The literature base is sparse. PMID:12508107 is the only seeded targeted study of MegBP/LRP2BP. It identifies MegBP as a TPR-containing scaffold associated with megalin/LRP2 and maps the receptor-side binding site to a proline-rich N-terminal region of the LRP2 cytoplasmic tail PMID:12508107 PMID:12508107. The cache is abstract-only and does not state the tested MegBP species or splice isoform. The reviewed human UniProt record maps this evidence to Q9P2M1, but that database assignment should not be mistaken for study-level proteoform resolution [file:human/LRP2BP/LRP2BP-uniprot.txt "CC -!- SUBUNIT: Interacts with LRP2. {ECO:0000269|PubMed:12508107}."].
The interaction did not block receptor endocytosis in the reported experiment PMID:12508107 and MegBP overexpression caused cellular lethality PMID:12508107. The proposed connection between megalin and transcriptional regulation is explicitly a model based on additional interaction-screen partners, including SKIP, rather than a demonstrated transcriptional activity of LRP2BP PMID:12508107 PMID:12508107.
PMID:28514442 is BioPlex 2.0 AP-MS; GOA/IntAct assigns an LRP2BP-GSTT1 co-association to it, but the pair is not named in the cached article body. The paper itself calls the network edges candidate interactions PMID:28514442. PMID:33961781 reports BioPlex 3.0 and a second cell-line-specific network; GOA/IntAct again assigns LRP2BP-GSTT1, providing screen-system recurrence but not targeted mechanistic validation PMID:33961781.
PMID:32296183 is the HuRI proteome-scale binary yeast-two-hybrid map. GOA/IntAct attributes 18 LRP2BP partners to this source, but none is named with LRP2BP in the cached article body. The screen used repeated search-space screens and pairwise verification PMID:32296183. These pairs are human binary-interaction leads; without targeted follow-up they do not establish stable complexes, shared pathways, localization, or adapter function.
Reviewed Q9P2M1 is a 347-residue cytoplasmic protein. UniProt records vesicular staining near the plasma membrane and throughout the cytoplasm [file:human/LRP2BP/LRP2BP-uniprot.txt "CC Note=Detected in a vesicular staining pattern close to the plasma"]. Its architecture is repeat-rich: one annotated TPR at residues 59–92 and six Sel1-like repeats extending through residues 297–332 [file:human/LRP2BP/LRP2BP-uniprot.txt "FT REPEAT 59..92"] [file:human/LRP2BP/LRP2BP-uniprot.txt "FT /note=\"Sel1-like 6\""]. These repeats are compatible with a protein-interaction scaffold, but the paper maps the binding site on LRP2, not an interface on LRP2BP.
UniProt names two splice isoforms [file:human/LRP2BP/LRP2BP-uniprot.txt "CC Event=Alternative splicing; Named isoforms=2;"]. Isoform 2 carries the small VSP_030664 change at residue 35, recorded as T to TKS [file:human/LRP2BP/LRP2BP-uniprot.txt "FT /note=\"T -> TKS (in isoform 2)\""]. No cited functional experiment is isoform-resolved, so the LRP2 interaction, localization, and screen results must not be assigned specifically to isoform 1 or 2.
The reviewed record has AlphaFoldDB and SMR cross-references but no PDB cross-reference [file:human/LRP2BP/LRP2BP-uniprot.txt "DR AlphaFoldDB; Q9P2M1; -."]. Thus there is a predicted structural model, but no curated experimental structure and no experimentally resolved LRP2BP-LRP2 interface.
The fetched PANTHER table places human Q9P2M1 with mouse Q9D4C6, rat Q569C2, macaque Q4R3N2, zebrafish A5PLI4, and Xenopus Q6IND7 in exact subfamily PTHR44554:SF1 [file:interpro/panther/PTHR44554/PTHR44554-entries.csv "Q9P2M1,LRP2-binding protein,protein,9606,Homo sapiens,Homo sapiens (Human),LRP2BP,347,PTHR44554:SF1,LRP2-BINDING PROTEIN,True"]. This supports a conserved vertebrate ortholog group and similar protein lengths, not automatic transfer of LRP2 binding or the proposed transcriptional model. No distinct human paralog is present in the fetched representative table; this does not prove that paralogs are absent from every broader database classification.