Adenylyl Cyclase Isoform–Selective Regulation of Vascular Smooth Muscle Proliferation and Cytoskeletal Reorganization
Compartmentation of cAMP signaling been demonstrated to be attributable to the structural association of protein kinase A (PKA) (via association with A-kinase anchoring proteins [AKAPs]) with phosphodiesterase and AKAP-dependent effector molecules. However, other mechanisms contributing to compartmentalization have not been rigorously explored, including the possibility that different isoforms of adenylyl cyclase (AC) may be functionally “compartmentalized” because of differential association with tethering or signaling molecules. To this end, we examined the effect of adenoviral transduction of representative AC isoforms (AC1, AC2, AC5, and AC6) on cellular cAMP production, PKA activation, extracellular signal-regulated kinase (ERK) activation, cell doubling and proliferation, as well as arborization responses (an index of cAMP-mediated cytoskeletal re-organization) in vascular smooth muscle cells. When isoforms were expressed at levels to achieve comparable forskolin-stimulated AC activity, only gene transfer of AC6 significantly enhanced PKA-dependent vasodilator-stimulated phosphoprotein (VASP) phosphorylation and arborization responses. Treatment of control cells, which express AC6 endogenously, as well as vascular smooth overexpressing the AC6 isoform with small interfering RNA directed against AC6, significantly suppressed both isoproterenol-stimulated cAMP accumulation and arborization. Notably, the selective effects of AC6 expression were abrogated in the presence of phosphodiesterase suppression. In contrast, only the expression of AC1 enhanced forskolin-stimulated association of ERK with AC, demonstrated by coimmuno-isolation of ERK with Flag-tagged AC1, but not with Flag-tagged AC6. To determine whether these isoform-selective effects of AC were unique to differentiated and morphologically compartmentalized vascular smooth muscle cells or were a general property of these isoforms, we examined the consequence of expression of these various isoforms in human embryonic kidney (HEK) cells. Indeed, we observed similar isoform-dependent association of AC1 with ERK, activation of ERK by stimulation of AC1 with forskolin, and AC1-dependent lengthening of doubling time, indicating that these properties of AC1 are cell autologous and likely result from AC1-dependent protein-protein interactions. In aggregate, these findings suggest that isoform-selective signaling complexes likely contribute to various functional consequences of cAMP elevation in vascular smooth muscle cells.
Cyclic AMP is the most ubiquitous second messenger used from single-cell organisms to mammalian cells and tissues. cAMP synthesis is catalyzed by the enzyme adenylyl cyclase (AC), which has been shown to be encoded by several genes. Nine membrane-bound isoforms of AC have been cloned and grouped into 3 major subfamilies comprising1: group 1 (AC1, AC3, and AC8); group 2 (AC2, AC4, and AC7); and group 3 (AC5 and AC6). Further, AC9 has been characterized as a distinct (and somewhat atypical) isoform.2 Additionally, a soluble AC, the predominant form in mammalian sperm, has been characterized.3 Each isoform has a specific pattern of tissue/organ distribution and a specific pattern of regulation by G proteins, calcium/calmodulin, and protein kinases.4 For example, differences in patterns of regulation by G protein subunits have been associated with isoform-specific differences in AC activation.5,6 Whereas AC1, AC5, and AC6 are inhibited by Gi, AC2 is not. Furthermore, whereas Gβγ subunits inhibit isoforms AC1, AC5, and AC6, they stimulate isoforms AC2 and AC4. Similarly, calcium inhibits the activation of isoforms AC5 and AC6 but activates isoforms AC1, AC2, and AC8.2
Variability in the synthesis of cAMP by AC is thought to be predominantly regulated either by the specific characteristics of the G protein–coupled receptors (GPCRs) linked to enzyme activation or by the catalytic and regulatory properties of the specific isoforms of AC expressed in any individual target cell. To date, differential effects of the product of AC, cAMP, are thought to be attributable to microcompartmentation via A-kinase anchoring proteins (AKAPs)7 and the particular phosphodiesterase (PDE) isoform that hydrolyzes the cAMP.8 However, whether different AC isoforms might have differential effects on cAMP-mediated responses independent of their rate of synthesis of cAMP has not previously been tested.
We examined the impact of introduction of AC isoforms AC1 (representative of the AC1, AC3, and AC8 subfamily), AC2 (representative of the AC2, AC4, and AC7 subfamily) and AC5, AC6 (the 2 members of the subfamily) on cAMP production, accumulation, and cAMP-dependent functions in vascular smooth muscle cells. The present studies demonstrate that there are significant differences in the regulation of vascular function mediated by AC1 versus the AC6 isoforms, with little impact of either the AC2 or AC5 isoform on any of the responses noted. These data support previously unappreciated roles for specific AC isoforms in directing the effect of elevation of intracellular cAMP in the regulation of vascular contractility and cellular proliferation.
Materials and Methods
Construction of Adenoviral Constructs Expressing AC and AC Isoform–Specific Small Interfering RNA
cDNAs encoding Flag-tagged AC1 and AC3 (representatives of the AC1, AC3, and AC8 subfamily), AC2 (representative of the AC2, AC4, and AC7 subfamily) and AC5, AC6 (the 2 members the subfamily) or a control cDNA encoding green fluorescence protein (GFP) were used as previously described9 to generate adenoviral constructs (AdMax) as per the instructions of the manufacturer (Microbix Biosystems Inc, Toronto, Canada).
AC Isoform–Specific Small Interfering RNAs
To generate adenoviral constructs encoding AC isoform-specific small interfering RNAs (siRNAs), we used previously published techniques.10 Briefly, a modified CMV promoter was generated by PCR using the peGFPN1 template plasmid (BD Biosciences Clontech, Mississauga, Ontario, Canada). The modified CMV product was cloned into the XbaI and HindIII sites of adenoviral shuttle vector PDC312 (Microbix Biosystems Inc). The minimal poly A was synthesized and ligated into the BglII and EcoRI sites of PDC312CMV. The resultant shuttle plasmid was used for construction of head-to-head hairpins of eGFP (bp 418 to 438), AC6 (bp 3127 to 47), or AC2 (bp 356 to 376). Shuttle plasmids were cotransfected into HEK293 cells along with adenovirus backbones for generation of adenoviral genomes. Virus was then harvested 2 weeks following transfection. The specificity for each siRNA of AC isoforms were determined by Western blotting and AC activity assays.
Vascular Smooth Muscle Cell Primary Cultures
Rat aortic vascular smooth muscle cell primary cultures were isolated by a modification of the methods of Touyz et al.11 Briefly, freshly isolated thoracic aortae from twelve male Wistar rats 10 to 12 weeks of age (Harlan, Indianapolis, Ind) were digested using collagenase and elastase incubations as previously described.11 Following digestion/isolation, vascular smooth muscle cells were resuspended in DMEM supplemented with 10% FBS, gentamicin, and Fungizone. Vascular smooth muscle cells were used between passages 4 to 12 for all experiments. We used both early and late passages of cells for all assays to negate the possibility of differential effects reflected because of passage-specific differences in vascular smooth muscle cells. The rats were cared for in accordance with the Canadian Council on Animal Care guidelines.
Gene Transfer in Vascular Smooth Muscle Cells by Adenovirus
Vascular smooth muscle cells were infected with adenoviral constructs for 16 hours at 37°C, following which, infection media was replaced with fresh DMEM culture media. Cells were used for experimentation 48 to 72 hours postinfection. Under these conditions, infection efficiency as assessed in GFP-infected cells was greater than 95%.
Cell Culture of HEK293 Cells and Their Transfection
HEK293 cells were grown at 37°C, 5% CO2 in minimum essential medium (MEM) (GIBCO, Grand Island, NY) containing 10% FBS. HEK293 cells were transiently transfected with pcDNA3 containing either GFP or Flag-tagged AC1 (as representative of the AC1, AC3, AC8 subfamily), AC2 (as representative of the AC2, AC4, AC7 subfamily), and AC5 or AC6 cDNA using calcium phosphate as previously described.9,12 Cells were harvested 48 to 72 hours after transfection for assessment of biological functions.
Immunoprecipitation of AC Isoforms and Western Blotting for AC and/or for Extracellular Signal-Regulated Kinase
Infected vascular smooth muscle cells or transiently transfected HEK293 cells were lysed in buffer A (containing 20 mmol/L Tris, pH 8.0, 1% NP-40, 0.1% SDS [sodium dodecyl sulfate], 140 mmol/L NaCl, 1 mmol/L phenylmethanesulfonyl fluoride). For immunoprecipitation experiments, 500 μg of cell lysates were incubated with 15 μL of anti-Flag beads (Sigma) and incubated overnight at 4°C.
For AC/extracellular signal-regulated kinase (ERK) interaction experiments, whole cell extracts were subjected to immunoprecipitation with either anti-Flag M2 (Sigma, St Louis, Mo) or anti-ERK antibodies. Indistinguishable results were obtained using these 2 strategies, but we report only our findings using the anti-Flag M2 antibody.
The resulting immunoprecipitation complexes (or whole cell lysates to provide estimates of total protein content for proteins of interest) were resolved on 8% SDS-PAGE and electrophoretically blotted onto Immun-Blot polyvinylidene fluoride (PVDF) membrane (Bio-Rad, Hercules, Calif). The membranes were blocked with 5% skim milk for 1 hour at room temperature, then incubated either with an anti-AC polyclonal antibody (1:5000) that demonstrates cross-reactivity with all AC isoforms studied (the epitope was generated against a 14 amino acid peptide of the C-terminal region of AC, common to cloned AC isoforms13) or with an anti-ERK1/2 (1:1000 from Upstate) or anti-ERK5 (1:200 from Santa Cruz Biotechnology, Santa Cruz, Calif) antibody.
Blots were washed 3 times with Tris-buffered saline and incubated with the appropriate secondary antibodies for 1 hour at room temperature, and immunoreactive bands were detected by chemiluminescence as described by the protocol of the manufacturer (NEN, Boston, Mass).
Assessment of cAMP Production
cAMP accumulation was measured in 2 ways. One method involved examining cAMP synthesis in both vascular smooth muscle cells and HEK293 cells by monitoring the conversion of [α-32P]ATP to [32P]cAMP (following digitonin-dependent permeabilization of the cells) as previously described.9,12 AC activity was assessed in response to GTP (100 μmol/L), GTP plus isoproterenol (100 μmol/L), or forskolin (10 μmol/L). Suppression of PDE in these experiments was achieved by including excess unlabeled cAMP in our permeabilization medium, as described previously.9,12
Alternatively, assessment of cAMP production was determined via monitoring the conversion of [3H]ATP to [3H]cAMP with modifications as previously described.14 Briefly, AC isoform-infected smooth muscle cells or HEK293 cells were incubated for 4 hours with [3H]adenine (≈3 μCi/mL; NEN). Cells were washed twice with DMEM and trypsinized. Cells were incubated in the absence or presence of forskolin (100 μmol/L) for 10 minutes at 37°C. The reaction was stopped by the addition of an ice-cold solution containing 2.5% perchloric acid, 100 μmol/L cAMP, and 1500 counts per minute of [14C]-cAMP (NEN). cAMP was isolated by sequential Dowex and alumina chromatography using [14C]-cAMP as an internal standard. When we wished to inhibit intracellular PDE activity in the [3H]adenine-labeling experiments, IBMX (500 μmol/L) was included in the incubation.
Assessment of cAMP-Mediated Vasodilator-Stimulated Phosphoprotein Phosphorylation
AC isoform-infected smooth muscle cells were incubated in the absence or presence of forskolin (100 μmol/L) for 10 minutes, following which cells were washed twice with ice-cold PBS and lysed in buffer A (as described above). The resulting whole cell lysates were resolved on SDS-PAGE and blotted electrophoretically onto an Immun-Blot PVDF membrane (Bio-Rad). Membranes were blocked overnight at 4°C with 5% skim milk and incubated with anti–vasodilator-stimulated phosphoprotein (anti-VASP) polyclonal antibody (1:3000, Calbiochem) overnight at 4°C. Blots were washed in Tris-buffered saline for 1 hour, followed by incubation of secondary anti-rabbit antibody (1:5000, Sigma) for 1 hour at room temperature. Proteins were detected by chemiluminescence as described by the protocol of the manufacturer (NEN).
Arborization of Vascular Smooth Muscle Cells
The arborization response mediated by elevations of cAMP has been linked to cytoskeletal changes including reorganization of actin fibers15,16 and assembly of microtubules.17 The effect of AC activation on vascular smooth muscle cell arborization was assessed as recently described.18 Briefly, vascular smooth muscle cells were cultured onto 35-mm dishes and infected with the various adenoviral constructs as described above. Twenty-four hours postinfection, the media was replaced with DMEM containing 0.1% BSA. Dishes were placed mounted in a temperature-controlled chamber (Bionomic controller, 20/20 Technology Inc) on an inverted microscope (Zeiss, Axiovert S100). Smooth muscle cell arborization was induced by the addition of either forskolin (10 μmol/L) or isoproterenol (10 μmol/L). Progression of arborization was evaluated using time-lapse video microscopy with a digital recording system. Images were obtained every minute and the extent of arborization was determined by the change in image intensity (Northern Eclipse 6.0, Empix Imaging, Toronto, Canada). The change in image intensity was expressed as a percent of basal intensity (before the addition of drug). The change in image intensity was plotted against time and slopes were determined from linear regression analysis using Prism 4.0 (GraphPad Software, San Diego, Calif).
Vascular Smooth Muscle Cell Proliferation Assays
Vascular smooth muscle cells were platted into 24-well plates and infected with various adenoviral constructs as described above. Vascular smooth muscle cells were incubated in the absence or presence of isoproterenol (100 μmol/L) for 21 hours. [3H]Thymidine (1 μCi/mL; ICN Pharmaceuticals) was then added for the remaining 3 hours before the harvest of vascular smooth muscle cells. Following incubation, the medium was aspirated and the cells washed three times with ice-cold PBS and 10% trichloroacetic acid and then distilled water and allowed to air dry. Cells were solubilized with 1 mL of 1% SDS and radioactivity of each sample was determined by liquid scintillation spectrometry.
Assessment of Cell-Doubling Times
Vascular smooth muscle cells or HEK293 cells were plated into 6-well plates and transduced/transfected with AC isoforms or GFP as described above. Cells were harvested by trypsinization at various times following transfection (0, 6, 24, and 48 hours) and cell numbers determined using a hemocytometer. Cell-doubling times were determined from growth curve data using Prism 4.0 (GraphPad Software).
For 2-group comparisons, the statistical significance of differences was determined by Student’s t test for paired data with Welch’s correction. For assessing statistical differences between data expressed in bar graph format, the significance of differences from control was determined by 1-sample t tests. For multiple group comparisons, initial analysis by ANOVA was followed by Dunnett’s multiple comparison tests. P<0.05 on a 2-sided test was taken as a minimum level of significance.
Properties of cAMP Accumulation of the Various AC Isoforms
As depicted in Figure 1, under conditions of comparable AC protein expression (Figure 1A), forskolin-stimulated cAMP production was significantly increased to levels that did not differ between isoforms following the expression of AC1, AC2, AC5, or AC6 when cAMP accumulation was assessed by conversion of [α-32P]ATP to [32P]cAMP in digitonin-permeabilized cells (Figure 1B).
Isoform-Specific Effects on Forskolin-Stimulated Protein Kinase A–Dependent VASP Phosphorylation
To determine whether differing AC isoform expression led to altered cAMP-dependent protein kinase activation, we examined the activation of protein kinase A following forskolin stimulation of AC using an indirect measure of protein kinase A (PKA) activation, ie, the phosphorylation of VASP. In these studies, the viral doses used for AC isoform gene transfer were identical to those used in assessment of AC activity in Figure 1B. Interestingly, only expression of AC6 significantly increased forskolin-stimulated VASP phosphorylation (Figure 1C).
Isoform-Specific Effects of AC Activation on Arborization of Vascular Smooth Muscle Cells
We also examined the impact of the expression of various isoforms of AC on smooth muscle cell arborization, a response that reflects acute cytoskeletal reorganization and has been proposed as an index of vasodilator reactivity.17 The inset in Figure 2A demonstrates that isoproterenol-stimulated elevations of cAMP in control cells leads to readily detectable changes in cell morphology, which are paralleled by changes in image intensity, that are measured as outlined in Materials and Methods and as we have described previously.18 Analogous to observations for VASP phosphorylation, only gene transfer of AC6 resulted in a significant increase in the arborization process as compared with control-infected (or other AC isoform-infected) smooth muscle cells (Figure 2A). Similar changes were observed with forskolin exposure as well (data not shown).
Because the “control” rat vascular smooth muscle cells endogenously express the AC6 isoform, but not the AC1, AC2, AC4, AC7, AC8, or AC9 isoforms,19,20 we used short interfering RNA (siRNA) strategies directed against the AC6 isoform to explore the impact of silencing of AC6 expression on control as well as AC6-overexpressing cells. As shown in Figure 2B, siRNA targeted against AC6 selectively reduces the heterologous expression (detected by Western blots for the Flag-tagged AC introduced by adenoviral transduction) of AC6 without impact on the expression of AC2, indicating the specificity of the siAC6 reagent. Similarly, the siRNA directed against AC6 also reduces forskolin-activated AC activity in control cells (consistent with the previously demonstrated endogenous expression of some AC6 in these cells19), as well as cells overexpressing AC6 because of adenoviral transduction (Figure 2C). Consistent with the data in Figure 2A, implicating AC6 in arborization of smooth muscle cells, siRNA “knockdown” of AC6 diminishes forskolin-induced arborization in control cells and in cells overexpressing the AC6 isoform (Figure 2C).
Effect of PDE Suppression on cAMP Accumulation for AC Isoforms
The apparent selective effect of AC6 on VASP phosphorylation and arborization was hard to reconcile with the comparable production of cAMP by the various AC isoforms, when PDE activity was masked during the incubation period (of Figure 1B). Because PDE activity was not suppressed during our assessment of VASP phosphorylation or arborization, we examined the impact of not suppressing PDE activity on cAMP accumulation and observed that only AC6 expressing smooth muscle cells demonstrated significantly increased levels of cAMP (Figure 3, open bars). This is inconsistent to our findings under conditions of PDE suppression, where comparable levels of cAMP accumulation were observed for all AC isoforms examined (Figure 3, solid bars). These finding suggest that cAMP synthesis by the AC6 isoform is not tightly coupled to cAMP hydrolysis, whereas for the other AC isoforms evaluated cAMP synthesis and hydrolysis are much more tightly linked. These data indicate that VASP phosphorylation and the arborization process in vascular smooth muscle cells require the sustained cAMP accumulation characteristic of the AC6 isoform.
Isoform-Specific Effects of AC Activation on Cell Doubling, Cellular Proliferation, and ERK Association With AC Isoforms
Interestingly, examination of functional changes in vascular smooth muscle cells in response to expression of different AC isoforms revealed that the AC1 isoform contributed to a statistically significant increases in cell doubling times (Figure 4A) and parallel reductions in cell proliferation ([3H]thymidine uptake; Figure 4B). In fact, expression of AC1 alone (Figure 4B), without any parallel stimulation by forskolin (not shown) or isoproterenol, caused a 50% reduction in thymidine incorporation. Notably, the impact of the expression of AC1 is equieffective with the isoproterenol-stimulated inhibition of proliferation (Figure 4B) in control and following expression of other isoforms of AC. In follow-up experiments, we examined whether the effects of AC1 overexpression were unique to AC1 or were shared with a member of the subfamily (ie, AC3) endogenously expressed in vascular smooth muscle cells. Similar to the results obtained with AC1 and AC3 overexpression resulted in a significant reduction in thymidine incorporation (to 56±5% of that in GFP-infected controls, n=3, P<0.05).
The activation of the ERK signaling pathway, and its regulation by cAMP and PKA, has profound effects on growth of any number of target cells. Furthermore, ERK1 and ERK2 have been shown to be associated with other GPCR-related “signalosomes” (eg, in conjunction with GPCRs and β-arrestins).21 Consequently, we examined the effect of expression of various isoforms of AC on the ERK signaling pathway. We were not able to study ERK activity directly on smooth muscle cells because of the inherent variability of the assay for phospho-ERK in these cells following AC isoform gene transfer, but we were able to examine the ability of ERK to associate with AC in our smooth muscle cells following expression of the various AC isoforms. This association has not been described previously, although activation of ERK has been demonstrated to parallel changes in the association of the enzyme with other regulatory molecules.21,22 As shown in Figure 4C, vascular smooth muscle cells expressing AC1, even without hormonal or forskolin stimulation, display a readily detectable association of ERK1/2 with Flag-tagged AC1. Total ERK expression was not altered with AC1 expression (101±15% of GFP-infected control cells, n=3). Similar to the results obtained with AC1, an association of AC3 with ERK1/2 was also readily detected (data not shown).
Effect of Varying AC Isoform Expression on cAMP Production, Cell-Doubling Time, AC/ERK Interactions and ERK Activation in HEK293 Cells
Our data in smooth muscle cells are consistent with the interpretation that the different isoforms of participate in varying protein/protein interactions to create “signalosomes” that represent “functional compartments” within cells, such that the AC1 isoform contributes to modulation of ERK signaling, proliferation, and control of cell division, whereas AC6, at least partly because of uncoupling of cAMP synthesis from cAMP breakdown, results in sustained cAMP accumulation, VASP phosphorylation, and control of cytoskeletal rearrangements that contribute to vascular arborization. Because some of this functional compartmentalization detected in smooth muscle cells could result directly from morphological compartmentalization characteristic of highly differentiated cells, we examined whether AC isoform selectivity in regulation of the ERK/growth regulation pathway might also exist in nondifferentiated cells, such as HEK293 cells. Because HEK293 cells do not undergo arborization, only the growth-related functional properties of various AC isoforms could be assessed in HEK293 cells.
As in our initial studies of vascular smooth muscle cells, we assessed the protein expression and cAMP production mediated by various AC isoforms transiently expressed in HEK293 cells. Comparable protein expression of AC1, AC2, AC5, and AC6 was detected in HEK293 cells using our anti-AC antibody (Figure 5A). As in smooth muscle cells, comparable expression of all AC isoforms studied resulted in comparable increases in cAMP production (Figure 5B). Despite the elevations in cAMP achieved by all of the isoforms studies, only the AC1 isoform showed a dramatic increase in cell-doubling time in response to forskolin-mediated cAMP accumulation, with overexpression of the AC2, AC5, and AC6 isoforms causing no increase in doubling time beyond that characteristic of nontransfected HEK293 cells (Figure 5C).
We wondered whether the expression of differing isoforms of AC might be accompanied by differential effects on ERK activation and the novel finding of ERK association with AC that we had observed in smooth muscle cells (see Figure 4C). As shown in Figure 5D, the AC1 isoform, but not the comparably expressed AC2, AC5, or AC6 isoforms, is associated with ERK1/2 protein in HEK293 cells and a simultaneous activation of ERK1/2 (Figure 5E) with expression of the AC1 isoform as well.
Because ERK5 has been shown to be associated with and regulated by a PKA/AKAP complex,23 we attempted to coimmunoprecipitate ERK5 with the various AC isoforms in both HEK293 cells (with AC isoform transfection) and vascular smooth muscle cells (with AC isoform gene transfer). However, in neither model system could an association be demonstrated between ERK5 and any of the AC isoforms examined (data not shown) despite the ready association of AC1 with ERK1/2. Thus, the ERK/AC interactions appear to be selective both for the AC isoform and for the companion ERK isoform.
It has been widely believed that the “downstream” consequences of cAMP generation were common to all isoforms of AC and was determined to a greater extent by the repertoire of AKAPs and PKA substrates in various target cells. The current studies indicate that even in the setting of comparable abilities to elevate intracellular cAMP levels, specific isoforms of AC play distinct roles in the regulation of cellular proliferation versus contractility in vascular smooth muscle cells.
Perhaps the most striking finding of these studies is the observation that the growth and proliferative responses and changes in ERK phosphorylation are attributable to the AC1 protein and its catalytic activity, whereas the “arborization” response in vascular smooth muscle cells is highly dependent on (and specific to) expression of AC6 and can be diminished in control cells and in AC6 overexpressing cells by siRNA-mediated suppression of AC6 expression. The specific effect of overexpression of AC6 (and not AC5, AC2, or AC1) to enhance arborization responses is consistent with a model of selective coupling of AC6 activation to regulation of cytoskeletal re-organization. This differential effect of AC6 on arborization, like the selective effects of the group 1 subfamily of AC isoforms on ERK association and decreased proliferative response, cannot be attributed to global enhancement of cAMP content in the AC isoform–expressing cells, because the AC isoform gene transfer resulted in comparable increases in forskolin-stimulated cAMP content for all isoforms studied. Thus, the selective effect of group 1 family of isoforms to reduce the rate of cell doubling and to inhibit proliferative responses without any enhanced effect on arborization, whereas, in the same cell type, expression of AC6 increased arborization without effects on growth would suggest either that these 2 isoforms selectively elevate cAMP generation in particular cellular subcompartments or that these AC isoforms interact in selective manners with other downstream effectors/regulators to create different functional subcompartments within these cells. Because the AC6 effect on arborization is paralleled by an AC6-selective uncoupling of cAMP synthesis and breakdown (Figure 1C) and an AC6-selective effect on phosphorylation of the PKA substrate, VASP, the AC 6 isoform may participate in molecular complexes devoid of PDE but enriched in VASP as well as in intermediates in cytoskeletal rearrangements. The fact that the AC1-selective ERK association and reduced rate of proliferation occur not only in highly differentiated cells, like the vascular smooth muscle cells, but also in HEK293 cells argues against morphological compartmentalization but rather functional compartmentalization achieved by selective AC isoform–enriched complexes caused by isoform-selective protein/protein interactions, or signalosomes. The widespread effects of AC1 on these responses may mean that the identification of the molecular partners may be feasible in cells that are easily enriched for AC1 expression, cellular expansion, and biochemical harvesting, such as HEK293 cells.
In summary, these data support previously unappreciated differences in the association of specific isoforms of AC in discrete “functional compartments” that result in isoform-selective regulation of cellular growth versus cytoskeletal organization.
Sources of Funding
These studies were supported by grants-in-aid to R.D.F from the Canadian Institutes of Health Research and Heart and Stroke Foundation of Ontario.
↵*Both authors contributed equally to this work.
Original received June 1, 2005; first resubmission received December 27, 2005; second resubmission received July 12, 2006; third resubmission received August 21, 2006; accepted August 30, 2006.
Harry A, Chen Y, Magnusson R, Iyengar R, Weng G. Differential regulation of adenylyl cyclases by Galphas. J Biol Chem. 1997; 272: 19017–19021.
Thomas JM, Hoffman BB. Isoform-specific sensitization of adenylyl cyclase activity by prior activation of inhibitory receptors: role of beta gamma subunits in transducing enhanced activity of the type VI isoform. Mol Pharmacol. 1996; 49: 907–914.
Dodge-Kafka KL, Langeberg L, Scott JD. Compartmentation of cyclic nucleotide signaling in the heart: the role of A-kinase anchoring proteins. Circ Res. 2006; 98: 993–1001.
Georget M, Mateo P, Vandecasteele G, Lipskaia L, Defer N, Hanoune J, Hoerter J, Lugnier C, Fischmeister R. Cyclic AMP compartmentation due to increased cAMP-phosphodiesterase activity in transgenic mice with a cardiac-directed expression of the human adenylyl cyclase type 8 (AC8). FASEB J. 2003; 17: 1380–1391.
Ding Q, Gros R, Gray ID, Taussig R, Ferguson SS, Feldman RD. Raf kinase activation of adenylyl cyclases: isoform-selective regulation. Mol Pharmacol. 2004; 66: 921–928.
Ding Q, Gros R, Chorazyczewski J, Ferguson SS, Feldman RD. Isoform-specific regulation of adenylyl cyclase function by disruption of membrane trafficking. Mol Pharmacol. 2005; 67: 564–571.
Pieroni JP, Harry A, Chen J, Jacobowitz O, Magnusson RP, Iyengar R. Distinct characteristics of the basal activities of adenylyl cyclases 2 and 6. J Biol Chem. 1995; 270: 21368–21373.
Ben-Ze’ev A, Amsterdam A. In vitro regulation of granulosa cell differentiation. Involvement of cytoskeletal protein expression. J Biol Chem. 1987; 262: 5366–5376.
Westermark B, Portor KR. Hormonally induced changes in the cytoskeleton of human thyroid cells in culture. J Cell Biol. 1982; 94: 42–50.
Gros R, Ding Q, Chorazyczewski J, Andrews J, Pickering JG, Hegele RA, Feldman RD. The impact of blunted beta-adrenergic responsiveness on growth regulatory pathways in hypertension. Mol Pharmacol. 2006; 69: 317–327.
Ostrom RS, Liu X, Head BP, Gregorian C, Seasholtz TM, Insel PA. Localization of adenylyl cyclase isoforms and G protein-coupled receptors in vascular smooth muscle cells: expression in caveolin-rich and noncaveolin domains. Mol Pharmacol. 2002; 62: 983–992.