|
|
|
|
|
|
|
|
|||||
| Home Help Feedback Subscriptions Archive Search Table of Contents | |||||
First published online 29 August 2006
doi: 10.1242/jcs.03106
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
1 Molecular Pharmacology Group, Division of Biochemistry and Molecular Biology, Institute of Biomedical and Life Sciences
2 Bioinformatics Research Centre, Davidson Building, University of Glasgow, Glasgow, G12 8QQ, Scotland, UK
* Author for correspondence (e-mail: m.houslay{at}bio.gla.ac.uk)
Accepted 19 June 2006
 |
Summary |
|---|
 Top Summary IntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
Key words: PDE4A, Cyclic AMP, Trafficking, Ca2+, Phosphatidic acid, Golgi
|
Introduction |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
; Wong and Scott, 2004). It has become clear over recent years that this pathway is subject to complex spatial and temporal regulation. The cAMP signalling pathway is subject to regulation by cross-talk through interaction with a number of other distinct signalling cascades at many points, and also to the carefully controlled intracellular localisation of the signalling components, which allow gradients of cAMP to be produced in the cell in response to distinct stimuli and environmental conditions (Baillie et al., 2005; Houslay and Adams, 2003; Houslay and Kolch, 2000; Tasken and Aandahl, 2004; Wong and Scott, 2004; Zaccolo et al., 2002).
Once generated by adenylyl cyclases, the only way in which cAMP can be inactivated is by its hydrolysis through the action of cyclic nucleotide phosphodiesterases (PDEs) (Houslay and Adams, 2003). Enzymes of the PDE4 family, which specifically hydrolyse cAMP, are of particular interest because PDE4-selective inhibitors have been implicated in the treatment of a number of disease states such as asthma, COPD, rheumatoid arthritis and depression (Houslay et al., 2005). Four genes encode the family of PDE4 enzymes (A, B, C and D) and from these are generated 
20 different isoforms (Houslay and Adams, 2003). Differential regulation of cAMP signalling by PDE4 is achieved by the ability of their distinct N-terminal regions to confer differential intracellular localisations, and also by their ability to interact with components of signalling systems or intracellular substrates (Baillie and Houslay, 2005; Houslay and Adams, 2003). Thus interaction of PDE4D5 with ß-arrestin and its subsequent recruitment to the ß2-adrenergic receptor on agonist stimulation (Bolger et al., 2003; Perry et al., 2002) drives the ability of these receptors to switch their signalling from the Gs-driven adenylyl cyclase pathway to Gi-coupled activation of ERK (Baillie et al., 2003). In addition PDE4D5 also interacts with the signalling scaffold protein, RACK1 (Yarwood et al., 1999); PDE4A4/5 and PDE4D4 interact with SH3 domains of src-family tyrosyl kinases (Beard et al., 1999; McPhee et al., 1999); PDE4D3 interacts with a variety of proteins including AKAP450 (Tasken et al., 2001), mAKAP (Dodge et al., 2001) and myomegalin (Verde et al., 2001) and PDE4B1 with DISC1 (Millar et al., 2005).
Of the 20 or so PDE4 isoforms, the brain-specific (McPhee et al., 2001) super-short PDE4A1 is unique in that it is exclusively membrane associated, this being conferred by its isoform-specific 25 amino acid N-terminal region (Shakur et al., 1993). Deletion of this region has been shown to yield a fully active, entirely soluble enzyme, whereas the fusion of this unique N-terminal region to various proteins that were normally soluble rendered them membrane associated with identical distributions to the full-length PDE4A1 (Scotland and Houslay, 1995; Shakur et al., 1993; Smith et al., 1996). NMR analyses of the unique N-terminal region of PDE4A1 showed it to be composed of two distinct helical structures separated by a mobile `hinge' region (Smith et al., 1996). The first of these helices (Helix-1; amino acids 1-8) comprises a N-terminal amphipathic 
-helix whereas the second (Helix-2; amino acids 14-25) is a compact helical structure that is highly hydrophobic and tryptophan-rich (Smith et al., 1996). More recently we have identified a novel microdomain within helix-2, called TAPAS-1, which allows association with membranes and lipid vesicles (Baillie et al., 2002). It is the binding of Ca2+ to Asp21 in TAPAS-1 that serves as a trigger for both membrane insertion of this domain and its preference for interaction with phosphatidic acid (PA) (Baillie et al., 2002).
Phospholipase D (PLD) hydrolyses phosphatidylcholine (PC) and through its action generates PA. Both PLD and PA have been implicated in a range of cellular functions such as exo- and endocytosis, cellular proliferation, membrane trafficking and cytoskeletal organisation. PLD is widely expressed (Meier et al., 1999) and PLD activity has been shown to be stimulated through the action of various cell surface receptors (Exton, 2002). Two mammalian PLD genes (PLD1 and PLD2) have been described, both of which generate splice variants (Colley et al., 1997; Hammond et al., 1995; Hammond et al., 1997; Steed et al., 1998) and which differ in their subcellular distribution and regulation (Brown et al., 1998; Exton, 2002; Freyberg et al., 2003). However, downstream targets of PLD have proved difficult to elucidate, although a number of candidates have been proposed, such as Raf-1 (Rizzo et al., 2000), mTOR (Fang et al., 2001), RGS proteins (Ouyang et al., 2003), various traffic-related proteins (Manifava et al., 2001), the p47phox domain of NADPH oxidase (Karathanassis et al., 2002) and sphingosine kinase-1 (Delon et al., 2004). It appears that such proteins can interact with PA in very different ways. Thus, in PDE4A1, TAPAS-1 inserts in the lipid bilayer where it shows selectivity for association with PA rather than other acidic phospholipids, such as phosphatidyl serine (PS) and phosphatidylinositol-4,5-bisphosphate (PIP2). In contrast to this, other proteins, such as Raf, do not insert into bilayers but appear to interact electrostatically with the PA headgroup and, invariably, that of PS and other acidic phospholipids. Our previous in vitro analyses have shown that PA binding to PDE4A1 is intrinsic to its membrane association. Here we show that the presence of the TAPAS1 domain alone is insufficient for the fidelity of membrane targeting of PDE4A1 in situ in intact cells. We identify a novel key role that helix-1 plays in achieving not only membrane association and targeting of PDE4A1 but conferring a dynamic ability to alter intracellular distribution of PDE4A1. This dynamic relocalisation occurs in response to changes in intracellular Ca2+ levels upon its release from intracellular stores and is, additionally, dependent on the ability of PLD to generate PA.
|
Results |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
). NMR analyses have shown that this region forms two distinct helical regions, namely helix-1 formed from the first seven amino acids of PDE4A1 and helix-2, formed from amino acids 14-25 of PDE4A1 (Smith et al., 1996). These two regions are separated by a mobile hinge region, provided by amino acids 8-13 of PDE4A1 (Smith et al., 1996) (Fig. 1). The second helical region of (1-25) PDE4A1 contains within it a microdomain, named TAPAS1, which confers insertion of this unit into lipid vesicles (Baillie et al., 2002). This irreversible association occurs in a Ca2+-dependent manner with the TAPAS-1 microdomain showing a preference for interaction with PA (Baillie et al., 2002).
|
; Pooley et al., 1997). We aimed here to explore the relevance of helix-1 and the hinge region of (1-25) PDE4A1 to the intracellular targeting of PDE4A1 in living cells. To do this we used confocal immunofluorescence microscopy to examine the intracellular targeting of a series of deletion and amino acid substitutions in a chimera [(1-25)-4A1-GFP] formed between GFP and the first 25 amino acids of PDE4A1 [(1-25) PDE4A1] (Fig. 1). Expressed in COS1 cells, (1-25)-4A1-GFP showed a tight localisation to the Golgi, co-localising with the Golgi marker BODIPY ceramide. Indeed, (1-25)-4A1-GFP showed a 97±1% co-localisation with the Golgi probe, BODIPY ceramide with a correlation factor of 0.970±0.004 (n=31) (Fig. 2A). Intriguingly, however, only 58±3% of this Golgi marker overlaps with (1-25)-4A1-GFP suggesting that, although PDE4A1 co-localises entirely within the Golgi apparatus, it is only present within a specific region of this structure. To examine this further, we used markers specific for the cis (GM130) and trans (TGN46) regions of the Golgi, which allowed us to show that (1-25)-4A1-GFP localises to the trans-Golgi (Fig. 2A), with a correlation factor of 0.86±0.04 (n=27). By contrast, localisation of (1-25)-4A1-GFP showed no correlation with GM130 cis-Golgi marker staining (0.71±0.06%; n=27) (Fig. 2A).
|
To identify any influence of helix-1 on PDE4A1 targeting in situ we examined deletion constructs of the (1-25)-4A1-GFP chimera (Fig. 1). To our surprise, we found that the targeting of (1-25)-4A1-GFP to the Golgi was disrupted by the deletion of helix-1, as seen in the construct 
2-7(1-25)-4A1-GFP with only 20±2% of this construct co-localising with the Golgi marker (n=25) (Fig. 2B,C). This paralleled the effect of deletion of helix-2, as in the 14-25(1-25)-4A1-GFP chimera (Fig. 2B,C) co-localisation with the Golgi was reduced to 21±4% (n=25). Failure to target to the Golgi occurred in spite of the fact that our previous analyses showed the cognate helix-1 deletion chimera, 2-7(1-25) PDE4A1-CAT, when synthesised in vitro, was capable of binding to isolated membranes (Smith et al., 1996). One possibility is that helix-1 serves to stabilise membrane interaction in intact cells, perhaps because an additional membrane interaction site is required in intact cells whereas a single site sufficed in the high membrane input used for in vitro assays. However, as our previous analyses have shown, single point mutations in helix-2 served to ablate in vitro binding to membranes and lipid vesicles of 1-25PDE4A1-CAT, a construct that contained helix-1 (Baillie et al., 2002). Thus if helix-1 does contain an additional membrane-association site we suggest that such a site may operate consequent to prior membrane insertion of TAPAS-1 domain in helix-2 and the resultant conformational change that occurs in the unique N-terminal region of PDE4A1 (Baillie et al., 2002).
Two binding sites are required for the correct targeting of PDE4A1 in living cells
Many essentially `soluble' proteins that are found associated with membranes have two points of interaction, such as the key signalling protein Ras (Goodwin et al., 2005; Hancock et al., 1990; Zhang and Casey, 1996). We set out to examine the concept that two membrane-association sites may be needed to achieve efficient membrane targeting of PDE4A1 in intact cells. Our hypothesis is that helix-1 serves to stabilise a prior helix-2 membrane insertion. One way of evaluating a need for a second site would be to compare a GFP chimera of helix-2-hinge-helix-2 (H2-H2-GFP) with one that was composed of helix-1-hinge-helix-1 (H1-H1-GFP). Indeed, doing this we found that the H2-H2-GFP chimera exhibited complete membrane-association in COS1 cells (Fig. 3A). However, the fidelity of its targeting was clearly compromised compared with (1-25)-4A1-GFP (Fig. 3A), with now only 39±5% (n=25 cells) of its immunofluorescence being Golgi associated. Furthermore, unlike PDE4A1, H2-H2-GFP was not constrained to the trans-Golgi as the majority (86±4%; n=25 cells) of the Golgi marker, BODIPY ceramide now overlapped with H2-H2-GFP fluorescence. H2-H2-GFP was also found associated with other membrane structures (Fig. 3A,B), with 70±4.8% (n=25 cells) of the ER marker overlapping with H2-H2-GFP fluorescence and 20±2%, (n=25 cells) of the mitochondria marker overlapping with H2-H2-GFP fluorescence.
|
A mobile hinge region separates helix-1 from helix-2 in (1-25)-4A1-GFP, which we surmise may act to ensure the correct orientation of the two helical regions relative to each other. Our previous analyses have shown that when a chimera lacking the hinge region, 8-13(1-25)PDE4A1-CAT, was constructed and expressed in vitro, then this construct was still able to bind to membranes (Smith et al., 1996). Consistent with this, we show here that the 8-13(1-25)-4A1-GFP chimera is still membrane associated when expressed in living cells (Fig. 3C). However, what is apparent is that such a chimera now fails to exhibit such tight localisation to the Golgi and fluorescence is now also associated with other vesicles and membranous structures within the cell, including mitochondria (Fig. 3C). This may indicate that correct targeting of PDE4A1 depends upon the appropriate positioning of helix-1 relative to the TAPAS1-containing helix-2, a function provided by the mobile hinge region.
Helix-1 contains a distinct binding motif that facilitates Golgi targeting
The TAPAS-1 domain of helix-2 contains a LVxWW motif that provides its central bilayer insertion unit (Baillie et al., 2002). The function of this motif is destroyed when any of these various residues are mutated to either alanine or aspartate (Baillie et al., 2002). Interestingly, helix-1 contains what might be construed as a similar motif (aliphatic-aliphatic-X-aromatic-aromatic), comprised of the residues LVDFF, with the noticeable difference of an acidic aspartate residue (Asp5) at position 3 instead of the neutral glycine (Gly18) found at position 3 in helix-2. In the case of the TAPAS1 domain, substitution of Gly18 with Asp would preclude membrane insertion. However, in helix-2 an Asp residue is present immediately following the bilayer insertion motif, as in LVGWWD. This Asp21 residue provides a Ca2+-operated switch, which gates insertion of LVGWW into the lipid bilayer. When Asp21 is mutated to Ala this allows the TAPAS1 domain to insert into lipid bilayers in a Ca2+-independent manner (Baillie et al., 2002). To examine whether the Asp5 residue in LVDFF of helix-1 plays a functional role, we mutated this residue to Ala and examined the distribution of such a mutant species in living cells. In doing this we saw that Asp5 plays an important role in the fidelity of Golgi targeting. Thus its simple mutation to Ala caused a profound redistribution of the (1-25)-4A1-GFP chimera such that now it was localised to dense puncta throughout the cytoplasm in addition to a fraction being associated with the Golgi (Fig. 4A,B). The LV and FF pairings in LVDFF of helix-1 are also pivotal in conferring Golgi association as evidenced from analyses of the L3A,V4A double mutant and the F6A, F7A double mutant, which both showed loss of targeting of (1-25)-4A1-GFP to the Golgi (Fig. 4C,E). Further analysis showed that Golgi targeting was severely reduced in the L3A mutant but not the V4A mutant with 76±4% (n=25) of V4A being associated with the Golgi in contrast to only 5±0.7% (n=25) of the L3A mutant (Fig. 4D,E), identifying Leu3, rather than Val4 as being the key residue of this pairing involved in the fidelity of Golgi targeting. In contrast to this, the single mutation of either F6A or F7A caused profound loss of membrane targeting of (1-25)-4A1-GFP (Fig. 4D,E) indicating the importance of both of these phenylalanine residues in helix-1 to the targeting of PDE4A1. Thus the key motif required for helix-1 to confer Golgi targeting on PDE4A1 is LxDFF.
|
). As others have shown for PLD2 (Freyberg et al., 2002), we also failed to observe any co-localisation of the punctate distribution of D5A(1-25)-4A1-GFP with markers of ER (BiP and GRP78), lysosomes (Lysotracker), mitochondria (Mitotracker) or early endosomes (EEA1) (Fig. 4A,B). PLD is responsible for catalysing the hydrolysis of phospholipids, primarily phosphatidylcholine, to produce PA (Exton, 2002), to which the TAPAS-1 domain in helix-2 of PDE4A1 preferentially binds (Baillie et al., 2002). We therefore examined the effect of inhibiting the ability of PLD to produce PA on the localisation of (1-25)-4A1-GFP and the D5A(1-25)-4A1-GFP mutant. To do this we took advantage of the well-tried procedure using butan-1-ol as a specific inhibitor of PA generation. This is based upon the fact that primary alcohols compete with water to be the hydroxyl donor in the hydrolysis of PA by PLD, resulting in the production of phosphatidylalcohol and thereby ablating PA formation through PLD (Exton, 2002). When cells overexpressing D5A(1-25)-4A1-GFP were pre-incubated with butan-1-ol the distribution of this mutant protein lost its punctate-distributed component and reverted to an exclusively Golgi-associated form (Fig. 5A). Butan-2-ol, which is unable to substitute for water in the hydrolysis of PA from PC, and which therefore acts as a stringent negative control for the specificity of the action of butan-1-ol at these concentrations of alcohol (Exton, 2002), caused no change in the distribution of D5A(1-25)-4A1-GFP (Fig. 5A). Calphostin-c, has recently been shown to be a direct and potent inhibitor of PLD1 and PLD2 (Sciorra et al., 2001). This inhibitory effect on PLD has been shown to be independent of any action on PKC (Sciorra et al., 2001). We therefore looked at the effect of 1 µM calphostin-c on the distribution of D5A(1-25)-4A1-GFP where, in agreement with results using butan-1-ol, we found it constrained the localisation of this mutant protein to the Golgi (Fig. 5A). This demonstrates that the distribution of D5A(1-25)-4A1-GFP mutant is influenced by PLD activity, specifically in allowing this mutant chimera to redistribute out of the Golgi and into punctate structures located throughout the cell cytosol.
|
Mobilisation of intracellular Ca2+ stores by thapsigargin causes redistribution of PDE4A1 from the Golgi in a PLD-dependent manner
In contrast to the wild-type (1-25)-4A1-GFP chimera, the Asp5Ala mutant displayed a profound redistribution to punctate structures throughout the cytosol rather than being confined to the trans-Golgi stack (Fig. 5A). We have previously shown that the trigger for the irreversible insertion of TAPAS-1 into the lipid bilayer is due to the interaction of Ca2+ with Asp21 (Baillie et al., 2002). On this basis we hypothesised that the intracellular redistribution of (1-25)-4A1-GFP, as seen in the D5A mutants, might in fact be mimicking a dynamic physiological process regulated by changes in intracellular Ca2+ levels in intact cells and mediated by Ca2+ interaction with Asp5. To assess this, we treated cells with the Ca2+-ATPase inhibitor, thapsigargin in order to increase intracellular Ca2+ levels. We observed that 2 µM thapsigargin (37°C, 30 minutes) caused the redistribution of (1-25)-4A1-GFP to punctate structures throughout the cytoplasm similar to that seen in the D5A(1-25)-4A1-GFP mutant (Fig. 6A). Indeed, as with the D5A(1-25)-4A1-GFP mutant, the redistribution of (1-25)-4A1-GFP caused by thapsigargin was reversed by pre-incubation of the cells with butan-1-ol or calphostin-c, but not butan-2-ol (Fig. 6A) indicating that the redistribution of (1-25)-4A1-GFP in response to thapsigargin is a dynamic and reversible process.
|
To investigate if full-length PDE4A1 was relocalised by increasing intracellular Ca2+ we examined the effect of thapsigargin on the distribution of PDE4A1 in paraformaldehyde-fixed COS1 cells. As with the (1-25)-4A1-GFP chimera, PDE4A1 was relocalised from an exclusively Golgi localisation to a punctuate distribution throughout the cytosol following thapsigargin treatment (Fig. 6A). This paralleled the distribution seen with the (1-25)-4A1-GFP upon such treatment. Such a punctuate distribution of PDE4A1 was reversed following butan-1-ol or calphostin-c but not butan-2-ol treatment (Fig. 6A).
As for the intracellular distribution of D5A(1-25)-4A1-GFP, no co-localisation of the redistributed mutant PDE4A1 could be seen with markers of the Golgi apparatus, ER, lysosomes or mitochondria and no effect on the integrity of these structures was observed following thapsigargin treatment (Fig. 6B). Thus mobilisation of intracellular Ca2+ stores by thapsigargin causes the dynamic intracellular redistribution of PDE4A1 from an exclusive trans-Golgi localisation to a punctate distribution in cells through a process that is dependent upon PLD activation.
|
Discussion |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
; Cooper, 2003; Dumaz and Marais, 2005; Zaccolo et al., 2002).
Although PLD is recognised as being an important signalling protein, the functional consequences of PA generation are only just beginning to be understood. Like other acidic phospholipids, PA can bind electrostatically to functional surfaces of various proteins and exert regulatory effects, such as that seen with Raf-1 (Ghosh et al., 1996), sphingosine kinase-1 (Delon et al., 2004), mTOR (Fang et al., 2001) and also the UCR1 regulatory region of long PDE4 isoforms (Grange et al., 2000). However, the TAPAS-1 domain found in the short PDE4A1 isoform lacks the PA/PS-binding UCR1 region and, instead, provides a Ca2+-triggered bilayer integration module which shows a preference for interacting with PA over other phospholipids, including the acidic PS and PIP2 (Baillie et al., 2002). Once exposed to elevated Ca2+ levels, TAPAS-1 irreversibly integrates into the lipid bilayer and PDE4A1 localises to the trans-Golgi stack (Baillie et al., 2002). Such an integration of PDE4A1 into the lipid bilayer is thus a permanent reflection of a cell having experienced an elevation in intracellular Ca2+ levels. However, as we show here for the first time, in the intact cell, such a finely targeted PDE4A1 does not remain insensitive to further changes in intracellular Ca2+. Thus a Ca2+-gated process, involving Asp5 within helix-1, together with the action of PLD act to dynamically and reversibly determine the spatial and temporal localisation of PDE4A1 within the cell.
More than one binding domain is required for membrane association and Golgi targeting of PDE4A1
An increasing body of evidence indicates that essentially soluble proteins that are able to anchor to membranes invariably require more than one binding site or lipid modification to facilitate membrane association in living cells. For example, the key signalling protein Ras needs to be both farnesylated and palmitoylated to achieve fidelity of membrane targeting and signalling (Goodwin et al., 2005; Hancock et al., 1990; Zhang and Casey, 1996), the src-family tyrosyl kinase, LYN requires two hydrophobic interactions supplied by palmitoylation and myristoylation in order to localise within lipid rafts in the plasma membrane (Xu et al., 2005) and, in the case of p47phox, simultaneous occupancy of two phospholipid binding pockets, one of which binds PIP2 and the other either PA or PS, radically increases membrane association (Karathanassis et al., 2002).
|
), implying that it may associate with a protein. If so, then we presume that this occurs subsequent to helix-2 insertion, as the helix-1-hinge-helix-1 chimera does not target to membranes (Fig. 3A). Resolution of the action of helix-1 in the targeting of PDE4A1 will be a considerable challenge, because, not only does membrane insertion of helix-2 serve as a prerequisite for helix-1 to express membrane-association properties, but the release of PDE4A1 from membranes is exquisitely sensitive to detergents (Shakur et al., 1995). Thus solubilising detergents are likely to displace putative proteins interacting with the hydrophobic amino acids we have identified here as being involved in the membrane-association properties of helix-1. We infer that the nature of the association of helix-1 to its binding partner is likely to be exquisitely dependent on the TAPAS1-PA interaction and thus is readily disrupted by detergents.
We have previously determined (Smith et al., 1996) the NMR structure of the unique N-terminal region of PDE4A1, which reveals that the negatively charged Ca2+-binding Asp21 is found on the aqueous-facing surface of helix-2 where it juxtaposes the membrane insertion module formed by the W19-W20 located on the opposite face of helix-2 (Fig. 7) (Baillie et al., 2002). However, the Ca2+-interacting Asp5 of helix-1, in contrast to Asp21 of helix-2, lies on the membrane-facing surface of helix-2 (Fig. 7). Such an orientation of the negatively charged Asp5 is likely to prevent any possible membrane-insertion of helix-1 concomitant with membrane insertion of the W19-W20 in helix-2. Indeed, coupled with the mobile hinge region, the presence of acidic Asp5 at this surface is likely to orientate helix-1 away from the membrane surface. This may be exacerbated further through electrostatic repulsion in the presence of head groups of negatively charged phospholipids until, presumably, it is negated by the binding of Ca2+ to Asp5 (see below).
Intriguingly, the key residues of helix-1 that are crucial for the fidelity of localisation of (1-25)-4A1-GFP (Fig. 4C), namely Phe6, Phe7 and Leu3, form a compact hydrophobic pocket (Fig. 7). This hydrophobic pocket in helix-1 is located on the opposite face of the helix to both Asp5 and the established bilayer insertion unit located in helix-2 (Fig. 7), making it unlikely that they integrate into the membrane bilayer. On this basis we suggest that the amino acids forming this hydrophobic pocket in helix-1 may confer fidelity of targeting of PDE4A1 by interaction with a Golgi-located docking protein rather than by membrane insertion.
That (1-25)-4A1-GFP was found to be co-localised with the trans-Golgi marker TGN46 is particularly interesting. The trans-Golgi or late Golgi functions to sort proteins, lipids and membranes of the secretory pathway into distinct carriers to be delivered to the plasma membrane and endosomes (Gleeson et al., 2004). It is interesting to speculate that the presence of PDE4A1 at the TGN may have functional relevance in regulating localised cAMP gradients there.
In proposing that the membrane insertion of TAPAS-1 in helix-2 is required to allow helix-1 to perform its trans-Golgi targeting role, this suggests that helix-1 and helix-2 need to be connected and orientated appropriately. Consistent with this, deletion of the mobile hinge region that links these two helices clearly compromises the fidelity of Golgi targeting of (1-25)-4A1-GFP (Fig. 3B). Such a hinge-deficient construct was found associated with various membrane structures including mitochondria (Fig. 3B). This hinge region is highly flexible (Smith et al., 1996) and so its functional role may be in allowing helix-1 and helix-2 to align such that TAPAS-1 can insert appropriately into the bilayer and then allow helix-1 to interact with a docking protein so as to enhance the efficiency of membrane association through helix-2 action and to confer localisation to the trans-Golgi stack.
PDE4A1 localises in a PLD-dependent manner to structures outside the Golgi following release of Ca2+ from intracellular stores
In the TAPAS-1 domain, it is the binding of Ca2+ to Asp21 in the LVGWWD motif that allows membrane insertion of the tryptophan pairing (Baillie et al., 2002). Like the Asp21 residue within the LVGWWD motif of TAPAS-1 we observe that, in helix-1, the key Asp5 residue is positioned so as to juxtapose the hydrophobic pocket formed by Leu3-Phe5-Phe6 (Fig. 7). In the LVDFF motif of helix-1, the mutation of Asp5, although not affecting the ability of (1-25)-4A1-GFP to associate with membranes, caused a profound change in the intracellular distribution of this chimera (Fig. 4A). Thus Asp5Ala(1-25)-4A1-GFP, unlike the wild-type form, was not confined to the trans-Golgi stack, but now distributed throughout the various Golgi stacks and into punctate structures found throughout the cytosol (Fig. 4A). Given that mutation of Asp21 to Ala negates the requirement for Ca2+ to achieve membrane association of TAPAS1 in helix-2 (Baillie et al., 2002), we surmised that elevated Ca2+ might also interact with Asp5 and trigger a similar relocalisation of (1-25)-4A1-GFP, as seen with the Asp5Ala mutant. This was seemingly the case, as elevating intracellular Ca2+, using thapsigargin, conferred on (1-25)-4A1-GFP a similar localisation to that shown by Asp5Ala(1-25)-4A1-GFP in resting cells. Our proposal is that `neutralisation' of Asp5 by either mutation to an uncharged amino acid or by interaction with Ca2+ ablates the trans-Golgi retention signal while retaining a membrane association signal.
A number of predictions arise from such a proposal. One is that amino acids in helix-1 other than Asp5 are involved in membrane association. Indeed, this is likely to be the case as the functioning of helix-1 to confer membrane association of (1-25)-4A1-GFP in living cells is clearly compromised upon alanine mutation of any of Leu3, Phe6 and Phe7 (Fig. 4C). The second is that ablation of the trans-Golgi retention signal may now allow the preference of TAPAS-1 for interaction with PA to dominate the intracellular distribution of (1-25)-4A1-GFP. Indeed, this appears likely to be the case as the redistribution of wild-type (1-25)-4A1-GFP subsequent to thapsigargin treatment, as well as the punctate distribution of the D5A(1-25)-4A1GFP mutant, are both interrupted (Figs 5,6) upon treatment with butan-1-ol, which prevents the PLD-mediated generation of PA (Exton, 2002), or calphostin-c (Sciorra et al., 2001). However, the distribution of the dual helix-2 chimera is somewhat different from that of the Asp5Ala(1-25)-4A1-GFP, indicating that even though Golgi retention is aborted, residues in helix-1 are able to influence intracellular distribution of PDE4A1 in addition to any purported influence of PA exerted through helix-2. In performing these studies we also ascertained that full-length PDE4A1 was able to relocalise from the Golgi to punctate structures throughout the cytosol following treatment with thapsigargin (Fig. 6). This indicates, as shown previously (Baillie et al., 2002), that the (1-25)-4A1-GFP construct is a true indicator of PDE4A1 distribution. Thus we uncover a novel regulatory process where an elevation in the intracellular levels of Ca2+ would serve first, through interaction with Asp21 in helix-2, to cause the irreversible membrane association of PDE4A1, providing a `long-term' memory of Ca2+ action. Subsequent changes in the intracellular levels of Ca2+ would, by interacting with Asp5 in helix-1, trigger dynamic changes in the intracellular distribution of PDE4A1 that are determined by PLD activity, providing `short-term' memory of Ca2+ action.
Our studies attribute distinct functional significance to each of the helical regions identified in the unique N-terminal region of PDE4A1. These are a Ca2+-gated membrane insertion domain (TAPAS-1) within helix-2 and a trans-Golgi retention function associated with helix-1, which can be released upon elevation of intracellular Ca2+. Although the functional role of such redistribution poses a major challenge to be addressed we provide here, for the first time, evidence for the dynamic redistribution of a cAMP-specific phosphodiesterase in response to elevated Ca2+ levels in the cell. As such we identify a potential novel point of cross-talk between the cAMP, Ca2+ and PLD signalling systems.
|
Materials and Methods |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
Constructs
Site-directed mutagenesis was performed using a QuikChange DNA mutagenesis kit according to the manufacturer's instructions (Stratagene, CA). All mutagenesis and deletion constructs were confirmed by DNA sequencing.
Full-length PDE4A1 (GenBankTM accession number M26715) (Davis et al., 1989) was used. This was cloned into the pcDNA3 vector for COS1 cell expression studies.
The N-terminal 25 amino acids of PDE4A1 (RD1) were cloned as an in-frame fusion with GFP in the vector pEGFP-N1, via KpnI-BamHI sites to create NT-4A1-GFP. The methionine residue at the start of eGFP was mutated to alanine in order to prevent any `false' initiation at this point leading to the expression of eGFP alone. This modified construct was used as the template for all mutations and deletions in NT-4A1-GFP.
Transfection of COS1 cells
COS1 cells were maintained and transfected essentially as described previously (Huston et al., 1996). COS1 cells were plated out onto coverslips (21 mm diameter) at 30% confluency, 24 hours before transfection with the required plasmid DNA. Cells were transfected with 1 µg of plasmid DNA using 5 µl FuGENE 6 transfection reagent.
Microscopy
Cells were visualised using a Zeiss Pascal laser-scanning confocal microscope and an Axiovert 100 microscope with a 63x /1.4 NA plan apochromat lens, as described (Huston et al., 1996). Where quantification was to be carried out, a Nikon Diaphot inverted microscope was used equipped with a 40x oil-immersion Fluor objective lens (1.3 NA). A monochromator (Optoscan) was used to excite cells and fluorescence emission was detected by a cooled digital charge-coupled device camera (Cool Snap-HQ; Photometrics, Tucson, AZ), as described previously by others (Pediani et al., 2005). Cells were analysed 24-48 hours after transfection. In some instances, as indicated, cells were fixed for 10 minutes in TBS containing 4% paraformaldehyde and 5% sucrose. They were then permeabilised by performing three 5-minute incubations with 200 µl of 0.2% Triton X-100 in Tris-buffered saline (TBS; 150 mM NaCl, 20 mM Tris-HCl, pH 7.4), followed by a blocking step comprising three 5-minute incubations with blocking buffer (10% appropriate serum and 2% BSA). Primary antibodies were diluted to the required concentration in dilutant buffer (blocking buffer diluted 1:1 with TBS) and 200 µl of this mix was added to the slide and incubated at 20°C for 2 hours. Slides were then washed with three changes of blocking buffer and incubated with 200 µl of 1:400 diluted Alexa Fluor conjugated IgG for 1 hour. Cells transfected with full-length PDE4A1 were labeled for 2 hours with antibodies raised against specific peptide sequences of the C-terminal region of PDE4A1 as described (Huston et al., 2000; Shakur et al., 1995). Where co-localisation of organelles was analysed, the appropriate antibody was used, as indicated, and detected with Alexa Fluor 594 for 1 hour. All incubations were carried out at room temperature. Where Mitotracker, ERtracker or Lysotracker were used, these were added to living cells at a concentration of 50 nM for 15 minutes before fresh medium was added and the cells observed.
Image analysis of co-localisation
To analyse the degree of co-localisation between proteins of interest and various subcellular markers, MetaMorph imaging software (MetaMorph 6.2.6 software, Universal Imaging Corporation) was used. A region lacking fluorescence adjacent to the cell was used to determine the average background level of fluorescence in the red and green channels. The background amount was then subtracted from each pixel in each channel. Cells were chosen at random from the transfected population of cells.
Where correlation plots were generated, a region of interest was manually drawn around positive staining for the BODIPY® Fl C5 Golgi marker (BODIPY ceramide), which was then selected and transferred to the matched image acquired in the GFP channel. Correlation coefficients were evaluated by comparing the amounts of fluorescence measured in each matched pixel of the two different channels using the MetaMorph `correlation plot' application. Where pixel-by-pixel evaluation of co-localisation within the whole cell was to be analysed, MetaMorph imaging software was used to highlight a region of interest, this time around the whole cell in the GFP channel. This region was then transferred to the red channel and individual cells extracted. The threshold pixel intensity tool of Metamorph was then used to pick out areas of fluorescence on both the GFP and red image. The corresponding images were compared using the co-localisation tool of Metamorph and expressed as a percentage of overlapping fluorescence. All quantification was carried out using cells from at least three independent transfections.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopSummaryIntroductionResultsDiscussionMaterials and MethodsReferences |
|---|
Baillie, G. S. and Houslay, M. D. (2005). Arrestin times for compartmentalised cAMP signalling and phosphodiesterase-4 enzymes. Curr. Opin. Cell Biol. 17, 129-134.[CrossRef][Medline]
Baillie, G. S., Huston, E., Scotland, G., Hodgkin, M., Gall, I., Peden, A. H., MacKenzie, C., Houslay, E. S., Currie, R., Pettitt, T. R. et al. (2002). TAPAS-1, a novel microdomain within the unique N-terminal region of the PDE4A1 cAMP-specific phosphodiesterase that allows rapid, Ca2+-triggered membrane association with selectivity for interaction with phosphatidic acid. J. Biol. Chem. 277, 28298-28309.
Baillie, G. S., Sood, A., McPhee, I., Gall, I., Perry, S. J., Lefkowitz, R. J. and Houslay, M. D. (2003). beta-Arrestin-mediated PDE4 cAMP phosphodiesterase recruitment regulates beta-adrenoceptor switching from Gs to Gi. Proc. Natl. Acad. Sci. USA 100, 940-945.
Baillie, G. S., Scott, J. D. and Houslay, M. D. (2005). Compartmentalisation of phosphodiesterases and protein kinase A: opposites attract. FEBS Lett. 579, 3264-3270.[CrossRef][Medline]
Beard, M. B., O'Connell, J. C., Bolger, G. B. and Houslay, M. D. (1999). The unique N-terminal domain of the cAMP phosphodiesterase PDE4D4 allows for interaction with specific SH3 domains. FEBS Lett. 460, 173-177.[CrossRef][Medline]
Bolger, G. B., McCahill, A., Huston, E., Cheung, Y. F., McSorley, T., Baillie, G. S. and Houslay, M. D. (2003). The unique amino-terminal region of the PDE4D5 cAMP phosphodiesterase isoform confers preferential interaction with beta-arrestins. J. Biol. Chem. 278, 49230-49238.
Brown, F. D., Thompson, N., Saqib, K. M., Clark, J. M., Powner, D., Thompson, N. T., Solari, R. and Wakelam, M. J. (1998). Phospholipase D1 localises to secretory granules and lysosomes and is plasma-membrane translocated on cellular stimulation. Curr. Biol. 8, 835-838.[CrossRef][Medline]
Colley, W. C., Sung, T. C., Roll, R., Jenco, J., Hammond, S. M., Altshuller, Y., Bar-Sagi, D., Morris, A. J. and Frohman, M. A. (1997). Phospholipase D2, a distinct phospholipase D isoform with novel regulatory properties that provokes cytoskeletal reorganization. Curr. Biol. 7, 191-201.[CrossRef][Medline]
Cooper, D. M. (2003). Regulation and organization of adenylyl cyclases and cAMP. Biochem. J. 375, 517-529.[CrossRef][Medline]
Davis, R. L., Takayasu, H., Eberwine, M. and Myres, J. (1989). Cloning and characterization of mammalian homologs of the Drosophila dunce+ gene. Proc. Natl. Acad. Sci. USA 86, 3604-3608.
Delon, C., Manifava, M., Wood, E., Thompson, D., Krugmann, S., Pyne, S. and Ktistakis, N. T. (2004). Sphingosine kinase 1 is an intracellular effector of phosphatidic acid. J. Biol. Chem. 279, 44763-44774.
Dodge, K. L., Khouangsathiene, S., Kapiloff, M. S., Mouton, R., Hill, E. V., Houslay, M. D., Langeberg, L. K. and Scott, J. D. (2001). mAKAP assembles a protein kinase A/PDE4 phosphodiesterase cAMP signaling module. EMBO J. 20, 1921-1930.[CrossRef][Medline]
Dumaz, N. and Marais, R. (2005). Integrating signals between cAMP and the RAS/RAF/MEK/ERK signalling pathways. Based on the anniversary prize of the Gesellschaft fur Biochemie und Molekularbiologie Lecture delivered on 5 July 2003 at the Special FEBS Meeting in Brussels. FEBS J. 272, 3491-3504.[CrossRef][Medline]
Exton, J. H. (2002). Regulation of phospholipase D. FEBS Lett. 531, 58-61.[CrossRef][Medline]
Fang, Y., Vilella-Bach, M., Bachmann, R., Flanigan, A. and Chen, J. (2001). Phosphatidic acid-mediated mitogenic activation of mTOR signaling. Science 294, 1942-1945.
Freyberg, Z., Bourgoin, S. and Shields, D. (2002). Phospholipase D2 is localized to the rims of the Golgi apparatus in mammalian cells. Mol. Biol. Cell 13, 3930-3942.
Freyberg, Z., Siddhanta, A. and Shields, D. (2003). "Slip, sliding away": phospholipase D and the Golgi apparatus. Trends Cell Biol. 13, 540-546.[CrossRef][Medline]
Ghosh, S., Strum, J. C., Sciorra, V. A., Daniel, L. and Bell, R. M. (1996). Raf-1 kinase possesses distinct binding domains for phosphatidylserine and phosphatidic acid. Phosphatidic acid regulates the translocation of Raf-1 in 12-O-tetradecanoylphorbol-13-acetate-stimulated Madin-Darby canine kidney cells. J. Biol. Chem. 271, 8472-8480.
Gleeson, P. A., Lock, J. G., Luke, M. R. and Stow, J. L. (2004). Domains of the TGN: coats, tethers and G proteins. Traffic 5, 315-326.[CrossRef][Medline]
Goodwin, J. S., Drake, K. R., Rogers, C., Wright, L., Lippincott-Schwartz, J., Philips, M. R. and Kenworthy, A. K. (2005). Depalmitoylated Ras traffics to and from the Golgi complex via a nonvesicular pathway. J. Cell Biol. 170, 261-272.
Grange, M., Sette, C., Cuomo, M., Conti, M., Lagarde, M., Prigent, A. F. and Nemoz, G. (2000). The cAMP-specific phosphodiesterase PDE4D3 is regulated by phosphatidic acid binding. Consequences for cAMP signaling pathway and characterization of a phosphatidic acid binding site. J. Biol. Chem. 275, 33379-33387.
Hammond, S. M., Altshuller, Y. M., Sung, T. C., Rudge, S. A., Rose, K., Engebrecht, J., Morris, A. J. and Frohman, M. A. (1995). Human ADP-ribosylation factor-activated phosphatidylcholine-specific phospholipase D defines a new and highly conserved gene family. J. Biol. Chem. 270, 29640-29643.
Hammond, S. M., Jenco, J. M., Nakashima, S., Cadwallader, K., Gu, Q., Cook, S., Nozawa, Y., Prestwich, G. D., Frohman, M. A. and Morris, A. J. (1997). Characterization of two alternately spliced forms of phospholipase D1. Activation of the purified enzymes by phosphatidylinositol 4,5-bisphosphate, ADP-ribosylation factor, and Rho family monomeric GTP-binding proteins and protein kinase C-alpha. J. Biol. Chem. 272, 3860-3868.
Hancock, J. F., Paterson, H. and Marshall, C. J. (1990). A polybasic domain or palmitoylation is required in addition to the CAAX motif to localize p21ras to the plasma membrane. Cell 63, 133-139.[CrossRef][Medline]
Houslay, M. D. and Kolch, W. (2000). Cell-type specific integration of cross-talk between extracellular signal-regulated kinase and cAMP signaling. Mol. Pharmacol. 58, 659-668.
Houslay, M. D. and Adams, D. R. (2003). PDE4 cAMP phosphodiesterases: modular enzymes that orchestrate signalling cross-talk, desensitization and compartmentalization. Biochem. J. 370, 1-18.[CrossRef][Medline]
Houslay, M. D., Schafer, P. and Zhang, K. Y. (2005). Keynote review: phosphodiesterase-4 as a therapeutic target. Drug Discov. Today 10, 1503-1519.[CrossRef][Medline]
Huston, E., Pooley, L., Julien, P., Scotland, G., McPhee, I., Sullivan, M., Bolger, G. and Houslay, M. D. (1996). The human cyclic AMP-specific phosphodiesterase PDE-46 (HSPDE4A4B) expressed in transfected COS7 cells occurs as both particulate and cytosolic species that exhibit distinct kinetics of inhibition by the antidepressant rolipram. J. Biol. Chem. 271, 31334-31344.
Huston, E., Beard, M., McCallum, F., Pyne, N. J., Vandenabeele, P., Scotland, G. and Houslay, M. D. (2000). The cAMP-specific phosphodiesterase PDE4A5 is cleaved downstream of its SH3 interaction domain by caspase-3. Consequences for altered intracellular distribution. J. Biol. Chem. 275, 28063-28074.
Karathanassis, D., Stahelin, R. V., Bravo, J., Perisic, O., Pacold, C. M., Cho, W. and Williams, R. L. (2002). Binding of the PX domain of p47(phox) to phosphatidylinositol 3,4-bisphosphate and phosphatidic acid is masked by an intramolecular interaction. EMBO J. 21, 5057-5068.[CrossRef][Medline]
Manifava, M., Thuring, J. W., Lim, Z. Y., Packman, L., Holmes, A. B. and Ktistakis, N. T. (2001). Differential binding of traffic-related proteins to phosphatidic acid- or phosphatidylinositol (4,5)-bisphosphate-coupled affinity reagents. J. Biol. Chem. 276, 8987-8994.
McPhee, I., Yarwood, S. J., Scotland, G., Huston, E., Beard, M. B., Ross, A. H., Houslay, E. S. and Houslay, M. D. (1999). Association with the SRC family tyrosyl kinase LYN triggers a conformational change in the catalytic region of human cAMP-specific phosphodiesterase HSPDE4A4B. Consequences for rolipram inhibition. J. Biol. Chem. 274, 11796-11810.
McPhee, I., Cochran, S. and Houslay, M. D. (2001). The novel long PDE4A10 cyclic AMP phosphodiesterase shows a pattern of expression within brain that is distinct from the long PDE4A5 and short PDE4A1 isoforms. Cell. Signal. 13, 911-918.[CrossRef][Medline]
Meier, K. E., Gibbs, T. C., Knoepp, S. M. and Ella, K. M. (1999). Expression of phospholipase D isoforms in mammalian cells. Biochim. Biophys. Acta 1439, 199-213.[Medline]
Millar, J. K., Pickard, B. S., Mackie, S., James, R., Christie, S., Buchanan, S. R., Malloy, M. P., Chubb, J. E., Huston, E., Baillie, G. S. et al. (2005). DISC1 and PDE4B are interacting genetic factors in schizophrenia that regulate cAMP signaling. Science 310, 1187-1191.
Ouyang, Y. S., Tu, Y., Barker, S. A. and Yang, F. (2003). Regulators of G-protein signaling (RGS) 4, insertion into model membranes and inhibition of activity by phosphatidic acid. J. Biol. Chem. 278, 11115-11122.
Pediani, J. D., Colston, J. F., Caldwell, D., Milligan, G., Daly, C. J. and McGrath, J. C. (2005). Beta-arrestin-dependent spontaneous alpha1a-adrenoceptor endocytosis causes intracellular transportation of alpha-blockers via recycling compartments. Mol. Pharmacol. 67, 992-1004.
Perry, S. J., Baillie, G. S., Kohout, T. A., McPhee, I., Magiera, M. M., Ang, K. L., Miller, W. E., McLean, A. J., Conti, M., Houslay, M. D. et al. (2002). Targeting of cyclic AMP degradation to beta 2-adrenergic receptors by beta-arrestins. Science 298, 834-836.
Pooley, L., Shakur, Y., Rena, G. and Houslay, M. D. (1997). Intracellular localization of the PDE4A cAMP-specific phosphodiesterase splice variant RD1 (RNPDE4A1A) in stably transfected human thyroid carcinoma FTC cell lines. Biochem. J. 321, 177-185.[Medline]
Rizzo, M. A., Shome, K., Watkins, S. C. and Romero, G. (2000). The recruitment of Raf-1 to membranes is mediated by direct interaction with phosphatidic acid and is independent of association with Ras. J. Biol. Chem. 275, 23911-23918.
Sciorra, V. A., Hammond, S. M. and Morris, A. J. (2001). Potent direct inhibition of mammalian phospholipase D isoenzymes by calphostin-c. Biochemistry 40, 2640-2646.[CrossRef][Medline]
Scotland, G. and Houslay, M. D. (1995). Chimeric constructs show that the unique N-terminal domain of the cyclic AMP phosphodiesterase RD1 (RNPDE4A1A; rPDE-IVA1) can confer membrane association upon the normally cytosolic protein chloramphenicol acetyltransferase. Biochem. J. 308, 673-681.[Medline]
Shakur, Y., Pryde, J. G. and Houslay, M. D. (1993). Engineered deletion of the unique N-terminal domain of the cyclic AMP-specific phosphodiesterase RD1 prevents plasma membrane association and the attainment of enhanced thermostability without altering its sensitivity to inhibition by rolipram. Biochem. J. 292, 677-686.[Medline]
Shakur, Y., Wilson, M., Pooley, L., Lobban, M., Griffiths, S. L., Campbell, A. M., Beattie, J., Daly, C. and Houslay, M. D. (1995). Identification and characterization of the type-IVA cyclic AMP-specific phosphodiesterase RD1 as a membrane-bound protein expressed in cerebellum. Biochem. J. 306, 801-809.[Medline]
Smith, K. J., Scotland, G., Beattie, J., Trayer, I. P. and Houslay, M. D. (1996). Determination of the structure of the N-terminal splice region of the cyclic AMP-specific phosphodiesterase RD1 (RNPDE4A1) by 1H NMR and identification of the membrane association domain using chimeric constructs. J. Biol. Chem. 271, 16703-16711.
Steed, P. M., Clark, K. L., Boyar, W. C. and Lasala, D. J. (1998). Characterization of human PLD2 and the analysis of PLD isoform splice variants. FASEB J. 12, 1309-1317.
Tasken, K. and Aandahl, E. M. (2004). Localized effects of cAMP mediated by distinct routes of protein kinase A. Physiol. Rev. 84, 137-167.
Tasken, K. A., Collas, P., Kemmner, W. A., Witczak, O., Conti, M. and Tasken, K. (2001). Phosphodiesterase 4D and protein kinase a type II constitute a signaling unit in the centrosomal area. J. Biol. Chem. 276, 21999-22002.
Verde, I., Pahlke, G., Salanova, M., Zhang, G., Wang, S., Coletti, D., Onuffer, J., Jin, S. L. and Conti, M. (2001). Myomegalin is a novel protein of the golgi/centrosome that interacts with a cyclic nucleotide phosphodiesterase. J. Biol. Chem. 276, 11189-11198.
Wong, W. and Scott, J. D. (2004). AKAP signalling complexes: focal points in space and time. Nat. Rev. Mol. Cell Biol. 5, 959-970.[CrossRef][Medline]
Xu, Y., Harder, K. W., Huntington, N. D., Hibbs, M. L. and Tarlinton, D. M. (2005). Lyn tyrosine kinase: accentuating the positive and the negative. Immunity 22, 9-18.[Medline]
Yarwood, S. J., Steele, M. R., Scotland, G., Houslay, M. D. and Bolger, G. B. (1999). The RACK1 signaling scaffold protein selectively interacts with the cAMP-specific phosphodiesterase PDE4D5 isoform. J. Biol. Chem. 274, 14909-14917.
Zaccolo, M., Magalhaes, P. and Pozzan, T. (2002). Compartmentalisation of cAMP and Ca(2+) signals. Curr. Opin. Cell Biol. 14, 160-166.[CrossRef][Medline]
Zhang, F. L. and Casey, P. J. (1996). Protein prenylation: molecular mechanisms and functional consequences. Annu. Rev. Biochem. 65, 241-269.[CrossRef][Medline]
This article has been cited by other articles:
|
|
