This guide distinguishes original experiments, database assertions and access limits. It provides locators, not an independent experimental citation. Publication caches retain their fetched bytes and metadata; a cached full_text_available flag does not guarantee that a required figure, method or supplementary target table is present. No publication quotations were added.
| Source | Original location inspected | Scope and limit |
|---|---|---|
| PMID:34431498(https://pmc.ncbi.nlm.nih.gov/articles/PMC8454701/) | Protein expression/purification and phosphorylation Methods; Results and Figure 4, including the actual Figure 4 image | Purified human EPHB4 residues 588–987, WT and YY590/596FF reagents, undergo ATP-dependent phosphorylation and phosphorylate EPHB6. This is a kinase-domain-containing preparation, not an intact ligand-stimulated receptor. Reported substrate sites belong to EPHB6. The downloaded supplement was not independently read. |
| PMID:27400125(https://pmc.ncbi.nlm.nih.gov/articles/PMC4966301/) | Methods for EPHB4 constructs and endothelial assays; Results on receptor phosphorylation and lymphovenous valves | Human receptor constructs and ephrin-B2 stimulation support receptor kinase activity. Separate conditional mouse experiments concern developmental lymphatic/venous anatomy. No equivalence of the four human products follows. |
| PMID:22555806(https://www.tandfonline.com/doi/full/10.4161/cbt.20080) | Publisher-indexed original Results and Figures 1–5 captions; separately indexed Methods subsections on cell lines/reagents, proliferation and siRNA | Human HUVEC and MCF-7 cells show receptor-specific, opposite ERK responses to ephrin-B2 or agonist antibody, with EPHB4/EPHB2 knockdown controls. The reagent is murine ephrin-B2 extracellular domain fused to human Fc. RasGAP knockdown supports functional coupling, not isolated EPHB4–RASA1 binding. Downloadable full main text, complete Methods and figure pixels were unavailable. |
| PMID:16424904(https://www.embopress.org/doi/full/10.1038/sj.emboj.7600949) | Original Results, Figures 2–5 and 7–8 captions; Methods on viral constructs and endothelial targeting | Ecotropic delivery targets proliferating mouse endothelial cells in tumor models. Full-length and artificial kinase-truncated EphB4 affect branching, sprouting and vessel morphology; the evidence includes ephrin-B2 reverse signaling and does not isolate every phenotype to forward receptor catalysis. The publisher main Results and operative Methods were read independently of the abstract-only normal cache. Figure captions were read; figure pixels and supplementary bodies were not. The Methods specify an artificial 1–611 truncation but do not identify the receptor cDNA species, so mouse host and construct species are kept distinct. |
| PMID:28687708(https://www.researchgate.net/publication/318288597_Germline_Loss-of-Function_Mutations_in_EPHB4_Cause_a_Second_Form_of_Capillary_Malformation-Arteriovenous_Malformation_CM-AVM2_Deregulating_RAS-MAPK_Signaling) | Author-uploaded original Methods, Results, Figure 2 legend and supplementary Methods page 4 | EGFP-EPHB4 in COS-7 shows WT membranous staining and variant-dependent abundance/localization changes. Exact construct species/isoform was not specified in the inspected methods. No figure pixels or full downloadable PDF were inspected. Disease association is not by itself a particular angiogenesis assay. |
| PMID:30578106(https://pmc.ncbi.nlm.nih.gov/articles/PMC10292091/) | STAR Methods, Mutagenesis of Eph-B4 and Plasmid Transfection; Co-immunoprecipitation and Western Blotting; Results and actual Figure 4D–E image | The HA-tagged receptor construct is mouse Ephb4, expressed in human HEK293T. WT receptor associates with Ras-GAP; kinase-domain mutants alter phosphorylation/association. Human cell host does not make the construct human. The normal cache contains only part of this original; the operative methods were recovered independently. |
| PMID:10518221(https://doi.org/10.1016/S1097-2765(00)80342-1) | Verified abstract and mouse donor annotations | The abstract supports EphB4/ephrin-B2 vascular signaling. Full original and heart-specific morphogenesis experiments were unavailable; the two transferred heart annotations remain undecided. |
The older sources PMID:12734395(https://doi.org/10.1242/jcs.00426), PMID:15930280(https://doi.org/10.1158/0008-5472.CAN-04-2667), PMID:16867992(https://doi.org/10.1074/jbc.M605766200) and PMID:8188704(https://doi.org/10.1016/S0021-9258(17)36776-5) were not recovered as complete operative originals. Abstracts, introductory/partial original text and, for PMID:16867992, the linked human EphB4–ephrin-B2 structure 2HLE were distinguished from full Methods/Results. Independent originals support the general receptor and plasma-membrane claims. The specific sprouting-migration and cell-adhesion claims from PMID:12734395 remain uncertain.
| Source | Exact location | Interpretation |
|---|---|---|
| PMID:26481148(https://pmc.ncbi.nlm.nih.gov/articles/PMC4643364/) | Main Methods for MST/SPR and ligand-binding assays; Results/Figures 1–3 and captions | Recombinant binding and receptor-dependent human tumor-cell responses support a contextual erythropoietin receptor role. EPO–EphB4 apparent affinity is weaker than EPO–EPOR in the reported comparison. Commercial recombinant boundaries were not fully specified in the inspected main Methods. This is not evidence of a general physiological erythropoiesis function. Current PubMed XML had no linked correction/retraction entry; this is a bounded metadata check, not a claim of universal agreement. |
| PMID:28514442(https://pmc.ncbi.nlm.nih.gov/articles/PMC5531611/) | Original Methods on tagged-bait AP-MS and scoring; official BioPlex 2 interaction list | The actual list contains EFNB2/P52799 bait and EPHB4/P54760 prey. Complex recovery does not alone establish a binary interface. |
| PMID:33961781(https://pmc.ncbi.nlm.nih.gov/articles/PMC8165030/) | Official HEK293T BioPlex 3 network, EFNB2–EPHB4 row | Target pair recovered from the data; complete original Methods were not present in the inspected normal cache. |
| PMID:35384245(https://pmc.ncbi.nlm.nih.gov/articles/PMC9171411/) | Main Methods; actual Figure EV2 localization image; Dataset EV1, sheet EV1A, header row 8 | Excel rows 1751 (EPHA7 WT AP-MS→EPHB4), 1792 (EPHA5 WT AP-MS→EPHB4), 5493 (EPHB4 WT AP-MS→EFNB2) and 5498 (EPHB4 WT AP-MS→EPHA6) support the named associations. Human tagged receptors were expressed in HEK293 cells; AP-MS used pervanadate. BioID rows are distinct proximity evidence. Pair-specific co-IP validation was not located in EV1C. |
| PMID:40205054(https://pmc.ncbi.nlm.nih.gov/articles/PMC12137143/) | Original U2OS AP-MS Methods; Supplementary Tables 1–10 workbook | Methods were read, but the EPHB4–EFNB2 target pair was not located in the recovered workbook. This source-specific row is UNDECIDED; the limit does not negate the independently established ligand/receptor function. |
PMID:18593464(https://pmc.ncbi.nlm.nih.gov/articles/PMC2575619/), Methods and Table 1, plus Additional file 1 entries EU826608/012D11, EU826609/012C08 and EU826610/012E11, identifies EPHB4 products of 516, 306 and 414 residues. Table 2's ten functionally tested variants do not include these EPHB4 products. Reconstructing the current UniProt VSP edits gives sequences identical to the independently downloaded official P54760-1/-2/-3/-4 FASTAs (987/516/306/414 residues). This proves source sequence consistency, not secretion, ligand trapping, or equal function.
The source UniProt P54760 describes the displayed receptor's extracellular region, transmembrane helix at 540–560 and cytoplasmic kinase. Each shorter product lacks the displayed transmembrane and kinase regions. Proteolytic fragments in Reactome are a separate product category from these splice products.
PMID:18713744(https://pmc.ncbi.nlm.nih.gov/articles/PMC2570862/), Experimental Procedures, Results and the cleavage-site discussion, directly examines EphB2 with metalloproteinases and notes variable family susceptibility. Its double-site construct designation must not be mistaken for EPHB4. PMID:17428795(https://pmc.ncbi.nlm.nih.gov/articles/PMC4005067/), Experimental Procedures and Results/Figures 1–2 captions, examines murine EphB2 processing, with HEK293T as a host and separate mouse/rat material. No EPHB4 cleavage-site measurement was recovered in these originals.
Current Reactome R-HSA-3928657 and R-HSA-3928656 explicitly contain human EPHB4 fragment members: extracellular set 3928488→member 3928420 and cytosolic set 3928494→member 3928406, each linked to P54760 reference entity 54382. The uncertainty concerns experimental transfer of the cleavage mechanism to EPHB4, not missing target membership or the compartment label of an unrelated event.
PMID:23533145(https://pmc.ncbi.nlm.nih.gov/articles/PMC3773505/), main Methods/Results, describes expressed-prostatic-secretion urine pools, vesicle isolation and LC-MS/MS. EPHB4-specific peptides in Supplemental Table 2 were not independently recovered. Vesicle detection and exosome biogenesis are different assertions.
The actual PANTHER v19 human leaf PTN002521330 descends from both imported IBD nodes PTN001230349 and PTN000700012. The target path contains the relevant positive assertions and no inspected NOT assertion. Official human/mouse sequences match the MSA leaves. The current UniProt family label and archived v19 family label differ by release; this is not evidence that the target lies outside the IBD clade. Mouse P54761 has experimental PMID:10518221 assertions, so its pathway transfer is not merely reciprocal electronic propagation.
The actual InterPro2GO mapping, Rhea 10596 and UniProt SL-0039 support protein-tyrosine kinase chemistry, ATP binding and plasma membrane. Current ARBA00090097 and ARBA00086040 emit angiogenesis and ephrin receptor activity, respectively. Target-specific FunFam eligibility was not fully replayed because those assignments were absent from the inspected target JSON; their biological conclusions were assessed against independent evidence.
The specific receptor replacement GO:0005005 is transmembrane-ephrin receptor activity, not the GPI-linked ephrin receptor term. Its current ancestry includes the broader kinase/receptor terms it replaces. Partner refinements GO:0032794 and GO:0046875 describe GAP and Eph-receptor binding without asserting their partners' catalytic activities.
For PMID:26481148, Figure 1 and the MST/SPR Methods establish recombinant EPO–EphB4 binding; Figure 2 and its Results separate A2780 EPOR knockdown from combined EPOR/EPHB4 knockdown and receptor addition in COS-1 cells. Figure 3 separates EPHB4-dependent SRC/STAT3 from EPOR-dependent JAK2/STAT5 signaling. The positive binding and signal-transmission evidence supports the contextual receptor activity. The exact commercial-protein boundaries and general physiological erythropoiesis are not established by these observations.
For PMID:16424904, Results “Generation of recombinant retrovirus-producing cell lines,” “Interference with endothelial EphB4 signaling,” “Expression of endothelial EphB4 results in circumferential blood vessel growth,” and the retinal-vascularization section were read with Methods on retroviral generation, mouse models, intravital microscopy, immunohistochemistry and retinal preparation. The historical wording about caption access did not mean that the main Results or Methods were unavailable. These data identify remodeling and branching effects with both full-length and kinase-truncated reagents; initial vessel formation and subsequent patterning are distinguished.