLN1 rs1435166rs1435166 (Table 2)had highly pairwise correlations with seven SNPs reported to be significantlyassociated with HAPE-s in the Indian population(cid:50891)(cid:50894)(cid:50903)All three SNPs are located in introns. (cid:50775) Genotyping GenomicDNA samplesfromHAPE-sandHAPE-rsubjects were extracted from venous blood leuko-cytes as described previously(cid:50895)(cid:50903)(cid:50891)(cid:50 Genotyping Assay M ix containing the forward and reverse primers and FAM(cid:50979)(cid:50947)and VIC(cid:50979)(cid:50947)dye-minor groove binder-labeled probes for the six SNPs were pur- chased from Applied Biosystems Inc. (Tokyo,Japan). Allelic discrimination was performed using a pre-designed 5′nuclease assay (TaqM an(cid:51378) SNP Genotyping Assay)and theApplied Biosystems 7500 Fast Real-timePCR System according to themanu-lation from Tokyo, Japan (JPT) (Table 1). In addi-s instructions (Applied Biosystems Inc,tion, these SNPs are significantly associated with hypoxia tolerance in indigenous high-altitude Tibetan and Sherpa populations(cid:50897)(cid:50903)(cid:50891)(cid:5 three SNPsFoster City, CA, USA). Following thermal cycling,genotype data were acquired automatically and analyzed using sequence detection software (SDSTable 1 Tagging efficiency of single nucleotide polymorphisms (SNPs) rs13419896 (G/A), rs4953354 (A/G), andrs4953388 (G/A)in EPAS1 in the Tokyo, Japan (JPT)population(cid:50924). HapMap-JPTTest SNP Alleles captured(cid:50924)(cid:50924)rs13419896 rs4952819, rs4953342, rs4952820, and rs4953348 rs4953354 rs10187368, rs4953355, rs1868085, rs1868086, rs1868087, rs1374748, and rs13409493rs4953388rs1447563, rs4953372, rs11125075, rs1530624, rs4396767, rs12986899, rs13003074, rs2121700, rs718849, rs13020043, rs13002880, rs2121698, rs4953387, rs4953390, rs13030579, rs12986653, rs13001507, rs13007688,rs13011721, rs6544898, rs13005507, rs13024277, rs13024546, rs13025211, rs17817004, rs4953396, rs2084679, rs13010097, rs2346414, rs2346415, rs880671, and rs4952830(cid:50924)Checked by tagger software based on the genotype data in the International HapMap Consortium 2005 (http:// hapmap.ncbi.nlm.nih.gov/)with Japanese from the JPT.(cid:50924)(cid:50924)The threshold of pairwise correction (r(cid:50892))was set at 1.0, at which all alleles are to be captured. No. 3, 2015Horiuchi.Ota.Droma et al.Table 2 Tagging efficiency of single nucleotide polymorphisms (SNPs) rs1435166 (A/G), rs7542797 (A/C), andrs2153364 (A/G)in EGLN1 in the Tokyo, Japan (JPT)population(cid:50924). HapMap-JPTTest SNP Alleles captured(cid:50924)(cid:50924)rs479200,rs2486729,r rs2790889, rs2486736, rs2749699, rs2808616, rs2808584, rs2808586, rs2790890, rs2486731, rs2486742,rs2790887,rs2749710,rs480902,rs2066140,r rs2491405, rs2491404, rs2790891, rs2790882, and rs2486745rs7542797 rs7542797 rs2153364 rs2153364 (cid:50924)Checked by tagger software based on the genotype data in the International HapMap Consortium 2005 (http:// hapmap.ncbi.nlm.nih.gov/)with Japanese from the JPT. (cid:50924)(cid:50924)Thethreshold ofpairwisecorrection (r(cid:50892))was set at 1.0,at which allalleles areto becaptured.TheSNPs in bold were reported to be significantly associated with HAPE-s in the Indian population (14).v1.3.1, Applied Biosystems Inc.).(cid:50801) Statistical analysis The chi-square test was used for between-group comparisons,but Fisher’s exact probabilitytest was used for comparisons when the number of subjects was less than five. The exact test of Hardy-Wein-berg equilibrium (HWE) was performed by the Markov chain method,which is reported to havethe advantage ofobtaining completeenumeration when the number of alleles and sample size are small(cid:50892)(cid:50894)(cid:50903).Therefore,t possibilityoftypeII error dueto our relatively small sample sizes was minimized.Two-tailed P-values ＜0.05 indicated significance. The current geneticanalysis wastrustworthydespitethe relativelysmallsamplesizes in both groups because the studied cohorts were ethnically homogenous Japanese.Gene-gene interactions were evaluated by multi-dimensional reduction (MDR,version 3.0.2)with the variable alleles of the SNPs in the two genes.(cid:50305) Results(cid:50771) Single gene model The genotype distributions of the six SNPs con-formed to HWE (P＞0.05) in both the case and control groups. No significant differences were detected between the case and control groups in terms of genotype distributions or allelic fre-quencies for the six SNPs (Table 3),suggesting that theseSNPs werenot associated with HAPE suscep-tibility in the single gene model.(cid:50772) Gene-gene interaction analysis Because three SNPs (EPAS1 rs4953354 andrs4953388 and EGLN1 rs7542797rs7542797) were not derived homozygous genotypes in the HAPE-s and HAPE- r groups (Table 3), the EGLN1 rs2153364 (A/G) x EPAS1 rs13419896 (G/A)model was the best model of random pairwise combinations for the six SNPs across the two genes,based on a balanced accuracy of 63.23% and a cross-validation consistency of 6/Random combinations of pairwise SNPs were anal-10 in the performance of MDR in the whole datasetyzed by M DR (http://www.multifactordimen-(Fig. 1). This best pairwise interaction model (Fig.sionalityreduction.org/) to detect and characterize the pairwise interactions of these SNPs across the two genes. The relative risk (RR) and 95%1) showed a significant association with HAPE-s［P＝0.0049 ;95% confidence interval (CI)3.16 (1.40-7.14)］. Regarding HAPE susceptibility, the interac-confidence interval (95% CI) were calculated accordingly.tion between EGLN1 rs2153364 AArs2153364 AA (ancestral genotype)and EPAS1 rs13419896 GArs13419896 GA (heterozygous genotype) was categorized as the most high-risk combination (9 subjects ofHAPE-s vs.3subjects of Shinshu Med J Vol.63EPAS1 and EGLN1 interaction in high-altitude pulmonary edema Table 3 Genotype distributions and allele frequencies of rs13419896, rs4953354, and rs4953388 in EPAS1 and rs1435166, rs7542797, and rs2153364 in EGLN1 in HAPE-s (n＝59)and HAPE-r subjects (n＝67) Alleles(cid:50924)Frequency 1/2 HAPE-s HAPE-rGenotype distributionP(cid:50323)P(cid:50115)P(cid:50322)HA HAPE-r HAPE-s HAPE-r HAPE-s HAPE-r rs13419896 G/A EPAS1 rs4953354 A/G rs4953388 G/Ars1435166 A/GA/G rs4953388 G/Ars1435166 A/G EGLN1 rs7542797 A/C rs2153364 A/Gn＝total number of subjects ;SNP＝single nucleotidepolymorphism ;HAPE-s＝subjects susceptibleto high-altitudepulmonaryedema ; HAPE-r＝subjects resistant to high-altitude pulmonary edema. (cid:50924)1/2 indicates ancestral allele/derived allele according to the NCBI dbSNP database. (cid:50322)Chi-square test (2×2 contingency table). (cid:50323)2×2 contingency table assuming dominant mode (11/12＋22)of inheritance in HAPE-s. (cid:50115)2×2 contingency table assuming recessive mode (11＋12/22)of inheritance in HAPE-s. Fig. 1 Thebest modelasdetermined bymulti-dimensionalreduction (M DR)forEPAS1 rs13419896(G/ A)x EGLN1 rs2153364 (A/G). The numbers within each square represent the numbers of cases (left) and controls (right). For each square, dark-grey shading indicates a high risk of high- altitude pulmonary edema (HAPE), whereas light-grey shading represents a low risk of HAPE and white shading indicates no combination.The interaction between EGLN1 rs2153364 AArs2153364 AA and EPAS1 rs13419896 GArs13419896 GA was categorized as the most high-risk combination for HAPE susceptibil - ity［relative risk (RR)＝3.41 with 95% confidence interval (95% CI)from 0.97 to 11.99］,whereas the interaction between EGLN1 rs2153364 AG and EPAS1 rs13419896 AA was the most low-risk combination for HAPE susceptibilityin the Japanese population (RR＝0.23 with 95% CI from 0. 03to 1.89).Theother high-risk combinations wereEGLN1 rs2153364AA x EPAS1 rs13419896GG (RR＝2.04 with 95% CI from 0.73 to 5.76) and EGLN1 rs2153364 GG x EPAS1 rs13419896 GG (RR＝1.70with 95% CI from 0.064to 4.50),whereastheotherlow-risk combinationswereEGLN1 rs2153364 AG x EPAS1 rs13419896 GG (RR＝0.70 with 95% IC from 0.39 to 1.28), EGLN1 rs2153364 AG x EPAS1 rs13419896 GA (RR＝0.71 with 95% IC from 0.25 to 2.05) and EGLN1 rs2153364 GG x EPAS1 rs13419896 GA (RR＝0.97 with 95% IC from 0.35 to 2.73). No. 3, 2015Horiuchi.Ota.Droma et al. HAPE-r,RR＝3.41 with 95% CI from 0.97 to 11.99)erythropoietin, among others, to challenge hypoxia among the detected three high-risk combinations, stress physiologically(cid:50896)(cid:50903)(cid:50892)(c the interaction between EGLN1 rs2153364rs2153364 AG (heterozygous genotype)and EPAS1 rs13419896 AA (ancestral genotype) was the most low-risk combination (1 subjects ofHAPE-s vs.5subjects of HAPE-r, RR＝0.23 with 95% CI from 0.03 to 1.89)among the four detected low-risk combinations of genotypes in the present Japanese population (Fig. (cid:50306) DiscussionThe main result of the present study was the detection of a potential association of the interac-tion between EPAS1 and EGLN1 with HAPE sus- ceptibility in the Japanese population.The subjects carrying the AA ancestral genotype of EGLN1 rs2153364 (A/G)and the GA heterozygous genotype of EPAS1 rs13419896 (G/A) were at high risk for HAPE.The present studyprovides somesuggestive information about the EPAS1 and EGLN1 genes in association with HAPE susceptibility in regards to the HIF signaling pathway. However, these two genes did not show an association with HAPE-s in the Japanese population in the single gene model. EPAS1 encodes the hypoxia-inducible-factor-2alpha (HIF2A)protein,which shares 48% sequence identitywith HIF1A(cid:50892)(cid:50893)(cid:50903).Similar to HIF1A,EPAS1 is stimulated by hypoxic conditions and heterodimer-izes with HIF2B for transcriptional activation of target genes(cid:50896)(cid:50903)In contrast, EGLN1 encodes prolylprolyl hydroxylase domain protein 2(PHD2),which senses hypoxia and negatively affects the activity of HIF2A and plays a regulatory role in mammalian oxygen homeostasis(cid:50892)(cid:50895)(cid:50903) In.normoxia, HIF2Asubunits are marked for the ubiquitin-proteasome degradation pathway through hydroxylation of proline by PHD2. When exposed to hypoxia, the PHD2-dependent hydroxylation is inadequate and HIF2A marks for ubiquitin-proteasome degrada-Over theyears,a bodyofevidencehas accumulat-ed for the genetic loci of EPAS1 and EGLN1 being associated with human adaptation to high-altitude hypoxia through generations(cid:50897)(cid:50903)(cid:50873)(cid:5 three SNPs［rs13419896 (G/A), rs4953354 (A/G), rs4953388 (G/A)］ in EPAS1 are significantly associated with hypoxia tolerance in indigenous high-altitude Tibetan(cid:50897)(cid:50903)and Sherpa(cid:50891)(cid:50896)(cid:50903)populations addition,the three SNPs ［rs1435166 (A/G), rs7542797 (A/C),rs2153364 (A/G)］in EGLN1 highly correlate with rs479200, rs2486729, rs2486736, rs480902, rs2790879,rs973253, and rs1538664 in EGLN1 in the Japanese population ; these SNPs have been significantly associated with HAPE-s subjects in theIndo-Aryan population(cid:50891)(cid:50894)(cid:50903)However the present study did not show any associations between the six SNPs and HAPE susceptibility in the Japanese population in the single gene model.On the other hand,the gene- geneinteraction analysis suggested a potentialasso-ciation of the EPAS1-EGLN1 interaction with HAPE susceptibility in the Japanese population.The adaptive physiological features of Tibetans,such as a low hypoxic pulmonary vasoconstrictor response and relativelylowhemoglobin levels,were acquired through naturalselection overgenerations,as the ancestors of Tibetans occupied the high-altitude region as early as 30,000 years ago(cid:50891)(cid:50897)(cid:50903)In. contrast, the pathophysiological features of HAPE characterized byacute hypoxia-induced pulmonary vasoconstriction and permeable pulmonary vascu-lar endothelium occur acutelyin lowlanders rapidly exposed to high altitudes and hypoxia(cid:50891)(cid:50903)(cid:50873)(cid:50894 process of natural selection in the Tibetan adapta-tion to hypoxia is distinct from the pathogenesis of an acute challenge to hypoxic stress in lowlanders rapidly exposed to high altitudes. Regarding the significant associations of EGLN1 loci with suscep-tibility to HAPE in the Indo-Aryan population(cid:50891)(cid:50894)(cid:50903),tion fade away, thereby enhancing the HIF signa-ling pathwayand inducing theexpression ofa series of genes that activate glycolytic enzymes, hemeox-the probability of replicating such results is only occasional in the Japanese population because of the polygenic architecture of the human genome(cid:50892)(cid:50896)(cid:50903)ygenase, vascular endothelial growth factor, andand the characteristics of SNPs(cid:50892)(cid:50897)(cid:50903). Nevertheless,Shinshu Med J Vol.63EPAS1 and EGLN1 interaction in high-altitude pulmonary edema tureand function ofgeneticpathways and theevolu-EGLN1 rs480902rs480902 did not show a significant associa-tion with acute mountain sickness in the Han Chinese population(cid:50891)(cid:50895)(cid:50903).Epista interactions between genes,is funda- mentallyimportant to understanding both thestruc-tionary dynamics of complex genetic systems(cid:50892)(cid:50898)(cid:50903).The methodology of gene-gene interaction analysis was eventually carried out to assess the multi-dimensional relations of the EPAS1 and EGLN1 genes of the HIF signaling pathway with HAPE susceptibility in the Japanese population in the present study,despite the fact that numerous candi- date genes were individually identified as being significantly associated with HAPE-s populations worldwide(cid:50892)(cid:50899)(cid:50903)(cid:508 The current superficial results are not direct evidenceofthefunctionsofand communi- cations between the geneticvarieties for thephysio-bility.However,thepresent resultsshowing that the ancestral homozygous AA genotype of EGLN1 rs2153364 (A/G) interacts with EPAS1 rs13419896 (G/A)as a high-risk genotype in HAPE-s subjects is in agreement with the previous report by M ishar et al.; they reported a 4.55-fold up-regulation of EGLN1 expression in HAPE-s subjects compared to HAPE-r subjects(cid:50891)(cid:50894)(cid:50903).Therefor proposed that the up-regulation of the negative regulator of EGLN1 causes down-regulation ofproteasome deg-radation of HIF-2 in a hypoxic environment, lead-dothelial growth factor, and erythropoietin, among others,to participate in the pathogenesis of HAPE.Detailed investigations are expected to determine the functions of the gene-gene interactions in the HIF signaling pathway in the pathogenesis of The power of the present study was inadequatedue to the limitations of the M DR method, which reduced high-dimensional genetic data into a single dimension so that gene-gene interactions could be detected in a sample size smaller than the original sample size. HAPE is actually a relatively rare disease and occurs only in lowlanders rapidly exposed to high altitudes, so the sample size is always smaller than other common diseases. Thus, the present gene-gene interaction analysis needs to be replicated in other populations with HAPE and proven by other technical analyses of gene-gene In conclusion, the present study demonstrated apotential association of the EPAS1×EGLN1 inter-action with HAPE susceptibility in the Japanese population, despite no associations detected in the single gene model.This finding is small,but encour-aging,and needs to bevalidated through replication studies with large sample sizes and different eth- nicities to establish the global efficacy of these genes in the HIF signaling pathway regarding hypoxic response in HAPE susceptibility.logicalresponseto acutehypoxia in HAPE suscepti-interactions, such as BOOST(cid:50893)(cid:50893)(cid:50903).ing to HIF-2 accumulation and inducing over-The authors state that thay have no conflict ofinterest in the present study. expressions of downstream genes that activate glycolytic enzymes, hemeoxygenase, vascular en-References1) Smith CA,DempseyJA,Hornbein TF :Control ofbreathing at high altitude.In :Hornbein TF,SchoeneRB (eds),High altitude:an exploration of human adaptation, pp 525-568, Marcel Dekker, New York, 20012) Hohenhaus E,Paul A,McCullough RE,Kucherer H,Bartsch P :Ventilatoryand pulmonaryvascular response tohypoxia and susceptibility to high altitude pulmonary oedema. Eur Respir J 8:1825-1833, 19953) Maggiorini M :High altitude-induced pulmonary oedema. Cardiovasc Res 72:41-50, 20064) Hackett PH, Roach RC :High-altitude illness. N Engl J Med 345:107-114, 2001 5) Kobayashi N,Hanaoka M,Droma Y,Ito M,Katsuyama Y,Kubo K,Ota M :Polymorphisms ofthetissueinhibitor of metalloproteinase 3 gene are associated with resistance to high-altitude pulmonary edema (HAPE) in a Japanese population :A case control studyusing polymorphic microsatellite markers.PLoS ONE,8:e71993.doi:No. 3, 2015Horiuchi.Ota.Droma et al.6) Semenza GL :Regulation ofoxygen homeostasis byhypoxia-induciblefactor 1.Physiology(Bethesda)24:97-106,10.1371/journal. pone. 0071993, 20137) Beall CM,Cavalleri GL,Deng L,Elston RC,Gao Y,Knight J,Li C,Li JC,Liang Y,M cCormack M ,MontgomeryiHE,Pan H,Robbins PA,Shianna KV,Tam SC,Tsering N,Veeramah KR,Wang W,Wangdui P,Weale ME,Xu Y, Xu Z, Yang L, Zaman MJ, Zeng C, Zhango L, Zhang X, Zhanxi P, Zheng YT :Natural selection on EPAS1(HIF2α)associated with low hemoglobin concentration in Tibetan highlanders. Proc Natl Acad Sci U S A 107:11459-11464, 20108) Simonson TS,Yang Y,HuffCD,Yun H,Qin G,Witherspoon DJ,Bai Z,Lorenzo FR,Xing J,JordeLB,Prchal JT,Ge R :Genetic evidence for high-altitude adaptation in Tibet. Science 329 :72-75, 20109) Simonson TS,McClain DA,JordeLB,PrchalJT :Geneticdeterminants ofTibetan high-altitudeadaptation.Hum Genet 131:527-533, 201210) Yi X, Liang Y, Huerta-Sanchez E, Jin X, Cuo ZX, Pool JE, Xu X, Jiang H, Vinckenbosch N, Korneliussen TS,Zheng H,Liu T,HeW,Li K,Luo R,NieX,Wu H,Zhao M ,Cao H,Zou J,Shan Y,Li S,Yang Q,Asan,NiP,Tian G, Xu J, Liu X, Jiang T, Wu R, Zhou G, Tang M , Qin J, Wang T, Feng S, Li G, Huasang, Luosang J, Wang W,Chen F, Wang Y, Zheng X, Li Z, Bianba Z, Yang G, Wang X, Tang S, Gao G, Chen Y, Luo Z, Gusang L, Cao Z,Zhang Q, Ouyang W, Ren X, Liang H, Zheng H, Huang Y, Li J, Bolund L, Kristiansen K, Li Y, Zhang Y, Zhang X, Li R, Li S, Yang H, Nielsen R, Wang J, Wang J :Sequencing of 50 human exomes reveals adaptation to high altitude. Science 329 :75-78, 2010 11) Bigham AW,Lee FS :Human high-altitude adaptation :forward genetics meets the HIF pathway.Genes Dev 28:2189-2204, 201412) Song D, Li LS, Arsenault PR, Tan Q, Bigham AW, Heaton-Johnson KJ, M aster SR, Lee FS :Defective TibetanPHD2 binding to p23 links high altitude adaption to altered oxygen sensing. J Biol Chem 289 :14656-14665, 201413) Xiang K, Ouzhuluobu, Peng Y, Yang Z,Zhang X,Cui C,Zhang H,Li M ,Zhang Y,Bianba,Gonggalanzi,Basang, Ciwangsangbu,Wu T,Chen H,ShiH,QiX,Su B :Identification ofa Tibetan-specificmutation in thehypoxicgene EGLN1 and its contribution to high-altitude adaptation. M ol Biol Evol 30:1889-1898, 201314) M ishra A, Mohammad G, Thinlas T, Pasha MA : EGLN1 variants influence expression and SaO2 levels toassociate with high-altitude pulmonary oedema and adaptation. Clin Sci (Lond)124:479-489, 201315) Buroker NE, Ning XH, Zhou ZN, Li K, Cen WJ, Wu XF, Zhu WZ, Scott CR, Chen SH : EPAS1 and EGLN1 associations with high altitude sickness in Han and Tibetan Chinese at the Qinghai-Tibetan Plateau.Blood Cells M ol Dis 49 :67-73, 201216) Hanaoka M,Droma Y,Basnyat B,Ito M ,KobayashiN,Katsuyama Y,Kubo K,Ota M :Geneticvariantsin EPAS1contribute to adaptation to high-altitude hypoxia in Sherpas. PLoS ONE 7:e50566. doi:10.1371/journal. pone.0050566, 201217) Zhao M,Kong QP,Wang HW,Peng MS,Xie XD,Wang WZ,Jiayang,Duan JG,Cai MC,Zhao SN,Cidanpingcuo,Tu YQ, Wu SF, Yao YG, Bandelt HJ, Zhang YP : M itochondrial genome evidence reveals successful Late Paleolithic settlement on the Tibetan Plateau. Proc Natl Acad Sci U S A 106:21230-21235, 200918) Gilbert-Kawai ET,Milledge JS,Grocott M P,M artin DS :King ofthemountains:Tibetan and Sherpa physiologi-cal adaptations for life at high altitude. Physiology (Bethesda)29 :388-402, 201419) World Medical Association General Assembly: World M edical Association Declaration of Helsinki: ethicalprinciples for medical research involving human subjects. J Int Bioethique 15:124-129, 200420) Kobayashi T, Koyama S, Kubo K, Fukushima M , Kusama S : Clinical features of patients with high-altitudepulmonary edema in Japan. Chest 92:814-821, 198721) Savourey G, Guinet A, Besnard Y, Garcia N, Hanniquet AM, Bittel J : Evaluation of the Lake Louise acute Shinshu Med J Vol.63EPAS1 and EGLN1 interaction in high-altitude pulmonary edema mountain sickness scoring system in a hypobaric chamber. Aviat Space Environ Med 66:963-967, 199522) de Bakker PI,YelenskyR,Pe er I,Gabriel SB,DalyMJ,Altshuler D :Efficiencyand power in geneticassociationstudies. Nat Genet 37:1217-1223, 2005 23) Tian H,McKnight SL,Russell DW :Endothelial PAS domain protein 1(EPAS1),a transcription factor selectivelyexpressed in endothelial cells. Genes Dev 11:72-82, 199724) Guo SW,Thompson EA :Performing theexact test ofHardy-Weinberg proportion formultiplealleles.Biometrics48:361-372, 199225) Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O Rourke J, M ole DR, Mukherji M, M etzen E, Wilson MI, Dhanda A, Tian YM, Masson N, Hamilton DL, Jaakkola P, Barstead R, Hodgkin J, Maxwell PH, Pugh CW,Schofield CJ, Ratcliffe PJ : C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell 107:43-54, 200126) 1000Genomes Project Consortium,Abecasis GR,Altshuler D,Auton A,Brooks LD,Durbin RM ,Gibbs RA,HurlesME, McVean GA :A map of human genome variation from population-scale sequencing.Nature 467:1061-1073,826-830, 200227) Gorlov IP, Moore JH, Peng B, Jin JL, Gorlova OY, Amos CI :SNP characteristics predict replication success inassociation studies. Hum Genet 133:1477-1486, 201428) Phillips PC :Epistasis-theessential roleofgeneinteractions in thestructureand evolution ofgeneticsystems.NatRev Genet 9 :855-867, 200829) Droma Y,Hanaoka M,Ota M,Katsuyama Y,KoizumiT,Fujimoto K,KobayashiT,Kubo K :Positiveassociation oftheendothelialnitricoxidesynthasegenepolymorphis with high-altitudepulmonaryedema.Circulation 106:30) Qi Y,Sun J,Zhu T,Wang W,Liu J,Zhou W,Qiu C,Zhao D :Association ofangiotensin-converting enzyme geneinsertion/deletion polymorphism with high-altitude pulmonary oedema : a meta-analysis. JRAAS 12: 617-623,31) Woods DR,Pollard AJ,Collier DJ,Jamshidi Y,Vassiliou V,HaweE,Humphries SE,MontgomeryHE :Insertion/deletion polymorphism of the angiotensin I-converting enzyme gene and arterial oxygen saturation at high altitude. Am J Respir Crit Care Med 166:362-366, 200232) Luo Y, Zou Y, Gao Y :Gene polymorphisms and high-altitude pulmonary edema susceptibility:a 2011 update.Respiration 84:155-162, 201233) Wan X, Yang C, Yang Q, Xue H, Fan X, Tang NL, Yu W :BOOST :A fast approach to detecting gene-geneinteractions in genome-wide case-control studies. Am J Hum Genet 87:325-340, 2010 （2015 . 1.29 received ;2015. 2.17 accepted）No. 3, 2015