Intravenous adeno-associated virus serotype 8 encoding urocortin-2 provides sustained augmentation of left ventricular function in mice

Abstract

Urocortin-2 (UCn2) peptide infusion increases cardiac function in patients with heart failure, but chronic peptide infusion is cumbersome, costly, and provides only short-term benefits. Gene transfer would circumvent these shortcomings. Here we ask whether a single intravenous injection of adeno-associated virus type 8 encoding murine urocortin-2 (AAV8.UCn2) could provide long-term elevation in plasma UCn2 levels and increased left ventricular (LV) function. Normal mice received AAV8.UCn2 (5×10¹¹ genome copies, intravenous). Plasma UCn2 increased 15-fold 6 weeks and >11-fold 7 months after delivery. AAV8 DNA and UCn2 mRNA expression was persistent in LV and liver up to 7 months after a single intravenous injection of AAV8.UCn2. Physiological studies conducted both in situ and ex vivo showed increases in LV +dP/dt and in LV -dP/dt, findings that endured unchanged for 7 months. SERCA2a mRNA and protein expression was increased in LV samples and Ca²⁺ transient studies showed an increased rate of Ca²⁺ decline in cardiac myocytes from mice that had received UCn2 gene transfer. We conclude that a single intravenous injection of AAV8.UCn2 increases plasma UCn2 and increases LV systolic and diastolic function for at least 7 months. The simplicity of intravenous injection of a long-term expression vector encoding a gene with paracrine activity to increase cardiac function is a potentially attractive strategy in clinical settings. Future studies will determine the usefulness of this approach in the treatment of heart failure.

FIG. 1.

AAV8.CBA.UCn2 map and urocortin-2 expression after intravenous delivery (5×10¹¹ genome copies [GC]). AAV8.CBA.UCn2 vector map: ITR, inverted terminal repeat; SVpA, poly(A) from the simian virus 40 (SV40) viral genome; UCn2, urocortin-2; CBA, chicken β-actin promoter; CMV.en, human cytomegalovirus enhancer. Plasma: Three AAV vectors were compared in their ability to increase plasma UCn2 levels 6 weeks after intravenous delivery (5×10¹¹ GC): AAV9.CMV.UCn2 (9.CMV, n=8), AAV9.CBA.UCn2 (9.CBA, n=8), and AAV8.CBA.UCn2 (8.CBA, n=19). To obtain an estimate of variability among animals, among 4 subgroups of pooled samples in the 17 control (C) mice, the SE was 17% of the mean. Plasma UCn2 was increased 15-fold over control (n=17) after AAV8.CBA.UCn2 delivery. Columns denote mean values of pooled samples; numbers above and within columns denote group size. Liver: Time course of UCn2 mRNA expression in liver. Livers were obtained sequentially 1–6 weeks after AAV8.UCn2 gene transfer (n=2 per time point). Copies of UCn2 mRNA are presented as fold increase versus endogenous UCn2 mRNA from saline-injected control mice. Left ventricle: LV samples were obtained 6 weeks after UCn2 gene transfer. Copies of UCn2 mRNA are presented as fold increase versus endogenous UCn2 mRNA from saline-injected control mice. UCn2 transgene mRNA was increased 120-fold versus control. Values represent means±SE. Values of p were determined by Student t test (unpaired, two-tailed). Numbers in columns denote group size.

FIG. 2.

LV function in situ and AAV8.UCn2 dose–response effect. (A and B) Six weeks after AAV8.UCn2 (5×10¹¹ GC, intravenous) or saline (CON) administration in vivo studies were performed to measure the rates of LV pressure development (LV +dP/dt) and decay (LV −dP/dt), which were used to assess LV systolic and diastolic function. AAV8.UCn2 increased LV +dP/dt at 6 weeks (p=0.002), 4 months (p=0.001), and 7 months (p=0.0001) after gene transfer, indicating that UCn2 gene transfer increases LV contractile function. Similarly, LV −dP/dt, a measure of LV diastolic function, was increased 4 months (p=0.009) and 7 months (p<0.02) after AAV8.UCn2 delivery. Studies were performed without knowledge of group identity. Values of p in (A) and (B) were determined by Student t test (unpaired, two-tailed). (C) To determine whether there was a relationship between the amount of AAV8.UCn2 delivered and its effect on LV function, LV +dP/dt was measured 6 weeks after delivery of AAV8.UCn2 in various doses: 5×10⁹ GC (n=3); 5×10¹⁰ GC (n=3); and 5×10¹¹ GC (n=7) versus saline-injected control mice (C) (n=6). Increased LV +dP/dt showed a dose-related effect; values of p were determined by ANOVA. 6W, 6 weeks; 4M, 4 months; 7M, 7 months; GC, genome copies; IV, intravenous; LV, left ventricle; UCn2, urocortin-2. In all graphs, data represent means±SE, and numbers in columns denote group size.

FIG. 3.

LV function ex vivo. To ensure that differences in LV function measured in vivo were not influenced by the effects of anesthesia, vasodilation, or autonomic reflexes, isolated perfused hearts were studied. Six weeks after AAV8.UCn2 (5×10¹¹ GC, intravenous) or saline (CON) administration, hearts were isolated and perfused and LV pressure was measured, blinded to group identity. (A) AAV8.UCn2 increased LV developed pressure. (B and C) LV +dP/dt (B), and LV −dP/dt (C) were increased in hearts from mice that had received AAV8.UCn2, confirming data obtained in vivo (Fig. 2 and Table 2). (D) Heart rates were not different between groups. These data (A–D) confirm that alterations in systolic and diastolic function are intrinsic to the heart, and do not reflect vasodilation due to increased levels of urocortin-2 alone. In all graphs, data represent means±SE, and numbers in columns denote group size. Values of p were determined by Student t test (unpaired, two-tailed).

FIG. 4.

Left ventricular Ca²⁺ handling. (A) SERCA2a protein expression was increased in LV samples 4 weeks after AAV8.CBA.UCn2 gene transfer (5×10¹¹ GC, intravenous) (p=0.02). A significant increase in SERCA2a mRNA was also seen (data not shown). Columns and error bars denote means and SE; numbers in columns denote group size; the number above the column on the right indicates the p value determined by Student t test (unpaired, two-tailed). (B) Phospholamban and troponin I. Phosphorylation and expression of phospholamban and troponin I were unaltered in LV samples 6 weeks after AAV8.UCn2 gene transfer. (C) Cytosolic Ca²⁺ transients in cardiac myocytes isolated from mice 6 weeks after UCn2 administration versus saline-treated control mice. The rate of Ca²⁺ decline was increased in cardiac myocytes from mice that had received AAV8.UCn2 gene transfer. (D) Summary of Ca²⁺ transient experiments. Time to Ca²⁺ decline was shortened in cardiac myocytes obtained from mice 6 weeks after gene transfer. Experiments were repeated three times and are derived from studies of 112 cardiac myocytes. Columns and error bars denote means and SE; numbers in columns denote number of cardiac myocytes obtained from a subset of n=3 mice per group; the number above the column on the right indicates the p value determined by Student t test (unpaired, two-tailed), confirming a group difference in these cardiac myocytes.

FIG. 5.

Cardiac myocyte cAMP–PKA signaling. Cardiac myocytes were isolated from control (C) mice and from mice that had received AAV8.UCn2 (UC) gene transfer 6 weeks previously. Cyclic AMP and protein kinase A (PKA) activity were assesses in the unstimulated (basal) state and after stimulation with isoproterenol (Iso, 10 μM, 10 min) and, in separate experiments, NKH477 (NKH, 10 μM, 10 min). NKH477 is a water-soluble forskolin analog that stimulates adenylate cyclase independently of β-adrenergic receptors. Numbers in columns denote group size. Left: cAMP production. No group differences were seen in basal or NKH477-stimulated cAMP production. However, β-adrenergic receptor-dependent cAMP production was increased after UCn2 gene transfer (Iso; p=0.04). Columns and error bars denote means and SE; numbers in columns denote group size; numbers above the columns indicate p values determined by Student t test (unpaired, two-tailed). Middle: PKA activity. PKA activity was assessed in cardiac myocytes pooled from each group; the means of five replicates are shown. PKA activity appeared to show small increases (22–34%) in PKA activity after UCn2 gene transfer. Right: PKA expression. No group differences in the expression or phosphorylation of the PKA catalytic subunit (PKA-C) were seen.

FIG. 6.

Histology: Liver and left ventricle. Hematoxylin and eosin (H&E) and Masson's trichrome staining was performed on sections of liver (original magnification,×10) and transmural sections of left ventricle (LV; original magnification,×40), 7 months after intravenous delivery of AAV8.UCn2 (5×10¹¹ GC). AAV8.UCn2 delivery and sustained urocortin-2 expression had no adverse effects even after 7 months. St, standard control stain for Masson's trichrome. Color images available online at www.liebertpub.com/hum