The GAF-Tandem Domain of Phosphodiesterase 5 as a Potential Drug Target - Phosphodiesterases as Drug Targets - page 153

this discrepancy was recently provided with the publication of the structure of the
human PDE2A holoenzyme, which is a dimer (Pandit et al. 2009 ). This structure
has provided novel and exciting insights into our understanding of the regulation
of PDE2 activity. It demonstrated that, in the homodimer, the catalytic domains
of PDE2 occlude each other, thus impeding access of substrate to the catalytic site.
PDE2 has an N-terminal tandem GAF domain, GAF-A and GAF-B. Binding of
cGMP to GAF-B in PDE2 is proposed to trigger rotation of the monomers against
each other and, thereby, open the catalytic sites for substrate binding, i.e., the
catalytic center is active as a monomer and dimerization is an instrument for
regulation by controlling the position of the catalytic domains and solvent access
of the catalytic sites (Pandit et al. 2009 ). Albeit it is debatable whether the insights
with human PDE2 can be extended across different PDE families, this certainly
provides a very attractive hypothesis considering the similarity of several architec-
tural features in mammalian PDEs (Beavo and Brunton 2002 ; Conti and Beavo
2007 ; Francis et al. 2001 ).
The variety of different PDEs recognized during the last two decades has resulted
in a renaissance for PDEs as drug targets. Probably, we are only at the beginning
of exploiting this potential as just a few family-specific PDE inhibitors have reached
the market and a few others are in clinical development (Aversa et al. 2006 ; Croom
and Curran 2008 ; Houslay et al. 2005 ; Wang et al. 2007 ). The suitability of PDEs as
drug targets appears to be due to several prominent features. (1) Twenty-one distinct
gene families exist, which share certain structural similarities - regulatory domains
generally are located toward the N-terminus and the catalytic domains are located
toward the C-terminus (Conti and Beavo 2007 ; Handa et al. 2008 ); regulatory and
catalytic domains of different PDE families have distinct molecular properties,
and hence, may be amenable to development of specific drugs; (2) within individual
PDE families several distinctive isoforms exist, which appear either differentially
regulated or localized to peculiar subcellular compartments or multiprotein com-
plexes (Houslay 2010 ); (3) the X-ray structures of most catalytic PDE domains
(exceptions are PDE6 and 11) show defined structural variations and the develop-
ment of rather specific inhibitors of individual catalytic domains appears possible
(Conti and Beavo 2007 ; Pandit et al. 2009 ; Zhang et al. 2005 ). In fact, a number of
compounds have been designed and developed, which show a very high degree of
PDE subfamily specificity (Lugnier 2006 ). Enzymatic assays using recombinant
catalytic domains that are very active as monomers have been adapted to an
industrial high-throughput format. Complications that may arise with the multi-
domain PDE holoenzymes are thereby effectively circumnavigated.
So far, the regulatory domains of PDEs have not been described as drug targets,
although, in principle, they should be quite attractive. PDE2, PDE5, PDE6, PDE10,
and PDE11 contain an N-terminal tandem GAF domain that, upon binding of
cGMP (cAMP for PDE10), enhances catalytic activity (see above). The acronym
GAF is derived from the first identified GAF proteins, namely mammalian cGMP-
regulated PDEs, Anabaena adenylyl cyclase, and the E. coli transcription factor Fhl
A (Aravind and Ponting 1997 ). On the other hand, PDE1 isoforms that are encoded
by three separate genes possess an N-terminal tandem of two calmodulin-binding