Heart failure (HF) is a major cause of premature mortality and increased morbidity. One of the hallmarks of HF is a decrease in responsiveness to beta-adrenergic (β-AR) stimulation. β-AR dysfunction is characterised by alterations to key components of excitation contraction coupling (ECC), such as reduced L-type calcium current and attenuated rises in calcium transient amplitude. As HF patients have elevated levels of circulating catecholamines, it is possible that changes to β-AR signalling are caused by desensitisation of the β-adrenoreceptors. The β-AR signalling pathway influences ECC through phosphorylation by downstream effector protein kinases (PKA and PKG), which are activated by cyclic nucleotide second messengers (cAMP and cGMP). Therefore, the bioavailability of these second messengers directly influences the efficacy of β-AR signalling. A key group of enzymes, phosphodiesterases (PDE), hydrolyse cyclic nucleotides and strictly control their bioavailability in the cytosol. Thus, the current study sought to elucidate whether changes in PDE expression in HF contributes to the observed β-AR dysfunction. We induced heart failure in sheep using tachypacing. Animals were anaesthetised for pacemaker implantation (isoflurane, 1-3% in oxygen) and perioperative analgesia provided (meloxicam, 0.5 mg/kg). After 7 days recovery, RVP (210-220 bpm) was applied until clinical symptoms of HF were evident. Animals were sacrificed by I.V. injection of pentobarbitone (200mg/kg) and left ventricular tissue homogenised for analysis of PDE expression using Western blotting. PDE isoforms blotted for included the cAMP-hydrolysing PDE isoforms: PDE3A, PDE4B and PDE4D; and the cGMP-hydrolysing isoform: PDE5A. PDE expression was quantified by measurement of band intensity normalised to an internal control and data expressed as an average of three repeats. Protein expression analysis revealed that the cGMP-inhibited, cAMP-hydrolysing PDE3A was significantly reduced in HF (58% decrease control (n=7) vs failure (n=6), P=0.005). A reduced expression in failing tissue was also observed for the two isoforms of PDE4 (PDE4B, 45% decrease control (n=7) vs failure (n=6), P=0.05; PDE4D, 36% decrease control (n=7) vs failure (n=6), P=0.001). Furthermore, the cGMP-hydrolysing PDE5A showed a tendency for reduction in HF ventricular tissue (24% decrease control (n=7) vs failure (n=6), P=0.08). Statistical significance compared by student’s t-test. This study shows that in a large animal model of HF in which β-AR signalling is disrupted, PDEs are down regulated. This suggests a reduced control over cyclic nucleotide bioavailability in failing cardiac myocytes. In turn this may reveal a mechanism, which may contribute to the changes in calcium homeostasis observed in failing cardiac myocytes. Experiments were carried out in accordance with the UK Home Office Animals (Scientific Procedures) Act 1986.