Abstract
138
Objectives: Non-coding ribonucleic acids (RNA) are upregulated in heart failure, during which they modulate ventricle remodeling. Selective inhibitors of microRNA-21 (miR-21) interfere with fibrotic signaling and improve cardiac function in small and large animal models. But the transient and non-tissue-specific upregulation of miR-21 may require precise timing for optimal therapeutic efficacy. We aimed to image cardiac miR-21 in vivo to characterize its temporal and spatial distribution in pressure overload heart failure.
Methods: The locked nucleic acid (LNA) miR-21 complementary 15mer strand (anti-miR-21) bearing a hexyl-disulfide moiety was used for coupling or labeling via maleimide-thiol-coupling. The subsequent derivative NODAGA-anti-miR-21 was utilized for labeling with gallium-68; the prosthetic group 18F-FPyME allowed labeling with fluorine-18 for positron emission tomography (PET). Tracer stability was assessed in human serum. C57Bl6 mice (n=14) underwent dynamic PET image acquisition to delineate kinetics and biodistribution of the labeled anti-miR-21-LNA constructs. An additional group of mice underwent transverse aortic constriction (TAC, n=6) to induce pressure overload heart failure, or sham surgery (n=4). Serial imaging was conducted at 7d and 21d after injury. Mice were administered 68Ga-anti-miR-21-LNA (11.7±1.5MBq) and a static image was acquired over 30-60min after injection. Tracer distribution was validated by ex vivo autoradiography and histopathology in adjacent ventricle sections.
Results: Both 68Ga- and 18F-anti-miR-21-LNA exhibited high stability in human serum over 120min (85-90% authentic compound). In healthy mice, 68Ga-anti-miR-21-LNA displayed rapid clearance from blood and accumulation in the liver. Low signal was observed in the myocardium, as confirmed by ex vivo biodistribution (% injected dose (ID)/g; 2.1±0.6). Pretreatment with unlabeled anti-miR-21-LNA (20mg/kg ip) did not affect tracer accumulation in target organs. Thin layer chromatography revealed predominantly unchanged tracer in blood and urine at 60min after injection and some hydrophilic metabolites. In heart failure mice, in vivo imaging did not reveal differences in the cardiac tracer accumulation between TAC and sham at 7d (%ID/g: 2.7±1.0 vs 2.7±0.7, p=0.982) or 21d (2.6±1.4 vs 3.3±2.0, p=0.581), due to significant spillover of liver activity (12-17%ID/g). However, ex vivo autoradiography revealed diffuse global enrichment of 68Ga-anti-miR-21-LNA throughout the left ventricle at 21d after TAC compared to sham (~3-fold). Masson trichrome histology identified interstitial collagen deposition throughout the left ventricle, consistent with pressure overload-induced remodeling. Quantitative PCR confirmed upregulation of miR21 in TAC hearts at 7d compared to sham (1.21±0.48 vs 0.52±0.08, p=0.026) at a similar magnitude to the autoradiography signal increase. Initial images from healthy mice with 18F-anti-miR-21-LNA displayed lower liver spillover to the myocardium region, which may facilitate isolation of the myocardial signal in heart failure. Conclusions: Radiolabeled anti-miR-21-LNA delineated increased miR-21 expression in failing hearts. High hepatic signal is consistent with the well known accumulation of the LNA in this organ, but may complicate quantitative interpretation of in vivo cardiac images. Accordingly, strategies to suppress liver uptake may enable clearer visualization of non-coding RNA upregulation in heart failure and refine therapeutic strategies.