ABSTRACT

The archetypical second messengers, cAMP and cGMP, act mainly via binding to the cyclic nucleotide (cN) binding domain (cNBD), rst found in the catabolite activator protein of Escherichia coli. The cNs can also act via the cN binding GAF domain initially discovered in some cN phosphodiesterases (PDEs). An isolated case is provided by the Dictyostelium cAMP receptor of the seven transmembrane G protein-coupled receptor family, which responds to extracellular cAMP. In general, cN activation is driven by a higher cN afnity for the active rather than the inactive receptor conformation. The structural details of the molecular mechanisms involved in the activation process are emerging.1,2

Many cell types signal through both cAMP and cGMP, and additional binding site restraints seem to be required to ensure discrimination. The cAMP-and cGMPdependent protein kinases (PKAs and PKGs) contain two tandem CNBDs. This arrangement limits cross-signaling because both cN sites must be occupied to ensure efcient activation.3,4 Therefore, a 30-fold binding site preference for each site is theoretically sufcient to obtain a 900-fold preference for the two sites combined. Tandem CNBDs also permit intrachain allosteric interactions, as observed during the activation of PKA and PKG.3,5

Two genes, PRKG1 and PRKG2, give rise to PKG type I (PKG-I) and type II (PKG-II). Their two CNBDs differ in cGMP afnity. A missense mutation (p.Arg177Gln) of the high-afnity site produces a constitutively active PKG-I and segregates with aortic disease in humans.6 The PKG-Iα and PKG-Iβ isoforms are formed through alternative splicing and differ only in their far N-terminal region that contains the autoinhibitory motif. Despite having the same cGMP binding domains, PKG-Iα and PKG-Iβ have different cGMP activation constants (Ka = 0.1 and 0.9 μM, respectively) due to differential interaction between the CNBDs and the variable N-terminal.7