Inhibition of phosphodiesterases 1 and 4 prevents myofibroblast transformation in Peyronie's disease

Abstract

Objectives: To investigate which phosphodiesterase (PDE) isoforms are expressed in fibroblasts isolated from the tunica albuginea (TA) of patients with Peyronie's disease (PD), and to measure the potency of PDE inhibitors in preventing transformation of these fibroblasts to profibrotic myofibroblasts.

Materials and methods: Fibroblasts isolated from the TA of men undergoing surgery for correction of PD curvature were transformed to myofibroblasts using transforming growth factor beta-1. The expression of 21 PDE isoforms was investigated using quantitative reverse transcriptase-polymerase chain reaction and protein analysis, as were the effects of various PDE inhibitors on prevention of myofibroblast transformation. Intracellular cAMP and cGMP in the presence of PDE inhibitors were quantified using cGMP/cAMP enzyme-linked immunosorbant assay assays.

Results: We found that PDE1A, 1C, 4, 5A, 7B and 8B were expressed at mRNA and protein levels. Selective inhibitors of these enzymes prevented myofibroblast transformation in a concentration-dependent manner, with PDE1 inhibitor ITI-214 and PDE4 inhibitors roflumilast and roflumilast N-oxide showing greatest potency. ITI-214 and roflumilast N-oxide increased intracellular cAMP, but not cGMP, in a concentration-dependent manner.

Conclusions: This is the first demonstration of the expression of PDE1, 7 and 8 isoforms, and the function of PDE1 and PDE4 in human TA fibroblasts. The ability of inhibitors of these enzymes to prevent myofibroblast transformation suggests that such inhibitors can be developed to treat acute PD.

Keywords: Peyronie's disease; fibroblast; fibrosis; myofibroblast; phosphodiesterase; tunica albuginea.

Fig. 1

Protein expression of phosphodiesterase (PDE) proteins in tunica albuginea‐derived cells from patients with Peyronie's disease. Fibroblasts were exposed to transforming growth factor beta‐1 (TGF‐β1; 10 ng/mL) or blank media for 72 h and stained with secondary antibody raised against PDE1A, PDE1C, PDE3A, PDE3B, PDE4, PDE5A, PDE7A, PDE7B, PDE8A or PDE8B. Data were normalised as a ratio of relevant protein/DNA staining (700 nm fluorescence intensity / 800 nm fluorescence intensity). Data points are plotted as mean ± SEM, N = 3. *P < 0.05 significant statistical difference vs no primary antibody control using Dunnett's multiple comparisons test.

Fig. 2

Effect of phosphodiesterase (PDE) inhibitors on transforming growth factor beta‐1 (TGF‐β1)‐induced myofibroblast transformation in tunica albuginea‐derived fibroblasts. (A) Cells were exposed to 10 ng/mL TGF‐β1 and a range of concentrations of PDE inhibitors for 72 h. Data were normalised as a ratio of α‐smooth muscle actin (SMA)/DNA staining (700 nm fluorescence intensity/800 nm fluorescence intensity) and presented as a percentage of TGF‐β1‐induced α‐SMA expression. (A) Concentration–response curves (CRCs) of selective PDE1 inhibitors DSR‐141562, ITI‐214 and vinpocetine. (B) CRCs of selective PDE4 inhibitors apremilast, roflumilast, roflumilast N‐oxide and rolipram. (C) CRCs of selective PDE5 inhibitors vardenafil, sildenafil and tadalafil. (D) CRCs of pan‐PDE inhibitor pentoxifylline, PDE7 inhibitor BRL‐50481, and PDE8 inhibitor PF‐04957325. Data points are plotted as mean ± SEM, N = 3, n = 9.

Fig. 3

cGMP and cAMP detected from lysed tunica albuginea (TA)‐derived fibroblasts treated with transforming growth factor beta‐1 (TGF‐β1), plus spermine NONOate, forskolin, or phosphodiesterase (PDE) inhibitors. (A) cGMP levels in fibroblasts, which were treated with blank media (no treatment [NT]), 10 ng/mL TGF‐β1 (treatment [T]), 10 ng/mL TGF‐β1 plus nitric oxide donor 3 μM spermine NONOate (T + 3S), 10 μM spermine NONOate (T + 10S) or 30 μM spermine NONOate (T + 30S) and in rhabdomyosarcoma cells, which were treated with 10 μM spermine NONOate (rhabdomyosarcoma). (B) cGMP levels in fibroblasts were treated with blank media (NT), 10 ng/mL TGF‐β1 (T), 10 ng/mL TGF‐β1 plus 0.3 μM PDE1 inhibitor ITI‐214 (T + 0.3I), 3 μM ITI‐214 (T + 3I) or 30 μM ITI‐214 (T + 30I). (C) cGMP levels in fibroblasts were treated with blank media (NT), 10 ng/mL TGF‐β1 (T), 10 ng/mL TGF‐β1 plus 0.03 μM PDE4 inhibitor roflumilast N‐oxide (T + 0.03R), 0.3 μM roflumilast N‐oxide (T + 0.3R) or 3 μM roflumilast N‐oxide (T + 3R). (D) cAMP levels in fibroblasts were treated with blank media (NT), 10 ng/mL TGF‐β1 (T), 10 ng/mL TGF‐β1 plus 1 μM forskolin (T + 1F), 3 μM forskolin (T + 3F) or 10 μM forskolin (T + 10F). (E) cAMP levels in fibroblasts were treated with blank media (NT), 10 ng/mL TGF‐β1 (T), 10 ng/mL TGF‐β1 plus 0.3 μM PDE1 inhibitor ITI‐214 (T + 0.3I), 3 μM ITI‐214 (T + 3I) or 30 μM ITI‐214 (T + 30I). (F) cAMP levels in fibroblasts were treated with blank media (NT), 10 ng/mL TGF‐β1 (T), 10 ng/mL TGF‐β1 plus 0.03 μM PDE4 inhibitor roflumilast N‐oxide (T + 0.03R), 0.3 μM roflumilast N‐oxide (T + 0.3R) or 3 μM roflumilast N‐oxide (T + 3R). Data points were plotted as mean ± SEM, N = 3, n = 9. *P < 0.05 significant increase in cGMP or cAMP detected vs TGF‐β1‐treated TA‐derived fibroblasts, using Student's t‐test.