RESUMO
The principal regulator of cellular response to low oxygen is hypoxia-inducible factor (HIF)-1, which is stabilized in several forms of heart failure. Our laboratory developed a mouse strain in which a stable form of HIF-1 can be inducibly expressed in cardiomyocytes. Strikingly, these mice show a rapid decrease in cardiac contractility and a rapid loss of SERCA2 protein, which is also seen in heart failure. Interestingly, while the SERCA2 transcript decreased, it did not fully account for the observed decrease in protein. We therefore investigated whether HIF-1-regulated microRNA could impair SERCA translation. Multiple screening analyses identified the microRNA miR-29c to be substantially upregulated upon HIF-1 induction and to have complementarity to SERCA, and therefore be a potential regulator of SERCA2 expression in hypoxia. Subsequent evaluation confirmed that miR-29c reduced SERCA2 expression and Ca2+ reuptake. Additionally, administration of an antagonist sequence (antimir) improved cardiac contractility and SERCA2 expression in HIF transgenic mice. To extend the significance of these findings, we examined miR-29c expression in physiological hypoxia. Surprisingly, miR-29c decreased in these settings. We also treated mice with antimir before infarction to see if further suppression of miR-29c could improve cardiac function. While no improvement in contractility or SERCA2 was observed, reduction of heart size after infarction indicated that the antimir could modulate cardiac physiology. These results demonstrate that while a HIF-1-regulated microRNA, miR-29c, can reduce SERCA2 expression and contractility, additional factors in the ischemic milieu may limit these effects. Efforts to develop miRNA-based therapies will need to explore and account for these additional countervailing effects. NEW & NOTEWORTHY Our study demonstrated hypoxia-inducible factor-1-dependent upregulation of miR-29c, which, in turn, inhibited SERCA2 expression and reduced cardiac contractility in a transgenic overexpression system. Interestingly, these results were not recapitulated in a murine myocardial infarction model. These results underscore the complexity of the pathological environment and highlight the need for therapeutic target validation in physiologically relevant models.