HMGN3

HMGN3 is a nucleosome-binding chromatin architectural protein that modulates transcription in a tissue-specific manner by reducing chromatin compaction and facilitating histone modification. The full-length HMGN3a isoform stimulates PCAF-mediated acetylation of histone H3K14 on nucleosomal substrates, promoting transcription elongation, while the truncated HMGN3b splice variant lacks this activity (PMID:22150271). HMGN3 cooperates with transcription factors—reinforcing PDX1 occupancy at the Glut2 promoter to sustain glucose-stimulated insulin secretion in pancreatic beta cells (PMID:19651901), partnering with SNAI2 to repress epithelial genes via HDAC-dependent mechanisms in cholangiocarcinoma (PMID:35635715), and acting downstream of Hoxa10 to direct uterine decidualization through Hand2 regulation (PMID:26112184). HMGN3 mRNA is itself stabilized by WTAP-mediated m6A modification at specific 3′-UTR sites recognized by the reader IGF2BP1, linking epitranscriptomic regulation to HMGN3 protein abundance in trophoblasts and microglia (PMID:36412513).

1. 2002 Medium

   Establishing that HMGN3 has a tissue-restricted expression pattern—highest in brain astrocytes—distinguished it from the ubiquitous HMGN2 and raised the question of what tissue-specific chromatin functions it performs.

   Evidence Western/Northern blotting and immunohistochemistry in mouse brain sections

   PMID:12185205

   Open questions at the time

   • No functional consequence of astrocyte expression demonstrated
   • Brain-specific target genes unknown
   • Expression pattern not confirmed by single-cell approaches
2. 2009 High

   Demonstrating that HMGN3 loss in knockout mice impaired glucose-stimulated insulin secretion and GLUT2 levels, while ChIP showed mutual reinforcement of HMGN3 and PDX1 binding at the Glut2 promoter, established HMGN3 as a functionally required chromatin factor in beta-cell gene regulation.

   Evidence Hmgn3−/− knockout mouse model, ChIP, and transcription profiling in pancreatic islets

   PMID:19651901

   Open questions at the time

   • Mechanism by which HMGN3 and PDX1 reinforce each other's binding is unclear
   • Genome-wide targets in beta cells not mapped
   • Whether the insulin secretion defect is solely GLUT2-dependent was not resolved
3. 2010 Medium

   Showing that HMGN3 deletion reduced plasma glucagon without a cell-autonomous effect in alpha cells clarified that HMGN3's endocrine role extends beyond beta cells but acts through indirect, non-cell-autonomous mechanisms on glucagon.

   Evidence Hmgn3−/− mice combined with siRNA knockdown in alphaTC1-9 cells and glucagon secretion assays

   PMID:19885867

   Open questions at the time

   • Identity of the indirect signal from beta cells or other tissues not determined
   • Only one alpha-cell line tested
   • No rescue experiment performed
4. 2012 High

   Reconstituting HMGN3a-dependent stimulation of PCAF acetylation on nucleosomal H3K14 in vitro, and showing this promotes transcription elongation while the HMGN3b isoform cannot, defined the molecular activity of the C-terminal regulatory domain as a cofactor for histone acetyltransferase activity.

   Evidence In vitro acetylation assay with PCAF and nucleosomal substrates, ChIP across the Glyt1 locus, splice variant comparison

   PMID:22150271

   Open questions at the time

   • Structural basis for C-terminal domain interaction with PCAF unknown
   • Whether other HATs are similarly stimulated not tested
   • In vivo isoform-specific contributions not separated genetically
5. 2015 Medium

   Demonstrating that HMGN3 mediates progesterone- and Hoxa10-driven decidualization by regulating Hand2 extended its role to a reproductive chromatin effector that integrates hormonal and homeobox transcription factor signals.

   Evidence Overexpression and siRNA knockdown in uterine stromal cells, progesterone treatment in ovariectomized mice, Hoxa10 epistasis

   PMID:26112184

   Open questions at the time

   • Direct versus indirect regulation of Hand2 by HMGN3 not resolved
   • Genome-wide chromatin occupancy in decidualizing cells not mapped
   • In vivo fertility or implantation phenotype of Hmgn3−/− females not reported
6. 2022 Medium

   ChIP-seq revealing that HMGN3 requires SNAI2 for binding to epithelial gene loci (CDH1, TJAP1) and that repression depends on HDACs established a mechanism by which HMGN3 can repress rather than activate transcription, contingent on its transcription factor partner.

   Evidence HMGN3 knockdown, ChIP-seq, HDAC inhibitor rescue, migration/invasion assays in cholangiocarcinoma cells

   PMID:35635715

   Open questions at the time

   • Whether HMGN3 directly recruits HDACs or acts through SNAI2-dependent corepressors is unresolved
   • Observations limited to cholangiocarcinoma lines; generality across EMT contexts unknown
   • No structural or biochemical detail on HMGN3–SNAI2 interaction
7. 2022 Medium

   Identifying WTAP-dependent m6A modification at specific 3′-UTR sites of HMGN3 mRNA, read by IGF2BP1 to stabilize the transcript, revealed an epitranscriptomic layer controlling HMGN3 protein abundance relevant to trophoblast invasion.

   Evidence MeRIP-microarray, RIP for IGF2BP1, luciferase reporter with mutated m6A sites, WTAP knockdown in trophoblast cells

   PMID:36412513

   Open questions at the time

   • In vivo relevance in preeclampsia not established with animal models
   • Whether m6A regulation of HMGN3 operates in other HMGN3-expressing tissues is unknown
   • Contribution of other m6A readers besides IGF2BP1 not excluded
8. 2022 Medium

   Showing that Hmgn3 overexpression facilitates ESC-to-trophoblast stem cell conversion downstream of Rif1 deletion implicated HMGN3 in the earliest lineage decisions of mammalian development.

   Evidence Rif1 knockout ESCs, transcriptomics, Hmgn3 overexpression, chimeric embryo assays

   PMID:35354046

   Open questions at the time

   • Whether HMGN3 is sufficient or only contributory to trophoblast conversion unclear
   • Mechanism linking HMGN3 chromatin remodeling to trophectoderm gene activation not defined
   • Single-lab finding without independent replication

• How HMGN3 switches between transcriptional activation (via PCAF/H3K14ac) and repression (via SNAI2/HDACs) at different loci, and whether its isoforms partition these opposing functions in vivo, remains unresolved.
• No genome-wide occupancy map exists in primary tissues
• Isoform-specific knockout models have not been generated
• Structural basis for partner-dependent functional switching is unknown