PARD3B (partitioning defective 3 homolog B; PAR3beta, PAR3L) is the vertebrate paralog of the cell-polarity scaffold PARD3 in the PAR3 family (PANTHER subfamily PTHR16484:SF4). Like PARD3 it has an N-terminal Par3/HAL oligomerization domain and three PDZ domains, but it diverges from PARD3 in its partners. In mammalian cells it does not bind atypical protein kinase C, and binding to PAR6 is weak or absent depending on the isoform and the study. Even so, it localizes to epithelial tight junctions, where it colocalizes with TJP1/ZO-1, independently of the PAR3-PAR6-aPKC complex. It also binds the spindle-orientation adaptor Inscuteable. Its best-defined activity comes from mouse. There PARD3B binds the polarity kinase LKB1 (STK11) through a PARD3B-specific sequence between PDZ2 and PDZ3, and it suppresses LKB1 kinase activity, measured as phosphorylation of the LKB1 target AMPK. This restraint on LKB1 is needed to keep mammary stem cells alive and self-renewing. Whether PARD3B takes part in apical-basal polarity in the way PARD3 does has not been established. The donkey protein is a full-length ortholog, 88% identical to human PARD3B, and its PDZ domains and the LKB1-binding linker are near-identical to the human sequence.
Summary: Microtubule cytoskeleton organization (TreeGrafter). A PAR3-family property that PAINT places at the family root (PTN001046835; donors Drosophila Bazooka and mouse Pard3), so PARD3B inherits it formally. No PARD3B-specific evidence exists.
Reason: The PAINT node sits above the PARD3/PARD3B duplication, so the donkey protein is inside the clade. But there is target-specific evidence that PARD3B has diverged from PARD3. It does not bind aPKC or PAR6, the partners through which PAR3 couples polarity to the cytoskeleton, and its one defined function (restraining LKB1) is PARD3B-specific. The donor evidence comes only from PARD3-type orthologs, and nothing shows that PARD3B organizes microtubules, so the inherited term is likely an over-annotation for this paralog.
Summary: Adherens junction (TreeGrafter, from the PAR3 family root). PARD3B is reported along cell-cell contacts and at tight junctions; adherens-junction localization specifically has not been shown.
Reason: This is a family-level location next to the tight junction where PARD3B has been observed. PARD3B reaches junctions without the PAR3-PAR6-aPKC complex, so junctional targeting is a property it shares with PARD3. No adherens-junction-specific data exist for PARD3B, so the row is kept as non-core.
Summary: Bicellular tight junction (UniProt SubCell mapping). Human PARD3B, expressed endogenously or exogenously, localizes to epithelial tight junctions with TJP1. This is the best-supported location of the protein.
Reason: Two independent groups showed tight-junction localization of human PAR3beta/Par3L in epithelial cells, and it does not need the ternary complex. The donkey ortholog is full length and 88% identical, with conserved PDZ domains.
Nevertheless PAR3beta, expressed intrinsically or extrinsically, localizes to the tight junctions, indicating that the localization does not require the ternary complex formation.
Summary: Cell cortex (TreeGrafter, from the PAR3 family root; donors mouse Pard3 and C. elegans PAR-3). PARD3B lies at the cortex in the broad sense, at cell-cell contacts.
Reason: Peripheral, membrane-proximal localization is consistent with the PARD3B data (cell-cell contact region, tight junctions). It is a family-level location with no PARD3B-specific cortex data, so it is kept as non-core.
Summary: Cell adhesion (TreeGrafter, from the PAR3 family root). PAR3-family scaffolds organize apical junctions but do not themselves mediate adhesion.
Reason: Overexpressed N-terminal PARD3B fragments disrupt tight-junction formation, as PARD3 fragments do. That links the protein to junction assembly, not to adhesion itself. No PARD3B adhesion data exist, and the donor evidence is from PARD3-type orthologs.
the N-terminal region of Par3L, like that of Par3, can disrupt the formation of tight junctions when ectopically expressed in Madin-Darby canine kidney (MDCK) cells.
Summary: Intracellular protein localization (TreeGrafter, from the PAR3 family root; donors Bazooka and PAR-3). In PAR3 proteins this function is the recruitment and positioning of PAR6-aPKC.
Reason: The family function rests on PAR3 positioning PAR6-aPKC, and PARD3B does not bind aPKC or PAR6. The previous review cited the PARD6B interaction as support, but two studies contradict it (Kohjima 2002; Huo and Macara 2014), and Gao 2002 reports only isoform-dependent Par6 binding. No protein-localization function has been shown for PARD3B.
Summary: Endomembrane system (UniProt SubCell mapping). UniProt reports it for the human ortholog.
Reason: This is a reported location for human PARD3B, but it is vague and secondary to the junctional localization. The primary data behind it are not in the local cache.
Summary: Apical plasma membrane (TreeGrafter, from the PAR3 family root; donor Bazooka). This is a family-level apical localization.
Reason: The term is consistent with apical-junctional (tight-junction) localization of PARD3B, but no PARD3B data show it on the apical membrane itself. It is kept as non-core.
Summary: Establishment of cell polarity (TreeGrafter, from the PAR3 family root). PARD3B binds the polarity adaptor Inscuteable and the polarity kinase LKB1/PAR-4. A role in cell polarity has not been demonstrated, however; the authors of the definitive functional study say so explicitly.
Reason: The previous review accepted this as core. Re-examination shows that PARD3B lacks the aPKC/PAR6 binding on which PAR3 polarity function depends. Huo and Macara state that whether PARD3B plays any role in cell polarity had not been addressed, and they then defined a polarity-independent function, LKB1 restraint in stem-cell survival. The broad term is kept because PARD3B sits inside the PAINT clade and binds two polarity proteins, INSC and LKB1. It is not core for this paralog.
Whether the Par3L isoform, which does not bind to Par6 or aPKC, plays any role either in cell polarity or in stem cell function has not been previously addressed.
Summary: Protein-containing complex (Ensembl Compara, from the human IDA in PMID:16458856, a complex with Inscuteable).
Reason: True, since PARD3B co-precipitates with INSC (human) and LKB1 (mouse), but uninformative. No specific GO complex term describes these assemblies, so there is no more precise replacement.
Summary: Phosphatidylinositol binding (TreeGrafter, from the PAR3 family root; sole donor Drosophila Bazooka). Phosphoinositide binding is documented for PAR3 proteins but has not been tested for PARD3B.
Reason: PAINT placed this at the family root on Bazooka evidence. PARD3B PDZ2, the lipid-binding PDZ in PAR3, is only 52/86 identical to human PARD3 PDZ2. The basic residues needed for lipid binding have not been checked, and no PARD3B lipid-binding data exist. Neither accepting nor removing the row is justified.
Summary: Apical junction complex (TreeGrafter). True, but entailed by the accepted bicellular tight junction, which is part of the apical junction complex.
Reason: Bicellular tight junction, accepted in this review, is part_of apical junction complex.
Nevertheless PAR3beta, expressed intrinsically or extrinsically, localizes to the tight junctions
GO:0045197 establishment or maintenance of epithelial cell apical/basal polarity
IEA GO_REF:0000118
MARK AS OVER ANNOTATED
Summary: Establishment or maintenance of epithelial apical/basal polarity (TreeGrafter, from the PAR3 family root; donors Bazooka and human PARD3). PARD3B localizes to epithelial tight junctions, but no role in apical/basal polarity has been shown.
Reason: The previous review accepted this. The donor evidence is from PARD3 and Bazooka, which act through PAR6-aPKC, and PARD3B does not bind aPKC or PAR6. Junctional localization does not need the ternary complex, so it is not evidence of polarity function. The only PARD3B perturbation data are dominant-negative N-terminal fragments shared with PARD3, and those disrupt tight-junction formation; they do not show a requirement for PARD3B. The inherited term is likely an over-annotation for the PARD3B paralog.
Whether the Par3L isoform, which does not bind to Par6 or aPKC, plays any role either in cell polarity or in stem cell function has not been previously addressed.
GO:0051660 establishment of centrosome localization
IEA GO_REF:0000118
REMOVE
Summary: Establishment of centrosome localization (TreeGrafter). Current PAINT does not propagate this term in the PAR3 family. Human PARD3, human PARD3B and mouse Pard3b carry no IBA for it, and the only current IBA carriers of GO:0051660 are CEP83 orthologs.
Reason: This TreeGrafter transfer has no counterpart in the current PAINT annotation of the PAR3 family. The family-root node PTN001046835 gives PARD3, PARD3B and Pard3b ten IBA terms, and centrosome localization is not one of them. That omission is a curator decision. Centrosome positioning by PAR3 depends on its PARD3-type partner set, which PARD3B lacks, and no PARD3B centrosome data exist.
Supporting Evidence:
file:EQUAS/PARD3B/PARD3B-notes.md
Current GO_Central IBA rows for human PARD3B (Q8TEW8), human PARD3 (Q8TEW0) and mouse Pard3b (Q9CSB4) come from PANTHER node PTN001046835 and do not include GO:0051660
ISS PMID:24859006 The Par3-like polarity protein Par3L is essential for mammar...
NEW
Summary: Proposed molecular function. Mouse Par3L binds LKB1 (endogenous co-IP from kidney) and suppresses LKB1 kinase output in cells, measured as reduced AMPK phosphorylation. A mutant in the PARD3B-specific PDZ2-PDZ3 linker cannot bind LKB1, is much less inhibitory, and fails to rescue. Donkey PARD3B is in the same PANTHER subfamily (SF4) and keeps that linker (26/29 identical to human, against 15/29 for PARD3).
Reason: This is the one activity defined for PARD3B, and it is subfamily-specific. PARD3 does not bind LKB1, and the binding sequence is not shared with PARD3. The transfer is therefore grounded in SF4 membership plus conservation of the mapped determinant, not in family-level similarity. The evidence is cell-based (phospho-AMPK and a binding-defective mutant), not an in vitro kinase assay, so direct inhibition remains to be confirmed. ISS is used.
Notably, the expression of wild type Par3L in HEK293T cells inhibited AMPK phosphorylation, with no change in expression levels of either LKB1 or AMPK, while the mutant had significantly less inhibitory effect on phospho-AMPK
donkey: human PARD3B 469-497: 29 of 29 aligned, 26 identical
Core Functions
PARD3B binds the polarity kinase LKB1 (STK11) through a PARD3B-specific sequence between PDZ2 and PDZ3 and restrains its kinase activity. In mouse, this protects mammary stem cells from LKB1-driven apoptosis and loss of self-renewal.
donkey: human PARD3B 469-497: 29 of 29 aligned, 26 identical
PDZ-domain scaffold of the PAR3 family that localizes to epithelial tight junctions without the PAR6-aPKC complex. It binds the spindle-orientation adaptor Inscuteable.
Nevertheless PAR3beta, expressed intrinsically or extrinsically, localizes to the tight junctions, indicating that the localization does not require the ternary complex formation.
Q: Does PARD3B take part in epithelial apical-basal polarity or tight-junction assembly at all, given that it binds neither aPKC nor PAR6? If it does, through which partners (INSC, LKB1, others)?
Q: Does PARD3B inhibit LKB1 directly, for example by blocking the LKB1-STRAD-MO25 complex or access to substrates, or does it act indirectly by sequestering or relocalizing LKB1?
Q: The PAINT family-root node (PTN001046835) propagates PARD3/Bazooka-derived process terms to PARD3B. Should PAINT curators add a PARD3B-subfamily node that blocks the polarity-machinery terms PARD3B is not known to share?
Suggested Experiments
Experiment: Run in vitro kinase assays with purified LKB1-STRAD-MO25 and recombinant PARD3B (wild type or the PDZ2-PDZ3 linker swap), with AMPK or a peptide substrate, to test direct inhibition of LKB1 by PARD3B.
Experiment: Knock out PARD3B in polarized epithelial cells (e.g. MDCK, or kidney epithelium where PARD3B is abundant), and read out tight-junction assembly, apical markers and phospho-AMPK. This tests whether the family-level polarity functions apply to PARD3B.
External Prediction Reviews
These computational predictions are reviewed separately from the GOA annotation set used for this review. The assessments below are from this project and do not constitute official GO annotations or endorsement by GO/UniProt. They are not included in the existing annotation review above.
Assessment of 2 ProtNLM2 GO predictions for donkey PARD3B (A0A9L0JEY6). Anchoring junction is correct but less precise than the bicellular tight junction annotation already in donkey GOA (LSP). Cell cycle is uncertain (UNC): a mouse Par3L knockout shows increased mitotic cells with aberrant spindles, but this loss-of-function phenotype does not establish a direct cell-cycle role. Non-GO ProtNLM2 outputs, not scored: subcellular locations 'Endomembrane system' and 'Cell junction', both matching locations reported for human PARD3B.
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-10-14
Review rationale: The broad term cell cycle is not in donkey GOA, and no experimental GO annotation for human or mouse PARD3B places it in a cell-cycle process. Human PARD3B is described as a putative adapter in asymmetric cell division through its interaction with Inscuteable. The only direct cell-division evidence is a mouse knockout phenotype, in which Par3L-null intestinal epithelium has more mitotic cells and a higher fraction of defective spindles, accompanied by tumours. This is a loss-of-function phenotype in a tumour-prone tissue; it does not show which step of the cell cycle PARD3B performs or whether the spindle defects are direct, and the PARD3B-specific mechanism defined so far (LKB1 binding and restraint of its kinase activity) is a signalling role, not a cell-cycle step. The gene review removes the related TreeGrafter centrosome-localization term because it is a PARD3-derived transfer. The prediction is plausible but neither established nor refuted by the available evidence.
PMID:41274508: "Further analysis of the gastrointestinal track in the Par3L KO mice found increased mitotic cells, which have significantly higher ratios of aberrant spindles compared to that of the WT mice."
PMID:24859006: "Unexpectedly, Par3L, but not Par3, binds to the tumour suppressor protein Lkb1 and inhibits its kinase activity."
LSP — Less precise than existing annotation Review score: 2/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-10-14
Review rationale: Human PARD3B colocalizes with TJP1 at epithelial tight junctions, and donkey GOA already contains bicellular tight junction (GO:0005923), which is a GO descendant of anchoring junction. The prediction is therefore correct but less precise than an existing supported annotation.