Doxorubicin is a highly effective chemotherapy drug commonly used in approximately 60% of pediatric patients with metastatic solid tumors (sarcomas), leukemia, and lymphoma. Treatments using doxorubicin are complicated by its well-established cardiotoxic adverse drug reaction (ADR), which can lead to heart failure and affects approximately 16% of pediatric patients. At present, there is little potential for predicting which patients will experience doxorubicin-induced cardiotoxicity (DIC). If DIC could be genomically predicted, patients could be given an alternative treatment or additional therapy in the form of a genotype-specific cardioprotectant to prevent their cardiotoxicity. Recently we confirmed, for the first time that patient-specific human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CM) are efficient predictors of a patient's likelihood of developing DIC. This exciting finding proves that DIC has a genomic basis and our next question is: what are the specific genomic variants that control this predilection? >40 candidate gene association studies and 3 genome-wide association studies (GWAS) have been conducted, of varying rigor and depth, that have identified many single-nucleotide polymorphisms (SNPs) that are statistically correlated with DIC. There is a great need to validate these findings in a human model, a process not previously possible due the difficulty of working with primary human cardiomyocytes in vitro. To this end, we have validated two highly statistically significant SNPs in RARG and SLC28A3 using hiPSC-CMs (see Preliminary Data), proving their correlation and discovering the molecular mechanisms involved. The major disadvantage of GWAS is that it detects statistical correlation not causation and frequently identifies loci in non-coding sequence or that are synonymous, and identified SNPs commonly fail to be validated. We will therefore look to identifying gene expression regulatory variants as potential drivers of drug response by mapping expression quantitative trait loci (eQTL). Commonly, eQTLs are mapped from patient organ biopsies or lymphoblastoid cell lines which have proven successful in primary cells even in small populations (60 patients). Here, we intend to take this eQTL bioinformatics approach a stage further and compare the differential (i.e. with and without drug) changes in gene expression in response to doxorubicin in hiPSC-CMs generated from 100 patients who either did or did not experience cardiotoxicity. It is our hypothesis that this patient-specific differential eQTL (deQTL) mapping will be uniquely effective in discovering SNPs causatively associated with DIC. When we have identified the deQTL, we will then combine hiPSC-CM DIC phenotyping assays with CRISPR/Cas9-mediated gene editing to correct the SNPs of interest, validate these SNPs in a human model, and further probe their mechanisms of influence, and discover cardioprotectants. This will be a first of its kind study, and the methodology may be applied to many other drug ADRs.
|Effective start/end date||7/1/18 → 6/30/21|
- American Heart Association (18TPA34230105)