Effects of phosphodiesterase inhibitors on the contraction of myocytes from rat, guinea pig and human ventricle

William B. Johnson, *Sidath Katugampola, Sian E. Harding, Myocyte Systems Biology, National Heart and Lung Institute, Imperial College, Dovehouse St, London, SW3 6LY and *Investigative Pharmacology, Candidate Research Group, Pfizer Ltd, Kent, CT13 9NJ

Cardiotoxicity of PDE inhibitors (PDEIs) has become evident from clinical trials (Nony et al., 1994), and it is therefore important to screen for this effect. Species differences in PDE expression are known to exist, so the aim of this study was to compare the responses to PDEIs between human and animal ventricular myocytes to determine the optimal animal model.

Ventricular myocytes were isolated from adult male Sprague Dawley rats (250-350g), adult Dunkin Hartley guinea pigs (550-750g), or human heart tissue (failing heart or HOCM (Hypertrophic Obstructive Cardiomyopathy) biopsy). Cells were placed in a bath on the stage of an inverted microscope and superfused with Krebs-Henseleit buffer containing Ca2+ at 1mM (rat, guinea pig) or 2-4mM (human), at 37°C (rat, guinea pig) or 32°C (human), and paced electrically. Contraction amplitude of myocytes was measured via a video camera using software from IonOptix.

In the rat, the non-specific PDE inhibitor IBMX (3-isobutyl-1-methylxanthine) caused a significant concentration-dependent increase in amplitude (% shortening) of myocytes from a basal of 4.6 ± 0.8% to 10.7 ± 1.4% (P<0.02, n=6, EC50 3.9x10-5M CI 4.9x10-6-3.1x10-4). Milrinone, a PDE3 inhibitor, and rolipram, a PDE4 inhibitor, did not affect percentage shortening at concentrations from 10-7 M to 10-4 M. Sub-maximal concentrations of IBMX (1-3x10-5M) sensitised myocytes to the β-adrenoceptor agonist isoprenaline, causing a left-shift in the concentration-response curve (EC50 5.1x10-9M, CI 3.8-6.7x10-9M without IBMX, 1.1x10-9M, CI 0.44-2.7x10-9M with IBMX, P<0.0001, n=6). Rolipram (10-5 M) was not able to shift the isoprenaline concentration-response curve. In the guinea pig, IBMX also caused a significant concentration-dependent increase in myocyte amplitude from 3.8 ± 0.6% to a maximum of 8.0 ± 1.1% (P<0.02, n=6, EC50 1.29x10-5 M, CI 0.5-3.3x10-5). In contrast to the rat, inhibition of PDE3 with milrinone also increased percentage shortening from a basal of 4.1 ± 0.9% to 9.7 ± 0.8% (P<0.0001, n=7, EC50 4.8x10-7 M, CI 0.13-1.8x10-6M). Rolipram did not increase amplitude. A combination of milrinone and rolipram did not cause a significant increase in percentage shortening over that seen with milrinone alone. In human myocytes, either from failing or HOCM tissue, IBMX and milrinone both caused an increase in percentage shortening (IBMX 3.2 ± 1.1% to 7.9 ± 2.5%, P<0.05, n=5; Milrinone 2.8 ± 0.4% to 5.8 ± 0.9%, P<0.01, n=8). Rolipram at concentrations up to 10 -4 M did not increase amplitude.

In conclusion, while rat ventricular myocytes are insensitive to PDE3 inhibition, both guinea pig and human cells show an increase in contraction when PDE3 is inhibited. All species were insensitive to PDE4 inhibition. The rat is a common model for drug testing and development, yet we have shown that the guinea pig may well be more appropriate when considering the effect of phosphodiesterase inhibitors in humans.

Nony et al., Eur J Clin Pharmacol. 1994; 46(3):191-6.