PBX2

PBX2 is a TALE-class homeodomain transcription factor that functions as a context-dependent transcriptional activator or repressor by forming heterodimeric complexes with HOX proteins (via the HOX hexapeptide motif) and with PREP1/MEIS family members, which also protect PBX2 from proteasomal degradation (PMID:7577680, PMID:12871956, PMID:10082572). In myeloid cells PBX2 participates in trimeric HOXA9–PBX2–MEIS1 complexes that co-localize in nuclear speckles, while in various cell types HOX–PBX2 dimers regulate targets including miR-221/222, CDK6, circTLK1, and HOX11 (PMID:10082572, PMID:23400877, PMID:37272544, PMID:12054735). Although Pbx2-null mice are viable owing to functional redundancy with Pbx1 and Pbx3, compound Pbx1/Pbx2 loss-of-function reveals essential roles in limb morphogenesis and axial skeletal patterning through hierarchical control of Hox, Polycomb, Shh, and Pax gene expression (PMID:15169896, PMID:16672333, PMID:18691704). PBX2 activity is post-translationally regulated by mTORC1-dependent dephosphorylation via PP1, opposing GSK3-mediated phosphorylation, establishing PBX2 as a unique mTORC1 effector among PBX family members (PMID:36477205).

1. 1991 High

   Identification of PBX2 as a new TALE homeodomain protein closely related to PBX1 established the existence of a multigene PBX family with potential functional overlap, and its ubiquitous expression pattern distinguished it from more tissue-restricted paralogs.

   Evidence cDNA cloning, sequence analysis, chromosomal mapping, and Northern blotting across fetal and adult tissues

   PMID:1682799

   Open questions at the time

   • Transcriptional targets and in vivo functions of PBX2 were unknown
   • Whether PBX2 was functionally redundant with PBX1/PBX3 was untested
2. 1995 High

   Demonstration that PBX2 cooperatively binds DNA with HOX proteins (Hoxb-7, Hoxb-8) through the hexapeptide motif established the molecular basis for HOX–PBX heterodimer formation and composite DNA recognition, a paradigm for TALE homeodomain partnerships.

   Evidence EMSA with deletion and point mutagenesis of the HOX hexapeptide

   PMID:7568094 PMID:7577680

   Open questions at the time

   • Whether HOX–PBX2 complexes regulate specific endogenous target genes was unknown
   • The structural basis of the hexapeptide–PBX interaction awaited crystallographic analysis
3. 1999 High

   Discovery of trimeric HOXA9–PBX2–MEIS1 complexes in myeloid cells, with MEIS1 enhancing HOXA9–PBX2 complex formation even without DNA, revealed a higher-order combinatorial logic for TALE–HOX transcription factor assemblies and their subnuclear organization.

   Evidence Co-immunoprecipitation, EMSA, and immunofluorescence co-localization in nuclear speckles in myeloid cells

   PMID:10082572

   Open questions at the time

   • Specific genomic targets of the trimeric complex were not identified
   • Functional significance of nuclear speckle localization was not determined
4. 2002 Medium

   Identification of PBX2–PREP1 as both a transcriptional repressor (at the UGT2B17 promoter, interfering with HNF1α) and an activator (at the HOX11 promoter) established PBX2 as a context-dependent transcriptional regulator whose output depends on the promoter architecture.

   Evidence EMSA, supershift assays, and promoter-reporter assays in K562 and other cell lines

   PMID:12054735 PMID:12065766

   Open questions at the time

   • Direct ChIP evidence for PBX2 occupancy at these endogenous promoters was not provided
   • Mechanism switching between activation and repression was not characterized
5. 2003 High

   Demonstrating that PREP1 dimerization stabilizes PBX2 by preventing its proteasomal degradation defined a key post-translational regulatory axis in which PBX2 protein levels are controlled by the availability of TALE cofactors.

   Evidence Cycloheximide chase, MG132 proteasome inhibition, EMSA, and immunoblotting in stably transfected F9 cells

   PMID:12871956

   Open questions at the time

   • The ubiquitin ligase responsible for PBX2 degradation was not identified
   • Whether MEIS proteins stabilize PBX2 through the same mechanism was not formally tested
6. 2004 High

   Retinoic acid was shown to increase PBX2 both transcriptionally (as a primary response) and post-translationally via MEIS-mediated stabilization, linking signaling pathways to PBX2 protein accumulation; meanwhile, Pbx2-null mice proved viable, revealing that Pbx2 is functionally redundant with other Pbx members despite being the predominant isoform in postnatal bone marrow and thymus.

   Evidence RT-PCR and cycloheximide chase in P19 cells; germline Pbx2 knockout mouse with developmental, hematopoietic, and expression analysis

   PMID:15095411 PMID:15169896

   Open questions at the time

   • Which Pbx paralog(s) compensate for Pbx2 loss in specific tissues was not resolved
   • Direct RA-responsive elements in the PBX2 promoter were not mapped
7. 2006 High

   Compound Pbx1/Pbx2 knockout mice revealed that PBX2 cooperates with PBX1 to pattern the limb skeleton by hierarchically controlling Hox gene domains and Shh expression in the zone of polarizing activity, establishing redundant but essential roles for Pbx factors in appendicular development.

   Evidence Compound knockout mouse with skeletal analysis and in situ hybridization of Hox, Shh targets

   PMID:16672333

   Open questions at the time

   • Direct PBX2 binding at limb enhancers was not shown
   • Whether PBX3 further compensates was not fully addressed
8. 2008 High

   Extension of compound Pbx1/Pbx2 analysis to axial skeleton demonstrated hierarchical control of Polycomb group genes, Hox gene spatial domains, and Pax1/Pax9 in sclerotome, explaining the complete loss of vertebral identity; separately, Prep1/Pbx2 was shown to mediate allele-specific repression of CCL2 in astrocytes, illustrating how genetic variation creates PBX2-dependent regulatory switches.

   Evidence Compound knockout mouse skeletal and molecular analysis; EMSA and reporter assays on CCL2 promoter polymorphism in astrocytes

   PMID:18480829 PMID:18691704

   Open questions at the time

   • Genome-wide binding profiles of PBX2 in paraxial mesoderm were not generated
   • The mechanism by which IL-1β relieves Prep1/Pbx2 repression at CCL2 was not defined
9. 2013 Medium

   The HOXB7–PBX2 dimer was identified as a positive transcriptional regulator of miR-221/222 in melanoma, with dimer disruption inducing apoptosis, providing direct evidence that HOX–PBX2 complexes control non-coding RNA targets relevant to cancer cell survival.

   Evidence HOX/PBX dimerization antagonist HXR9, miRNA expression analysis, apoptosis assays in melanoma cells

   PMID:23400877

   Open questions at the time

   • Direct PBX2 ChIP at the miR-221/222 locus was not performed
   • Whether the apoptotic effect is PBX2-specific or general to all PBX members was not tested
10. 2021 Medium

    PBX2 was shown to directly activate circTLK1 transcription in glioma and to interact with HOXA6 to promote gastric cancer metastasis, broadening the repertoire of HOX–PBX2 target genes in oncogenic contexts and demonstrating that HOXA6 stabilizes PBX2 protein.

    Evidence ChIP, luciferase reporter, AGO2-RIP, RNA pulldown in glioma cells; Co-IP, siRNA knockdown, orthotopic implantation in gastric cancer models

    PMID:33535170 PMID:34721518

    Open questions at the time

    • Whether PBX2 contributes to tumorigenesis in vivo beyond xenograft models is unknown
    • Global target gene repertoire of PBX2 in cancer cells has not been defined by genome-wide approaches
11. 2023 Medium

    Two studies identified HOXB9–PBX2 as a transcriptional activator of CDK6 promoting cell cycle progression in gastric cancer, and demonstrated that PBX2 is uniquely regulated among PBX family members by mTORC1-dependent dephosphorylation via PP1 opposing GSK3-mediated phosphorylation, linking PBX2 activity to growth factor signaling.

    Evidence Co-IP, ChIP, reporter and cell cycle assays for HOXB9–PBX2–CDK6 axis; phosphoproteomics, rapamycin treatment, GSK3/PP1 siRNA knockdown for mTORC1 regulation

    PMID:36477205 PMID:37272544

    Open questions at the time

    • The phosphorylation sites on PBX2 targeted by GSK3/PP1 were not mapped to individual residues
    • Functional consequences of PBX2 phosphorylation on DNA binding or target gene selection were not determined
    • Whether mTORC1-dependent PBX2 regulation operates in non-cancer contexts is untested

• Key unresolved questions include the genome-wide binding landscape of PBX2, the identity of the E3 ubiquitin ligase mediating its proteasomal degradation, and the specific phosphorylation sites and their functional impact on PBX2 transcriptional output.
• No genome-wide ChIP-seq or CUT&RUN data for PBX2 have been reported
• E3 ligase targeting PBX2 for degradation is unidentified
• Functional consequence of mTORC1-regulated phosphorylation on specific PBX2 target genes is unknown