Regulation of PDE4

PDE4 activity appears to be regulated at a number of levels, including post-translationally by multi-site phosphorylation (Lim et al. 1999; MacKenzie et al. 2000; MacKenzie et al. 2002), at the transcriptional level (Vicini and Conti 1997; Rena et al. 2001; Wallace et al. 2005) and through regulation of mRNA stability (Liu et al. 2000). Central to coordinating the functional output of regulation by phosphorylation are the UCR1

and UCR2 domains. These interact to, presumably, alter the conformation of the catalytic unit and mediate the consequences of phosphorylation by PKA and ERK (Houslay 2001). Their role in regulation of enzymatic activity first came to light when truncation studies revealed that removal of UCR2 led to an increase in catalytic activity (Jin et al. 1992; Lim et al. 1999). The constitutive inhibitory effect of UCR2 on PDE activity can be relieved by PKA phosphorylation at a site in the N-terminal region of UCR1 (Sette and Conti 1996; Hoffmann et al. 1998; MacKenzie et al. 2002). This site is present in all long PDE4 isoforms, but absent in short forms, giving insight into the importance of alternative splice variants (MacKenzie et al. 2002). The phosphorylation of UCR1 by PKA has also been shown to disrupt its intramolecular interaction with UCR2 (Beard et al. 2000). Interaction between UCR1 and UCR2 has been demonstrated by a variety of methods and is thought to occur via electrostatic interactions between the hydropho-bic C-terminal portion of UCR1 and the hydrophilic N-terminal region of UCR2 (Beard et al. 2000). That PKA phosphorylation can activate all long PDE4 isoforms accounts for the early observation that elevation of intracellular cAMP leads to increased PDE4 activity (Marchmont and Houslay 1980) and also provides a feedback mechanism to reset the cellular cAMP levels after stimulation. PKA phosphorylation of UCR1 is thought to cause a conformational change, which activates PDE4 long forms by attenuating the interaction of UCR1 with UCR2 (Beard et al., 2000). In PDE4D3, the PKA target residue is serine 54 (S54), and mutation of this residue to a negatively charged aspartate or glutamate residue can mimic PKA activation (Hoffmann et al. 1998). Mutation of the conserved neighboring glutamate residue (E53) also mimics PKA activation, leading to the hypothesis that this negatively charged residue is required for an ion pair interaction, which holds the enzyme in a low activity state. Disruption of this ion pair interaction by E53 mutation or PKA phosphorylation of S54 shifts the PDE to an active conformation. The increase in the activity of PDE4 long forms observed on addition of the negatively charged phospholipid, phosphatidic acid (PA) is thought to work via the same mechanism (Nemoz et al. 1997). In PDE4D3 alone, the PKA phosphorylation of S54 also leads to an increase in sensitivity to rolipram inhibition (Alvarez et al. 1995; Hoffmann et al. 1998). Mutations in S54 and E53 lead to a range of effects on PKA activation and rolipram inhibition, showing that small changes in configuration or charge of UCR1 can lead to measurable activity changes via conformational changes to the catalytic unit (Hoffmann et al. 1998).

On the basis of studies done with PDE4D3, it has recently been proposed that a further means by which the UCR1 and UCR2 regions can regulate PDE4 enzymes is by mediating PDE4 homo-dimerization. Long forms of PDE4 have been proposed to exist as dimers within cells, while short forms have been proposed to exist as monomers, again highlighting the differential regulation of short and long forms. Deletion analysis indicated that dimerization requires the C-terminal half of UCR1 and the N-terminal half of UCR2, but does not involve the same charged residues reported to mediate the intramolecular interaction (Richter and Conti 2002). Mutations that prevent the dimerization of PDE4D3 also ablated its activation by PKA or PA, demonstrating the significance of dimerization for enzyme regulation (Richter and Conti 2004). Mutations that abolished PDE4D3 dimerization also reduced the sensitivity of this isoform to rolipram inhibition.

PDE4s are generally accepted to exist in a number of different states that have different affinities for the inhibitor, rolipram (Souness and Rao 1997). It now appears that dimerization via UCR1 and UCR2 may be involved in the generation and stabilization of certain of these different rolipram binding conformers (Bolger et al. 2007; Richter and Conti 2004). Other means of generating such conformers include protein-protein interactions, where changes in rolipram affinity have been seen with RACK1 binding to PDE4D5 (Yarwood et al. 1999), and SH3-domain-containing proteins with PDE4A4 (McPhee et al. 1999).

In addition to PKA phosphorylation, PDE4s can also be regulated via phosphorylation by the MAP kinase, ERK (Hoffmann et al. 1999; MacKenzie et al. 2000; Baillie et al. 2001; Hill et al. 2006). All PDE4 subfamilies, except for PDE4A, contain a single ERK consensus motif (P-X-S-P) within the third subdomain of their catalytic unit. This serine residue is subject to phosphorylation by ERK both in vitro and in vivo. In order to phosphorylate PDE4 isoforms, ERK must bind to the PDE4 via two docking sites (see Sect. 2.5.8), which flank the phosphorylation site (MacKenzie et al. 2000). The functional consequences of ERK phosphorylation are dependent on the presence of UCR1 and UCR2, such that long forms are profoundly inhibited, short forms are activated, and super-short forms are weakly inhibited (Hoffmann et al. 1999; MacKenzie et al. 2000). However, ERK inhibition of long isoforms can be negated by PKA phosphorylation (Hoffmann et al. 1999) and switched to activation by additional phosphorylation of a site within the catalytic unit through activation of reactive oxygen (ROS) signalling cascades (Hill et al. 2006). The complement of long or short isoforms generated by alternative RNA splicing will therefore determine cellular response to crosstalk between cAMP signaling and the ERK pathway. The importance of this is emphasized in U937 cells as they undergo remodeling of their PDE4 profile upon differentiation from a monocyte-like to macrophage-like pheno-type. In monocytic U937 cells, long PDE4D isoforms predominate, so activation of ERK has an overall inhibitory effect, but upon differentiation to macrophages, the short form PDE4B2 is up-regulated to become predominant, and so activation of ERK elicits an overall increase in PDE4 activity (Shepherd et al. 2004). In the case of PDE4 long forms, a novel feedback system is in operation, whereby the inhibition induced by ERK phosphorylation causes a localized increase in cAMP levels, which activates PKA that, in turn, phosphorylates UCR1 to activate the enzyme and abolish ERK inhibition. Long forms of PDE4B, PDE4C and PDE4D can therefore affect a transient programmed rise in cAMP levels in response to activation of the ERK pathway (Hoffmann et al. 1999; Houslay and Kolch 2000).

Regulation of PDE4 also occurs at the level of transcription and translation (Conti 2002). As discussed above, the cell-type-specific expression pattern of different PDE4 isoforms is of crucial importance in determining response to their phosphorylation by PKA and ERK; however, the mechanisms behind this differential expression are only beginning to be elucidated. As would be expected for such large and complex genes, multiple mechanisms of regulation appear to be in operation. Intronic promoters have been identified for a number of specific PDE4 isoforms (Vicini and Conti 1997; Olsen and Bolger 2000; Rena et al. 2001; Le Jeune et al. 2002; Wallace et al. 2005). The presence of multiple promoters allows different combinations of long and short isoforms to be expressed as required. Expression of a number of PDE4 isoforms was observed to be sensitive to cAMP, such that chronic stimulation with the adenylyl cyclase activator forskolin, or cAMP analogues, results in their up-regulation (Swinnen et al. 1991; Vicini and Conti 1997; Seybold et al. 1998; Le Jeune et al. 2002). This effect is now known to be mediated by cAMP response elements (CREs) in the promoter regions of PDE genes (D'Sa et al. 2002; Le Jeune et al. 2002). After phosphorylation by PKA, the transcription factor, CREB (CRE binding protein), can bind to the CRE regions and modulate gene transcription (Mayr and Montminy 2001). This feedback mechanism, whereby cAMP modulates the expression of the enzyme that degrades it, is though to be a long-term adaptive desensitization response that complements the short-term desensitization response accomplished by increased PDE4 activity on the PKA phosphorylation of long isoforms. A number of other agents have been reported to regulate PDE4 expression by mechanisms that remain undetermined (Houslay 2001). Of these, the most significant is the pro-inflammatory agent LPS (lipopolysaccharide), which specifically up-regulates PDE4B in monocytes and macrophages (Ma et al. 1999; Wang et al. 1999). This suggests that PDE4B, and in particular PDE4B2, may play a key role in inflammatory responses and is therefore the most appropriate target for the development of anti-inflammatory drugs (Ma et al. 1999). Finally, the expression of PDE4 isoforms is also likely to be regulated at the level of mRNA stability (Swinnen et al. 1991), as has been demonstrated to be the case for a number of PDE4D isoforms in vascular smooth muscle cells (Liu et al. 2000). In these cells, activation of the cAMP-PKA pathway results in the induction of the PDE4D1 and PDE4D2 isoforms; however, simultaneous activation of the cAMP-PKA and the ERK MAPK pathways attenuates induction of these two short forms by a mechanism involving altered mRNA stability.

0 0

Post a comment