Our previous studies have demonstrated that ischemically-injured cardiomyocytes release DAMPs into the circulation which activate the pDC – IFN-I pathway, exacerbate the inflammatory response, and induce post-ischemic reperfusion injury. The present study further demonstrates that the inflammatory response is triggered by the TLR9 – IFN-I pathway both inside ischemic myocardium and systemically during post-ischemic reperfusion. HCQ exerts cardioproctective effects against both initial ischemic injury and subsequent post-ischemic reperfusion injury by inhibiting the TLR9 – IFN-I pathway.
Total myocardial infarction during ischemia and post-ischemic reperfusion is a function of ischemic infarction; i.e. there will be no cardiomyocytic necrosis during reperfusion if there is no initial ischemic myocardial necrosis11,27. Myocardial ischemic injury has been investigated, to some extent, by isolated Langendorff heart models as they demonstrate de novo inflammatory responses inside the heart. Hypoxia-injured cardiomyocytes release DAMPs including mitochondrial DNA, which activate the TLR9 – IFN-I pathway and mediate the inflammatory response inside the myocardium28. In vivo myocardial IRI activates not only the intrinsic myocardial inflammatory response during ischemia, but also the extrinsic inflammatory responses during post-ischemic reperfusion, which collectively culminates in overall intramyocardial inflammation4,11,22. This helps explain why HCQ has demonstrated protection against myocardial IR injury when administered before the onset of ischemia21. In the current study, we found that both IFNα and IFNβ levels were significantly elevated in 40’/0’ CP (Fig. 2), 40’/60’ plasma (Fig. 5) and 40’/0’ CP-treated splenocytes (Fig. 7). Both intrinsic deficiency of TLR9 (TLR9−/− mice) or blocking the effects of TLR9 (HCQ treatment) significantly reduced the production of IFNα and IFNβ. These results demonstrate that cfDNA and HMGB1 released from ischemically-injured cardiomyocyes activates the TLR9 – IFN-I pathway both inside the myocardium during ischemia and outside the myocardium during post-ischemic reperfusion. During ischemia, cfDNA/HMGB1 activates the TLR9 – IFN-I pathway, which leads to more ischemic injury and further cfDNA/HMGB1 release in addition to hypoxic injury, resulting in a vicious cycle that exacerbates ischemic myocardial injury. Deficiency of TLR9 or HCQ treatment before ischemia significantly reduced the production of IFNα and IFNβ (Fig. 2B) and attenuated ischemic myocardial infarction (Fig. 2A). Taken together, these results demonstrated that TLR9 – IFN-I pathway inside myocardium was activated during ischemia and mediated ischemic myocardial infarction.
It has been demonstrated that post-ischemic reperfusion injury is induced by inflammatory responses4,11,12, 22–24,27 that are triggered by DAMPs released from ischemically-injured cardiomyocytes11,12,29. DAMPs thus are thought to exacerbate myocardial infarction during reperfusion11. We have demonstrated that two important DAMPs in particular, cfDNA and HMGB1, play a critical role in mediating inflammatory responses during reperfusion. Specifically, HMGB1 binds to RAGEs of inflammatory cells and facilitates intracellular migration of cfDNA11,12. Increased levels of cytosolic cfDNA then activates TLR9 and enhances production of IFN-Is13. The role of TLR9 in mediating ischemia/reperfusion injury has been explored in the liver30 and heart28,31. In the current study, we found that TLR9−/− mice had significantly smaller IS following 40’/60’ IR injury compared to WT control mice; HCQ failed to further decrease the IS in TLR9−/− mice (Fig. 4A&B). The role of TLR9 was further defined by using the TLR9 antagonists ODN-2088 and HCQ. Administration of a TLR9 antagonist (ODN-2088 or high-dose HCQ) before occlusion of the LCA similarly attenuated myocardial IS after 40’/60’ IR injury in WT mice (Fig. 4C). Interestingly, ODN-2088 failed to attenuate IS if administered before reperfusion (data not shown). However, HCQ administered either before or after LCA occlusion similarly attenuated IS (Fig. 5A and 4C). HCQ significantly decreased the plasma level of IFNα and IFNβ (Fig. 5B), again supporting its inhibitory effects on the TLR9 – IFN-I pathway.
We further defined the role of TLR9 in mediating myocardial IRI with the use of a selective TLR9 agonist, ODN-1826. ODN-1826 and 40’/0’ WT CP significantly exacerbated IS in WT mice that underwent 20’/60’ IRI (Fig. 6). Our results are consistent with a recent report demonstrating that ODN-1826 increases myocardial IRI31. WT 40’/0’ CP contained high levels of cfDNA and HMGB1 (Fig. 3). We have demonstrated that cfDNA and HMGB1 in CP trigger the inflammatory response by activating the RAGE-TLR9 pathway11,12. Using 40’/0’ CP to treat isolated splenocytes, we found that CP stimulated splenocytes to secrete IFNα and IFNβ (Fig. 7). These results further ellucidate the role of the TLR9 – IFN-I pathway in mediating myocardial IRI.
There are multiple limitations to this study. Though HCQ was recently found to inhibit both TLR7 and TLR9, the role of TLR7 was not directly studied in this set of experiments. This might help to explain why ODN-2088 failed to attenuate IS when administered before reperfusion. However, the role of TLR7 in myocardial ischemia and IR injury may be less significant than TLR9 given that IS in TLR9−/− mice did not change with administration of HCQ. There are additional limitations with the clinical applicability of pre-ischemic treatments in particular. Treatment with HCQ before the onset of ischemia could likely protect the heart both during ischemia and during post-ischemic reperfusion. However, this is clinically not applicable unless patients are already taking HCQ regularly for another indication. Additionally, the dose required for attenuation of ischemia and IR injury may be different than doses used in current practice. In the present study, high-dose HCQ tended to have the most significant impact on IS, which raises concerns about toxicity once translated clinically. There are numerous cardiovascular, renal, and metabolic effects that can be detrimental with high levels of HCQ32. Regarding the mode of administration, HCQ has previously been reported to be cardioprotective against myocardial IRI when administered orally for several days before the IRI21. However, all experiments in this study utilized intravenous HCQ. Thus, further research is required to determine the appropriate dosage of HCQ as well as the ideal mode of administration.