DGKE performs ATP-dependent phosphorylation of DAG to PA. Human recombinant experiments support this chemistry directly; mouse knockout lipidomics supply additional pathway context. One catalytic core combines ATP-dependent DAG kinase activity with PA production and upstream participation in PI recycling. The enzyme is not a PI synthase. DAG metabolism and lipid phosphorylation describe the same chemistry, rather than additional catalytic units. The current typed ontology does not make lipid phosphorylation an ancestor of PA biosynthesis (PMID:18004883(https://doi.org/10.1021/bi701584v), PMID:19744926(https://doi.org/10.1074/jbc.M109.050617), PMID:21477596(https://pmc.ncbi.nlm.nih.gov/articles/PMC3095698/), PMID:22108654(https://pmc.ncbi.nlm.nih.gov/articles/PMC3237191/)).
The four-source independent chemistry assessment follows. The two abstract-only normal caches retain that flag even though original PDFs were recovered separately. Original-only quotations remain in these notes; YAML anchors use literal text in the normal caches.
PMID:8626589(https://pubmed.ncbi.nlm.nih.gov/8626589/) — Tang et al., J Biol Chem 271, 10237–10241 (1996). The original PDF identifies an open reading frame of “567 residues” (p. 10237). Human endothelial-cell cDNA was expressed in COS-7 cells, and vector-controlled lysate assays measured radiolabeled PA production with a preference for arachidonoyl-containing DAG (Methods p. 10238; Results and Figure 3 caption pp. 10239–10240). The cDNA truncation removed untranslated regions, not the protein's catalytic domain. The reported supernatant activity is qualified by poor total activity recovery and does not establish exclusively soluble localization.
PMID:19744926(https://pubmed.ncbi.nlm.nih.gov/19744926/) — Lung et al., doi:10.1074/jbc.M109.050617. The original PDF describes human full-length/engineered Δ58 proteins in COS-7 cells and His6-DGKE in Sf21 cells. Enriched detergent lysates, mock controls and ATP-to-PA assays support human catalytic activity and acyl-chain selectivity (Methods pp. 31064–31065; Results/Figures 4 and 7, pp. 31067–31071). Separate knockout mouse-fibroblast lipidomics found 29% less stearoyl-containing PI; the method “does not determine their position on the glycerol backbone” (p. 31065). This supports PI-cycle participation without proving direct human PI synthesis, positional lipid flux or a physiological feedback circuit. Δ58 is not the natural short isoform.
PMID:21477596(https://pubmed.ncbi.nlm.nih.gov/21477596/) — Shulga et al., doi:10.1016/j.jmb.2011.03.071, PMC3095698. The Materials and Methods subsection on DGKε constructs states: “All correspond to human forms of the respective enzymes.” FLAG-DGKE wild type and mutants were expressed in COS-7 cells; detergent-lysate assays measured ATP-dependent PA formation with mock subtraction, lipid dependence and a DGKalpha comparator (Materials and Methods, Enzyme Preparation and Enzymatic Activity). Mutations near the proposed selectivity motif alter activity and acyl-chain preference. This is functional mutagenesis, not a solved substrate-binding structure. PIP5Kalpha experiments in the paper do not establish that activity for DGKE.
PMID:22108654(https://pubmed.ncbi.nlm.nih.gov/22108654/) — Shulga et al., doi:10.1016/j.febslet.2011.11.016, PMC3237191. Human His6-DGKE expressed in Sf21 cells was assayed in detergent extracts and mixed micelles (Methods §§2.1, 2.2 and 2.4). Results and Discussion, including the Figure 1/Table 1 discussion, show that arachidonoyl chains at both positions can support activity similar to stearoyl/arachidonoyl DAG: “18:0/20:4-DAG is not unique in being a good substrate for DGKε”. Selectivity is stricter at sn-2 than at sn-1; linoleoyl at sn-2 is a poorer substrate. The membrane-penetration explanation is a hypothesis. None of these studies tests the natural 177-residue product.
The separately recovered PMID:18004883 original, published pp. 14225–14231, was read at Methods/Results pp. 14227–14230, Table 1 and Figure 5. Human C-terminal His6-DGKE in Sf21-cell detergent extracts phosphorylates exogenous DAG with radiolabeled ATP. No-enzyme/mock controls and substrate comparisons support the reaction and stereoselectivity. The small 2,3-DAG signal has a possible 1,2-isomer contamination explanation; the extracts are not a purified native membrane system. The normal cache remains abstract-only.
PMID:23542698(https://pmc.ncbi.nlm.nih.gov/articles/PMC3719402/), Methods and Figure 4b, identifies endogenous DGKE in both membrane and cytoplasmic fractions of unstimulated human platelets, with Na,K-ATPase and β-tubulin controls. Human endothelial and renal expression are distinct from the rat kidney experiments. The platelet fractionation does not identify a unique membrane compartment. The official Human Protein Atlas record reports approved cytosol/nucleoplasm staining with antibody-based ICC/IF in A-431, U-251MG and U2OS. This supports the existing cytosol annotation and refinement of broad cytoplasm; it does not justify a new nucleoplasm assertion or product-specific localization. The general membrane location remains appropriate for the catalytic core.
PMID:23274426(https://pmc.ncbi.nlm.nih.gov/articles/PMC3582208/), Methods and Results, tests full-length human myc-DGKE, c.610delA and siRNA in HEK293T cells with M3-receptor-driven TRPC6 currents. This supports intracellular lipid signaling. The current is a DAG-sensitive functional readout, not direct DAG quantification or an endogenous podocyte recording. Platelet activation remains non-core pathway context: the human disease study proposes a thrombosis mechanism, and nonselective DGK inhibitor observations do not establish a selective DGKE platelet mechanism.
The full mouse original PMID:11287665 shows weaker high-frequency-stimulation-induced EPSP potentiation in knockout perforant path–dentate granule-cell synapses, with basic membrane properties preserved. It does not image DGKE in a pre- or postsynaptic compartment. The exact Q9R1C6/ENSMUSP00000000285 donor and SynGO annotations ground the non-core orthology annotations; their is_active_in relation is preserved. Mouse results are not presented as a direct human synaptic assay.
The abstract of PMID:15544348(https://pubmed.ncbi.nlm.nih.gov/15544348/) explicitly includes epsilon activity despite its alpha-focused title. PMID:23949095(https://pubmed.ncbi.nlm.nih.gov/23949095/) explicitly includes epsilon in a ten-isozyme comparison. Both full originals remain inaccessible. Their established catalytic annotations are retained with the target-positive abstracts and independently read human DGKE assays; no inaccessible construct detail is invented. In contrast, the PMID:19946888 NK-cell membrane screen could not be checked at the DGKE peptide/table level, so that source-specific membrane annotation remains UNDECIDED rather than being rejected as a wrong target.
The current PTHR11255 tree places human P52429 in the epsilon subfamily below IBD node PTN007499105. The literal root IBDs cover ATP-dependent DAG kinase activity, membrane, glycerolipid metabolism and intracellular signal transduction. Human target evidence among the descendant sources is valid grounding, not circularity. The signaling IBD has DGKZ, DGKK, DGKH and Dictyostelium descendant evidence; donor-specific RAF/RAS or immune biology is not transferred to DGKE. Target human receptor-current experiments independently support the general intracellular-signaling context.
The exact Reactome route is platelet pathway 76002 → DAG/IP3 signaling 114508 → reaction 426240 → catalyst 426239 → set 426070 → DGKE member 426061. That member identifies human P52429 residues 1–567 at plasma membrane. It is an explicit catalyst for DAG-to-PA conversion, not mere pathway membership. The curated plasma-membrane and platelet contexts remain non-core because the independently read native fractionation does not resolve a specific membrane. The model does not transfer activity to the natural 177-residue product.
Literal InterPro and ARBA mappings were inspected. IPR000756/IPR037607 and ARBA00028046 support the specific reaction; IPR001206 supplies broader kinase activity. The primate/domain clause of ARBA00027406 supplies platelet context. The current ARBA00028351 output is glycerophospholipid biosynthesis, but its exact DGKE-matching clause was unresolved. Direct target chemistry supports refinement to PA biosynthesis without inventing a historical rule condition. Broad supported terms use MODIFY even where the specific assertion already occurs elsewhere, under the explicit project instructions.
All 41 imported source dictionaries and both product identities/sequence notes are preserved. The proposed actions are 25 ACCEPT, 8 MODIFY, 7 KEEP_AS_NON_CORE and 1 UNDECIDED, with no NEW annotations. The natural short product lacks the canonical catalytic-domain segment, but this does not establish its inactivity, stability or cellular role. Normal gene/publication retrieval succeeded. Falcon authorization and fallback quota failures produced no provider research document; this manually authored record preserves that distinction. The source guide provides public original and database locators.
The broad membrane assignment from the NK-cell screen is retained as ACCEPT on curator deference and independent endogenous human platelet membrane evidence (PMID:19946888(https://pubmed.ncbi.nlm.nih.gov/19946888/), PMID:23542698(https://pmc.ncbi.nlm.nih.gov/articles/PMC3719402/), Methods/Figure 4b). The particular NK-cell peptide/table entry was not independently verified. This access limitation does not negate established DGKE membrane association and does not justify a specific membrane subcompartment. This supersedes the earlier UNDECIDED proposal above. Current decisions are 26 ACCEPT, 8 MODIFY and 7 KEEP_AS_NON_CORE, with no NEW annotation.
The exact public Reactome records make the modeled catalytic role reproducible: platelet pathway 76002 → DAG/IP3 signaling 114508 → DAG-to-PA reaction 426240 → catalyst activity 426239 → enzyme set 426070 → DGKE member 426061. The last record identifies P52429 residues 1–567 at plasma membrane. The pathway context is retained without inferring activity of the natural short product or a selective platelet-activation mechanism.