{"gene":"PDE4A","run_date":"2026-06-10T05:19:53","timeline":{"discoveries":[{"year":1996,"finding":"The N-terminal splice region of PDE4A splice variant RPDE-6 (RNPDE4A5) interacts with SH3 domains of Src-family tyrosine kinases (v-Src, Lyn, Fyn, c-Abl, Crk, Csk, Lck). Interaction requires the proline-rich N-terminal unique region of RPDE-6; splice variant RPDE-39 (lacking this region) and met26RD1 (N-terminal deletion) fail to associate. Binding to SH3 domains of Crk, Csk, and Lck reduces PDE4A catalytic activity.","method":"GST pulldown with v-Src-SH3 fusion protein, co-immunoprecipitation from transfected COS7 cells, competition with N-terminal fusion protein, analysis of deletion and splice variants","journal":"The Biochemical journal","confidence":"High","confidence_rationale":"Tier 2 / Strong — reciprocal Co-IP, GST pulldown, multiple splice variant controls, functional activity readout, all in one rigorous study","pmids":["8761480"],"is_preprint":false},{"year":1995,"finding":"PDE4A splice variants RD1 (RNPDE4A1A) and RPDE-6 (RNPDE4A5) have distinct subcellular distributions determined by their unique N-terminal splice regions: RD1 is exclusively in the high-speed pellet (P2) membrane fraction, while RPDE-6 distributes between pellet (~25%) and cytosol (~75%) fractions. Pellet RPDE-6 is resistant to high NaCl and Triton X-100 solubilization. Soluble and pellet RPDE-6 show different rolipram IC50 values (~0.16 µM vs ~1.2 µM), indicating conformationally distinct pools.","method":"Subcellular fractionation of transfected COS-7 cells and brain tissue, immunoprecipitation with anti-C-terminal antisera, enzymatic activity assays with rolipram and cAMP Km measurements","journal":"The Biochemical journal","confidence":"High","confidence_rationale":"Tier 2 / Strong — direct fractionation with functional activity measurements, validated in both transfected cells and native brain tissue","pmids":["7575434"],"is_preprint":false},{"year":1997,"finding":"PDE4A splice variant RD1 (RNPDE4A1A) localizes to the Golgi complex in stably transfected human follicular thyroid carcinoma cells. RD1 immunoreactivity colocalizes with Golgi marker Tex1, and redistribution upon treatment with Golgi-perturbing agents monensin and brefeldin A confirmed Golgi targeting. RD1 is membrane-associated (detergent-soluble, not salt-extractable) and located exclusively in the membrane fraction.","method":"Laser scanning confocal immunofluorescence, subcellular fractionation, Golgi-disrupting drug treatment (monensin, brefeldin A) in stably transfected FTC cell lines","journal":"The Biochemical journal","confidence":"High","confidence_rationale":"Tier 2 / Strong — direct imaging with organelle marker colocalization, pharmacological disruption confirming Golgi identity, fractionation data","pmids":["9003417"],"is_preprint":false},{"year":1998,"finding":"Growth hormone activates PDE4A5 in 3T3-F442A preadipocytes via a JAK2-dependent pathway through phosphatidylinositol 3-kinase and p70S6 kinase, resulting in decreased SDS-PAGE mobility (consistent with phosphorylation) and increased catalytic activity. This activation lowers intracellular cAMP. Antisense depletion of PDE4A5 mimicked rolipram in enhancing growth hormone-stimulated adipocyte differentiation. Activation was independent of ERK2, PKC, or transcriptional effects.","method":"Kinase inhibitor epistasis, antisense depletion, intracellular cAMP measurement, SDS-PAGE mobility shift assay in 3T3-F442A cells","journal":"Proceedings of the National Academy of Sciences of the United States of America","confidence":"High","confidence_rationale":"Tier 2 / Strong — pathway epistasis with multiple inhibitors, antisense loss-of-function with defined phenotype, cAMP functional readout, multiple orthogonal methods","pmids":["9520403"],"is_preprint":false},{"year":2005,"finding":"AKAP3 selectively binds PDE4A5 but not PDE4D in bovine spermatozoa. Co-immunoprecipitation in COS cells co-transfected with AKAP3 and Pde4a5 or Pde4d4 confirmed selectivity. Pulldown from sperm lysates confirmed the in vivo interaction. PDE4A5 localization shifts from Triton X-100-soluble fraction in cauda epididymal sperm to SDS-soluble (insoluble) fraction in ejaculated sperm during capacitation.","method":"Co-immunoprecipitation in co-transfected COS cells, pulldown from sperm lysates, immunolocalization, subcellular fractionation","journal":"Biology of reproduction","confidence":"High","confidence_rationale":"Tier 2 / Strong — reciprocal pulldown in two systems (transfected cells + native sperm lysate), selectivity demonstrated with PDE4D as negative comparator, localization shift documented","pmids":["16177223"],"is_preprint":false},{"year":2005,"finding":"PDE4A11, a novel long-isoform splice variant of human PDE4A, is activated by PKA-mediated phosphorylation at Ser119. PDE4A11 localizes predominantly around the nucleus and in membrane ruffles when expressed in COS-7 cells. It hydrolyzes cAMP with Km ~4 µM. Unlike PDE4A4, PDE4A11 shows differential sensitivity to caspase-3 cleavage and to PDE4 inhibitors, and has distinct rolipram redistribution behavior. All three PDE4A long isoforms (PDE4A4, PDE4A10, PDE4A11) can interact with beta-arrestin.","method":"Recombinant expression in COS-7 cells, kinase activity assay with PKA, subcellular fractionation, immunofluorescence localization, inhibitor IC50 measurements, cAMP hydrolysis kinetics","journal":"Molecular pharmacology","confidence":"High","confidence_rationale":"Tier 2 / Moderate — multiple orthogonal methods (enzymatic assay, PKA phosphorylation, localization imaging, inhibitor profiling) in one study; single lab","pmids":["15738310"],"is_preprint":false},{"year":2001,"finding":"The human PDE4A catalytic domain (residues 330-723) expressed in Sf9 cells exists as a tetramer at ~1 mg/ml (by light scattering) and is heavily phosphorylated on both serines of the conserved SPS motif (by mass spectrometry). Despite this phosphorylation, the SPS motif is not part of the active site but is positioned near it, as shown by covalent labeling of an adjacent peptide by an electrophilic cAMP analogue. Km for cAMP hydrolysis is ~2 µM.","method":"Mass spectrometry for phosphorylation site mapping, light scattering for oligomeric state, covalent labeling with electrophilic cAMP analogue, enzymatic kinetics","journal":"Archives of biochemistry and biophysics","confidence":"Medium","confidence_rationale":"Tier 1 / Weak — in vitro biochemical characterization with MS and affinity labeling, single lab, single study","pmids":["11566027"],"is_preprint":false},{"year":2004,"finding":"PDE4A7, an isoform encoded by the human PDE4A gene, lacks catalytic activity due to its unique C-terminal region, not its N-terminal region. Chimera analysis showed that replacing the C-terminal unique portion of PDE4A7 with the conserved C-terminal sequence of active PDE4 isoforms (Hyb1) restored full catalytic activity, whereas replacing the N-terminal portion (Hyb2) did not. Three functional regions within PDE4A isoforms govern catalytic activity, subcellular targeting, and conformational status. PDE4A7 is exclusively in the P1 particulate fraction, and a region in the conserved C-terminal of active PDE4A isoforms prevents this exclusive targeting.","method":"Chimeric protein construction and expression, enzymatic activity assays, subcellular fractionation, SDS-PAGE analysis in transfected cells","journal":"The Biochemical journal","confidence":"High","confidence_rationale":"Tier 1 / Moderate — reconstitution through chimera/mutagenesis approach with enzymatic activity as direct readout, multiple constructs tested","pmids":["15025561"],"is_preprint":false},{"year":1997,"finding":"In Jurkat T-cells, forskolin (via cAMP elevation) selectively down-regulates a novel ~118 kDa PDE4A splice variant (distinct from PDE4A4B) at the transcriptional level, while inducing PDE4D1 and PDE4D2. Immunoprecipitation showed the ~118 kDa PDE4A species provides all PDE4 activity in control cells. This down-regulation is blocked by actinomycin D, confirming transcriptional dependence. The effect is mimicked by cholera toxin and 8-bromo-cAMP.","method":"RT-PCR, immunoblotting, immunoprecipitation with PDE4-selective antisera, pharmacological inhibition of transcription with actinomycin D in Jurkat T-cells","journal":"The Biochemical journal","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — immunoprecipitation activity assay plus transcriptional inhibitor epistasis, multiple pharmacological tools, single lab","pmids":["9003416"],"is_preprint":false},{"year":1999,"finding":"Inhibition of T-cell proliferation and LPS-stimulated TNFα release from monocytes by subtype-selective PDE4 inhibitors correlates significantly with inhibition of recombinant human PDE4A or PDE4B catalytic activity, but not PDE4D. This establishes that PDE4A (and/or PDE4B) plays the major functional role in regulating these inflammatory cell functions.","method":"Recombinant human PDE4 subtype enzymatic assays correlated by linear regression and Spearman's rank-order with cellular functional assays (T-cell proliferation, TNFα release) using 10 subtype-selective inhibitors","journal":"British journal of pharmacology","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — pharmacological correlation across 10 compounds with two orthogonal functional readouts and statistical analysis; indirect (no genetic KO), single study","pmids":["10602317"],"is_preprint":false},{"year":2014,"finding":"AKAP149-PKA-PDE4A complex redistributes within K-562 cells following YTX treatment: the complex decreases in cytosol and increases in plasma membrane (at 24 h, associated with apoptosis/caspase activation) and then in the nucleus (at 48 h, associated with non-apoptotic cell death). Silencing of either AKAP149 or PDE4A prevented YTX-induced cell death, establishing the complex as required for YTX cytotoxicity.","method":"Subcellular fractionation, AKAP149/PDE4A siRNA silencing, caspase activity assays, western blotting in K-562 cells","journal":"The international journal of biochemistry & cell biology","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — loss-of-function (siRNA) with defined cell death phenotype and subcellular fractionation; single lab, single study","pmids":["24813785"],"is_preprint":false},{"year":2015,"finding":"PDE4A is required for autophagy triggered by yessotoxin (YTX) in K-562 cells. PDE4A silencing experiments showed that PDE4A regulates distinct steps of the autophagic process induced by YTX versus classical autophagy inducers (e.g., rapamycin), establishing PDE4A as a key mediator of YTX-induced autophagy after 48 h treatment.","method":"PDE4A siRNA silencing, autophagy marker analysis (mTOR, LC3B), rapamycin as comparator in K-562 cells","journal":"Toxicology","confidence":"Medium","confidence_rationale":"Tier 3 / Moderate — siRNA loss-of-function with autophagy pathway markers; pathway placement partially established; single lab","pmids":["25576684"],"is_preprint":false},{"year":2007,"finding":"NMDA receptor activity regulates PDE4A1 and PDE4A5 expression in rat primary cortical cultures. Chronic blockade of NMDA receptors with MK-801 reduces PDE4A1 and PDE4A5 expression/activity in a time-dependent manner, reversed by the PKA activator dbr-cAMP. NR1/NR2B-induced cGMP signaling (via PDE4) negatively cross-regulates NR1/NR2A-induced cAMP levels. GABA receptor inhibition increases NMDA-induced cAMP and PDE4A expression in mature but not young cultures.","method":"Pharmacological manipulation (MK-801, ifenprodil, bicuculline, dbr-cAMP) with PDE4 activity assays and immunoblot in rat primary cortical/hippocampal cultures","journal":"Brain research","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — multiple pharmacological tools with enzymatic activity and protein expression readouts; no genetic KO, single lab","pmids":["17407767"],"is_preprint":false},{"year":1998,"finding":"Human PDE4A short isoform RD1 (homologue of rat RNPDE4A1A), when transiently expressed in COS-7 cells, appears as an 83 kDa species primarily in the high-speed membrane fraction. It exhibits Km for cAMP of ~3 µM and IC50 for rolipram of ~0.3 µM. In vitro transcription-translation shows RD1 is produced as an 80 kDa species capable of binding to membranes. The gene spans 50 kb with at least 17 exons, located at chromosome 19p13.2.","method":"Transient expression in COS-7 cells, subcellular fractionation, in vitro transcription-translation, enzymatic activity assay","journal":"The Biochemical journal","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — direct expression and fractionation with functional enzymatic assay and in vitro translation; single lab","pmids":["9677330"],"is_preprint":false},{"year":2024,"finding":"GATA4 transcription factor directly binds the PDE4A promoter and upregulates PDE4A expression in Aβ1-42-stimulated BV2 microglial cells (confirmed by Jaspar prediction and dual-luciferase reporter assay). Increased PDE4A expression downstream of GATA4 inactivates the PI3K/AKT pathway, promoting microglial apoptosis and inflammation.","method":"Dual-luciferase reporter assay for GATA4-PDE4A promoter interaction, GATA4 knockdown/overexpression, PDE4A overexpression, western blot for PI3K/AKT pathway in BV2 cells","journal":"Neuroscience","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — luciferase reporter for direct transcription factor-promoter interaction, epistasis with PDE4A overexpression; single lab","pmids":["39653247"],"is_preprint":false},{"year":2024,"finding":"Knockdown of PDE4A in ovarian cancer cells promotes proliferation, migration, and invasion, while overexpression suppresses these processes. PDE4A loss induces epithelial-mesenchymal transition (EMT) and nuclear translocation of Snail. In vivo, PDE4A-overexpressing OVCAR3 cells formed fewer and smaller metastatic foci. Rolipram (PDE4 inhibitor) mimicked PDE4A deletion effects.","method":"PDE4A knockdown and overexpression in OC cell lines, in vitro proliferation/migration/invasion assays, in vivo mouse metastasis model, EMT marker analysis, Snail nuclear localization by western blot","journal":"Experimental cell research","confidence":"Medium","confidence_rationale":"Tier 2 / Moderate — loss- and gain-of-function with defined cellular and molecular phenotype (Snail/EMT) in vitro and in vivo; single lab","pmids":["38797258"],"is_preprint":false}],"current_model":"PDE4A encodes a cAMP-specific phosphodiesterase whose multiple splice variants (e.g., PDE4A1/RD1, PDE4A5/RPDE-6, PDE4A4, PDE4A7, PDE4A11) are differentiated by unique N- or C-terminal regions that govern subcellular targeting (Golgi, membranes, cytosol, perinuclear), protein interactions (SH3 domains of Src-family kinases via a proline-rich N-terminal motif; AKAP3 in sperm; AKAP149 in somatic cells; beta-arrestin), and catalytic regulation (PKA phosphorylation at Ser119 activates long isoforms; p70S6K pathway activated by growth hormone increases PDE4A5 activity; GATA4 transcriptionally upregulates PDE4A); PDE4A activity degrades cAMP in inflammatory cells, neurons, and cancer cells, thereby modulating PKA signaling, Snail/EMT-driven cancer invasion, and autophagy."},"narrative":{"mechanistic_narrative":"PDE4A encodes a cAMP-specific phosphodiesterase whose multiple splice variants share a conserved catalytic core but are functionally individualized by unique N- and C-terminal regions that dictate subcellular targeting, protein interactions, and catalytic regulation [PMID:7575434, PMID:15025561]. The N-terminal splice region defines localization—the short RD1 isoform partitions exclusively into membranes and concentrates at the Golgi complex (detergent-soluble, salt-resistant), whereas RPDE-6 distributes between membrane and cytosol as conformationally distinct pools [PMID:7575434, PMID:9003417]. The unique C-terminal region is equally decisive: it can abolish catalytic activity, as in PDE4A7, where swapping the variant C-terminus for the conserved active sequence restores both catalysis and normal targeting [PMID:15025561]. The N-terminal proline-rich region also mediates protein interactions, binding the SH3 domains of Src-family kinases (v-Src, Lyn, Fyn, Lck, Csk, Crk) in a manner that can suppress catalytic activity [PMID:8761480], while long isoforms engage scaffolds including AKAP3 in sperm, AKAP149 in somatic cells, and beta-arrestin [PMID:16177223, PMID:15738310, PMID:24813785]. Catalytic output is tuned by phosphorylation: PKA phosphorylation at Ser119 activates the long isoform PDE4A11 [PMID:15738310], and a growth hormone–driven JAK2/PI3K/p70S6 kinase cascade activates PDE4A5 to lower cAMP and restrain adipocyte differentiation [PMID:9520403]. Through cAMP hydrolysis, PDE4A shapes physiological outputs across cell types—it provides the dominant PDE4 activity controlling T-cell proliferation and monocyte TNFα release [PMID:10602317], is regulated by NMDA-receptor signaling in cortical neurons [PMID:17407767], mediates yessotoxin-induced cell death and autophagy via the AKAP149–PKA–PDE4A complex [PMID:24813785, PMID:25576684], and acts as a suppressor of Snail/EMT-driven invasion in ovarian cancer [PMID:38797258]. Transcriptionally, GATA4 directly upregulates PDE4A, which in turn dampens PI3K/AKT signaling to promote microglial apoptosis and inflammation [PMID:39653247].","teleology":[{"year":1995,"claim":"Established that PDE4A splice variants are not functionally redundant but are differentially targeted within the cell by their unique N-terminal regions, defining the principle of isoform-specific compartmentalization.","evidence":"Subcellular fractionation of transfected COS-7 cells and brain tissue with rolipram/cAMP activity assays comparing RD1 and RPDE-6","pmids":["7575434"],"confidence":"High","gaps":["Did not identify the membrane anchor or docking partner retaining RD1 in the pellet fraction","Mechanism producing two conformationally distinct RPDE-6 pools unresolved"]},{"year":1996,"claim":"Identified the first protein-interaction function for the PDE4A N-terminus, showing the proline-rich splice region binds SH3 domains of Src-family kinases and that binding modulates catalytic activity—linking PDE4A to tyrosine-kinase signaling scaffolds.","evidence":"GST pulldown with v-Src-SH3, reciprocal Co-IP from COS7 cells, deletion/splice-variant controls with activity readout","pmids":["8761480"],"confidence":"High","gaps":["Physiological consequence of kinase scaffolding in a native cell unestablished","Whether PDE4A is a tyrosine-kinase substrate not addressed"]},{"year":1997,"claim":"Resolved the specific organelle target of the RD1 isoform, demonstrating Golgi localization and thereby connecting a defined PDE4A variant to a discrete subcellular cAMP microdomain.","evidence":"Confocal immunofluorescence with Golgi marker colocalization, monensin/brefeldin A disruption, and fractionation in stably transfected FTC cells","pmids":["9003417"],"confidence":"High","gaps":["Golgi-targeting determinant within the N-terminus not mapped to residues","Functional role of Golgi-localized cAMP hydrolysis not tested"]},{"year":1997,"claim":"Showed PDE4A is itself a transcriptional target of cAMP signaling, with a ~118 kDa variant providing all PDE4 activity in T-cells and being down-regulated upon cAMP elevation, revealing a feedback architecture.","evidence":"RT-PCR, immunoprecipitation activity assay, and actinomycin D epistasis with forskolin/cholera toxin/8-Br-cAMP in Jurkat T-cells","pmids":["9003416"],"confidence":"Medium","gaps":["Identity of the transcriptional regulator mediating cAMP-driven repression unknown","Single cell line; generality across T-cell states untested"]},{"year":1998,"claim":"Defined a hormone-driven signaling pathway that activates a specific PDE4A variant, establishing PDE4A5 as a regulated effector that lowers cAMP to restrain adipocyte differentiation.","evidence":"Kinase-inhibitor epistasis (JAK2/PI3K/p70S6K), antisense depletion, cAMP measurement, and mobility-shift in 3T3-F442A preadipocytes","pmids":["9520403"],"confidence":"High","gaps":["Direct phosphorylation site(s) on PDE4A5 not mapped","Whether p70S6K phosphorylates PDE4A directly or via an intermediate kinase unresolved"]},{"year":1998,"claim":"Characterized the human RD1 short isoform biochemically and placed the gene physically at chromosome 19p13.2, providing the human counterpart to rat targeting/activity findings.","evidence":"Transient expression, fractionation, in vitro transcription-translation, and enzyme kinetics in COS-7 cells with gene structure mapping","pmids":["9677330"],"confidence":"Medium","gaps":["Membrane-binding determinant in human RD1 not defined at residue level"]},{"year":1999,"claim":"Assigned the functional inflammatory role of PDE4A by correlating subtype-selective inhibitor potency against recombinant enzyme with suppression of T-cell proliferation and monocyte TNFα release.","evidence":"Recombinant PDE4 subtype enzymatic assays correlated by regression/Spearman analysis with two cellular readouts across 10 inhibitors","pmids":["10602317"],"confidence":"Medium","gaps":["Pharmacological correlation cannot distinguish PDE4A from PDE4B contribution","No genetic loss-of-function confirmation"]},{"year":2001,"claim":"Provided biochemical/structural insight into the catalytic domain, showing it is tetrameric and phosphorylated on the SPS motif positioned adjacent to—but not part of—the active site.","evidence":"Mass spectrometry phosphosite mapping, light scattering, and covalent labeling with an electrophilic cAMP analogue on the Sf9-expressed catalytic domain","pmids":["11566027"],"confidence":"Medium","gaps":["Functional consequence of SPS phosphorylation on activity not demonstrated","No high-resolution crystal structure reported here"]},{"year":2004,"claim":"Established that the unique C-terminal region, not the N-terminus, can govern catalytic competence, refining the modular model: three functional regions control activity, targeting, and conformation.","evidence":"Chimera/mutagenesis (Hyb1/Hyb2) with activity assays and fractionation of the catalytically dead PDE4A7 isoform","pmids":["15025561"],"confidence":"High","gaps":["Physiological function of catalytically inactive PDE4A7 unknown","Mechanism by which the C-terminus suppresses activity not structurally defined"]},{"year":2005,"claim":"Identified the long isoform PDE4A11 and showed PKA phosphorylation at Ser119 activates it, defining a PKA-feedback node and extending the beta-arrestin interaction to all PDE4A long isoforms.","evidence":"Recombinant expression, PKA phosphorylation/activity assay, fractionation, immunofluorescence, and inhibitor profiling in COS-7 cells","pmids":["15738310"],"confidence":"High","gaps":["Functional consequence of beta-arrestin binding for PDE4A11 not tested","In vivo relevance of perinuclear/membrane-ruffle localization unestablished"]},{"year":2005,"claim":"Demonstrated isoform-selective scaffolding in a specialized cell type, with AKAP3 binding PDE4A5 (but not PDE4D) in sperm and PDE4A5 solubility shifting during capacitation, tying PDE4A to compartmentalized cAMP control in fertilization.","evidence":"Reciprocal Co-IP in transfected COS cells, pulldown from sperm lysates with PDE4D negative control, and fractionation across capacitation states","pmids":["16177223"],"confidence":"High","gaps":["Functional effect of AKAP3 anchoring on sperm cAMP signaling not directly measured","Binding interface on PDE4A5 not mapped"]},{"year":2007,"claim":"Placed PDE4A within neuronal signaling, showing NMDA-receptor activity regulates PDE4A1/PDE4A5 expression and that PDE4 mediates cGMP-dependent cross-regulation of cAMP between NMDA receptor subtypes.","evidence":"Pharmacological manipulation (MK-801, ifenprodil, bicuculline, dbr-cAMP) with PDE4 activity and immunoblot in rat cortical cultures","pmids":["17407767"],"confidence":"Medium","gaps":["Transcriptional mechanism linking NMDA receptor to PDE4A expression unknown","No genetic confirmation of PDE4A-specific contribution"]},{"year":2014,"claim":"Defined a functional AKAP149-PKA-PDE4A complex whose dynamic redistribution is required for yessotoxin-induced cell death, demonstrating that anchored PDE4A controls a cytotoxic cAMP signaling outcome.","evidence":"Subcellular fractionation, AKAP149/PDE4A siRNA, and caspase assays in YTX-treated K-562 cells","pmids":["24813785"],"confidence":"Medium","gaps":["Mechanism coupling complex relocation to apoptosis vs non-apoptotic death unresolved","Single cell line and single stimulus"]},{"year":2015,"claim":"Extended PDE4A's role to autophagy regulation, showing it is required for yessotoxin-induced autophagy and acts at steps distinct from classical inducers.","evidence":"PDE4A siRNA with mTOR/LC3B autophagy markers and rapamycin comparator in K-562 cells","pmids":["25576684"],"confidence":"Medium","gaps":["Molecular step in autophagy controlled by PDE4A not defined","Dependence on cAMP/PKA versus scaffolding not separated"]},{"year":2024,"claim":"Identified PDE4A as a tumor-suppressive regulator of cancer cell invasion, with loss promoting EMT and Snail nuclear translocation and overexpression suppressing metastasis.","evidence":"PDE4A knockdown/overexpression with proliferation/migration/invasion assays, EMT/Snail analysis, and an in vivo OVCAR3 metastasis model","pmids":["38797258"],"confidence":"Medium","gaps":["Mechanistic link between cAMP hydrolysis and Snail regulation not defined","Single cancer type; clinical correlation absent"]},{"year":2024,"claim":"Placed PDE4A downstream of GATA4 in neuroinflammation, showing GATA4 directly drives PDE4A transcription to inactivate PI3K/AKT and promote microglial apoptosis and inflammation.","evidence":"Dual-luciferase promoter assay, GATA4 knockdown/overexpression, PDE4A overexpression, and PI3K/AKT western blot in BV2 microglia","pmids":["39653247"],"confidence":"Medium","gaps":["Whether PDE4A acts via cAMP hydrolysis or a non-catalytic mechanism on PI3K/AKT untested","In vivo relevance to neurodegeneration not established"]},{"year":null,"claim":"How PDE4A isoform-specific compartmentalization and scaffold binding are integrated with its disease roles in cancer and neuroinflammation, and whether these depend on cAMP catalysis or non-catalytic functions, remains unresolved.","evidence":"","pmids":[],"confidence":"Medium","gaps":["No structural model linking N/C-terminal targeting to substrate-level cAMP microdomain control in disease cells","Catalytic versus scaffolding contributions to EMT and PI3K/AKT effects not dissected","No genetic in vivo loss-of-function model defining endogenous PDE4A function"]}],"mechanism_profile":{"molecular_activity":[{"term_id":"GO:0140098","term_label":"catalytic activity, acting on RNA","supporting_discovery_ids":[1,5,6,7,13]},{"term_id":"GO:0016787","term_label":"hydrolase activity","supporting_discovery_ids":[5,6,13]}],"localization":[{"term_id":"GO:0005794","term_label":"Golgi apparatus","supporting_discovery_ids":[2]},{"term_id":"GO:0005886","term_label":"plasma membrane","supporting_discovery_ids":[1,5,10,13]},{"term_id":"GO:0005829","term_label":"cytosol","supporting_discovery_ids":[1,10]},{"term_id":"GO:0005634","term_label":"nucleus","supporting_discovery_ids":[5,10]}],"pathway":[{"term_id":"R-HSA-162582","term_label":"Signal Transduction","supporting_discovery_ids":[0,3,5,12]},{"term_id":"R-HSA-168256","term_label":"Immune System","supporting_discovery_ids":[9,14]},{"term_id":"R-HSA-9612973","term_label":"Autophagy","supporting_discovery_ids":[11]},{"term_id":"R-HSA-5357801","term_label":"Programmed Cell Death","supporting_discovery_ids":[10,14]},{"term_id":"R-HSA-1643685","term_label":"Disease","supporting_discovery_ids":[14,15]}],"complexes":["AKAP149-PKA-PDE4A complex"],"partners":["AKAP3","AKAP149","ARRB","SRC","LYN","FYN","LCK","CSK"],"other_free_text":[]}},"prefetch_data":{"uniprot":{"accession":"P27815","full_name":"3',5'-cyclic-AMP phosphodiesterase 4A","aliases":["DPDE2","PDE46","cAMP-specific phosphodiesterase 4A"],"length_aa":886,"mass_kda":98.1,"function":"Hydrolyzes the second messenger 3',5'-cyclic AMP (cAMP), which is a key regulator of many important physiological processes Efficiently hydrolyzes cAMP Efficiently hydrolyzes cAMP Efficiently hydrolyzes cAMP. The phosphodiesterase activity is not affected by calcium, calmodulin or cyclic GMP (cGMP) levels. Does not hydrolyze cGMP Efficiently hydrolyzes cAMP Efficiently hydrolyzes cAMP Efficiently hydrolyzes cAMP","subcellular_location":"Cytoplasm, cytosol; Membrane","url":"https://www.uniprot.org/uniprotkb/P27815/entry"},"depmap":{"release":"DepMap","has_data":true,"is_common_essential":false,"resolved_as":"","url":"https://depmap.org/portal/gene/PDE4A","classification":"Not Classified","n_dependent_lines":6,"n_total_lines":1208,"dependency_fraction":0.004966887417218543},"opencell":{"profiled":false,"resolved_as":"","ensg_id":"","cell_line_id":"","localizations":[],"interactors":[],"url":"https://opencell.sf.czbiohub.org/search/PDE4A","total_profiled":1310},"omim":[{"mim_id":"602987","title":"PHOSPHODIESTERASE 1C; PDE1C","url":"https://www.omim.org/entry/602987"},{"mim_id":"600129","title":"PHOSPHODIESTERASE 4D; PDE4D","url":"https://www.omim.org/entry/600129"},{"mim_id":"600127","title":"PHOSPHODIESTERASE 4B; PDE4B","url":"https://www.omim.org/entry/600127"},{"mim_id":"600126","title":"PHOSPHODIESTERASE 4A; PDE4A","url":"https://www.omim.org/entry/600126"}],"hpa":{"profiled":true,"resolved_as":"","reliability":"Supported","locations":[{"location":"Plasma membrane","reliability":"Supported"}],"tissue_specificity":"Low tissue specificity","tissue_distribution":"Detected in all","driving_tissues":[],"url":"https://www.proteinatlas.org/search/PDE4A"},"hgnc":{"alias_symbol":[],"prev_symbol":["DPDE2"]},"alphafold":{"accession":"P27815","domains":[{"cath_id":"-","chopping":"181-197_224-290","consensus_level":"medium","plddt":87.6845,"start":181,"end":290},{"cath_id":"1.10.1300.10","chopping":"407-684","consensus_level":"high","plddt":96.4393,"start":407,"end":684}],"viewer_url":"https://alphafold.ebi.ac.uk/entry/P27815","model_url":"https://alphafold.ebi.ac.uk/files/AF-P27815-F1-model_v6.cif","pae_url":"https://alphafold.ebi.ac.uk/files/AF-P27815-F1-predicted_aligned_error_v6.png","plddt_mean":64.5},"mouse_models":{"mgi_url":"https://www.informatics.jax.org/marker/summary?nomen=PDE4A","jax_strain_url":"https://www.jax.org/strain/search?query=PDE4A"},"sequence":{"accession":"P27815","fasta_url":"https://rest.uniprot.org/uniprotkb/P27815.fasta","uniprot_url":"https://www.uniprot.org/uniprotkb/P27815/entry","alphafold_viewer_url":"https://alphafold.ebi.ac.uk/entry/P27815"}},"corpus_meta":[{"pmid":"10602317","id":"PMC_10602317","title":"Suppression of human inflammatory cell function by subtype-selective PDE4 inhibitors correlates with inhibition of PDE4A and PDE4B.","date":"1999","source":"British journal of pharmacology","url":"https://pubmed.ncbi.nlm.nih.gov/10602317","citation_count":117,"is_preprint":false},{"pmid":"9003416","id":"PMC_9003416","title":"Challenge of human Jurkat T-cells with the adenylate cyclase activator forskolin elicits major changes in cAMP phosphodiesterase (PDE) expression by up-regulating PDE3 and inducing PDE4D1 and PDE4D2 splice variants as well as down-regulating a novel PDE4A splice variant.","date":"1997","source":"The Biochemical journal","url":"https://pubmed.ncbi.nlm.nih.gov/9003416","citation_count":114,"is_preprint":false},{"pmid":"7575434","id":"PMC_7575434","title":"Identification, characterization and regional distribution in brain of RPDE-6 (RNPDE4A5), a novel splice variant of the PDE4A cyclic AMP phosphodiesterase family.","date":"1995","source":"The Biochemical journal","url":"https://pubmed.ncbi.nlm.nih.gov/7575434","citation_count":100,"is_preprint":false},{"pmid":"8761480","id":"PMC_8761480","title":"The SH3 domain of Src tyrosyl protein kinase interacts with the N-terminal splice region of the PDE4A cAMP-specific phosphodiesterase RPDE-6 (RNPDE4A5).","date":"1996","source":"The Biochemical journal","url":"https://pubmed.ncbi.nlm.nih.gov/8761480","citation_count":85,"is_preprint":false},{"pmid":"16177223","id":"PMC_16177223","title":"AKAP3 selectively binds PDE4A isoforms in bovine spermatozoa.","date":"2005","source":"Biology of reproduction","url":"https://pubmed.ncbi.nlm.nih.gov/16177223","citation_count":61,"is_preprint":false},{"pmid":"9520403","id":"PMC_9520403","title":"Stimulation of p70S6 kinase via a growth hormone-controlled phosphatidylinositol 3-kinase pathway leads to the activation of a PDE4A cyclic AMP-specific phosphodiesterase in 3T3-F442A preadipocytes.","date":"1998","source":"Proceedings of the National Academy of Sciences of the United States of America","url":"https://pubmed.ncbi.nlm.nih.gov/9520403","citation_count":59,"is_preprint":false},{"pmid":"10693959","id":"PMC_10693959","title":"Effects of repeated antidepressant treatment of type 4A phosphodiesterase (PDE4A) in rat brain.","date":"2000","source":"Journal of neurochemistry","url":"https://pubmed.ncbi.nlm.nih.gov/10693959","citation_count":47,"is_preprint":false},{"pmid":"15738310","id":"PMC_15738310","title":"Identification and characterization of PDE4A11, a novel, widely expressed long isoform encoded by the human PDE4A cAMP phosphodiesterase gene.","date":"2005","source":"Molecular pharmacology","url":"https://pubmed.ncbi.nlm.nih.gov/15738310","citation_count":42,"is_preprint":false},{"pmid":"9677330","id":"PMC_9677330","title":"Identification and characterization of the human homologue of the short PDE4A cAMP-specific phosphodiesterase RD1 (PDE4A1) by analysis of the human HSPDE4A gene locus located at chromosome 19p13.2.","date":"1998","source":"The Biochemical journal","url":"https://pubmed.ncbi.nlm.nih.gov/9677330","citation_count":40,"is_preprint":false},{"pmid":"35229272","id":"PMC_35229272","title":"CircAXL Knockdown Alleviates Aβ1-42-Induced Neurotoxicity in Alzheimer's Disease via Repressing PDE4A by Releasing miR-1306-5p.","date":"2022","source":"Neurochemical research","url":"https://pubmed.ncbi.nlm.nih.gov/35229272","citation_count":32,"is_preprint":false},{"pmid":"9003417","id":"PMC_9003417","title":"Intracellular localization of the PDE4A cAMP-specific phosphodiesterase splice variant RD1 (RNPDE4A1A) in stably transfected human thyroid carcinoma FTC cell lines.","date":"1997","source":"The Biochemical journal","url":"https://pubmed.ncbi.nlm.nih.gov/9003417","citation_count":28,"is_preprint":false},{"pmid":"22223051","id":"PMC_22223051","title":"Fragment-based screening for inhibitors of PDE4A using enthalpy arrays and X-ray crystallography.","date":"2012","source":"Journal of biomolecular screening","url":"https://pubmed.ncbi.nlm.nih.gov/22223051","citation_count":19,"is_preprint":false},{"pmid":"25576684","id":"PMC_25576684","title":"Key role of phosphodiesterase 4A (PDE4A) in autophagy triggered by yessotoxin.","date":"2015","source":"Toxicology","url":"https://pubmed.ncbi.nlm.nih.gov/25576684","citation_count":18,"is_preprint":false},{"pmid":"15025561","id":"PMC_15025561","title":"Expression, intracellular distribution and basis for lack of catalytic activity of the PDE4A7 isoform encoded by the human PDE4A cAMP-specific phosphodiesterase gene.","date":"2004","source":"The Biochemical journal","url":"https://pubmed.ncbi.nlm.nih.gov/15025561","citation_count":18,"is_preprint":false},{"pmid":"31417563","id":"PMC_31417563","title":"Overexpression of PDE4A Acts as Checkpoint Inhibitor Against cAMP-Mediated Immunosuppression in vitro.","date":"2019","source":"Frontiers in immunology","url":"https://pubmed.ncbi.nlm.nih.gov/31417563","citation_count":17,"is_preprint":false},{"pmid":"10683824","id":"PMC_10683824","title":"Distribution of PDE4A and G(o) alpha immunoreactivity in the accessory olfactory system of the mouse.","date":"2000","source":"Neuroreport","url":"https://pubmed.ncbi.nlm.nih.gov/10683824","citation_count":17,"is_preprint":false},{"pmid":"21898905","id":"PMC_21898905","title":"Positive association of phencyclidine-responsive genes, PDE4A and PLAT, with schizophrenia.","date":"2011","source":"American journal of medical genetics. Part B, Neuropsychiatric genetics : the official publication of the International Society of Psychiatric Genetics","url":"https://pubmed.ncbi.nlm.nih.gov/21898905","citation_count":16,"is_preprint":false},{"pmid":"31737188","id":"PMC_31737188","title":"Pterostilbene attenuates amyloid-β induced neurotoxicity with regulating PDE4A-CREB-BDNF pathway.","date":"2019","source":"American journal of translational research","url":"https://pubmed.ncbi.nlm.nih.gov/31737188","citation_count":16,"is_preprint":false},{"pmid":"17407767","id":"PMC_17407767","title":"Changes in NMDA receptor-induced cyclic nucleotide synthesis regulate the age-dependent increase in PDE4A expression in primary cortical cultures.","date":"2007","source":"Brain research","url":"https://pubmed.ncbi.nlm.nih.gov/17407767","citation_count":15,"is_preprint":false},{"pmid":"24813785","id":"PMC_24813785","title":"Role of AKAP 149-PKA-PDE4A complex in cell survival and cell differentiation processes.","date":"2014","source":"The international journal of biochemistry & cell biology","url":"https://pubmed.ncbi.nlm.nih.gov/24813785","citation_count":14,"is_preprint":false},{"pmid":"10837797","id":"PMC_10837797","title":"Noradrenergic lesions differentially alter the expression of two subtypes of low Km cAMP-sensitive phosphodiesterase type 4 (PDE4A and PDE4B) in rat brain.","date":"2000","source":"Brain research","url":"https://pubmed.ncbi.nlm.nih.gov/10837797","citation_count":12,"is_preprint":false},{"pmid":"37620166","id":"PMC_37620166","title":"Resveratrol alleviates amyloid β-induced neuronal apoptosis, inflammation, and oxidative and endoplasmic reticulum stress by circ_0050263/miR-361-3p/PDE4A axis during Alzheimer's disease.","date":"2023","source":"Chemical biology & drug design","url":"https://pubmed.ncbi.nlm.nih.gov/37620166","citation_count":11,"is_preprint":false},{"pmid":"10602991","id":"PMC_10602991","title":"Physical mapping and promoter structure of the murine cAMP-specific phosphodiesterase pde4a gene.","date":"2000","source":"Mammalian genome : official journal of the International Mammalian Genome Society","url":"https://pubmed.ncbi.nlm.nih.gov/10602991","citation_count":10,"is_preprint":false},{"pmid":"31370708","id":"PMC_31370708","title":"Discovery of novel Schistosoma mansoni PDE4A inhibitors as potential agents against schistosomiasis.","date":"2019","source":"Future medicinal chemistry","url":"https://pubmed.ncbi.nlm.nih.gov/31370708","citation_count":10,"is_preprint":false},{"pmid":"11566027","id":"PMC_11566027","title":"Purification and characterization of the human PDE4A catalytic domain (PDE4A330-723) expressed in Sf9 cells.","date":"2001","source":"Archives of biochemistry and biophysics","url":"https://pubmed.ncbi.nlm.nih.gov/11566027","citation_count":9,"is_preprint":false},{"pmid":"39653247","id":"PMC_39653247","title":"Ginsenoside Rg1 attenuates Aβ1-42-induced microglial cell apoptosis and inflammation in Alzheimer's disease via the GATA4/PDE4A/PI3K/AKT axis.","date":"2024","source":"Neuroscience","url":"https://pubmed.ncbi.nlm.nih.gov/39653247","citation_count":7,"is_preprint":false},{"pmid":"38797258","id":"PMC_38797258","title":"Low PDE4A expression promoted the progression of ovarian cancer by inducing Snail nuclear translocation.","date":"2024","source":"Experimental cell research","url":"https://pubmed.ncbi.nlm.nih.gov/38797258","citation_count":2,"is_preprint":false},{"pmid":"39560930","id":"PMC_39560930","title":"Potential PDE4A inhibition-mediated neuroprotective effects of psoralidin.","date":"2024","source":"European review for medical and pharmacological sciences","url":"https://pubmed.ncbi.nlm.nih.gov/39560930","citation_count":1,"is_preprint":false},{"pmid":"40879806","id":"PMC_40879806","title":"Circ_0049472 downregulation relieves Amyloid-β-induced neuronal injury by modulating PDE4A expression via targeting miR-22-3p in Alzheimer's disease.","date":"2025","source":"Metabolic brain disease","url":"https://pubmed.ncbi.nlm.nih.gov/40879806","citation_count":1,"is_preprint":false},{"pmid":"41812758","id":"PMC_41812758","title":"A novel PDE4A inhibitor suppresses hepatocellular carcinoma progression through activating the cAMP-induced PKA pathway.","date":"2026","source":"Chemico-biological interactions","url":"https://pubmed.ncbi.nlm.nih.gov/41812758","citation_count":0,"is_preprint":false},{"pmid":"15988062","id":"PMC_15988062","title":"Renaturation of the catalytic domain of PDE4A expressed in Escherichia coli as inclusion bodies.","date":"2005","source":"Methods in molecular biology (Clifton, N.J.)","url":"https://pubmed.ncbi.nlm.nih.gov/15988062","citation_count":0,"is_preprint":false},{"pmid":null,"id":"bio_10.1101_2025.08.11.25333473","title":"Exploring the Potential of Antidepressants in Treating Coronary Artery Disease","date":"2025-08-12","source":"bioRxiv","url":"https://doi.org/10.1101/2025.08.11.25333473","citation_count":0,"is_preprint":true}],"cost":{"stage1":{"model":"claude-sonnet-4-6","input_tokens":20090,"output_tokens":4996,"usd":0.067605,"stage1_stop_reason":"end_turn"},"stage2":{"model":"claude-opus-4-8","input_tokens":13107,"output_tokens":5549,"usd":0.10213,"stage2_stop_reason":"end_turn"},"total_usd":0.169735,"stage1_batch_id":"msgbatch_013a8chtv3AWyTqv4z4sEF4C","stage2_batch_id":"msgbatch_01AHsuW63vinsvuTgzmQ3n62","note":"batch pricing = 50% of standard"},"stage1_raw":"```json\n{\n  \"discoveries\": [\n    {\n      \"year\": 1996,\n      \"finding\": \"The N-terminal splice region of PDE4A splice variant RPDE-6 (RNPDE4A5) interacts with SH3 domains of Src-family tyrosine kinases (v-Src, Lyn, Fyn, c-Abl, Crk, Csk, Lck). Interaction requires the proline-rich N-terminal unique region of RPDE-6; splice variant RPDE-39 (lacking this region) and met26RD1 (N-terminal deletion) fail to associate. Binding to SH3 domains of Crk, Csk, and Lck reduces PDE4A catalytic activity.\",\n      \"method\": \"GST pulldown with v-Src-SH3 fusion protein, co-immunoprecipitation from transfected COS7 cells, competition with N-terminal fusion protein, analysis of deletion and splice variants\",\n      \"journal\": \"The Biochemical journal\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — reciprocal Co-IP, GST pulldown, multiple splice variant controls, functional activity readout, all in one rigorous study\",\n      \"pmids\": [\"8761480\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 1995,\n      \"finding\": \"PDE4A splice variants RD1 (RNPDE4A1A) and RPDE-6 (RNPDE4A5) have distinct subcellular distributions determined by their unique N-terminal splice regions: RD1 is exclusively in the high-speed pellet (P2) membrane fraction, while RPDE-6 distributes between pellet (~25%) and cytosol (~75%) fractions. Pellet RPDE-6 is resistant to high NaCl and Triton X-100 solubilization. Soluble and pellet RPDE-6 show different rolipram IC50 values (~0.16 µM vs ~1.2 µM), indicating conformationally distinct pools.\",\n      \"method\": \"Subcellular fractionation of transfected COS-7 cells and brain tissue, immunoprecipitation with anti-C-terminal antisera, enzymatic activity assays with rolipram and cAMP Km measurements\",\n      \"journal\": \"The Biochemical journal\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — direct fractionation with functional activity measurements, validated in both transfected cells and native brain tissue\",\n      \"pmids\": [\"7575434\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 1997,\n      \"finding\": \"PDE4A splice variant RD1 (RNPDE4A1A) localizes to the Golgi complex in stably transfected human follicular thyroid carcinoma cells. RD1 immunoreactivity colocalizes with Golgi marker Tex1, and redistribution upon treatment with Golgi-perturbing agents monensin and brefeldin A confirmed Golgi targeting. RD1 is membrane-associated (detergent-soluble, not salt-extractable) and located exclusively in the membrane fraction.\",\n      \"method\": \"Laser scanning confocal immunofluorescence, subcellular fractionation, Golgi-disrupting drug treatment (monensin, brefeldin A) in stably transfected FTC cell lines\",\n      \"journal\": \"The Biochemical journal\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — direct imaging with organelle marker colocalization, pharmacological disruption confirming Golgi identity, fractionation data\",\n      \"pmids\": [\"9003417\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 1998,\n      \"finding\": \"Growth hormone activates PDE4A5 in 3T3-F442A preadipocytes via a JAK2-dependent pathway through phosphatidylinositol 3-kinase and p70S6 kinase, resulting in decreased SDS-PAGE mobility (consistent with phosphorylation) and increased catalytic activity. This activation lowers intracellular cAMP. Antisense depletion of PDE4A5 mimicked rolipram in enhancing growth hormone-stimulated adipocyte differentiation. Activation was independent of ERK2, PKC, or transcriptional effects.\",\n      \"method\": \"Kinase inhibitor epistasis, antisense depletion, intracellular cAMP measurement, SDS-PAGE mobility shift assay in 3T3-F442A cells\",\n      \"journal\": \"Proceedings of the National Academy of Sciences of the United States of America\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — pathway epistasis with multiple inhibitors, antisense loss-of-function with defined phenotype, cAMP functional readout, multiple orthogonal methods\",\n      \"pmids\": [\"9520403\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2005,\n      \"finding\": \"AKAP3 selectively binds PDE4A5 but not PDE4D in bovine spermatozoa. Co-immunoprecipitation in COS cells co-transfected with AKAP3 and Pde4a5 or Pde4d4 confirmed selectivity. Pulldown from sperm lysates confirmed the in vivo interaction. PDE4A5 localization shifts from Triton X-100-soluble fraction in cauda epididymal sperm to SDS-soluble (insoluble) fraction in ejaculated sperm during capacitation.\",\n      \"method\": \"Co-immunoprecipitation in co-transfected COS cells, pulldown from sperm lysates, immunolocalization, subcellular fractionation\",\n      \"journal\": \"Biology of reproduction\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Strong — reciprocal pulldown in two systems (transfected cells + native sperm lysate), selectivity demonstrated with PDE4D as negative comparator, localization shift documented\",\n      \"pmids\": [\"16177223\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2005,\n      \"finding\": \"PDE4A11, a novel long-isoform splice variant of human PDE4A, is activated by PKA-mediated phosphorylation at Ser119. PDE4A11 localizes predominantly around the nucleus and in membrane ruffles when expressed in COS-7 cells. It hydrolyzes cAMP with Km ~4 µM. Unlike PDE4A4, PDE4A11 shows differential sensitivity to caspase-3 cleavage and to PDE4 inhibitors, and has distinct rolipram redistribution behavior. All three PDE4A long isoforms (PDE4A4, PDE4A10, PDE4A11) can interact with beta-arrestin.\",\n      \"method\": \"Recombinant expression in COS-7 cells, kinase activity assay with PKA, subcellular fractionation, immunofluorescence localization, inhibitor IC50 measurements, cAMP hydrolysis kinetics\",\n      \"journal\": \"Molecular pharmacology\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — multiple orthogonal methods (enzymatic assay, PKA phosphorylation, localization imaging, inhibitor profiling) in one study; single lab\",\n      \"pmids\": [\"15738310\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2001,\n      \"finding\": \"The human PDE4A catalytic domain (residues 330-723) expressed in Sf9 cells exists as a tetramer at ~1 mg/ml (by light scattering) and is heavily phosphorylated on both serines of the conserved SPS motif (by mass spectrometry). Despite this phosphorylation, the SPS motif is not part of the active site but is positioned near it, as shown by covalent labeling of an adjacent peptide by an electrophilic cAMP analogue. Km for cAMP hydrolysis is ~2 µM.\",\n      \"method\": \"Mass spectrometry for phosphorylation site mapping, light scattering for oligomeric state, covalent labeling with electrophilic cAMP analogue, enzymatic kinetics\",\n      \"journal\": \"Archives of biochemistry and biophysics\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 1 / Weak — in vitro biochemical characterization with MS and affinity labeling, single lab, single study\",\n      \"pmids\": [\"11566027\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2004,\n      \"finding\": \"PDE4A7, an isoform encoded by the human PDE4A gene, lacks catalytic activity due to its unique C-terminal region, not its N-terminal region. Chimera analysis showed that replacing the C-terminal unique portion of PDE4A7 with the conserved C-terminal sequence of active PDE4 isoforms (Hyb1) restored full catalytic activity, whereas replacing the N-terminal portion (Hyb2) did not. Three functional regions within PDE4A isoforms govern catalytic activity, subcellular targeting, and conformational status. PDE4A7 is exclusively in the P1 particulate fraction, and a region in the conserved C-terminal of active PDE4A isoforms prevents this exclusive targeting.\",\n      \"method\": \"Chimeric protein construction and expression, enzymatic activity assays, subcellular fractionation, SDS-PAGE analysis in transfected cells\",\n      \"journal\": \"The Biochemical journal\",\n      \"confidence\": \"High\",\n      \"confidence_rationale\": \"Tier 1 / Moderate — reconstitution through chimera/mutagenesis approach with enzymatic activity as direct readout, multiple constructs tested\",\n      \"pmids\": [\"15025561\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 1997,\n      \"finding\": \"In Jurkat T-cells, forskolin (via cAMP elevation) selectively down-regulates a novel ~118 kDa PDE4A splice variant (distinct from PDE4A4B) at the transcriptional level, while inducing PDE4D1 and PDE4D2. Immunoprecipitation showed the ~118 kDa PDE4A species provides all PDE4 activity in control cells. This down-regulation is blocked by actinomycin D, confirming transcriptional dependence. The effect is mimicked by cholera toxin and 8-bromo-cAMP.\",\n      \"method\": \"RT-PCR, immunoblotting, immunoprecipitation with PDE4-selective antisera, pharmacological inhibition of transcription with actinomycin D in Jurkat T-cells\",\n      \"journal\": \"The Biochemical journal\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — immunoprecipitation activity assay plus transcriptional inhibitor epistasis, multiple pharmacological tools, single lab\",\n      \"pmids\": [\"9003416\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 1999,\n      \"finding\": \"Inhibition of T-cell proliferation and LPS-stimulated TNFα release from monocytes by subtype-selective PDE4 inhibitors correlates significantly with inhibition of recombinant human PDE4A or PDE4B catalytic activity, but not PDE4D. This establishes that PDE4A (and/or PDE4B) plays the major functional role in regulating these inflammatory cell functions.\",\n      \"method\": \"Recombinant human PDE4 subtype enzymatic assays correlated by linear regression and Spearman's rank-order with cellular functional assays (T-cell proliferation, TNFα release) using 10 subtype-selective inhibitors\",\n      \"journal\": \"British journal of pharmacology\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — pharmacological correlation across 10 compounds with two orthogonal functional readouts and statistical analysis; indirect (no genetic KO), single study\",\n      \"pmids\": [\"10602317\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2014,\n      \"finding\": \"AKAP149-PKA-PDE4A complex redistributes within K-562 cells following YTX treatment: the complex decreases in cytosol and increases in plasma membrane (at 24 h, associated with apoptosis/caspase activation) and then in the nucleus (at 48 h, associated with non-apoptotic cell death). Silencing of either AKAP149 or PDE4A prevented YTX-induced cell death, establishing the complex as required for YTX cytotoxicity.\",\n      \"method\": \"Subcellular fractionation, AKAP149/PDE4A siRNA silencing, caspase activity assays, western blotting in K-562 cells\",\n      \"journal\": \"The international journal of biochemistry & cell biology\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — loss-of-function (siRNA) with defined cell death phenotype and subcellular fractionation; single lab, single study\",\n      \"pmids\": [\"24813785\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2015,\n      \"finding\": \"PDE4A is required for autophagy triggered by yessotoxin (YTX) in K-562 cells. PDE4A silencing experiments showed that PDE4A regulates distinct steps of the autophagic process induced by YTX versus classical autophagy inducers (e.g., rapamycin), establishing PDE4A as a key mediator of YTX-induced autophagy after 48 h treatment.\",\n      \"method\": \"PDE4A siRNA silencing, autophagy marker analysis (mTOR, LC3B), rapamycin as comparator in K-562 cells\",\n      \"journal\": \"Toxicology\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 3 / Moderate — siRNA loss-of-function with autophagy pathway markers; pathway placement partially established; single lab\",\n      \"pmids\": [\"25576684\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2007,\n      \"finding\": \"NMDA receptor activity regulates PDE4A1 and PDE4A5 expression in rat primary cortical cultures. Chronic blockade of NMDA receptors with MK-801 reduces PDE4A1 and PDE4A5 expression/activity in a time-dependent manner, reversed by the PKA activator dbr-cAMP. NR1/NR2B-induced cGMP signaling (via PDE4) negatively cross-regulates NR1/NR2A-induced cAMP levels. GABA receptor inhibition increases NMDA-induced cAMP and PDE4A expression in mature but not young cultures.\",\n      \"method\": \"Pharmacological manipulation (MK-801, ifenprodil, bicuculline, dbr-cAMP) with PDE4 activity assays and immunoblot in rat primary cortical/hippocampal cultures\",\n      \"journal\": \"Brain research\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — multiple pharmacological tools with enzymatic activity and protein expression readouts; no genetic KO, single lab\",\n      \"pmids\": [\"17407767\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 1998,\n      \"finding\": \"Human PDE4A short isoform RD1 (homologue of rat RNPDE4A1A), when transiently expressed in COS-7 cells, appears as an 83 kDa species primarily in the high-speed membrane fraction. It exhibits Km for cAMP of ~3 µM and IC50 for rolipram of ~0.3 µM. In vitro transcription-translation shows RD1 is produced as an 80 kDa species capable of binding to membranes. The gene spans 50 kb with at least 17 exons, located at chromosome 19p13.2.\",\n      \"method\": \"Transient expression in COS-7 cells, subcellular fractionation, in vitro transcription-translation, enzymatic activity assay\",\n      \"journal\": \"The Biochemical journal\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — direct expression and fractionation with functional enzymatic assay and in vitro translation; single lab\",\n      \"pmids\": [\"9677330\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2024,\n      \"finding\": \"GATA4 transcription factor directly binds the PDE4A promoter and upregulates PDE4A expression in Aβ1-42-stimulated BV2 microglial cells (confirmed by Jaspar prediction and dual-luciferase reporter assay). Increased PDE4A expression downstream of GATA4 inactivates the PI3K/AKT pathway, promoting microglial apoptosis and inflammation.\",\n      \"method\": \"Dual-luciferase reporter assay for GATA4-PDE4A promoter interaction, GATA4 knockdown/overexpression, PDE4A overexpression, western blot for PI3K/AKT pathway in BV2 cells\",\n      \"journal\": \"Neuroscience\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — luciferase reporter for direct transcription factor-promoter interaction, epistasis with PDE4A overexpression; single lab\",\n      \"pmids\": [\"39653247\"],\n      \"is_preprint\": false\n    },\n    {\n      \"year\": 2024,\n      \"finding\": \"Knockdown of PDE4A in ovarian cancer cells promotes proliferation, migration, and invasion, while overexpression suppresses these processes. PDE4A loss induces epithelial-mesenchymal transition (EMT) and nuclear translocation of Snail. In vivo, PDE4A-overexpressing OVCAR3 cells formed fewer and smaller metastatic foci. Rolipram (PDE4 inhibitor) mimicked PDE4A deletion effects.\",\n      \"method\": \"PDE4A knockdown and overexpression in OC cell lines, in vitro proliferation/migration/invasion assays, in vivo mouse metastasis model, EMT marker analysis, Snail nuclear localization by western blot\",\n      \"journal\": \"Experimental cell research\",\n      \"confidence\": \"Medium\",\n      \"confidence_rationale\": \"Tier 2 / Moderate — loss- and gain-of-function with defined cellular and molecular phenotype (Snail/EMT) in vitro and in vivo; single lab\",\n      \"pmids\": [\"38797258\"],\n      \"is_preprint\": false\n    }\n  ],\n  \"current_model\": \"PDE4A encodes a cAMP-specific phosphodiesterase whose multiple splice variants (e.g., PDE4A1/RD1, PDE4A5/RPDE-6, PDE4A4, PDE4A7, PDE4A11) are differentiated by unique N- or C-terminal regions that govern subcellular targeting (Golgi, membranes, cytosol, perinuclear), protein interactions (SH3 domains of Src-family kinases via a proline-rich N-terminal motif; AKAP3 in sperm; AKAP149 in somatic cells; beta-arrestin), and catalytic regulation (PKA phosphorylation at Ser119 activates long isoforms; p70S6K pathway activated by growth hormone increases PDE4A5 activity; GATA4 transcriptionally upregulates PDE4A); PDE4A activity degrades cAMP in inflammatory cells, neurons, and cancer cells, thereby modulating PKA signaling, Snail/EMT-driven cancer invasion, and autophagy.\"\n}\n```","stage2_raw":"{\n  \"mechanistic_narrative\": \"PDE4A encodes a cAMP-specific phosphodiesterase whose multiple splice variants share a conserved catalytic core but are functionally individualized by unique N- and C-terminal regions that dictate subcellular targeting, protein interactions, and catalytic regulation [#1, #7]. The N-terminal splice region defines localization—the short RD1 isoform partitions exclusively into membranes and concentrates at the Golgi complex (detergent-soluble, salt-resistant), whereas RPDE-6 distributes between membrane and cytosol as conformationally distinct pools [#1, #2]. The unique C-terminal region is equally decisive: it can abolish catalytic activity, as in PDE4A7, where swapping the variant C-terminus for the conserved active sequence restores both catalysis and normal targeting [#7]. The N-terminal proline-rich region also mediates protein interactions, binding the SH3 domains of Src-family kinases (v-Src, Lyn, Fyn, Lck, Csk, Crk) in a manner that can suppress catalytic activity [#0], while long isoforms engage scaffolds including AKAP3 in sperm, AKAP149 in somatic cells, and beta-arrestin [#4, #5, #10]. Catalytic output is tuned by phosphorylation: PKA phosphorylation at Ser119 activates the long isoform PDE4A11 [#5], and a growth hormone–driven JAK2/PI3K/p70S6 kinase cascade activates PDE4A5 to lower cAMP and restrain adipocyte differentiation [#3]. Through cAMP hydrolysis, PDE4A shapes physiological outputs across cell types—it provides the dominant PDE4 activity controlling T-cell proliferation and monocyte TNFα release [#9], is regulated by NMDA-receptor signaling in cortical neurons [#12], mediates yessotoxin-induced cell death and autophagy via the AKAP149–PKA–PDE4A complex [#10, #11], and acts as a suppressor of Snail/EMT-driven invasion in ovarian cancer [#15]. Transcriptionally, GATA4 directly upregulates PDE4A, which in turn dampens PI3K/AKT signaling to promote microglial apoptosis and inflammation [#14].\",\n  \"teleology\": [\n    {\n      \"year\": 1995,\n      \"claim\": \"Established that PDE4A splice variants are not functionally redundant but are differentially targeted within the cell by their unique N-terminal regions, defining the principle of isoform-specific compartmentalization.\",\n      \"evidence\": \"Subcellular fractionation of transfected COS-7 cells and brain tissue with rolipram/cAMP activity assays comparing RD1 and RPDE-6\",\n      \"pmids\": [\"7575434\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Did not identify the membrane anchor or docking partner retaining RD1 in the pellet fraction\", \"Mechanism producing two conformationally distinct RPDE-6 pools unresolved\"]\n    },\n    {\n      \"year\": 1996,\n      \"claim\": \"Identified the first protein-interaction function for the PDE4A N-terminus, showing the proline-rich splice region binds SH3 domains of Src-family kinases and that binding modulates catalytic activity—linking PDE4A to tyrosine-kinase signaling scaffolds.\",\n      \"evidence\": \"GST pulldown with v-Src-SH3, reciprocal Co-IP from COS7 cells, deletion/splice-variant controls with activity readout\",\n      \"pmids\": [\"8761480\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Physiological consequence of kinase scaffolding in a native cell unestablished\", \"Whether PDE4A is a tyrosine-kinase substrate not addressed\"]\n    },\n    {\n      \"year\": 1997,\n      \"claim\": \"Resolved the specific organelle target of the RD1 isoform, demonstrating Golgi localization and thereby connecting a defined PDE4A variant to a discrete subcellular cAMP microdomain.\",\n      \"evidence\": \"Confocal immunofluorescence with Golgi marker colocalization, monensin/brefeldin A disruption, and fractionation in stably transfected FTC cells\",\n      \"pmids\": [\"9003417\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Golgi-targeting determinant within the N-terminus not mapped to residues\", \"Functional role of Golgi-localized cAMP hydrolysis not tested\"]\n    },\n    {\n      \"year\": 1997,\n      \"claim\": \"Showed PDE4A is itself a transcriptional target of cAMP signaling, with a ~118 kDa variant providing all PDE4 activity in T-cells and being down-regulated upon cAMP elevation, revealing a feedback architecture.\",\n      \"evidence\": \"RT-PCR, immunoprecipitation activity assay, and actinomycin D epistasis with forskolin/cholera toxin/8-Br-cAMP in Jurkat T-cells\",\n      \"pmids\": [\"9003416\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Identity of the transcriptional regulator mediating cAMP-driven repression unknown\", \"Single cell line; generality across T-cell states untested\"]\n    },\n    {\n      \"year\": 1998,\n      \"claim\": \"Defined a hormone-driven signaling pathway that activates a specific PDE4A variant, establishing PDE4A5 as a regulated effector that lowers cAMP to restrain adipocyte differentiation.\",\n      \"evidence\": \"Kinase-inhibitor epistasis (JAK2/PI3K/p70S6K), antisense depletion, cAMP measurement, and mobility-shift in 3T3-F442A preadipocytes\",\n      \"pmids\": [\"9520403\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Direct phosphorylation site(s) on PDE4A5 not mapped\", \"Whether p70S6K phosphorylates PDE4A directly or via an intermediate kinase unresolved\"]\n    },\n    {\n      \"year\": 1998,\n      \"claim\": \"Characterized the human RD1 short isoform biochemically and placed the gene physically at chromosome 19p13.2, providing the human counterpart to rat targeting/activity findings.\",\n      \"evidence\": \"Transient expression, fractionation, in vitro transcription-translation, and enzyme kinetics in COS-7 cells with gene structure mapping\",\n      \"pmids\": [\"9677330\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Membrane-binding determinant in human RD1 not defined at residue level\"]\n    },\n    {\n      \"year\": 1999,\n      \"claim\": \"Assigned the functional inflammatory role of PDE4A by correlating subtype-selective inhibitor potency against recombinant enzyme with suppression of T-cell proliferation and monocyte TNFα release.\",\n      \"evidence\": \"Recombinant PDE4 subtype enzymatic assays correlated by regression/Spearman analysis with two cellular readouts across 10 inhibitors\",\n      \"pmids\": [\"10602317\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Pharmacological correlation cannot distinguish PDE4A from PDE4B contribution\", \"No genetic loss-of-function confirmation\"]\n    },\n    {\n      \"year\": 2001,\n      \"claim\": \"Provided biochemical/structural insight into the catalytic domain, showing it is tetrameric and phosphorylated on the SPS motif positioned adjacent to—but not part of—the active site.\",\n      \"evidence\": \"Mass spectrometry phosphosite mapping, light scattering, and covalent labeling with an electrophilic cAMP analogue on the Sf9-expressed catalytic domain\",\n      \"pmids\": [\"11566027\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Functional consequence of SPS phosphorylation on activity not demonstrated\", \"No high-resolution crystal structure reported here\"]\n    },\n    {\n      \"year\": 2004,\n      \"claim\": \"Established that the unique C-terminal region, not the N-terminus, can govern catalytic competence, refining the modular model: three functional regions control activity, targeting, and conformation.\",\n      \"evidence\": \"Chimera/mutagenesis (Hyb1/Hyb2) with activity assays and fractionation of the catalytically dead PDE4A7 isoform\",\n      \"pmids\": [\"15025561\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Physiological function of catalytically inactive PDE4A7 unknown\", \"Mechanism by which the C-terminus suppresses activity not structurally defined\"]\n    },\n    {\n      \"year\": 2005,\n      \"claim\": \"Identified the long isoform PDE4A11 and showed PKA phosphorylation at Ser119 activates it, defining a PKA-feedback node and extending the beta-arrestin interaction to all PDE4A long isoforms.\",\n      \"evidence\": \"Recombinant expression, PKA phosphorylation/activity assay, fractionation, immunofluorescence, and inhibitor profiling in COS-7 cells\",\n      \"pmids\": [\"15738310\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Functional consequence of beta-arrestin binding for PDE4A11 not tested\", \"In vivo relevance of perinuclear/membrane-ruffle localization unestablished\"]\n    },\n    {\n      \"year\": 2005,\n      \"claim\": \"Demonstrated isoform-selective scaffolding in a specialized cell type, with AKAP3 binding PDE4A5 (but not PDE4D) in sperm and PDE4A5 solubility shifting during capacitation, tying PDE4A to compartmentalized cAMP control in fertilization.\",\n      \"evidence\": \"Reciprocal Co-IP in transfected COS cells, pulldown from sperm lysates with PDE4D negative control, and fractionation across capacitation states\",\n      \"pmids\": [\"16177223\"],\n      \"confidence\": \"High\",\n      \"gaps\": [\"Functional effect of AKAP3 anchoring on sperm cAMP signaling not directly measured\", \"Binding interface on PDE4A5 not mapped\"]\n    },\n    {\n      \"year\": 2007,\n      \"claim\": \"Placed PDE4A within neuronal signaling, showing NMDA-receptor activity regulates PDE4A1/PDE4A5 expression and that PDE4 mediates cGMP-dependent cross-regulation of cAMP between NMDA receptor subtypes.\",\n      \"evidence\": \"Pharmacological manipulation (MK-801, ifenprodil, bicuculline, dbr-cAMP) with PDE4 activity and immunoblot in rat cortical cultures\",\n      \"pmids\": [\"17407767\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Transcriptional mechanism linking NMDA receptor to PDE4A expression unknown\", \"No genetic confirmation of PDE4A-specific contribution\"]\n    },\n    {\n      \"year\": 2014,\n      \"claim\": \"Defined a functional AKAP149-PKA-PDE4A complex whose dynamic redistribution is required for yessotoxin-induced cell death, demonstrating that anchored PDE4A controls a cytotoxic cAMP signaling outcome.\",\n      \"evidence\": \"Subcellular fractionation, AKAP149/PDE4A siRNA, and caspase assays in YTX-treated K-562 cells\",\n      \"pmids\": [\"24813785\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Mechanism coupling complex relocation to apoptosis vs non-apoptotic death unresolved\", \"Single cell line and single stimulus\"]\n    },\n    {\n      \"year\": 2015,\n      \"claim\": \"Extended PDE4A's role to autophagy regulation, showing it is required for yessotoxin-induced autophagy and acts at steps distinct from classical inducers.\",\n      \"evidence\": \"PDE4A siRNA with mTOR/LC3B autophagy markers and rapamycin comparator in K-562 cells\",\n      \"pmids\": [\"25576684\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Molecular step in autophagy controlled by PDE4A not defined\", \"Dependence on cAMP/PKA versus scaffolding not separated\"]\n    },\n    {\n      \"year\": 2024,\n      \"claim\": \"Identified PDE4A as a tumor-suppressive regulator of cancer cell invasion, with loss promoting EMT and Snail nuclear translocation and overexpression suppressing metastasis.\",\n      \"evidence\": \"PDE4A knockdown/overexpression with proliferation/migration/invasion assays, EMT/Snail analysis, and an in vivo OVCAR3 metastasis model\",\n      \"pmids\": [\"38797258\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Mechanistic link between cAMP hydrolysis and Snail regulation not defined\", \"Single cancer type; clinical correlation absent\"]\n    },\n    {\n      \"year\": 2024,\n      \"claim\": \"Placed PDE4A downstream of GATA4 in neuroinflammation, showing GATA4 directly drives PDE4A transcription to inactivate PI3K/AKT and promote microglial apoptosis and inflammation.\",\n      \"evidence\": \"Dual-luciferase promoter assay, GATA4 knockdown/overexpression, PDE4A overexpression, and PI3K/AKT western blot in BV2 microglia\",\n      \"pmids\": [\"39653247\"],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"Whether PDE4A acts via cAMP hydrolysis or a non-catalytic mechanism on PI3K/AKT untested\", \"In vivo relevance to neurodegeneration not established\"]\n    },\n    {\n      \"year\": null,\n      \"claim\": \"How PDE4A isoform-specific compartmentalization and scaffold binding are integrated with its disease roles in cancer and neuroinflammation, and whether these depend on cAMP catalysis or non-catalytic functions, remains unresolved.\",\n      \"evidence\": \"\",\n      \"pmids\": [],\n      \"confidence\": \"Medium\",\n      \"gaps\": [\"No structural model linking N/C-terminal targeting to substrate-level cAMP microdomain control in disease cells\", \"Catalytic versus scaffolding contributions to EMT and PI3K/AKT effects not dissected\", \"No genetic in vivo loss-of-function model defining endogenous PDE4A function\"]\n    }\n  ],\n  \"mechanism_profile\": {\n    \"molecular_activity\": [\n      {\"term_id\": \"GO:0140098\", \"supporting_discovery_ids\": [1, 5, 6, 7, 13]},\n      {\"term_id\": \"GO:0016787\", \"supporting_discovery_ids\": [5, 6, 13]}\n    ],\n    \"localization\": [\n      {\"term_id\": \"GO:0005794\", \"supporting_discovery_ids\": [2]},\n      {\"term_id\": \"GO:0005886\", \"supporting_discovery_ids\": [1, 5, 10, 13]},\n      {\"term_id\": \"GO:0005829\", \"supporting_discovery_ids\": [1, 10]},\n      {\"term_id\": \"GO:0005634\", \"supporting_discovery_ids\": [5, 10]}\n    ],\n    \"pathway\": [\n      {\"term_id\": \"R-HSA-162582\", \"supporting_discovery_ids\": [0, 3, 5, 12]},\n      {\"term_id\": \"R-HSA-168256\", \"supporting_discovery_ids\": [9, 14]},\n      {\"term_id\": \"R-HSA-9612973\", \"supporting_discovery_ids\": [11]},\n      {\"term_id\": \"R-HSA-5357801\", \"supporting_discovery_ids\": [10, 14]},\n      {\"term_id\": \"R-HSA-1643685\", \"supporting_discovery_ids\": [14, 15]}\n    ],\n    \"complexes\": [\n      \"AKAP149-PKA-PDE4A complex\"\n    ],\n    \"partners\": [\n      \"AKAP3\",\n      \"AKAP149\",\n      \"ARRB\",\n      \"SRC\",\n      \"LYN\",\n      \"FYN\",\n      \"LCK\",\n      \"CSK\"\n    ],\n    \"other_free_text\": []\n  }\n}","audit_flag":null,"evaluation":{"pairwise":"win","faith_supported":6,"faith_total":7,"faith_pct":85.71428571428571}}