Background: Mechanical unloading before reperfusion limits infarct size in pre-clinical models and is currently under clinical investigation. Autophagy, an evolutionarily conserved catabolic cellular process, plays an important role in many diseases. However, in acute MI (AMI), the pathophysiological impact of autophagy remains controversial; it has been shown to promote either cardiomyocyte death or recovery. A major limitation is the inability to accurately detect and quantify autophagy within the area at risk (AAR) during AMI. We hypothesise that a recently developed autophagy-detecting nanoparticle (ADN), based on the FDA-approved drug ferumoxytol, to target lysosomal compartments during autophagy in vivo, could provide a robust fluorescent readout of autophagy levels in a swine model of AMI.
Methods: To validate the autophagy levels, we first quantified the autophagy gold standard, microtubule-associated light chain 3 (LC3) protein, in a swine model of myocardial ischaemia–reperfusion (IR) injury, with or without Impella or extracorporeal membrane oxygenation (ECMO) interventions. Compared to sham controls, IR injury with or without interventions did not alter cytosolic LC3-I levels. Lipidated LC3-II, a marker of autophagosome formation, was significantly upregulated in IR pigs (by 48.7%, p=0.014), consistent with literature reports. Impella restored the increased LC3-II to levels comparable to sham (27.0% reduction, p=0.36 versus sham). ECMO further increased LC3-II levels twofold compared to sham (p<0.0001). The autophagy perturbations and restoration were further confirmed by p62 and Beclin-1 levels.
Results: To directly image autophagy levels with ADN, AMI was induced by balloon occlusion of the left anterior descending artery (LAD) for 120 minutes in an adult male pig. After 90 minutes of occlusion, we delivered 0.7 mg of ADN probe via intracoronary injection into the LAD, followed by reperfusion 30 minutes later. After 180 minutes of reperfusion, infarct size was quantified by triphenyltetrazolium chloride staining, and Evans blue counterstaining delineated the non-ischaemic septum from the AAR. Imaging of the ADN probe was performed by fluorescence reflectance imaging. In the representative mid-LV slice, ADN fluorescence was highly activated in the AAR, which was not seen in the contralateral septum. ADN intensity was quantified by signal-to-noise ratio, which increased significantly to 166.4 ± 22.5 in the AAR, compared to the septum (109.9 ± 6.4, p=0.013).
Conclusion: We report for the first time that cardiac-protective LV unloading restores basal autophagy levels, and further demonstrates the feasibility of quantitative autophagy imaging in the heart during AMI.