Background RNA interference (RNAi) has the potential to be a novel therapeutic strategy in diverse areas of medicine. We report on targeted RNAi for the treatment of heart failure (HF), an important disorder in humans resulting from multiple etiologies. Successful treatment of HF is demonstrated in a rat model of transaortic banding by RNAi targeting of phospholamban (PLB), a key regulator of cardiac Ca2+ homeostasis. Whereas gene therapy rests on recombinant protein expression as its basic principle, RNAi therapy employs regulatory RNAs to achieve its effect. Methods and Results We describe structural requirements to obtain high RNAi activity from adenoviral (AdV) and adeno-associated virus (AAV9) vectors and show that an AdV short hairpin RNA vector (AdV-shRNA) silenced PLB in cardiomyocytes (NRCMs) and improved hemodynamics in HF rats 1 month after aortic root injection. For simplified long-term therapy we developed a dimeric cardiotropic AAV vector (rAAV9-shPLB) delivering RNAi activity to the heart via intravenous injection. Cardiac PLB protein was reduced to 25% and SERCA2a suppression in the HF groups was rescued. In contrast to traditional vectors rAAV9 shows high affinity for myocardium, but low affinity for liver and other organs. rAAV9-shPLB therapy restored diastolic (LVEDP, dp/dtmin, Tau) and systolic (fractional shortening) functional parameters to normal range. The massive cardiac dilation was normalized and the cardiac hypertrophy, cardiomyocyte diameter and cardiac fibrosis significantly reduced. Importantly, there was no evidence of microRNA deregulation or hepatotoxicity during these RNAi therapies. Conclusion Our data show, for the first time, high efficacy of an RNAi therapeutic strategy in a cardiac disease.

The clinical phenotype of human dilated cardiomyopathy (DCM) encompasses a broad spectrum of etiologically distinct disorders. As targeting of etiologyrelated pathogenic pathways may be more efficient than current standard heart failure treatment, we obtained the genomic expression profile of a DCM subtype characterized by cardiac inflammation to identify possible new therapeutic targets in humans. In this inflammatory cardiomyopathy (DCMi), a distinctive cardiac expression pattern not described in any previous study of cardiac disorders was observed. Two significantly altered gene networks of particular interest and possible interdependence centered around the cysteine-rich angiogenic inducer 61 (CYR61) and adiponectin (APN) gene. CYR61 overexpression, as in human DCMi hearts in situ, was similarly induced by inflammatory cytokines in vascular endothelial cells in vitro. APN was strongly downregulated in DCMi hearts and completely abolished cytokine-dependent CYR61 induction in vitro. Dysbalance between the CYR61 and APN networks may play a pathogenic role in DCMi and contain novel therapeutic targets. Multiple immune cellassociated genes were also deregulated (e.g., chemokine ligand 14, interleukin-17D, nuclear factors of activated T

These data indicate the existence of a local cardiac adiponectin system regulated independent of adiponectin and TNFalpha serum levels and its disturbance in cardiac pathology. The study suggests a role for adiponectin in the pathogenesis of DCM and implicates the adipocytokine as a possible future therapeutic target in DCM.

RNA interference (RNAi) has potential to be a novel therapeutic strategy in diverse areas of medicine. In this paper, we report on targeted RNAi for the treatment of a viral cardiomyopathy, which is a major cause of sudden cardiac death or terminal heart failure in children and young adults. RNAi therapy employs small regulatory RNAs to achieve its effect, but in vivo use of synthetic small interfering RNAs is limited by instability in plasma and low transfer into target cells. We instead evaluated an RNAi strategy using short hairpin RNA (shRdRp) directed at the RNA polymerase (RdRP) of coxsackievirus B3 (CoxB3) in HeLa cells, primary rat cardiomyocytes (PNCMs) and CoxB3-infected mice in vivo. A conventional AAV2 vector expressing shRdRp protected HeLa against virus-induced death, but this vector type was unable to transduce PNCMs. In contrast, an analogous pseudotyped AAV2.6 vector was protective also in PNCMs and reduced virus replication by >3 log10 steps. Finally, we evaluated the intravenous treatment of mice with an AAV2.9-shRdRp vector because AAV9 carries the most cardiotropic AAV capsid currently known for in vivo use. Mice with CoxB3 cardiomyopathy had disturbed left ventricular (LV) function with impaired parameters of contractility (dP/dtmax = 3,006 +/- 287 vs. 7,482 +/- 487 mmHg/s, p < 0.01) and diastolic relaxation (dP/dtmin = -2,224 +/- 195 vs. -6,456 +/- 356 mmHg/s, p < 0.01 and Tau = 16.2 +/- 1.1 vs. 10.7 +/- 0.6 ms, p < 0.01) compared to control mice. AAV2.9-shRdRp treatment significantly attenuated the cardiac dysfunction compared to control vector-treated mice on day 10 after CoxB3 infection: dP/dtmax = 3,865 +/- 354 vs. 3,006 +/- 287 mmHg/s (p < 0.05), dP/dtmin = -3,245 +/- 231 vs. -2,224 +/- 195 mmHg/s (p < 0.05) and Tau = 11.9 +/- 0.5 vs. 16.2 +/- 1.1 ms (p < 0.01). The data show, for the first time, that intravenously injected AAV9 has the potential to target RNAi to the heart and suggest AAV9-shRNA vectors as a novel therapeutic approach for cardiac disorders.

Impaired function of the phospholamban (PLB)-regulated sarcoplasmic reticulum Ca 2+ pump (SERCA2a) contributes to cardiac dysfunction in heart failure (HF). PLB downregulation may increase SERCA2a activity and improve cardiac function. Small interfering (si)RNAs mediate efficient gene silencing by RNA interference (RNAi). However, their use for in vivo gene therapy is limited by siRNA instability in plasma and tissues, and by low siRNA transfer rates into target cells. To address these problems, we developed an adenoviral vector (AdV) transcribing short hairpin (sh)RNAs against rat PLB and evaluated its potential to silence the PLB gene and to modulate SERCA2a-mediated Ca 2+ sequestration in primary neonatal rat cardiomyocytes (PNCMs). Over a period of 13 days, vector transduction resulted in stable 499.9% ablation of PLB-mRNA at a multiplicity of infection of 100. PLB protein gradually decreased until day 7 (772% left), whereas SERCA, Na + / Ca 2+ exchanger (NCX1), calsequestrin and troponin I protein remained unchanged. PLB silencing was associated with a marked increase in ATP-dependent oxalate-supported Ca 2+ uptake at 0.34 mM of free Ca 2+ , and rapid loss of responsiveness to protein kinase A-dependent stimulation of Ca 2+ uptake was maintained until day 7. In summary, these results indicate that AdV-derived PLB-shRNA mediates highly efficient, specific and stable PLB gene silencing and modulation of active Ca 2+ sequestration in PNCMs. The availability of the new vector now enables employment of RNAi for the treatment of HF in vivo.

No abstract

Recently it was shown that several new pseudotyped adenoassociated virus (AAV) vectors support cardioselective expression of transgenes. The molecular mechanisms underlying this propensity for cardiac cell transduction are not well understood. We comparatively analyzed AAV vector attachment, internalization, intracellular trafficking, and nuclear uncoating of recombinant self-complementary (sc) AAV2.2 versus pseudotyped scAAV2.6 vectors expressing green fluorescence protein (GFP) in cells of cardiac origin. In cardiac-derived HL-1 cells and primary neonatal rat cardiomyocytes (PNCMs), expression of GFP increased rapidly after incubation with scAAV2.6-GFP, but remained low after scAAV2.2-GFP. Internalization of scAAV2.6-GFP was more efficient than that of scAAV2.2-GFP. Nuclear translocation was similarly efficient for both, but differential nuclear uncoating rates emerged as a key additional determinant of transduction: 30% of all scAAV2.6-GFP genomes translocated to the nucleus became uncoated within 48 h, but only 16% of scAAV2.2-GFP genomes. In contrast to this situation in cells of cardiac origin, scAAV2.2-GFP displayed more efficient internalization and similar (tumor cell line HeLa) or higher (human microvascular endothelial cell (HMEC)) uncoating rates than scAAV.2.6-GFP in non-cardiac cell types. In summary, both internalization and nuclear uncoating are key determinants of cardiac transduction by scAAV2.6 vectors. Any in vitro screening for the AAV pseudotype most suitable for cardiac gene therapy -which is desirable since it may allow significant reductions in vector load in upcoming clinical trials -needs to quantitate both key steps in transduction.

No abstract