The multiple neoplasia syndrome Carney complex (CNC) is caused by heterozygote mutations in the gene, which codes for the RIα regulatory subunit (PRKAR1A) of protein kinase A. Inactivation of PRKAR1A and the additional loss of the normal allele lead to tumors in CNC patients and increased cyclic AMP signaling in their cells, but the oncogenetic mechanisms in affected tissues remain unknown. Previous studies suggested that PRKAR1A down-regulation may lead to increased mitogen-activated protein kinase (MAPK) signaling. Here, we show that, in lymphocytes with PRKAR1A-inactivating mutations, there is increased extracellular signal-regulated kinase (ERK) 1/2 and B-raf phosphorylation and MAPK/ERK kinase 1/2 and c-Myc activation, whereas c-Raf-1 is inhibited. These changes are accompanied by increased cell cycle rates and decreased apoptosis that result in an overall net gain in proliferation and survival. In conclusion, inactivation of PRKAR1A leads to widespread changes in molecular pathways that control cell cycle and apoptosis. This is the first study to show that human cells with partially inactivated RIα levels have increased proliferation and survival, suggesting that loss of the normal allele in these cells is not necessary for these changes to occur. (Cancer Res 2006; 66(21): 10603-12)
Carney complex (CNC) is a multiple endocrine neoplasia syndrome (1) predisposing to the development of a variety of endocrine tumors, such as adrenal and pituitary hyperplasia and adenomas, thyroid follicular adenomas, and gonadal neoplasms (2–4). CNC is also associated with spotty skin lesions and nevi, myxomas of the skin, heart, and breast, and psammomatous melanotic schwannomas (1, 5). Primary malignancies are relatively rare, but most CNC patients die of complications of tumors, such as heart myxomas that grow fast and recur frequently. Other causes of death include metastatic thyroid cancer and schwannomas and pancreatic and other carcinomas. Mutations of PRKAR1A, a gene located on chromosomal region 17q22-24 that codes for the RIα regulatory subunit of the cyclic AMP (cAMP)-dependent protein kinase A (PKA), are responsible for the disease in most CNC patients (4). Tumor-specific loss of heterozygosity within 17q22-24 (6) suggested that RIα could function as a tumor suppressor gene. Accordingly, loss of RIα leads to increases in total cAMP-activated kinase activity in CNC tumors (5–7) and results in tumors in mice (8–10).
PKA is a holoenzyme that, when inactive, consists of two isodimers of two regulatory subunits (a pair of RIα or RIβ and RIIα or RIIβ) and two catalytic subunits (a pair of Cα, Cβ, or Cγ). Binding of cAMP to the regulatory subunits releases the catalytic subunits to act as serine-threonine kinases that phosphorylate a variety of molecules that control several cellular functions (11–14). PKA can be activated by a G-protein coupled receptor (GPCR) bound by isoproterenol as well as by other receptor agonists (13, 15) or at adenyl cyclase by forskolin or cholera toxin (13). PKA interacts with the mitogen-activated protein kinase (MAPK) pathway mainly at the level of c-Raf-1. MAPK is composed of multiple and interacting signaling cascades that regulate various functions, such as cell proliferation, differentiation, survival, and apoptosis. Each cascade consists of a three-core member module that phosphorylates and activates the succeeding core member until downstream effectors are activated to induce a cell response (15, 16). The extracellular signal-regulated kinase (ERK) 1/2 cascade of MAPK is activated by a receptor tyrosine kinase that stimulates the small G-protein Ras, with the sequential phosphorylation/activation of c-Raf-1 or B-raf through Rap-1 (17) followed by the activation of MAPK/ERK kinase (MEK) 1/2 and ERK1/2 (15, 16). Phosphorylated ERK1/2 then dimerizes, translocates to the nucleus, and enhances cell proliferation by phosphorylating transcription factors, such as c-Myc that, in turn, induce the expression of cell cycle–regulating genes, such as cyclin-dependent kinases and cyclin D1 and others that may promote cell cycle progression (18, 19).
MAPK also interacts with the two major apoptotic pathways that regulate programmed cell death: the intrinsic, mitochondrial-induced pathway and the extrinsic or death receptor–dependent pathway. The first is activated by a variety of stimuli, including camptothecin-1, camptothecin-2, glucocorticoids, and staurosporine, which cause the release of mitochondrial cytochrome c into the cytoplasm (20, 21). There, cytochrome c interacts with apoptotic protease activating factor-1, procaspase-9, and dATP to form an apoptosome complex (22) that activates procaspase-9 and cleaves downstream caspase-3, caspase-6, and caspase-7 (23) to induce apoptosis. BAD and Bcl-XL, proapoptotic and antiapoptotic members of the Bcl-2 family of apoptotic proteins, enhance and inhibit mitochondrial-induced apoptosis, respectively (20, 21). The extrinsic apoptotic pathway is activated extracellularly by a family of transmembrane death receptors, of the tumor necrosis factor family (e.g., Fas/CD95, Fas/APO-1, and Fas/APO-2), of which Fas is the best studied (24, 25). On binding of the Fas ligand to its receptor, an adapter molecule, Fas-associated death domain (FADD), is recruited. The binding of FADD to the receptor recruits caspase-8 and leads to the formation of a death-inducing signaling complex (DISC). Caspase-8 then proteolytically autoactivates itself and initiates apoptosis by the subsequent cleavage of the downstream effector caspase-3, caspase-6, and caspase-7 (20, 23).
The MAPK ERK1/2 inhibits mitochondrial-induced apoptosis by phosphorylating and inhibiting the proapoptotic protein BAD (26), inducing the binding of BAD to the 14-3-3 chaperone protein for sequestration into the cytosol (27), and preventing the inhibition by BAD of the antiapoptotic protein Bcl-XL. ERK1/2, thus, indirectly prevents the release of cytochrome c from the mitochondria (19, 28). ERK1/2 also inhibits the Fas/CD95-induced apoptotic pathway at a point before caspase-8 activation by preventing DISC signaling from activating caspase-8 (27).
Our previous studies (13) showed increased lysophosphatidic acid (LPA)-induced ERK1/2 phosphorylation in transformed B lymphoblastoid cell lines (tBls) bearing one mutant (mt) PRKAR1A allele. These cell lines possessed half of the normal RIα levels as expected due to nonsense-mediated decay of the mutant allele PRKAR1A mRNA (5, 6) but still had abnormal total kinase activity in response to cAMP. In the present study, we used an expanded number of such cell lines from patients with CNC that had identical or similar PRKAR1A-inactivating mutations and reduced (by half) RIα levels (5, 6, 13). In these cells, we studied cell cycle, proliferation, and apoptosis by comparing them with normal tBls (nl-tBls) from matched controls. We correlated our findings with the status of ERK1/2 signaling components and focused on transcription factor c-Myc. The data show that partial PRKAR1A inactivation in human B lymphocytes is associated with specific alterations in MAPK signaling components that have the overall net effect of increased cell proliferation and survival.
Materials and Methods
Materials. Jurkat cells were from the American Type Culture Collection (Rockville, MD); RPMI 1640, AIM-V medium, fetal bovine serum (FBS), HEPES, antibiotics/antimycotics, l-glutamine, trypan blue, Novex 10% and 14% Tris-glycine gels were from Invitrogen (Carlsbad, CA); forskolin, isoproterenol, dexamethasone, staurosporine, camptothecin-1, camptothecin-2, protein A, l-α-lysophosphatidic acid (l-α-LPA), RNase, Triton X-100, and propidium iodide were from Sigma-Aldrich (St. Louis, MO); phosphorylated anti-ERK1/2, anti-ERK1/2, phosphorylated MEK1, phosphorylated anti-BAD monoclonal antibody (mAb), and cleaved caspase-3 mAb were from Cell Signaling (Beverly, MA); phosphorylated anti-c-Myc was from Santa Cruz Biotechnology (Santa Cruz, CA); phosphorylated c-Raf-1 and anti-Fas mAb clone CH11 were from Upstate (Lake Placid, NY); β-actin was from Abcam (Cambridge, MA); anti-mouse and anti-rabbit IgG were from Oncogene Research Products (Darmstadt, Germany); B-raf kinase assay kit was from United States Biological (Swampscott, MA); bromodeoxyuridine (BrdU) and anti-BrdU mAb were from Molecular Probes (Eugene, OR); APO-1 and APO-2 were from Kamiya Biomedical (Seattle, WA); PD98059 was from Calbiochem (La Jolla, CA); Annexin V-FITC, 7 aminoactinomycin D (7AAD), and DX2-PE mAb were from BD Pharmingen (San Jose, CA); ZB4 was from Beckman Coulter (Fullerton, CA); 10× Tris-glycine-SDS buffer was from Bio-Rad (Hercules, CA); Protran nitrocellulose membranes were from Schleicher & Schuell (Keene, NH); enhanced chemiluminescence (ECL) blotting detection reagent was from Amersham Pharmacia (Piscataway, NJ); and FACSCalibur flow cytometer and CellQuest software were from Becton Dickinson (San Jose, CA).
Subjects and cell lines. CNC patients were carriers of previously described PRKAR1A-inactivating mutations (5, 6) that lead to haploinsufficiency for the PKA subunit RIα. Subjects were diagnosed with primary pigmented nodular adrenocortical disease (PPNAD) in the context of CNC, CNC, and Cushing's syndrome/PPNAD/CNC (Supplementary Table S1). Samples were collected under a research protocol approved by the Institutional Review Board of the National Institute of Child Health and Human Development (Bethesda, MD), and written informed consent was obtained from each subject.
Cell culture conditions. B lymphocytes (8 nl-tBls and 10 mt-tBls) were obtained from the peripheral blood of CNC patients, their unaffected relatives, and other normal subjects. Cells were transformed by Epstein-Barr virus (EBV) and characterized by staining with antibodies to the B lymphocyte cell surface antigen CD23 (13). Cells were maintained in RPMI 1640 with 1% l-glutamine, 10% FBS, and 1% antibiotic/antimycotic agents.
For immunoassays, cells were suspended in PBS (pH 7.4) and transferred to 1 mL Eppendorf tubes before the addition of forskolin, isoproterenol, or LPA. Cells were plated in 24-well plates (5 × 104 per well for proliferation experiments and 5 × 105/mL for cell receptor assays) and in 12-well plates (2-3 × 106 cells per well for cell cycle assays and 1 × 106 cells per well for apoptosis studies) in complete medium followed by the addition of stimulants. In apoptosis studies in which LPA was used, AIM-V serum-free medium containing 1% l-glutamine and 1% antibiotic/antimycotics was used.
Cell proliferation assay. Cell proliferation was determined when cells were incubated with forskolin or isoproterenol, centrifuged (1,000 × g, 5 minutes), resuspended in PBS (pH 7.4), and kept on ice. Direct cell counts were made each day for 4 days by hemocytometer counting using trypan blue to assess viability (13).
Gel electrophoresis. Briefly, proteins in whole-cell lysates or in clarified supernatants (where indicated) were separated by SDS-PAGE on 10% or 14% acrylamide gels in Tris-glycine-SDS buffer (13). Proteins, transferred to nitrocellulose membranes, were probed with primary antibodies to phosphorylated, nonphosphorylated, and cleaved proteins. Horseradish peroxidase–conjugated antibodies against mouse or rabbit IgG were used as secondary antibodies. Bands were detected by ECL reagent and quantitated by densitometer scanning (Molecular Dynamics, Sunnyvale, CA). Equal loading of sample was confirmed, and arbitrary values were calculated when blots were stripped and reprobed with β-actin mAb.
B-raf kinase assay. B-raf kinase activity was measured using a B-raf kinase assay kit. Briefly, cell lysates in ice-cold lysis buffer were clarified (10,000 rpm for 20 minutes, 4°C). Clarified supernatants were incubated with Mg2+/ATP, unactive glutathione S-transferase (GST)-tagged MEK1 fusion protein, and assay dilution buffer according to the manufacturers' directions. Phosphorylated GST-tagged MEK1, released by the reaction of cell B-raf kinase and unactive GST-tagged MEK1, was detected using phosphorylated anti-MEK1/2 mAb by SDS-PAGE and immunoblot assays as stated above.
Quantitation of apoptosis by flow cytometry. Apoptosis was quantified and reported as the percentage of cells that excluded 7AAD, and exteriorized membrane phosphatidylserine was detected by the binding of Annexin V-FITC (29). Cultures were treated with stimulants and incubated (37°C) and then washed twice by centrifugation (1,500 rpm, 5 minutes) in PBS/0.1% bovine serum albumin (BSA; pH 7.4). Cultures were suspended in binding buffer (10 mmol/L HEPES, 140 mmol/L NaCl, 2.5 mmol/L CaCl2) and labeled with 10 μg 7AAD and Annexin V and incubated in the dark (23°C, 15 minutes) before analysis on a FACSCalibur flow cytometer using CellQuest software. A minimum of 10,000 events was collected and analyzed.
Cell cycle analysis by flow cytometry. Cell cycle analysis was done by modification of a previously described method (30). Cells, synchronized in low-serum (0.1% FBS/RPMI 1640, 72 hours) medium, were released from synchronization by centrifugation (1,500 rpm, 5 minutes, 4°C) and resuspension in 10% FBS/RPMI 1640. Nonlabeled BrdU was added to cells 1 hour before each time point. Cells were incubated for 1 hour, 37°C, centrifuged, washed with PBS (pH 7.4), and resuspended in PBS before the addition of 70% ethanol. Samples were processed by centrifugation and resuspension in 0.1 mol/L HCl/0.5% Triton X-100, 4°C, followed by centrifugation and resuspension in H2O and boiled (10 minutes). Ice-cold PBS/0.5% Triton X-100 was added followed by centrifugation. Anti-BrdUrd-FITC mAb (0.2 mg/mL) was added, and samples were kept in the dark (23°C, 2 hours). Propidium iodide was added, excluding BrdU-FITC control samples. A minimum of 10,000 events was collected and analyzed using a FACSCalibur flow cytometer and CellQuest software.
Quantitation of Fas/CD95 receptors by flow cytometry. PBS/0.1% BSA was added to cell pellets followed by the addition of propidium iodide–labeled DX2 anti-CD95 receptor mAb (2 hours, 4°C) in the dark. Cells were centrifuged, and PBS/0.1% BSA was added. Mean fluorescent intensity of DX2-PE was measured with a FACSCalibur flow cytometer using CellQuest software.
Statistics. Experiments were designed in a completely randomized manner. Results were analyzed by ANOVA using the PROC MIXED procedure of the Statistical Analysis System (SAS Institute, Cary, NC). If data were significantly different and sampling was done on >3 days or with more than three doses, polynomial response curves of appropriate order were fit and tested for heterogeneity of regression (31, 32) to evaluate treatment effects. In all other cases, the PDIFF procedure (SAS Institute) was used to compare differences between treatment means. Differences were considered significant at P < 0.05.
PKA pathway stimulants induce increased cell proliferation in mt-tBls. Cells were cultured with increasing forskolin or isoproterenol concentrations (0-100 μmol/L). Both forskolin and isoproterenol inhibited growth in nl-tBls (Fig. 1A and C) by 25% to 40% and 33% to 81%, respectively (P = 0.05). However, in mt-tBls (Fig. 1B and D), growth was stimulated by forskolin and isoproterenol by 10% to 50% and 13% to 150%, respectively (P < 0.05-0.075). The percentage of necrotic cells in all cultures even after 4 days was <8% (data not shown), excluding the possibility that cell necrosis accounted for the decrease in cell number in nl-tBls. The effect of isoproterenol was greater than forskolin by up to 32% in nl-tBls and by up to 75% in mt-tBls, indicating a greater effect of receptor-stimulated cAMP signaling on cell proliferation than that of adenyl cyclase–induced activity.
ERK1/2 pathway components and c-Myc are involved in cell proliferation. We looked at the effect of PRKAR1A inactivation on components of the ERK1/2 MAPK cascade and on c-Myc. B-raf kinase activity was determined after treatment with isoproterenol or forskolin. The reaction of B-raf kinase and inactive GST-tagged MEK1/2 resulted in the release of phosphorylated GST-tagged MEK1. Phosphorylated GST-MEK1, c-Raf-1, MEK1, and c-Myc were then detected by specific immunoblot assays. Isoproterenol caused increased B-raf kinase activity in mt-tBls (up to 175% of control values) but inhibited kinase activity by 25% in nl-tBls (P < 0.01; Fig. 2A). Cells incubated with forskolin and LPA showed inhibition of phosphorylated c-Raf-1 levels in both nl-tBls and mt-tBls (P < 0.05; Fig. 2B) but increased levels of phosphorylated MEK1 only in mt-tBls and inhibition of phosphorylated MEK1 only in nl-tBls (P < 0.05; Fig. 2C). Likewise, forskolin, at all concentrations, increased LPA-induced stimulation of phosphorylated c-Myc in mt-tBls yet inhibited phosphorylated c-Myc in nl-tBls (P < 0.01; Fig. 2D). Interestingly, the effects seen in all experiments (Fig. 2) were biphasic.
mt-tBls traverse the cell cycle at a faster rate than nl-tBls. To determine if the large increase in cell proliferation seen with PKA stimulants in mt-tBls (Fig. 1B and D) may be caused by cells that traverse the cell cycle at an accelerated rate, we analyzed G0-G1, S, and G2-M phases of the cell cycle in tBls. Cells, synchronized and then treated with BrdU, were cultured for up to 48 hours; data were analyzed by flow cytometry. The degree of synchronization was identical in both cell types as seen at the 0-hour time point (Fig. 3). However, after 15 hours in culture, mt-tBls left G0-G1 at a faster rate than nl-tBls; both cell types entered S phase after 9 hours, with up to 50% more mt-tBls in S phase than nl-tBls (P < 0.05). In G2-M, the percentage DNA content decreased in both cell types with time but was significantly less in mt-tBls (P < 0.01). The data indicated that mt-tBls entered and went through the cell cycle at a faster rate than nl-tBls even in the absence of PKA stimulants, consistent with the large differences in cell proliferation seen in the previous experiments (Fig. 1).
PRKAR1A inactivation leads to decreased intrinsic apoptotic pathway activity in mt-tBls. To determine if cells were sensitive to known intrinsic apoptotic pathway stimulants, cells were incubated with camptothecin-1 and camptothecin-2 (2-10 μmol/L), dexamethasone (50-100 μmol/L), and staurosporine (2 μmol/L) and analyzed for apoptosis using Annexin V-FITC/7AAD staining. Apoptosis was seen only with camptothecin-2 and staurosporine in nl-tBls (5% and 15%, respectively) and in mt-tBls with staurosporine only (5%). There was no effect of dexamethasone (50-100 μmol/L) on cell viability at 24 hours and minimal inhibition at 72 hours in nl-tBls or mt-tBls (data not shown). Therefore, camptothecin-2 and dexamethasone were not used in further studies; in all experiments, staurosporine was used instead. Jurkat T lymphocytes, cells that are highly sensitive to staurosporine, were also used as controls. The rates of apoptosis were in the order of Jurkat > nl-tBls > mt-tBls at up to 35%, 14%, and 3%, respectively (P < 0.01; Fig. 4A). Therefore, both nl-tBls and mt-tBls were more resistant to staurosporine-induced apoptosis than Jurkat cells, most likely an effect of EBV transformation of these cells (33, 34). Nevertheless, apoptosis by this mechanism was significantly less in mt-tBls.
Because isoproterenol had a greater effect than forskolin on cell proliferation (Fig. 1), it was also used to assess apoptosis in tBls. Although apoptosis with isoproterenol was substantially less than that seen with staurosporine, isoproterenol produced greater levels of apoptosis in nl-tBls (to 6%) than in mt-tBls (to 3%) as early as after 1 hour of incubation (P = 0.01; Fig. 4B). When cells were incubated with isoproterenol or forskolin (50 μmol/L) or LPA (80 nmol/L), alone or combined, for longer times (24 and 96 hours), there were no significant rates of apoptosis (data not shown). To validate the apoptosis induced by isoproterenol, levels of cleaved caspase-3 were assessed by immunoblot assay. Cleavage of caspase-3 was concentration dependent and greater in nl-tBls than in mt-tBls (P < 0.01; Fig. 4C).
The MEK1 inhibitor PD98059 reverses isoproterenol-induced apoptosis in mt-tBls. ERK1/2 is known to inhibit apoptosis in many cell types by the phosphorylation of proapoptotic BAD (19, 26). Inhibition of ERK1/2 by the MEK1 inhibitor PD98059 would reverse this effect, causing an increase in apoptosis and a decrease in BAD phosphorylation. We thus incubated cells with PD98059 and then exposed them to isoproterenol. Apoptosis was greater in mt-tBls (up to 17%) than in nl-tBls cells (up to 7%; P < 0.01; Fig. 5A), indeed reversing what we had seen earlier (Fig. 4B) without PD98059 and suggesting that ERK1/2 inhibits apoptosis to a greater extent in mt-tBls. The mechanism seems to be increased BAD phosphorylation by ERK1/2 in mt-tBls because levels of phosphorylated BAD in cells treated with isoproterenol alone were 2-fold higher in mt-tBls (P < 0.05). This difference disappeared after the application of PD98059 (P > 0.1). PD98059 decreased significantly BAD phosphorylation levels in both cell types (P < 0.01 for BAD phosphorylation levels in each cell type before and after PD98059 application; Fig. 5B). These data suggest that the effect of isoproterenol on apoptosis may be two pronged: a large inhibitory effect (Fig. 5A and B), most likely the result of high levels of ERK1/2 in mt-tBls, and a smaller stimulatory effect that we discussed earlier (Fig. 4B and C).
PRKAR1A inactivation leads to decreased extrinsic apoptotic pathway activity in mt-tBls. The above data suggested a possible difference in extrinsic mechanisms regulating apoptosis in mt-tBls. Cells were incubated with the known extrinsic pathway Fas/CD95 receptor stimulants, APO-1, or APO-2 plus protein A for 1 to 16 hours or with CH11 for up to 18 hours. Annexin V/7AAD staining indicated rapid apoptosis that resulted in a necrotic phenotype produced by APO-1 in nl-tBls, whereas CH11 induced 50% and 15% apoptosis in nl-tBls and mt-tBls, respectively (P < 0.05). No apoptosis was triggered with APO-2/protein A in either cell type (data not shown). Because of the levels of apoptosis obtained with CH11 in both nl-tBls and mt-tBls, CH11 was used to further study extrinsic Fas/CD95-induced apoptosis. Jurkat T lymphocytes served as control cells because these cells are highly sensitive to Fas/CD95-mediated apoptosis. The order with which apoptosis occurred was Jurkat > nl-tBls > mt-tBls at up to 70%, 30%, and 15%, respectively (P < 0.05 for all between-line comparisons; Fig. 6A). CH11-induced apoptosis was confirmed in tBls by immunoblot assays of cleaved caspase-3 using cleaved caspase-3 mAb. There was significantly more cleavage of caspase-3 in nl-tBls than in mt-tBls (P < 0.001 at 0.5-1 μg CH11; Fig. 6B). Although these data indicated a generalized resistance of the transformed cells to Fas/CD95-induced apoptosis, mt-tBls were significantly more resistant.
This resistance was not due to a low number of Fas/CD95 cell surface receptors: we stained tBls with anti-Fas/CD95 propidium iodide–labeled DX2 mouse mAb and assessed mean fluorescent intensity by flow cytometry. Cell surface staining was greater in mt-tBls (3.6-fold) than in nl-tBls (P = 0.032; Fig. 6C). We then incubated tBls in the presence and absence of the Fas/CD95 receptor antagonist ZB4 mAb (Fig. 6D). CH11-induced apoptosis and nonstimulated apoptosis were inhibited in both cell types by 15% to 47%. These data confirmed the action of CH11 on Fas/CD95 receptors in tBls and suggested that mt-tBls have in fact higher levels of Fas/CD95 receptors, although they are more resistant to apoptosis than nl-tBls.
In a previous study (13), we showed a reversal of the normal PKA-mediated inhibitory effect on the ERK1/2 cascade of MAPK in cells bearing heterozygote, PRKAR1A-inactivating mutations that are subject to nonsense mRNA-mediated decay (NMD; refs. 5, 6). In these studies, an increase in ERK1/2 phosphorylation and cell proliferation was seen in mt-tBls on activation of PKA by the PKA pathway-specific stimulants isoproterenol and forskolin. We hypothesized that the differential effect of PKA on MAPK may be due to a shift from an inhibition of c-Raf-1 (35) to a stimulation of B-raf via activation of the small G-protein Rap-1 (17) by a switch to type II PKA (36). In the present study, we looked in depth at the interaction of the PKA and MAPK pathways and extended our previous studies to an analysis of the effect of PKA stimulation on cell cycle entry and apoptosis in these cells that have a reduction of normal RIα levels by half due to NMD (5, 6, 13).
We first showed that the effects of isoproterenol and forskolin on cell proliferation were concentration-dependent with the effect of isoproterenol greater than that of forskolin (Fig. 1). These results mirror the effect of isoproterenol and forskolin on ERK1/2 phosphorylation (13) and suggest a difference in receptor-mediated versus nonreceptor-mediated stimulation of PKA. Because isoproterenol is known to activate receptors (GPCRs) that may stimulate other signaling pathways in several cell types, the possibility exists that other GPCR-induced signaling pathways may be stimulated by isoproterenol in tBls. Forskolin acting at the level of adenyl cyclase also produces cAMP by a variety of pathways (not always PKA dependent). However, a reduced effect of forskolin over isoproterenol is seen in the present studies. These data may reflect an enhanced interference of other cAMP-associated pathways that are both dependent and independent of PKA activity (37, 38). Our finding of only <8% dead cells in each cell population, regardless of treatment or time in culture, suggested that the large (25-81%) decrease in cell proliferation seen with isoproterenol and forskolin in nl-tBls is not due to gross cell necrosis.
Our analysis of the ERK1/2 cascade components and c-Myc phosphorylation is consistent with our previous studies (13) that showed increases in ERK1/2 phosphorylation in mt-tBls. We postulated that increased ERK1/2 phosphorylation and cell proliferation on activation of PKA is due to the simulation of B-raf in mt-tBls. Here, we have shown that phosphorylation of B-raf (Fig. 2A) is increased, whereas phosphorylation of c-Raf-1 is inhibited in mt-tBls (Fig. 2B). In nl-tBls, both B-raf and c-Raf-1 phosphorylation is inhibited (Fig. 2A and B). The data imply that a switch from inhibition to stimulation of the ERK1/2 cascade by PKA occurs in mt-tBls. Although c-Raf-1 is also inhibited in mt-tBls (Fig. 2B), these data indicated an incomplete switch to B-raf probably due to the presence of some RIα protein (up to 50% of the normal levels) in these lymphocytes (13). The increased phosphorylation of MEK1/2 (Fig. 2C) is apparently due to the high levels of phosphorylated B-raf in mt-tBls and to the fact that, of the three Raf proteins present in all cells, B-raf has the highest affinity for MEK1 (19, 39). The biphasic effect of isoproterenol, forskolin, and LPA on MAPK components and on c-Myc (Fig. 2) has been shown in our previous studies and may represent a saturation of pathway receptor (MAPK and PKA) and enzyme (adenyl cyclase, B-raf kinase, and MAPK phosphatases) kinetics at higher agonist concentrations (13).
Although the reaction of three mt-tBl cell lines (1, 3, and 5; Supplementary Table S1) was more robust than in other mt-tBl cell lines in all experiments (possibly due to other genetic influences), the increased phosphorylation of B-raf, MEK1, ERK1/2, and c-Myc in mt-tBls (Fig. 2) suggests that PRKAR1A inactivation leads to c-Myc activation. c-Myc, an early response gene, is post-translationally regulated through phosphorylation and interactions with other proteins. c-Myc can respond to mitogenic signals and move cells from G0 to G1 and/or from G1 to S phases of the cell cycle (40). Cyclin D1 expression, an early event in G0-G1 to S phase transition, is a target of MAPK regulation (41). The increased rate of cell cycle transition by synchronized mt-tBls into S phase (Fig. 3) and the above data (Fig. 2D) suggest a possible action of c-Myc activity on cyclin-D1, allowing mt-tBls to enter G1 and cross the G1 restriction point (42) at a faster rate than nl-tBls, leading to an increased rate of cell proliferation. Indeed, in primary mouse embryonic fibroblasts, knocking out prkar1a led to dysregulation of D-type cyclins (43), suggesting that these are the molecules that mediate c-Myc activity in RIα-deficient cells.
We also looked at the possibility that apoptosis could account for the decrease in cell number seen in nl-tBls with PKA stimulants (Fig. 1). We then looked at apoptosis in general to determine the role that apoptosis plays in the response of the cell to PRKAR1A inactivation. We established a baseline by which to judge apoptosis in tBls by testing known stimulants of both the intrinsic and extrinsic apoptotic pathways. Both cell types were poorly responsive to the intrinsic apoptotic pathway stimulants camptothecin-1, camptothecin-2, and dexamethasone and to the extrinsic pathway stimulant APO-2 but more responsive to staurosporine (intrinsic pathway), APO-1, and CH11 (extrinsic pathway; refs. 24, 25, 44, 45). Because APO-1 gave a very rapid response that resulted in necrosis (data not shown), we chose staurosporine and CH11 as control stimulants for the induction of intrinsic and extrinsic pathway apoptosis, respectively. Although staurosporine-induced apoptosis was concentration-dependent and high in Jurkat cells, tBls were relatively resistant, with significantly less apoptosis in mt-tBls (Fig. 4A). On stimulation by isoproterenol, apoptosis was again lower in mt-tBls (Fig. 4B). Apoptosis was confirmed by increased levels of cleaved caspase-3 in nl-tBls versus mt-tBls (Fig. 4C). It was not a surprise that even normal transformed cells were relatively resistant to apoptosis: this has been shown in EBV-transformed and immortalized B lymphoblastoid cell lines (33, 34) and for CH11-induced apoptosis in tumor cells from non–Hodgkin's lymphoma (46).
PD98059 (Fig. 5) was seen to reverse the isoproterenol-induced apoptotic effect (Fig. 4B). These data suggested that higher levels of ERK1/2 in mt-tBls inhibited apoptosis to a greater extent than in nl-tBls. The mediator of this inhibition was found to be BAD because ERK1/2 is known to phosphorylate/inactivate BAD. Indeed, BAD phosphorylation was induced by isoproterenol, and this effect as well as nonstimulated BAD phosphorylation were inhibited by PD98059 (Fig. 5B). These data suggested that higher levels of ERK1/2 activity in mt-tBls, induced by PRKAR1A deficiency, resulted in a small but significant inhibition of the intrinsic apoptotic pathway, contributing to greater cell survival.
Apoptosis was induced by CH11 through the extrinsic apoptotic pathway in a concentration-dependent manner. However, like with staurosporine and isoproterenol, apoptosis induced by CH11 was less pronounced in nl-tBls than in Jurkat cells, with even lower amounts in mt-tBls (Fig. 6A). These data were confirmed by increased levels of cleaved caspase-3 in nl-tBls versus mt-tBls (Fig. 6B). Levels of Fas/CD95 receptors (Fig. 6C), confirmed by receptor inhibitor studies with ZB4 Fas receptor inhibitor (Fig. 6D), were found to be 3.6-fold higher on mt-tBls than on nl-tBls, although these cells exhibited the least amount of apoptosis (Fig. 6A and B). High Fas receptor number accompanied by low CH11-induced apoptosis has also been seen in tumor B lymphocytes from non–Hodgkin's lymphoma (46). At the present time, we have no answer to this paradox. We hypothesize that PRKAR1A inactivation may cause reduced signaling through the Fas/CD95 extrinsic apoptotic pathway, allowing reduced sensitivity to Fas/CH11 and subsequently higher Fas expression. As seen in other cancer cell types where cell survival is favored over apoptosis (47, 48), a disruption of the normal DISC function/structure may occur with a recruitment of anti-caspase-8, FLIP, instead of the procaspase-8, FADD. Our preliminary experiments (data not shown) have indicated higher levels of FLIP in the presence of CH11 in mt-tBls and suggest that a switch from apoptosis to cell survival occurs in mt-tBls. Because ERK1/2 has been shown to inhibit CD95/Fas-mediated apoptosis downstream of the DISC assembly (27), further experiments examining the structure of the DISC, involvement of other prosurvival elements [e.g., the nuclear factor-κB pathway (49)], and the role of PRKAR1A and ERK1/2 in mt-tBl survival must be done to address this question. In general, the present data suggested that the large decreases in cell proliferation seen in nl-tBls when stimulated by isoproterenol and forskolin were not due to apoptosis or necrosis; on the other hand, it was also clear that PRKAR1A inactivation leads to significant changes in both the intrinsic and extrinsic apoptotic pathways that contribute to an overall enhancement of cell survival.
In conclusion, the present studies suggested that the balance between cell proliferation and death that leads to tissue homeostasis (50) may be disturbed in PRKAR1A-deficient cells. This is the first study to show increased proliferation in human cells that maintain 50% of normal RIα levels. It expands our previous observation that these cells have abnormalities in their PKA function (13) and suggests that, although loss of the normal allele is seen in CNC tumors (leading to complete loss of RIα activity), PRKAR1A inhibition by a mere half is sufficient for increased cell proliferation, a sine qua non of tumorigenic potential. Furthermore, this investigation presents a model by which partial inactivation of PRKAR1A may increase cell cycle progression, proliferation, and survival (Supplementary Fig. S1) by widespread, mostly strong, and occasionally subtle statistically significant changes in MAPK signaling, c-Myc, cell cycle activity, and survival.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Grant support: NIH intramural project Z01-HD-000642-04.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Dr. Michael A. Beaven (National Heart Lung and Blood Institute, NIH, Bethesda, MD) for his critical advice in the preparation of this article.