#PAGE_PARAMS# #ADS_HEAD_SCRIPTS# #MICRODATA#

Identification of loci of functional relevance to Barrett’s esophagus and esophageal adenocarcinoma: Cross-referencing of expression quantitative trait loci data from disease-relevant tissues with genetic association data


Authors: Julia Schröder aff001;  Vitalia Schüller aff002;  Andrea May aff003;  Christian Gerges aff004;  Mario Anders aff005;  Jessica Becker aff001;  Timo Hess aff001;  Nicole Kreuser aff008;  René Thieme aff008;  Kerstin U. Ludwig aff001;  Tania Noder aff005;  Marino Venerito aff009;  Lothar Veits aff010;  Thomas Schmidt aff011;  Claudia Fuchs aff012;  Jakob R. Izbicki aff013;  Arnulf H. Hölscher aff012;  Dani Dakkak aff014;  Boris Jansen-Winkeln aff008;  Yusef Moulla aff008;  Orestis Lyros aff008;  Stefan Niebisch aff008;  Matthias Mehdorn aff008;  Hauke Lang aff015;  Dietmar Lorenz aff016;  Brigitte Schumacher aff014;  Rupert Mayershofer aff017;  Yogesh Vashist aff013;  Katja Ott aff011;  Michael Vieth aff010;  Josef Weismüller aff020;  Elisabeth Mangold aff001;  Markus M. Nöthen aff001;  Susanne Moebus aff021;  Michael Knapp aff002;  Horst Neuhaus aff004;  Thomas Rösch aff005;  Christian Ell aff003;  Ines Gockel aff008;  Johannes Schumacher aff007;  Anne C. Böhmer aff001
Authors place of work: Institute of Human Genetics, University of Bonn, School of Medicine & University Hospital Bonn, Bonn, Germany aff001;  Institute for Medical Biometry, Informatics, and Epidemiology, University of Bonn, School of Medicine & University Hospital Bonn, Bonn, Germany aff002;  Department of Medicine II, Sana Klinikum, Offenbach, Germany aff003;  Department of Internal Medicine II, Evangelisches Krankenhaus, Düsseldorf, Germany aff004;  Department of Interdisciplinary Endoscopy, University Hospital Hamburg-Eppendorf, Hamburg, Germany aff005;  Department of Gastroenterology and Interdisciplinary Endoscopy, Vivantes Wenckebach-Klinikum, Berlin, Germany aff006;  Center for Human Genetics, University Hospital Marburg, Marburg, Germany aff007;  Department of Visceral, Transplant, Thoracic and Vascular Surgery, University Hospital of Leipzig, Leipzig, Germany aff008;  Department of Gastroenterology, Hepatology and Infectious Diseases, Otto-von-Guericke University Hospital, Magdeburg, Germany aff009;  Institute of Pathology, Klinikum Bayreuth, Bayreuth, Germany aff010;  Department of General, Visceral and Transplantation Surgery, University of Heidelberg, Heidelberg, Germany aff011;  Department of General, Visceral, and Cancer Surgery, University of Cologne, Cologne, Germany aff012;  Department of General, Visceral, and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, University of Hamburg, Hamburg, Germany aff013;  Department of Internal Medicine and Gastroenterology, Elisabeth Hospital, Essen, Germany aff014;  Department of General, Visceral, and Transplant Surgery, University Medical Center, University of Mainz, Mainz, Germany aff015;  Department of General, Visceral, and Thoracic Surgery, Klinikum Darmstadt, Darmstadt, Germany aff016;  Gastroenterologie am Burgweiher, Bonn, Germany aff017;  Kantonsspital Aarau, Aarau, Switzerland aff018;  Department of General, Visceral, and Thorax Surgery, RoMed Klinikum Rosenheim, Rosenheim, Germany aff019;  Gastroenterologische Gemeinschaftspraxis, Koblenz, Germany aff020;  Centre of Urban Epidemiology, Institute of Medical Informatics, Biometry, and Epidemiology, University of Essen, Essen, Germany aff021
Published in the journal: PLoS ONE 14(12)
Category: Research Article
doi: https://doi.org/10.1371/journal.pone.0227072

Summary

Esophageal adenocarcinoma (EA) and its precancerous condition Barrett’s esophagus (BE) are multifactorial diseases with rising prevalence rates in Western populations. A recent meta-analysis of genome-wide association studies (GWAS) data identified 14 BE/EA risk loci located in non-coding genomic regions. Knowledge about the impact of non-coding variation on disease pathology is incomplete and needs further investigation. The aim of the present study was (i) to identify candidate genes of functional relevance to BE/EA at known risk loci and (ii) to find novel risk loci among the suggestively associated variants through the integration of expression quantitative trait loci (eQTL) and genetic association data. eQTL data from two BE/EA-relevant tissues (esophageal mucosa and gastroesophageal junction) generated within the context of the GTEx project were cross-referenced with the GWAS meta-analysis data. Variants representing an eQTL in at least one of the two tissues were categorized into genome-wide significant loci (P < 5×10−8) and novel candidate loci (5×10−8 ≤ P ≤ 5×10−5). To follow up these novel candidate loci, a genetic association study was performed in a replication cohort comprising 1,993 cases and 967 controls followed by a combined analysis with the GWAS meta-analysis data. The cross-referencing of eQTL and genetic data yielded 2,180 variants that represented 25 loci. Among the previously reported genome-wide significant loci, 22 eQTLs were identified in esophageal mucosa and/or gastroesophageal junction tissue. The regulated genes, most of which have not been linked to BE/EA etiology so far, included C2orf43/LDAH, ZFP57, and SLC9A3. Among the novel candidate loci, replication was achieved for two variants (rs7754014, Pcombined = 3.16×10−7 and rs1540, Pcombined = 4.16×10−6) which represent eQTLs for CFDP1 and SLC22A3, respectively. In summary, the present approach identified candidate genes whose expression was regulated by risk variants in disease-relevant tissues. These findings may facilitate the elucidation of BE/EA pathophysiology.

Keywords:

Gene expression – Genetic loci – Gene regulation – Molecular genetics – Genome-wide association studies – Genetics of disease – Replication studies

Introduction

Esophageal adenocarcinoma (EA) represents one of the most rapidly increasing cancers in Western populations [1]. Despite new treatment strategies, mortality rates among EA patients remain high [1]. EA is preceded by the precancerous condition Barrett’s esophagus (BE), which is characterized by a metaplastic transformation of the squamous epithelium in the distal esophagus. Here, the normal stratified squamous epithelium at the gastroesophageal junction is replaced by columnar epithelium, commonly found in the lower gastrointestinal tract. The prevalence of BE in the general population of Western countries is 1.6% [2]. Reported non-genetic risk factors for BE/EA include gastroesophageal reflux, obesity, and age > 50 years [3]. Additionally, family studies of EA and BE have implicated genetic factors in disease development and progression, thus demonstrating that the etiology of BE/EA is multifactorial [4,5]. Furthermore, genetic research has shown that BE and EA display a polygenic overlap [6].

In a recent meta-analysis of data from genome-wide association studies (GWAS), separate (BE and EA) and combined (BE/EA) analyses identified 14 genetic risk loci for BE/EA [7]. The majority of the associated variants map to intergenic or intronic regions of the genome, which renders the identification of the disease-relevant genes and underlying pathomechanisms difficult. Since many non-coding GWAS risk variants exert their effects via gene regulatory mechanisms, expression quantitative trait loci (eQTL) analyses make an important contribution to the elucidation of multifactorial disease etiology [8,9]. In eQTL studies, the alleles or genotypes of genetic variants are correlated with the quantitative expression level of transcripts [10], thereby identifying genetic variants that influence the expression level of a gene. This method is useful for identification of candidate genes at risk loci for functional follow-up studies [1113].

The aim of the present study was (i) to identify novel candidate genes of functional relevance to BE/EA at known risk loci and (ii) to find novel risk loci among the suggestively associated variants. This was accomplished by integrating eQTL data from BE/EA-relevant tissues (esophageal mucosa and gastroesophageal junction) [14] and genetic data from the recent BE/EA GWAS meta-analysis [7]. Variants with suggestive evidence for association were further investigated in a genetic association analysis in an independent replication case-control cohort. To increase statistical power, association data were combined with the data of the previous GWAS meta-analysis [7].

Materials and methods

The study workflow is illustrated in Fig 1.

Study workflow.
Fig. 1. Study workflow.

Cross-referencing of eQTLs from BE/EA-relevant tissues with BE/EA genetic association data

The GTEx project [14] represents the largest eQTL database to date comprising 152,869 cis-eQTLs from 44 tissues (V6P). Tissue was sampled from donors post-mortem and eQTLs were mapped using tissue-specific RNA sequencing data and genotype data of DNA from whole blood. Two tissue types most relevant to BE/EA were selected from the GTEx database [14]: esophageal mucosa and gastroesophageal junction. Both datasets were restricted to eQTLs with a false discovery rate (FDR) of < 0.05. The esophageal mucosa sample comprised tissues of 241 individuals with 6,169 cis-eQTLs (eQTL-gene located < 1 Mb distance from genetic variant) and the gastroesophageal junction sample comprised tissues of 127 individuals with 2,237 cis-eQTLs.

The eQTLs were cross-referenced to variants that showed at least a suggestive association to BE/EA (P ≤ 5×10−05) in the BE/EA GWAS meta-analysis [7]. All variants showing an eQTL effect in one or both tissues were then assigned to distinct genomic loci: The index SNP of each locus was specified as the variant with the most significant BE/EA association and each locus was defined at ± 1 Mb around this variant. All loci on the same chromosome where reviewed for independence by analysis of linkage disequilibrium (LD). Where applicable, long-range LD was taken into account and the respective loci were marked accordingly. We tested the option of applying statistical colocalization analyses on the selected loci using COLOC [15] but found that the analysis was severely underpowered most likely due to a small sample size in the eQTL datasets. The resulting loci were categorized into genome-wide significant loci (P < 5×10−8) and novel candidate loci (5×10−8 < P ≤ 5×10−5). The index SNPs of these candidate loci were then included in the subsequent genetic replication study in an independent BE/EA case-control.

Replication sample

The case-control cohort for the replication study comprised: (i) 1,117 BE cases and 876 EA cases (total of 1,993 BE/EA patients); and (ii) 967 controls. The cases were recruited in an ongoing effort as described for the Bonn sample in Gharahkhani et al. [7]. All samples included in this replication cohort have not been part of the prior Gharahkhani et al. [7] GWAS and were recruited between November 2014 and February 2018. Patients with suspected BE/EA were recruited in hospitals and clinics where they underwent endoscopies or surgeries. Exclusion criteria were (i) a negative histopathological diagnosis that did not confirm BE/EA disease status and (ii) self-reported descent was non-European. The patients were recruited at 15 medical centers across Germany and blood samples were collected at the University Hospital Leipzig before being sent to Bonn for DNA extraction and genotyping. The control cohort was recruited at the University Hospital Bonn from blood donors, also of European descent. Relevant demographic details for both cohorts can be found in S1 Table.

The study was approved by the ethics committees of the Universities of Bonn and Leipzig. All participants provided written informed consent prior to inclusion.

Genotyping

Genotyping was performed using the multiplex MALDI-TOF mass spectrometer MassArray system by Agena (San Diego, USA). Of the 16 index SNPs representing the novel candidate loci, three variants (rs59341339, rs11145842, rs12985299) were excluded from the plex for technical reasons. No alternative SNP in high LD was found among the associated eQTLs, and thus the corresponding loci were excluded from the analyses. The index SNPs of three further loci, which were excluded due to the same technical reasons, were replaced by variants in high LD [r2 > 0.95; rs2442722 (P = 1.22×10−6) was replaced by rs36057735 (P = 5.13×10−6), rs76510925 (P = 7.86×10−6) by rs12112778 (P = 1.57×10−5), and rs11169302 (P = 1.05×10−5) by rs9364 (P = 2.23×10−5)]. Thus, a total of 13 variants were genotyped in the 1,993 BE/EA cases and 967 controls. Primers for amplification and genotyping were synthesized by Metabion (Martinsried, Germany). For the purposes of quality control (QC), negative controls (H2O) and intra- and inter-plate duplicates were added to each plate. After genotyping, clusters were visually inspected, and re-clustering was performed if necessary. Finally, genotype and SNP information files were extracted for the subsequent genetic association analysis.

Association analyses

Genotype QC and association calculations were performed using R and PLINK [16,17]. SNPs or samples were excluded on the basis of: (i) low call rate (SNPs: ≤ 95%, samples: > 1 missing SNP call); (ii) deviation from Hardy-Weinberg equilibrium (HWE; P < 0.05 in controls). For the replication study, association was calculated using the one-sided Cochran-Armitage trend test in the direction of effect established through the previous GWAS meta-analysis [7]. The effect size was estimated using logistic regression. Standard errors of the effect sizes were calculated with 95% confidence intervals.

All BE/EA association results from the replication sample were then combined with the association results from the BE/EA GWAS meta-analysis [7]. This was performed via a fixed-effect meta-analysis, as based on the standard-error in METAL (version 2011-03-25) [18].

Downstream analyses

Downstream analyses of the target genes of the reported variants were performed using the tool STRING [19]. The STRING database is a collection of protein-protein interaction information that also integrates tools for pathway analyses such as Gene Ontology and KEGG. A gene-set enrichment analysis was performed on the list of target genes (see Tables 1 and 2) and analyzed for possible protein-protein interactions and enrichment in pathways.

Tab. 1. Genome-wide significant loci resulting from the cross-referencing of eQTL and genetic data.
Genome-wide significant loci resulting from the cross-referencing of eQTL and genetic data.
Tab. 2. Novel loci resulting from the cross-referencing of eQTL and genetic data.
Novel loci resulting from the cross-referencing of eQTL and genetic data.

Results

Cross-referencing of eQTLs with genetic association data

In total, 6,387 SNPs in the GWAS meta-analysis [7] showed at least a suggestive association with BE/EA (P ≤ 5×10−5) and were cross-referenced to the cis-eQTL data from GTEx esophageal mucosa and gastroesophageal junction tissues [14]. Of these, 2,180 SNPs showed eQTL effects in at least one of the two tissues. These variants were assigned to 25 distinct genomic loci (see Materials and Methods, and S2 Table).

Nine of the 25 loci were reported with genome-wide significance by Gharahkhani et al. [7], but we here identified novel downstream target genes at these nine loci based on regulatory effects on gene expression (see Table 1). For three of these loci, the best-associated SNP reported by Gharahkhani et al. [7] was not present in the GTEx dataset, but the locus is represented by the next best SNP (marked with b in Table 1). Three more loci show long-range LD with another locus and have therefore not been reported as separate loci by Gharahkhani et al. [7] (marked with c or d in Table 1).

Replication of candidate loci

Variants at 16 loci presented both an eQTL effect in relevant tissues and suggestive evidence of association (5x10-8 < P ≤ 5x10-5) in the GWAS data, respectively. For 13 loci, the index SNP (or a proxy SNP in strong LD) was genotyped in the replication sample. Of the genotyped SNPs, six variants failed QC: five variants showed a call-rate < 95% and another SNP deviated from HWE (P < 0.05 in controls). In addition, 53 samples (32 cases, 21 controls) were excluded due to of the presence of > 1 missing genotype. Details of the final BE/EA replication analysis are shown in Table 2.

Upon statistical analysis, the variant rs1540 on 16q23 showed a nominally significant association to BE/EA in the independent replication study (Preplication = 0.019). In the combined analysis, a lower P-value was observed as compared to the meta-analysis data alone (Pmeta-analysis = 3.02×10−5, Pcombined = 4.16×10−6). According to the GTEx data, this variant represents an eQTL for CFDP1 in gastroesophageal junction tissue (P = 2.32×10−5). Here, the BE/EA risk allele leads to an increase in gene expression. Similarly, rs7754014 on 6q25 showed a nominally significant association to BE/EA in the replication study (Preplication = 0.028) and a lower P-value in the combined analysis (Pmeta-analysis = 2.07×10−6, Pcombined = 3.16×10−7). According to the GTEx data, this variant represents an eQTL for SLC22A3 in esophageal mucosa tissue (P = 6.61×10−5). Again, the BE/EA risk allele leads to an increase in gene expression.

Downstream analyses

The target genes of the index variants of the nine genome-wide significant and seven candidate loci were analyzed using STRING. Several genes could not be included in the analyses because they do not code for proteins (RNA genes, pseudogenes). The protein-protein interaction (PPI) analysis of the remaining 14 genes did not show any interactions between the proteins encoded by genes (PPI enrichment p = 1, see S1 Fig). Likewise, the pathway analyses did not yield any significant results.

Discussion

Previous GWAS have identified a total of 14 genetic risk loci for BE/EA [7,2022]. However, the mechanisms through which these risk variants exert their effects remain unclear. The aim of the present study was (i) to identify candidate genes of functional relevance to BE/EA at known risk loci and (ii) to find novel risk variants among the suggestively associated variants through the integration of eQTL- and genetic association data. Cross-referencing of eQTL data and genetic data from the recent GWAS meta-analysis yielded 2,180 variants at 25 loci (see S2 Table). Of these, nine loci were established BE/EA risk loci from the GWAS meta-analysis and 16 were novel candidate loci.

The replication study yielded two nominally significant BE/EA-associated variants: rs1540 and rs7754014. Variant rs1540 on 16q23 regulates the expression of the gene CFDP1 (craniofacial development protein 1) in the gastroesophageal junction. The biological function of CFDP1 remains unclear. However, research suggests that the protein is involved in both the maintenance of higher-order chromatin organization and cell cycle progression [23]. Variant rs7754014 on 6q25 represents an eQTL for the gene SLC22A3 (solute carrier family 22 member 3) in the esophageal mucosa. SLC22A3 encodes the protein OCT3 (organic cation transporter 3), which transports endogenous organic cations as well as drugs and toxins [24,25]. Interestingly, SLC22A3 expression plays a role in other esophageal disorders: downregulation of SLC22A3 was reported in patients with familial esophageal squamous cell cancer [26]. Previous authors have therefore proposed that suppression of SLC22A3 may be implicated in the progression of this cancer type [27]. It remains to be shown how these findings relate to the upregulation of SLC22A3 as it was observed in BE/EA risk allele carriers through our integrative analysis. The independent replication of these two loci gives evidence to their functional relevance for the BE/EA phenotype. This is further supported by the decrease of the P-value after the combined analysis by one order of magnitude. However, since the effect sizes are small, the P-value has not reached genome-wide significance in the combined sample. Larger patient cohorts are warranted to carry these variants over the threshold of genome-wide significance.

Among the established BE/EA risk loci from the GWAS meta-analysis [7], the present analyses identified three eQTLs with a regulating effect on biologically plausible genes. Most of these eQTLs have not been reported previously despite the fact that cross-referencing with eQTL analyses had been performed in the context of the original GWAS meta-analysis [7]. The reason is most likely the use of GTEx version 6 in the analysis by Gharahkhani et al. [7] for the cross-referencing with eQTL data, as opposed to GTEx version 6P used in the present study. While this new dataset does not differ in respect to sample size, it provides new eQTL results due to an improved gene-level annotation. The most significantly associated risk variant from the BE/EA GWAS meta-analysis was rs7255 on 2p24. This is an eQTL for the expression of the gene C2orf43 in tissue from the esophageal mucosa and the gastroesophageal junction. This gene encodes the protein LDAH (lipid droplet-associated hydrolase), which is a lipid droplet-associated serine lipid hydrolase [28]. The BE/EA risk variant rs92578209 on 6p22 regulates the expression of the gene ZFP57 (zinc finger protein 57) in both the esophageal mucosa and the gastroesophageal junction. Research has shown that among others, ZFP57 plays a key role in cell fate decisions during early mouse gastrulation [29]. The third BE/EA risk variant from the GWAS meta-analysis was rs147462972 on 5p15, which represents an eQTL for the expression of SLC9A3 (solute carrier family 9 member A3) in the esophageal mucosa. The BE/EA risk allele of this variant results in a structural change in the binding sites of the transcription factors CTCF and RAD21. Interestingly, research has demonstrated an enrichment of somatic mutations in the CTCF binding motif in patients with esophageal cancer [30]. SLC9A3 encodes the epithelial brush border Na/H-exchanger NHE3, which uses the inward sodium ion gradient to expel acids from the cell [31]. Importantly, an increase in SLC9A3 expression has been correlated with the severity of gastroesophageal reflux disease, which is a major risk factor for BE [32]. Future studies are warranted to generate further evidence for the involvement of SLC9A3 in BE/EA development.

The present study had four main limitations. First, the capacity of the GTEx and BE/EA GWAS meta-analysis data to determine whether the eQTLs and BE/EA risk SNPs referred to the same causal variants, or whether they were only correlated, was limited. A different approach using a colocalization analysis could not bring forward significant results due to a lack of power caused by a small sample size of the eQTL samples. The exploratory approach applied in this study may be prone to type I error. Nevertheless, the discovery of genes associated to related phenotypes, such as esophageal squamous cell cancer and GERD, show that our approach has merit. Further research is warranted to establish a causal relationship between these genes and their effect on BE/EA development. Second, the replication sample was too small to achieve a test-wide significant association level in the replication study and a genome-wide significant association level after combination with the previous meta-analysis for the investigated variants. Third, the tissue of origin for development of BE/EA is not completely understood. Several studies discuss the importance of tissue selection in order to detect tissue-specific eQTL effects relevant to disease etiology [3335]. However, the specific tissue or cell type relevant to a trait or disease is often unknown. In this study, we used eQTL effects in tissues drawn from esophageal mucosa and gastroesophageal junction. Wang [36] discusses the evidence for the squamous epithelium mucosa cell as a precursor for BE/EA, while Zhuang and Fitzgerald [37] debate the existence of a transitional layer at the gastroesophageal junction to be the origin of BE/EA. Thus, to our present knowledge, esophageal mucosa and gastroesophageal junction are the most likely of the currently available tissues to represent the true tissue of origin for BE/EA. Fourth, the highlighted genes have not been yet characterized in functional studies using cellular or animal models. The manner in which the genes are influencing the disease development is currently unclear and requires further investigation.

Conclusions

Altogether, this study provides a link between BE/EA-associated genetic variants and a regulatory effect on candidate genes in disease-relevant tissues. The present analyses identified biologically plausible candidate genes for BE/EA, such as SLC22A3 and SLC9A3. Notably, SLC9A3 has already been implicated with gastroesophageal reflux, rendering it an interesting candidate gene. Follow-up analyses are warranted to refine the regulatory annotation and to elucidate the mechanisms through which the implicated variants and genes influence BE/EA development.

Supporting information

S1 Table [xlsx]
Demographic details on replication cohort.

S2 Table [xlsx]
25 risk loci determined after cross-referencing of eQTL and genetic data.

S3 Table [xlsx]
Genotype counts of all 13 candidate loci in all analyzed cases and controls.

S1 Fig [tiff]
Results of protein-protein interaction analysis by STRING.


Zdroje

1. Reid BJ, Li X, Galipeau PC, Vaughan TL. Barrett’s oesophagus and oesophageal adenocarcinoma: time for a new synthesis. Nat Rev Cancer. 2010;10: 87–101. Available: doi: 10.1038/nrc2773 20094044

2. Ronkainen J, Aro P, Storskrubb T, Johansson SE, Lind T, Bolling-Sternevald E, et al. Prevalence of Barrett’s esophagus in the general population: An endoscopic study. Gastroenterology. 2005;129: 1825–1831. doi: 10.1053/j.gastro.2005.08.053 16344051

3. Spechler SJ, Souza RF. Barrett’s Esophagus. N Engl J Med. 2014;371: 836–845. doi: 10.1056/NEJMra1314704 25162890

4. Chak A, Ochs-Balcom H, Falk G, Grady WM, Kinnard M, Willis JE, et al. Familiality in Barrett’s esophagus, adenocarcinoma of the esophagus, and adenocarcinoma of the gastroesophageal junction. Cancer Epidemiol Biomarkers Prev. 2006;15: 1668–73. doi: 10.1158/1055-9965.EPI-06-0293 16985029

5. Poynton AR, Walsh TN, O’Sullivan G, Hennessy TP. Carcinoma arising in familial Barrett’s esophagus. Am J Gastroenterol. 1996;91: 1855–6. Available: http://www.ncbi.nlm.nih.gov/pubmed/8792715 8792715

6. Ek WE, Levine DM, D’Amato M, Pedersen NL, Magnusson PKE, Bresso F, et al. Germline genetic contributions to risk for esophageal adenocarcinoma, Barrett’s esophagus, and gastroesophageal reflux. J Natl Cancer Inst. 2013;105: 1711–8. doi: 10.1093/jnci/djt303 24168968

7. Gharahkhani P, Fitzgerald RC, Vaughan TL, Palles C, Gockel I, Tomlinson I, et al. Genome-wide association studies in oesophageal adenocarcinoma and Barrett’s oesophagus: a large-scale meta-analysis. Lancet Oncol. 2016;17: 1363–1373. doi: 10.1016/S1470-2045(16)30240-6 27527254

8. Morley M, Molony CM, Weber TM, Devlin JL, Ewens KG, Spielman RS, et al. Genetic analysis of genome-wide variation in human gene expression. Nature. 2004;430: 743–747. doi: 10.1038/nature02797 15269782

9. Ghoussaini M, French JD, Michailidou K, Nord S, Beesley J, Canisus S, et al. Evidence that the 5p12 Variant rs10941679 Confers Susceptibility to Estrogen-Receptor-Positive Breast Cancer through FGF10 and MRPS30 Regulation. Am J Hum Genet. 2016;99: 903–911. doi: 10.1016/j.ajhg.2016.07.017 27640304

10. Rockman M V., Kruglyak L. Genetics of global gene expression. Nat Rev Genet. 2006;7: 862–872. doi: 10.1038/nrg1964 17047685

11. Guo X, Lin W, Bao J, Cai Q, Pan X, Bai M, et al. A Comprehensive cis-eQTL Analysis Revealed Target Genes in Breast Cancer Susceptibility Loci Identified in Genome-wide Association Studies. Am J Hum Genet. 2018;102: 890–903. doi: 10.1016/j.ajhg.2018.03.016 29727689

12. Li Q, Stram A, Chen C, Kar S, Gayther S, Pharoah P, et al. Expression QTL-based analyses reveal candidate causal genes and loci across five tumor types. Hum Mol Genet. 2014;23: 5294–5302. doi: 10.1093/hmg/ddu228 24907074

13. Small KS, Hedman ÅK, Grundberg E, Nica AC, Thorleifsson G, Kong A, et al. Identification of an imprinted master trans regulator at the KLF14 locus related to multiple metabolic phenotypes. Nat Genet. 2011;43: 561–564. doi: 10.1038/ng.833 21572415

14. Aguet F, Ardlie KG, Cummings BB, Gelfand ET, Getz G, Hadley K, et al. Genetic effects on gene expression across human tissues. Nature. 2017;550: 204–213. doi: 10.1038/nature24277 29022597

15. Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, et al. Bayesian Test for Colocalisation between Pairs of Genetic Association Studies Using Summary Statistics. PLoS Genet. 2014;10: e1004383. doi: 10.1371/journal.pgen.1004383 24830394

16. R Core Team, R Foundation for Statistical Computing. R: A language and environment for statistical computing. Vienna, Austria; 2017. Available: http://www.r-project.org

17. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: A Tool Set for Whole-Genome Association and Population-Based Linkage Analyses. Am J Hum Genet. 2007;81: 559–575. doi: 10.1086/519795 17701901

18. Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26: 2190–2191. doi: 10.1093/bioinformatics/btq340 20616382

19. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRING v11: protein–protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. 2019;47: D607–D613. doi: 10.1093/nar/gky1131 30476243

20. Levine DM, Ek WE, Zhang R, Liu X, Onstad L, Sather C, et al. A genome-wide association study identifies new susceptibility loci for esophageal adenocarcinoma and Barrett’s esophagus. Nat Genet. 2013;45: 1487–1493. doi: 10.1038/ng.2796 24121790

21. Palles C, Chegwidden L, Li X, Findlay JM, Farnham G, Castro Giner F, et al. Polymorphisms Near TBX5 and GDF7 Are Associated With Increased Risk for Barrett’s Esophagus. Gastroenterology. 2015;148: 367–378. doi: 10.1053/j.gastro.2014.10.041 25447851

22. Su Z, Gay LJ, Strange A, Palles C, Band G, Whiteman DC, et al. Common variants at the MHC locus and at chromosome 16q24.1 predispose to Barrett’s esophagus. Nat Genet. 2012;44: 1131–1136. doi: 10.1038/ng.2408 22961001

23. Messina G, Atterrato MT, Prozzillo Y, Piacentini L, Losada A, Dimitri P. The human Cranio Facial Development Protein 1 (Cfdp1) gene encodes a protein required for the maintenance of higher-order chromatin organization. Sci Rep. 2017;7: 45022. doi: 10.1038/srep45022 28367969

24. Kekuda R, Prasad PD, Wu X, Wang H, Fei YJ, Leibach FH, et al. Cloning and functional characterization of a potential-sensitive, polyspecific organic cation transporter (OCT3) most abundantly expressed in placenta. J Biol Chem. 1998;273: 15971–15979. doi: 10.1074/jbc.273.26.15971 9632645

25. Wu X, Huang W, Ganapathy ME, Wang H, Kekuda R, Conway SJ, et al. Structure, function, and regional distribution of the organic cation transporter OCT3 in the kidney. Am J Physiol Renal Physiol. 2000;279: F449–58. doi: 10.1152/ajprenal.2000.279.3.F449 10966924

26. Fu L, Qin Y-R, Ming X-Y, Zuo X-B, Diao Y-W, Zhang L-Y, et al. RNA editing of SLC22A3 drives early tumor invasion and metastasis in familial esophageal cancer. Proc Natl Acad Sci U S A. 2017;114: E4631–E4640. doi: 10.1073/pnas.1703178114 28533408

27. Nishikura K. Oesophageal cancer: RNA editing of SLC22A3 mRNAs: causative relevance to familial ESCC? Nat Rev Gastroenterol Hepatol. 2017;14: 569–570. doi: 10.1038/nrgastro.2017.102 28743982

28. Goo Y-H, Son S-H, Kreienberg PB, Paul A. Novel lipid droplet-associated serine hydrolase regulates macrophage cholesterol mobilization. Arterioscler Thromb Vasc Biol. 2014;34: 386–96. doi: 10.1161/ATVBAHA.113.302448 24357060

29. Mohammed H, Hernando-Herraez I, Savino A, Scialdone A, Macaulay I, Mulas C, et al. Single-Cell Landscape of Transcriptional Heterogeneity and Cell Fate Decisions during Mouse Early Gastrulation. Cell Rep. 2017;20: 1215–1228. doi: 10.1016/j.celrep.2017.07.009 28768204

30. Umer HM, Cavalli M, Dabrowski MJ, Diamanti K, Kruczyk M, Pan G, et al. A Significant Regulatory Mutation Burden at a High-Affinity Position of the CTCF Motif in Gastrointestinal Cancers. Hum Mutat. 2016;37: 904–913. doi: 10.1002/humu.23014 27174533

31. Biemesderfer D, Pizzonia J, Abu-Alfa A, Exner M, Reilly R, Igarashi P, et al. NHE3: a Na+/H+ exchanger isoform of renal brush border. Am J Physiol. 1993;265: F736–F742. doi: 10.1152/ajprenal.1993.265.5.F736 8238556

32. Yang S-C, Chen C-L, Yi C-H, Liu T-T, Shieh K-R. Changes in Gene Expression Patterns of Circadian-Clock, Transient Receptor Potential Vanilloid-1 and Nerve Growth Factor in Inflamed Human Esophagus. Sci Rep. 2015;5: 13602. doi: 10.1038/srep13602 26337663

33. Zhu Z, Zhang F, Hu H, Bakshi A, Robinson MR, Powell JE, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016;48: 481–7. doi: 10.1038/ng.3538 27019110

34. Nica AC, Montgomery SB, Dimas AS, Stranger BE, Beazley C, Barroso I, et al. Candidate causal regulatory effects by integration of expression QTLs with complex trait genetic associations. PLoS Genet. 2010;6. doi: 10.1371/journal.pgen.1000895 20369022

35. Lusis AJ, Arnold AP, Schadt EE, Wang H, Ingram-Drake L, Wang S, et al. Tissue-specific expression and regulation of sexually dimorphic genes in mice. Genome Res. 2006;16: 995–1004. doi: 10.1101/gr.5217506 16825664

36. Wang DH. The Esophageal Squamous Epithelial Cell—Still a Reasonable Candidate for the Barrett’s Esophagus Cell of Origin? CMGH. Elsevier; 2017. pp. 157–160. doi: 10.1016/j.jcmgh.2017.01.015 28593187

37. Zhuang L, Fitzgerald RC. Cancer development: Origins in the oesophagus. Nature. 2017;550: 463–464. doi: 10.1038/nature24150 29019986


Článok vyšiel v časopise

PLOS One


2019 Číslo 12
Najčítanejšie tento týždeň
Najčítanejšie v tomto čísle
Kurzy

Zvýšte si kvalifikáciu online z pohodlia domova

Získaná hemofilie - Povědomí o nemoci a její diagnostika
nový kurz

Eozinofilní granulomatóza s polyangiitidou
Autori: doc. MUDr. Martina Doubková, Ph.D.

Všetky kurzy
Prihlásenie
Zabudnuté heslo

Zadajte e-mailovú adresu, s ktorou ste vytvárali účet. Budú Vám na ňu zasielané informácie k nastaveniu nového hesla.

Prihlásenie

Nemáte účet?  Registrujte sa

#ADS_BOTTOM_SCRIPTS#