Serum hepcidin concentration is regulated by iron status, inflammation, erythropoiesis and numerous other factors, but underlying processes are incompletely understood. We studied the association of common and rare single nucleotide variants (SNVs) with serum hepcidin in one Italian study and two large Dutch population-based studies. We genotyped common SNVs with genome-wide association study (GWAS) arrays and subsequently performed imputation using the 1000 Genomes reference panel. Cohort-specific GWAS were performed for log-transformed serum hepcidin, adjusted for age and gender, and results were combined in a fixed-effects meta-analysis (total N 6,096). Six top SNVs (p<5x10-6) were genotyped in 3,821 additional samples, but associations were not replicated. Furthermore, we meta-analyzed cohort-specific exome array association results of rare SNVs with serum hepcidin that were available for two of the three cohorts (total N 3,226), but no exome-wide significant signal (p<1.4x10-6) was identified. Gene-based meta-analyses revealed 19 genes that showed significant association with hepcidin. Our results suggest the absence of common SNVs and rare exonic SNVs explaining a large proportion of phenotypic variation in serum hepcidin. We recommend extension of our study once additional substantial cohorts with hepcidin measurements, GWAS and/or exome array data become available in order to increase power to identify variants that explain a smaller proportion of hepcidin variation. In addition, we encourage follow-up of the potentially interesting genes that resulted from the gene-based analysis of low-frequency and rare variants.
Citation: Galesloot TE, Verweij N, Traglia M, Barbieri C, van Dijk F, Geurts-Moespot AJ, et al. (2016) Meta-GWAS and Meta-Analysis of Exome Array Studies Do Not Reveal Genetic Determinants of Serum Hepcidin. PLoS ONE 11(11): e0166628. https://doi.org/10.1371/journal.pone.0166628
Editor: Kostas Pantopoulos, Lady Davis Institute for Medical Research, CANADA
Received: May 17, 2016; Accepted: November 1, 2016; Published: November 15, 2016
Copyright: © 2016 Galesloot et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data underlying our findings are available upon request without any costs. The data are managed by the Nijmegen Biomedical Study (NBS) project team; see www.nijmegenbiomedischestudie.nl for an overview of the data available in this study. We cannot make the data underlying the findings of the current manuscript freely available, as we signed a Data Transfer Agreement in order to receive the data in which it is also stated that we cannot distribute the data to other parties. Current practical coordinator of the NBS is the first author of this manuscript, Dr. TE Galesloot. Readers can contact her to request the data (Tessel.Galesloot@radboudumc.nl). All interested parties will be able to obtain the dataset in the same manner as the authors did.
Funding: This work was supported by the Stichting Nationale Computerfaciliteiten (National Computing Facilities Foundation, NCF) for the use of supercomputer facilities, with financial support from the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (Netherlands Organization for Scientific Research, NWO); PREVEND genetics is supported by the Dutch Kidney Foundation (grant E033), the National Institutes of Health (grant LM010098), The Netherlands Organization for Scientific Research (NWO-Groot 175.010.2007.006, NWO VENI grant 916.761.70, ZonMW 90.700.441), and the Dutch Inter University Cardiology Institute Netherlands; N. Verweij is supported by the Netherlands Heart Foundation (grant NHS2010B280); funds from Compagnia di San Paolo, Torino, Italy; Fondazione Cariplo, Italy and Ministry of Health, Ricerca Finalizzata 2008 and CCM 2010, PRIN 2009 and Telethon, Italy to DT. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. DWS is an employee of Radboudumc and not of the hepcidinanalysis.com initiative. The hepcidinanalysis.com initiative is thus not a funder and did not have any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of the authors are articulated in the ‘author contributions’ section.
Competing interests: DWS is an employee of Radboudumc, that offers hepcidin measurements to the scientific, medical and pharmaceutical communities at a fee-for-service bases via the www.hepcidinanalysis.com initiative. The remaining authors declare that they have no competing interests. The competing interests do not alter our adherence to PLOS ONE policies on sharing data and materials.
Iron is an essential trace element for fundamental metabolic processes in humans [1, 2]. Iron deficiency limits hemoglobin synthesis and leads to anemia, whereas an excess of free iron is toxic because it catalyzes the production of free radicals resulting in tissue damage [1, 2]. In addition, iron imbalances have been associated with other diseases, e.g. diabetes mellitus [3, 4], inflammation  and diseases of aging . Hence, the iron balance in the human body is tightly controlled, with the peptide hormone hepcidin as key regulator of systemic iron homeostasis . Hepcidin controls the absorption, storage and tissue distribution of iron by binding to the cellular iron exporter ferroportin and inducing its internalization and degradation . In this way, hepcidin regulates the uptake of dietary iron from the intestine and the release of iron from macrophages involved in recycling of iron from senescent erythrocytes .
In the last few years, several genome-wide association studies (GWAS) have revealed genetic variants associated with iron status in the general population, including common variants in the hereditary hemochromatosis gene (HFE), transferrin gene (TF), transferrin receptor and transferrin receptor 2 gene (TFRC, TFR2), solute carrier family 40 member 1 gene (SLC40A1), and transmembrane serine protease 6 gene (TMPRSS6) [9–14]. On the contrary, little is known about genetic determinants of hepcidin. Mutations in hepcidin antimicrobial peptide (HAMP), the hepcidin encoding gene, lead to strongly decreased hepcidin levels and a severe juvenile form of the iron storage disorder hereditary hemochromatosis (HH), but HAMP mutations are very rare . In addition, mutations in HFE, TFR2 and TMPRSS6 have been related to hepcidin expression [15–20]. Furthermore, a single GWAS for serum hepcidin has been published . This study among 1,657 family members from the Val Borbera genetic isolate was however underpowered to identify genome-wide significant associations . Here, we aimed to identify genetic determinants of serum hepcidin in a larger set of individuals from three large cohorts in order to unravel potential new pathways involved in hepcidin regulation. We also studied the ratios of hepcidin to ferritin (hepcidin/ferritin) and hepcidin to transferrin saturation (TS) (hepcidin/TS) given the known dependence of hepcidin on stored iron and circulating iron, respectively [1, 2].
Materials and Methods
We included three cohorts in our study: the Nijmegen Biomedical Study (NBS) (Nijmegen, The Netherlands) , Prevention of REnal and Vascular ENd-stage Disease (PREVEND) (Groningen, The Netherlands)  and Val Borbera (VB) (Milan, Italy) [24, 25] (S1 Table). Blood samples for DNA isolation and biochemical measurements were obtained fasting in the morning for PREVEND and VB, whereas blood samples were not fasting and sampled throughout the day between 8 AM and 9 PM for NBS. All three studies were approved by appropriate ethical committees (PREVEND: local medical ethics committee; NBS: Radboud university medical center Institutional Review Board; VB: institutional review boards of San Raffaele Hospital in Milan and by the Regione Piemonte ethical committee), and all participants gave informed consent.
Serum hepcidin concentration was measured with a competitive enzyme-linked immunosorbent assay in NBS and PREVEND samples as described before [26, 27]. In VB samples, serum hepcidin was measured with a validated mass spectrometry based method as described before . Serum ferritin, iron, transferrin, TS and C-reactive protein (CRP) were measured according to standard methods. See S2 Table for details. Phenotype information (median and 5th-95th percentiles) is presented in S3 Table.
All three cohorts were genotyped with a GWAS-chip: PREVEND with the Illumina Cyto SNP12 v2, NBS with the Illumina HumanHap370CNV-Duo BeadChip, and VB with the Illumina HumanHap370CNV-Quad BeadChip v3. Standard quality checks were performed (filters for sample yield, SNV yield, MAF and HWE) and data were imputed to increase SNV density and harmonize SNV data over the cohorts using 1000 Genomes phase 1 version 3 as reference panel (S4 Table). Quality control also included evaluation of population stratification. For PREVEND, principal component analysis was used and samples with a Z-score>3 for the first five principal components were excluded. For NBS, Structure analysis was used and samples with less than 89% Caucasian ancestry were excluded. For VB, principal components were used in the analysis to adjust for potential population stratification.
Genome-wide association analysis
GWAS were performed in each cohort separately according to a set protocol. Analyses were performed for all individuals (PREVEND: N = 2,902; NBS: N = 1,819; VB: N = 1,480), and also for a subset (PREVEND: N = 2,695; NBS: N = 1,495; VB: N = 1,206) from which individuals with ferritin <30 ng/mL and CRP ≥10 mg/L were excluded as to remove individuals with iron deficiency and clinical inflammation , respectively, since both of these acquired conditions are associated with altered iron stores and altered iron transport [1,2]. Hepcidin and the ratios hepcidin/ferritin and hepcidin/TS were log-transformed and thereafter adjusted for age and squared age, independent determinants of serum hepcidin [21, 26], separately for males and females. For NBS, blood sampling was performed throughout the day, and therefore time of blood sampling was used as an additional covariate to account for the circadian rhythm of hepcidin  (three categories: before 12 PM, between 12 and 5 PM and after 5 PM in line with previously reported hepcidin concentration patterns throughout the day [30, 31]). Sex‐specific residuals were calculated and merged into one variable. Outliers, defined as values that differed more than four times the SD from the mean, were excluded (Nmax = 12). For PREVEND and NBS, the association between the single nucleotide variants and the trait was tested by linear regression using genotype probabilities and an additive model on the standardized residuals (Z score). For VB, a linear mixed model was used with a kinship (relatedness) matrix to account for the relatedness in this sample , also using genotype probabilities and an additive model on the standardized residuals (Z score).
The GWAS results from the three cohorts were combined in a fixed-effects meta-analysis using the software package METAL . The standard-error based approach was used, which weighs effect size estimates using the inverse of the corresponding standard errors. Variants with a minor allele frequency <1% and a poor imputation quality (SNPtest info value or MACH RSQR <0.4) were excluded prior to the meta-analysis. To adjust for potential residual population stratification, we applied genomic control correction to the individual cohorts (genomic inflation factors [lambdas] for GWAS results ranged from 0.989 to 1.015 in the three cohorts, indicating a negligible amount of population stratification) . A heterogeneity analysis was performed to test whether observed effect sizes were homogeneous across cohorts. Resulting betas express the change in log-transformed hepcidin (or the ratios) that can be attributed to each copy of the effect allele (additive model).
Our financial budget allowed us to genotype six SNVs with single SNP assays in all additional independent samples that were available for PREVEND (N = 2,876) and NBS (N = 1018). For the VB cohort, no additional samples were available. Single-SNP genotyping in PREVEND samples was performed by KBiosciences (KBiosciences, Herts, UK) utilizing the SNPline system. Single-SNP genotyping in NBS samples was carried out by deCODE Genetics using the Centaurus (Nanogen) platform . The quality of each Centaurus SNP assay was evaluated by genotyping HapMap CEU samples with each assay and comparing the results with the HapMap data. All assays were reliable, as the mismatch rates were all <0.5%. Association analyses were performed using the same strategy as for the discovery meta-GWAS: cohort-specific association analyses and subsequent combination of summary statistics in a fixed-effects meta-analysis. We also meta-analyzed these replication results together with the discovery meta-analysis results.
We performed gene-based analysis on SNV association P-values from the meta-analysis of discovery samples using VEGAS2 [36, 37]. Statistical significance of gene-based analysis results was based on Bonferroni correction of testing ~21,000 genes (P<2.4x10-6).
Data-driven Expression Prioritized Integration for Complex Traits (DEPICT, “v1 beta version rel194 for 1KG imputed”) was applied to the meta-GWAS discovery results to highlight enriched pathways and identify tissues/cell types in which genes from associated loci are highly expressed . Meta-GWAS association results with p<1x10-5 were pruned using PLINK v1.09 to obtain independent SNVs (‘—clump-kb 500—clump-p1 1e-05—clump-r2 0.1’) using the CEU (Utah Residents [CEPH] with Northern and Western Ancestry), GBR (British in England and Scotland) and TSI (Toscani in Italy) 1000 Genomes populations as reference to obtain the correlation structure of the SNVs. These pruned data were used as input for the DEPICT analysis using default settings. Details of the DEPICT analysis can be found in S1 File.
Exome array association analysis
Exome array data measured with the Illumina HumanExome BeadChip were available for both NBS and VB. Genotype data were called with the default genotype caller in Genome Studio and uncalled genotypes were recalled using the zCall algorithm, which is specifically designed for calling rare variants . Filters for sample yield, SNV yield, and HWE were applied, resulting in a total N of1,753 for NBS and 1,473 for VB after QC (S5 Table). Phenotype information for the exome array samples is given in S6 Table. Analyses were performed using the phenotypes as described under ‘Genome-wide association analysis’. Single variant analyses were performed to obtain cohort specific score statistics and their covariance matrix (RvTests software version 20150629 [http://zhanxw.github.io/rvtests/]). For VB, a kinship matrix was included in the analyses to account for relatedness. Cohort specific results were meta-analyzed in RareMETAL software version 4.13.8 (http://genome.sph.umich.edu/wiki/RAREMETAL_Documentation) on single variant level and in a gene-level test [sequence kernel association test (SKAT)] ; only non-synonymous, splice site and stop variants with a MAF ≤5% were included in the gene-level test.
Single variant results were filtered for a pooled minor allele frequency between 0.001 and 0.05 to prevent repeated analysis of the common variants and at the same time reduce the number of statistical tests. The lower bound frequency of 0.001 corresponds to a minor allele count of 6 for the whole cohort with exome array data (N up to 3,226), and 5 for the subset (N up to 2,623).
SKAT meta-analysis in RareMETAL presents an exact and approximate p-value in the output. We used the exact test called “Davies”, which computes the p-value by inverting the characteristic function of the mixture chisq, which is often used as the default in statistical analysis packages.
The statistical significance of single variant and gene-based exome array meta-analysis results was based on Bonferroni correction of testing ~37,000 variants (p< 1.4 x10-6) and ~14,000 genes (p< 3.6 x10-6), respectively.
Meta-analysis of GWAS (total N 6,096) revealed two loci that were genome-wide significantly associated (p<5x10-8) with serum hepcidin (Table 1): one on chromosome 10 in all individuals (rs118031191, nearest gene FOXI2) and one on chromosome 2 in the subset (four SNVs in EML6 with lead SNV rs354202). S1 and S2 Figs show Manhattan and QQ plots for the meta-GWAS for hepcidin in all individuals and in the subset, respectively. S3 and S4 Figs show the regional association plots for rs118031191 and rs354202, respectively. We also performed conditional analysis for the chromosome 2 locus in NBS data by adjusting for lead SNV rs354202 to investigate whether the additional signals at this locus identified in the discovery analysis were independent from rs354202. Associations disappeared upon adjustment for this SNV, revealing that all other signals at the chromosome 2 locus were driven by rs354202 (S5 and S6 Figs). No novel significant loci were found for the ratios hepcidin/ferritin and hepcidin/TS (Manhattan and QQ plots in S7–S10 Figs), but previously identified associations were confirmed: for the hepcidin/ferritin ratio we observed genome-wide significant associations with variation in HFE and TMPRSS6 (S7 Table), and for hepcidin/TS with variation in TMPRSS6, but not with HFE (S8 Table).
Genome-wide significant SNVs (p-value<5x10-8) are indicated in bold.
Gene-based analysis (S9 Table for results with p<1x10-2) did only show significant association for hepcidin/ferritin with HFE (all p = 7.2x10-7; subset p = 8.2x10-7). No significant enriched gene sets or tissues were identified for all traits (S10 and S11 Tables).
Six SNVs were brought forward to replication: rs354202, rs118031191, rs56281245 and rs12289793 were selected based on p<1x10-6 for association with hepcidin, and rs1835473 and rs12441903 were additionally selected based on hepcidin association p-values close to 1x10-6, location inside genes, and MAF >10%. All SNVs selected for replication had a non-significant p-value for the heterogeneity test (p>0.05). Replication analysis (N 3,821) revealed no significant associations at p = 0.05 (Table 2).
Meta-analysis on single variant level of exome array variants (total N 3,226) revealed no significantly associated SNVs for hepcidin (S12 Table presents results with p<1x10-3). The most significant signal was located in PCDHB1 (exm485355, p = 4.4x10-5) for all individuals and in PTPN13 for the subset (exm411306, p = 2.1x10-5). Also for hepcidin/TS no significant signals were observed, but for hepcidin/ferritin we identified two signals that reached exome-wide significance: exm470499 in WDR36 (all; p = 9.9x10-7) and exm162358 in MTR (subset; p = 2.7x10-8); variants close to and in these genes have previously been associated with allergy  and homocysteine levels , respectively. Gene-based meta-analyses (S13 Table) revealed eight genes that showed significant association with hepcidin in all individuals, and 11 in the subset, but none of these genes overlapped. For hepcidin/ferritin, 10 and nine genes showed significant association in all individuals and in the subset, respectively, with one overlapping gene, namely PAPSS1. This gene was described as a candidate gene for telomere length based on a GWAS . Previously, elevated iron phenotype was associated with shortened telomeres , indicating a potential link between this gene and systemic iron homeostasis. For hepcidin/TS, nine and 16 genes were significant in all individuals and in the subset, respectively, but none overlapped.
This is the first meta-analysis of GWAS and exome array results for serum hepcidin. The fact that our meta-analysis revealed no SNVs that were significantly associated with serum hepcidin suggests that there are no common, low-frequency or rare variants that explain a large proportion of phenotypic variation in serum hepcidin. Indeed, with our meta-analysis of common variants (N 6,096, effective sample size N 5,331) we had 80.0% and 97.8% chance of detecting (at alpha 5x10-8) a SNV that explains 0.71% and 1.00% of hepcidin variance, respectively. For our meta-analysis of exome array results (N 3,226, effective N 2,467) we had 80.0% and 99.2% chance of detecting (at alpha 1.4x10-6) a SNV explaining 1.22% and 2.00% of hepcidin variance, respectively. For comparison, the well-known iron-related SNVs rs1800562 in HFE and rs855791 in TMPRSS6 explain ~1% of serum iron variation. In addition, (narrow-sense) heritability of hepcidin was previously estimated to be 9.8% but non-significant (p>0.05) in the VB population and genome-wide SNP explained variance adjusted for age and gender was estimated at ~37% (SE~20%) in the NBS (data not shown). Overall, this suggests that a large part of hepcidin variability is caused by variation in environmental factors, e.g. inflammation, body-mass index and body iron stores (dietary intake and blood losses).
Serum hepcidin was not associated with common variants in HFE and TMPRSS6, as we have previously shown in independent studies both in the VB and NBS population [21, 47]. Now, we have also shown that low-frequency or rare variants in these genes do not seem to contribute, neither at single variant level nor at gene level. We confirmed previously reported associations for rs1800562 in HFE and rs855791 in TMPRSS6 with the ratio hepcidin/ferritin for the VB and NBS population [21, 47]. As expected, we further substantiated these associations here and found an even stronger signal. The association signal of the ratio hepcidin/TS with common variants in HFE and TMPRSS6 was less strong. Of note, the association of rs1800562 in HFE with the ratio hepcidin/TS, previously found in the NBS , disappeared upon meta-analysis of results of NBS, PREVEND and VB in all individuals (p = 0.13), but still showed a weak signal in the subset (p = 3.7Ex10-4). Also rs855791 in TMPRSS6 showed a stronger signal for association with the ratio hepcidin/TS in the subset compared to analysis based on all individuals. The stronger signal of rs1800562 and rs855791 with the ratio hepcidin/ferritin compared to the ratio hepcidin/TS indicates that these SNVs have a larger influence on hepcidin response to body iron stores than on hepcidin response to circulating iron.
There are several reasons that can cause failure of identifying a true association signal [48, 49], e.g. issues of heterogeneity between studies, such as variability in outcome measurement or statistical analysis, or bias due to population stratification. We took those issues into account by using a set analysis protocol for cohort-specific analysis, testing for heterogeneity for the top hits, and adjusting for population stratification in each cohort next to application of genomic control adjustment. Two different hepcidin assays were applied in this study, but correlation between these assays was high as measured in international round robins (sample send-out studies) for hepcidin assay harmonization (DWS and DG, unpublished findings). In addition, we prevented bias in SNV-hepcidin associations due to differences in hepcidin assays by first performing cohort-specific GWAS and thereafter combining results in a meta-analysis. However, it is important to realize that there are additional reasons that may have masked true association results in our study, e.g. residual confounding due to population stratification, presence of comorbidities that associate with serum hepcidin levels, incomplete adjustment for relatedness in the VB cohort, genetic or environmental interactions [48, 49], or high biological variation in serum hepcidin levels [29, 30, 31, 50].
In conclusion, our results indicate that there are no common SNVs that explain more than 1% and no low-frequency and rare SNVs that explain more than 2% of phenotypic hepcidin variation. We recommend extension of our study once additional substantial cohorts with hepcidin measurements, GWAS and/or exome array data become available in order to increase power to identify variants that explain a smaller proportion of hepcidin variation. In addition, we encourage follow-up of the potentially interesting genes that resulted from the gene-based analysis of low-frequency and rare variants with candidate gene, fine mapping and functional studies to increase the level of evidence for association and obtain insight into the underlying mechanism of action.
S1 Table. Cohort information and acknowledgments.
S3 Table. Phenotype information [median (P5-P95)] of the samples included in the meta-GWAS.
S4 Table. Information about genotyping, imputation and quality control of the cohort-specific GWAS.
S5 Table. Information about genotyping and quality control of the cohort-specific exome array analysis.
S6 Table. Phenotype information of the samples included in the exome array analyses [median (5th percentile-95th percentile)].
S7 Table. Top hits (p-value < 1E-06) for hepcidin/ferritin in all individuals and in the subset.
S8 Table. Top hits (p-value < 1E-06) for hepcidin/TS in all individuals and in the subset.
S11 Table. DEPICT results gene set enrichment.
S12 Table. Exome array single variants results.
S13 Table. Exome array gene-based (SKAT) results.
S1 Fig. Manhattan plot and QQ plot for the meta-analysis results for hepcidin in all individuals.
S2 Fig. Manhattan plot and QQ plot for the meta-analysis results for hepcidin in the subset.
S3 Fig. Regional association plot for rs118031191 with serum hepcidin in all individuals.
S4 Fig. Regional association plot for rs354202 with serum hepcidin in the subset.
S5 Fig. Regional association plot for the chromosome 2 locus with serum hepcidin conditioned on rs354202 in all individuals (NBS data only).
S6 Fig. Regional association plot for the chromosome 2 locus with serum hepcidin conditioned on rs354202 in the subset (NBS data only).
S7 Fig. Manhattan plot and QQ plot for the meta-analysis results for the ratio hepcidin/ferritin in all individuals.
S8 Fig. Manhattan plot and QQ plot for the meta-analysis results for the ratio hepcidin/ferritin in the subset.
S9 Fig. Manhattan plot and QQ plot for the meta-analysis results for the ratio hepcidin/TS in all individuals.
S10 Fig. Manhattan plot and QQ plot for the meta-analysis results for the ratio hepcidin/TS in the subset.
- Conceptualization: TEG NV MT CB PM CC DT SHV PH DWS.
- Formal analysis: TEG NV MT CB.
- Funding acquisition: PM CC DT SHV PH DWS.
- Investigation: TEG NV MT CB PM CC DT SHV PH DWS.
- Methodology: TEG NV MT CB PM CC DT SHV PH DWS.
- Project administration: TEG SHV DWS.
- Resources: FD AJG DG LALMK FCGJS MAS PM CC DT PH DWS.
- Software: MAS FD.
- Supervision: PM CC DT SHV PH DWS.
- Validation: TEG NV MT AJG.
- Visualization: TEG NV.
- Writing – original draft: TEG.
- Writing – review & editing: NV MT CB FD AJG DG LALMK FCGJS MAS PM CC DT SHV PH DWS.
- 1. Hentze MW, Muckenthaler MU, Galy B, Camaschella C. Two to tango: regulation of Mammalian iron metabolism. Cell. 2010;142(1):24–38. pmid:20603012
- 2. Ganz T. Systemic iron homeostasis. Physiological reviews. 2013;93(4):1721–41. pmid:24137020
- 3. Montonen J, Boeing H, Steffen A, Lehmann R, Fritsche A, Joost HG, et al. Body iron stores and risk of type 2 diabetes: results from the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study. Diabetologia. 2012;55(10):2613–21. pmid:22752055
- 4. Jiang R, Manson JE, Meigs JB, Ma J, Rifai N, Hu FB. Body iron stores in relation to risk of type 2 diabetes in apparently healthy women. Journal of the American Medical Association. 2004;291(6):711–7. pmid:14871914
- 5. Cherayil BJ. Iron and immunity: immunological consequences of iron deficiency and overload. Archivum immunologiae et therapiae experimentalis. 2010;58(6):407–15. pmid:20878249
- 6. Altamura S, Muckenthaler MU. Iron toxicity in diseases of aging: Alzheimer's disease, Parkinson's disease and atherosclerosis. Journal of Alzheimer's disease: JAD. 2009;16(4):879–95. pmid:19387120
- 7. Girelli D, Nemeth E, Swinkels DW. Hepcidin in the diagnosis of iron disorders. Blood. 2016;9;127(23):2809–13. pmid:27044621
- 8. Nemeth E, Tuttle MS, Powelson J, Vaughn MB, Donovan A, Ward DM, et al. Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science. 2004;306(5704):2090–3. pmid:15514116
- 9. Benyamin B, Ferreira MA, Willemsen G, Gordon S, Middelberg RP, McEvoy BP, et al. Common variants in TMPRSS6 are associated with iron status and erythrocyte volume. Nature genetics. 2009;41(11):1173–5. pmid:19820699
- 10. Benyamin B, McRae AF, Zhu G, Gordon S, Henders AK, Palotie A, et al. Variants in TF and HFE explain approximately 40% of genetic variation in serum-transferrin levels. American journal of human genetics. 2009;84(1):60–5. pmid:19084217
- 11. Tanaka T, Roy CN, Yao W, Matteini A, Semba RD, Arking D, et al. A genome-wide association analysis of serum iron concentrations. Blood. 2010;115(1):94–6. pmid:19880490
- 12. Pichler I, Minelli C, Sanna S, Tanaka T, Schwienbacher C, Naitza S, et al. Identification of a common variant in the TFR2 gene implicated in the physiological regulation of serum iron levels. Human molecular genetics. 2011;20(6):1232–40. pmid:21208937
- 13. Benyamin B, Esko T, Ried JS, Radhakrishnan A, Vermeulen SH, Traglia M, et al. Novel loci affecting iron homeostasis and their effects in individuals at risk for hemochromatosis. Nature communications. 2014;5:4926. pmid:25352340
- 14. Middelberg RP, Ferreira MA, Henders AK, Heath AC, Madden PA, Montgomery GW, et al. Genetic variants in LPL, OASL and TOMM40/APOE-C1-C2-C4 genes are associated with multiple cardiovascular-related traits. BMC medical genetics. 2011;12:123. pmid:21943158
- 15. Swinkels DW, Janssen MC, Bergmans J, Marx JJ. Hereditary hemochromatosis: genetic complexity and new diagnostic approaches. Clinical chemistry. 2006;52(6):950–68. pmid:16627556
- 16. Nemeth E, Roetto A, Garozzo G, Ganz T, Camaschella C. Hepcidin is decreased in TFR2 hemochromatosis. Blood. 2005;105(4):1803–6. pmid:15486069
- 17. van Dijk BA, Laarakkers CM, Klaver SM, Jacobs EM, van Tits LJ, Janssen MC, et al. Serum hepcidin levels are innately low in HFE-related haemochromatosis but differ between C282Y-homozygotes with elevated and normal ferritin levels. British journal of haematology. 2008;142(6):979–85. pmid:18557745
- 18. Bridle KR, Frazer DM, Wilkins SJ, Dixon JL, Purdie DM, Crawford DH, et al. Disrupted hepcidin regulation in HFE-associated haemochromatosis and the liver as a regulator of body iron homoeostasis. Lancet. 2003;361(9358):669–73. pmid:12606179
- 19. Finberg KE, Heeney MM, Campagna DR, Aydinok Y, Pearson HA, Hartman KR, et al. Mutations in TMPRSS6 cause iron-refractory iron deficiency anemia (IRIDA). Nature genetics. 2008;40(5):569–71. pmid:18408718
- 20. Ahmad KA, Ahmann JR, Migas MC, Waheed A, Britton RS, Bacon BR, et al. Decreased liver hepcidin expression in the Hfe knockout mouse. Blood cells, molecules & diseases. 2002;29(3):361–6.
- 21. Traglia M, Girelli D, Biino G, Campostrini N, Corbella M, Sala C, et al. Association of HFE and TMPRSS6 genetic variants with iron and erythrocyte parameters is only in part dependent on serum hepcidin concentrations. Journal of medical genetics. 2011;48(9):629–34. pmid:21785125
- 22. Hoogendoorn EH, Hermus AR, de Vegt F, Ross HA, Verbeek AL, Kiemeney LA, et al. Thyroid function and prevalence of anti-thyroperoxidase antibodies in a population with borderline sufficient iodine intake: influences of age and sex. Clinical chemistry. 2006; 52(1):104–11. pmid:16254196
- 23. Hillege HL, Fidler V, Diercks GF, van Gilst WH, de Zeeuw D, van Veldhuisen DJ, et al. Urinary albumin excretion predicts cardiovascular and noncardiovascular mortality in general population. Prevention of Renal and Vascular End Stage Disease (PREVEND) Study Group. Circulation. 2002;106(14):1777–82. pmid:12356629
- 24. Traglia M, Sala C, Masciullo C, Cverhova V, Lori F, Pistis G, et al. Heritability and demographic analyses in the large isolated population of Val Borbera suggest advantages in mapping complex traits genes. PLoS ONE. 2009;4(10):e7554. pmid:19847309
- 25. Colonna V, Pistis G, Bomba L, Mona S, Matullo G, Boano R, et al. Small effective population size and genetic homogeneity in the Val Borbera isolate. European journal of human genetics. 2013;21(1):89–94. pmid:22713810
- 26. Galesloot TE, Vermeulen SH, Geurts-Moespot AJ, Klaver SM, Kroot JJ, van Tienoven D, et al. Serum hepcidin: reference ranges and biochemical correlates in the general population. Blood. 2011;117(25):e218–25. pmid:21527524
- 27. Kroot JJ, Laarakkers CM, Geurts-Moespot AJ, Grebenchtchikov N, Pickkers P, van Ede AE, et al. Immunochemical and mass-spectrometry-based serum hepcidin assays for iron metabolism disorders. Clinical chemistry. 2010;56(10):1570–9. pmid:20739637
- 28. Nai A, Pagani A, Silvestri L, Campostrini N, Corbella M, Girelli D, et al. TMPRSS6 rs855791 modulates hepcidin transcription in vitro and serum hepcidin levels in normal individuals. Blood. 2011;118(16):4459–62. pmid:21873547
- 29. Schaap CC1, Hendriks JC, Kortman GA, Klaver SM, Kroot JJ, Laarakkers CM, et al. Diurnal rhythm rather than dietary iron mediates daily hepcidin variations. Clinical chemistry. 2013;59(3):527–35. pmid:23232066
- 30. Ganz T, Olbina G, Girelli D, Nemeth E, Westerman M. Immunoassay for human serum hepcidin. Blood. 2008;112(10):4292–7. pmid:18689548
- 31. Kroot JJ, Hendriks JC, Laarakkers CM, Klaver SM, Kemna EH, Tjalsma H, et al. (Pre)analytical imprecision, between-subject variability, and daily variations in serum and urine hepcidin: implications for clinical studies. Analytical biochemistry. 2009;389(2):124–9. pmid:19341701
- 32. Zhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nature genetics. 2012;44(7):821–4. pmid:22706312
- 33. Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26(17):2190–1. pmid:20616382
- 34. Devlin B, Roeder K, Wasserman L. Genomic control, a new approach to genetic-based association studies. Theoretical population biology. 2001;60(3):155–66. pmid:11855950
- 35. Kutyavin IV, Milesi D, Belousov Y, Podyminogin M, Vorobiev A, Gorn V, et al. A novel endonuclease IV post-PCR genotyping system. Nucleic acids research. 2006;34(19):e128. pmid:17012270
- 36. Liu JZ, McRae AF, Nyholt DR, Medland SE, Wray NR, Brown KM, et al. A versatile gene-based test for genome-wide association studies. American journal of human genetics. 2010;87(1):139–45. pmid:20598278
- 37. Mishra A, Macgregor S. VEGAS2: Software for More Flexible Gene-Based Testing. Twin research and human genetics: the official journal of the International Society for Twin Studies. 2015;18(1):86–91.
- 38. Pers TH, Karjalainen JM, Chan Y, Westra HJ, Wood AR, Yang J, et al. Biological interpretation of genome-wide association studies using predicted gene functions. Nature communications. 2015;6:5890. pmid:25597830
- 39. Goldstein JI1, Crenshaw A, Carey J, Grant GB, Maguire J, Fromer M, et al. zCall: a rare variant caller for array-based genotyping: genetics and population analysis. Bioinformatics. 2012;28(19):2543–5. pmid:22843986
- 40. Liu DJ, Peloso GM, Zhan X, Holmen OL, Zawistowski M, Feng S, et al. Meta-analysis of gene-level tests for rare variant association. Nature genetics. 2014;46(2):200–4. pmid:24336170
- 41. Feng S, Wang S, Chen CC, Lan L. GWAPower: a statistical power calculation software for genome-wide association studies with quantitative traits. BMC genetics. 2011;12:12. pmid:21255436
- 42. Falchi M, Forabosco P, Mocci E, Borlino CC, Picciau A, Virdis E, et al. A genomewide search using an original pairwise sampling approach for large genealogies identifies a new locus for total and low-density lipoprotein cholesterol in two genetically differentiated isolates of Sardinia. American journal of human genetics. 2004;75: 1015–31. pmid:15478097
- 43. Hinds DA, McMahon G, Kiefer AK, Do CB, Eriksson N, Evans DM, et al. A genome-wide association meta-analysis of self-reported allergy identifies shared and allergy-specific susceptibility loci. Nature genetics. 2013;45(8):907–11. pmid:23817569
- 44. van Meurs JB, Pare G, Schwartz SM, Hazra A, Tanaka T, Vermeulen SH, et al. Common genetic loci influencing plasma homocysteine concentrations and their effect on risk of coronary artery disease. The American journal of clinical nutrition. 2013;98(3):668–76. pmid:23824729
- 45. Lee JH, Cheng R, Honig LS, Feitosa M, Kammerer CM, Kang MS, et al. Genome wide association and linkage analyses identified three loci-4q25, 17q23.2, and 10q11.21-associated with variation in leukocyte telomere length: the Long Life Family Study. Frontiers in genetics. 2013;4:310. pmid:24478790
- 46. Mainous AG 3rd, Wright RU, Hulihan MM, Twal WO, McLaren CE, Diaz VA, et al. Telomere length and elevated iron: the influence of phenotype and HFE genotype. American journal of hematology. 2013;88(6):492–6. pmid:23512844
- 47. Galesloot TE, Geurts-Moespot AJ, den Heijer M, Sweep FC, Fleming RE, Kiemeney LA, et al. Associations of common variants in HFE and TMPRSS6 with iron parameters are independent of serum hepcidin in a general population: a replication study. Journal of medical genetics. 2013;50(9):593–8. pmid:23794717
- 48. Greene CS, Penrod NM, Williams SM, Moore JH. Failure to replicate a genetic association may provide important clues about genetic architecture. PLoS ONE. 2009;4(6):e5639. pmid:19503614
- 49. Kraft P, Zeggini E, Ioannidis JP. Replication in genome-wide association studies. Statistical science. 2009;24(4):561–73. pmid:20454541
- 50. Murphy AT1, Witcher DR, Luan P, Wroblewski VJ. Quantitation of hepcidin from human and mouse serum using liquid chromatography tandem mass spectrometry. Blood. 2007;110(3):1048–54. pmid:17435114