Genome Wide Association Analysis of Iron Overload in the Trans-Omics for Precision Medicine (TOPMed) Sickle Cell Disease Cohorts

Introduction: Transfusional iron (Fe) overload is a significant problem among patients with chronic, transfusion-dependent anemias. Iron overload is an important problem in pediatric sickle cell disease (SCD) patients on chronic transfusion regimens predominantly for primary and secondary prevention of stroke. Coexistent hereditary iron overload conditions contribute to the iron overload phenotype in SCD. For example, the Q248H mutation (rs11568350) in SLC40A1, which encodes ferroportin (FPN), is associated with a mild tendency to increase serum ferritin in the general population and with increased ferritin levels in SCD patients. Nevertheless, the molecular mechanisms underlying the progression to iron overload in SCD patients are poorly understood and more sensitive markers for outcome prediction that can be applied at early clinical stages are lacking. We hypothesize that genetic variation modifies the risk for iron overload in SCD patients and seek to validate previously identified mutation and identify novel genetic markers of iron overload among participants from TOPMed SCD cohorts by performing whole genome sequencing (WGS) association analyses.

Methods: The WGS was performed by several national sequencing centers sponsored by NHLBI’s TOPMed program at an average depth of 30× using DNA from SCD patient blood samples. Variant calling was performed jointly across TOPMed studies for all samples using the GotCloud pipeline by the TOPMed Informatics Research Center. The TOPMed data Coordinating Center performed quality control for sample identity. The data across the following studies were shared through the database of Genotype and Phenotype (dbGaP) exchange area: Howard PUSH SCD (N=370), OMG SCD (N=636), Walk PHaSST SCD (N=381) and REDS-III Brazil SCD (N=2620) with a total sample size of 4007. The study was approved by the appropriate institutional review boards (IRB) and informed consent was obtained from all participants. Genome Wide Association Analysis of iron overload was carried out using the University of Michigan ENCORE server. We performed single variant tests to test the association of log-transformed serum ferritin levels with single nucleotide variants (SNVs) while adjusting for sex, age, self-reported race, numbers of lifetime red blood transfusions and genetic substructure (PC’s 1-10). We used a significance threshold of p<5×10^-8 to report an association as genome-wide significant for common and rare genetic variants.

Results: We first included PUSH SCD, OMG SCD and Walk PHaSST SCD cohorts with 840 serum ferritin samples in the WGS association analyses, which revealed at the genome-wide level a new rare variant rs137929759 (chr7:49538810 (GRCh38.p12), MAF=0.0043532, p=2.25×10⁻⁸). A few variants such as rs80097634 in gene AL163195.3 and RNASE11 (Chr14:20579417. MAF=0.050373, p=3.1×10^-7) were close to the genome-wide significance level. We confirmed previously identified associations in SLC40A1 for ferritin (rs11568350, Chr2:189565370, MAF = 0.16853, p = 5.2×10⁻⁴). We also found several variants in AC105411.1, TJP1 and DCC that were close to genome-wide significance level. Further analysis will be carried out on the cloud-based platform provided by NHLBI BioData Catalyst using data from all the four cohorts to validate the previous analysis and expand to related phenotype such as transferrin, iron-overload status.

Discussion: In this study we identified common and rare variants that associate with serum ferritin concentration. The results from this pilot study point to novel gene variants that might contribute to iron overload in SCD patients and serve as new biomarkers. Future analysis is needed to determine whether the identified variants can also help with therapeutics and outcome prediction for early stages of SCD-associated iron overload. Our findings will be useful for the future treatment of SCD patients and design of novel SCD therapeutics.

ACKNOWLEDGMENTS: This work was supported by NIH Research Grants (1P50HL118006, and 1R01HL125005). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.