Anthracycline exposure alters the response to future cardiac stress response driven by FOXK1

Abstract Background Anthracyclines are the most used and efficient chemotherapy drugs for the treatment of several cancers. Their use is limited by their cardiotoxicity which presents itself as dilatative cardiomyopathy often several years after exposure to anthracyclines and the mechanisms underlying this delayed response remain unclear. Purpose We hypothesized that exposure to anthracyclines increases cardiac susceptible to future adverse stress response and could be the reason behind the delayed response. The purpose of this study was to test if an exposure to anthracyclines changes the response of the heart to future cardiac stress events. Methods To test our hypothesis, we developed a ‘two-hit’ model where neonatal rat cardiomyocytes (NRVM’s) were treated with low doses of doxorubicin (Doxo, 100nM or 300nM) for 1 h followed by beta-adrenergic stimulation using Isoproterenol (Iso) 5 days later. The two-hit model was also applied in vivo. Briefly, mice were injected intraperitoneally with 8 injections of Doxo (3 mg/kg per bodyweight) over a period of 2 weeks. Following an 8-week waiting period, mice were subjected to daily injections of Iso for two weeks (10 mg/kg of bodyweight). In addition, we performed ATAC-seq of mitochondria isolated from hearts of mice injected with Doxo to study its effect of mitochondrial DNA accessibility. Results qPCR analysis revealed that an initial exposure to Doxo leads to increased expression of the pathological marker nppb when stimulated with Iso (Iso, fc=2.79; Doxo+Iso, fc=5.5 vs. control, p= 0.0241, 0.0001). The in vivo experiment revealed that mice receiving Doxo+Iso had a worse systolic function compared to mice receiving Doxo or Iso alone. Transcriptomic analyses unraveled a group of ‘two-hit’ regulated genes, dysregulated by Doxo+Iso but not by Doxo alone. A motif-discovery analysis revealed FOXK1 to be a potential transcriptional regulator of these ‘two-hit’ genes which include DUSP4 and PDK2. Interestingly, immunostaining of NRVM’s suggests that FOXK1 might localize to the mitochondria in cardiomyocytes. Epigenetic analysis of the mitochondrial genome using ATAC-seq from isolated mitochondria from hearts treated with Doxo led to the identification of an unrecognized regulatory region in the MT-RNR2 gene which displays reduced accessibility upon doxorubicin exposure. An unbiased in vitro proteomic-interaction screen of this region again identified FOXK1 as an interacting protein, suggesting a direct regulation of mitochondrial function. Conclusion Using our ‘two-hit’ model we were able to show that anthracycline exposure is indeed memorized by the heart and leads to a more severe response to future adverse cardiac events. Our data highlight FOXK1 as a potential mediator of long-term cardiac anthracycline cardiotoxicity and mitotoxicity by direct regulation of mitochondrial function and gene expression.