DPF2

DPF2 (BAF45D) is a dedicated subunit of the BAF (SWI/SNF) chromatin-remodeling complex that reads histone acylation marks to direct BAF occupancy at enhancers and promoters, thereby controlling gene expression programs in pluripotency, hematopoiesis, macrophage polarization, and oncogenesis. Its tandem PHD finger domain binds acetylated histone H3 and H4 tails, H3K14 lactylation, and H3K14 crotonylation through structurally defined bipartite pockets; mutagenesis of these pockets abolishes chromatin recruitment and disrupts downstream transcriptional outputs including myeloid differentiation blockade and oncogene activation (PMID:28533407, PMID:39636855, PMID:39184054). DPF2 co-occupies enhancers with Oct4, Sox2, p300, and BRG1 in embryonic stem cells, where it maintains H3K27ac at critical loci such as the Tbx3 distal enhancer to sustain pluripotency and enable mesendodermal differentiation (PMID:30609396). Beyond chromatin reading, DPF2 functions as an E3 ubiquitin ligase that promotes K48-linked polyubiquitination and proteasomal degradation of OCT4, acts as a co-repressor for ERRα by recruiting HDAC1, suppresses IFN-β production through the non-canonical NF-κB pathway, and drives pro-inflammatory macrophage polarization by maintaining enhancer activity at the CACNA1D locus (PMID:26417682, PMID:20400511, PMID:28404846, PMID:41223220).

1. 2010 Medium

   Establishing that DPF2 functions as a transcriptional co-repressor: DPF2 was shown to bind the nuclear receptor ERRα, recognize acetylated H3 via its PHD fingers, and recruit HDAC1, thereby repressing ERRα target genes — linking DPF2 for the first time to histone-mark reading and gene silencing.

   Evidence Co-IP, ChIP, and siRNA knockdown with RT-qPCR in HeLa cells

   PMID:20400511

   Open questions at the time

   • Single-lab study without independent replication
   • Genome-wide scope of ERRα co-repression by DPF2 not defined
   • Whether DPF2 recruits other deacetylases besides HDAC1 unknown
2. 2011 Medium

   Solving the structure of DPF2's C2H2 zinc finger domain revealed a canonical fold with positively charged surface residues, raising the possibility that DPF2 contacts DNA directly in addition to reading histone marks.

   Evidence X-ray crystallography of the C2H2 domain

   PMID:21888896

   Open questions at the time

   • No direct DNA-binding assay performed
   • Functional relevance of the zinc finger in full-length DPF2 untested
   • Whether this domain contributes to BAF complex targeting unknown
3. 2015 Medium

   Two studies expanded DPF2's functional repertoire: it was identified as an E3 ubiquitin ligase for OCT4 (promoting K48-linked polyubiquitination and proteasomal degradation via its PHD2 domain) and as a required SWI/SNF subunit for MLL-rearranged leukemia maintenance, where its knockdown impaired leukemic self-renewal with transcriptomic changes overlapping hematopoietic stem-cell programs.

   Evidence In vitro ubiquitination assay, Co-IP, GST pull-down, mutagenesis in 293 cells (OCT4 study); shRNA knockdown, gene expression profiling, in vivo leukemia assays (leukemia study)

   PMID:26417682 PMID:26571505

   Open questions at the time

   • E3 ligase activity demonstrated only in vitro and overexpression systems
   • Whether OCT4 ubiquitination is relevant in ESCs in vivo not addressed
   • Whether DPF2's leukemia role is separable from its histone-reading function unknown
4. 2017 High

   The crystal structure of DPF2's tandem PHD fingers at 1.6 Å resolution defined bipartite binding pockets for acetylated H3 and H4 tails, and mutagenesis of these pockets abolished chromatin recruitment and the ability to block myeloid differentiation in vivo — establishing the structural basis of DPF2's histone-reader function.

   Evidence X-ray crystallography, histone peptide pull-down, mutagenesis, ChIP, myeloid differentiation assay

   PMID:28533407

   Open questions at the time

   • Whether both H3 and H4 must be co-acetylated for in vivo function not resolved
   • Structural basis for selectivity among different acyl marks not yet known
5. 2017 Medium

   DPF2 was placed in innate immune regulation: its knockdown increased IFN-β production and STAT1 phosphorylation via the non-canonical NF-κB pathway, reducing influenza virus replication — revealing a role beyond chromatin remodeling in suppressing antiviral responses.

   Evidence siRNA knockdown, viral growth kinetics, RT-qPCR, western blot for phospho-STAT1

   PMID:28404846

   Open questions at the time

   • Single-lab knockdown study
   • Mechanism by which DPF2 restrains non-canonical NF-κB signaling not defined
   • Whether this function depends on BAF complex membership unknown
6. 2018 Medium

   Two studies extended DPF2 biology in disease: disease-associated PHD missense variants were shown to form nuclear aggregates that dominantly sequester wild-type DPF2 and BRG1, and a PTBP1-dependent splice isoform (BAF45d/6A) was found to maintain the undifferentiated state in glioblastoma through a reciprocal feedback loop with PTBP1.

   Evidence Recombinant pull-down, overexpression with immunofluorescence (variant study); splicing arrays on patient samples, siRNA knockdown, RT-PCR (glioblastoma study)

   PMID:29373718 PMID:29429572

   Open questions at the time

   • Dominant-negative mechanism demonstrated only in overexpression; endogenous confirmation lacking
   • Isoform-specific functions not mapped to individual DPF2 domains
   • Patient-level phenotypic spectrum for PHD variants not fully delineated
7. 2019 High

   Genomic approaches in ESCs established that DPF2 co-occupies enhancers with Oct4, Sox2, p300, and Brg1, and that its deletion reduces H3K27ac at the Tbx3 distal enhancer, repressing Tbx3 and impairing mesendodermal differentiation — directly linking DPF2-dependent BAF chromatin remodeling to pluripotency regulation.

   Evidence CRISPR knockout, ChIP-seq, ATAC-seq, RNA-seq, genetic rescue with Tbx3 re-expression in mouse ESCs

   PMID:30609396

   Open questions at the time

   • Whether DPF2 directly recruits p300 or follows it to enhancers not resolved
   • Roles at other lineage-specific enhancers beyond Tbx3 not individually validated
   • Interplay between DPF2's E3 ligase activity toward OCT4 and its enhancer co-occupancy with Oct4 unexplored
8. 2024 High

   DPF2's reader repertoire was expanded beyond acetylation: its PHD fingers were shown to bind H3K14 lactylation at oncogene promoters (driving cervical cancer cell survival) and H3K14 crotonylation (mapped by protein footprinting to both PHD1 and PHD2 pockets in the intact BAF complex), demonstrating that DPF2 integrates diverse histone acylation signals.

   Evidence Photoaffinity probe proteomics, CUT&Tag, structure-guided mutagenesis, cell viability (lactylation study); protein footprinting MS with reconstituted BAF complex, direct binding assays (crotonylation study)

   PMID:39184054 PMID:39636855

   Open questions at the time

   • Relative affinities for acetylation, lactylation, and crotonylation not quantitatively compared under identical conditions
   • Whether acyl-mark preference varies by cell type or metabolic state unknown
   • Structural basis for mark discrimination at atomic level not yet solved
9. 2025 High

   Conditional DPF2 knockout in macrophages revealed that DPF2 maintains H3K27ac and H3K4me1 at the CACNA1D enhancer, sustaining calcium influx and MAPK signaling to promote pro-inflammatory polarization; DPF2 loss shifts macrophages toward an anti-inflammatory state and protects against intestinal injury.

   Evidence Conditional knockout mouse, scRNA-seq, spatial transcriptomics, ChIP for histone marks, calcium imaging, MAPK pathway assays

   PMID:41223220

   Open questions at the time

   • Whether DPF2 directly binds the CACNA1D enhancer or acts indirectly through BAF not distinguished
   • Generalizability beyond intestinal macrophages not tested
   • Interaction between DPF2's histone-mark reading and enhancer activation at this locus not mechanistically resolved

• It remains unknown how DPF2 discriminates among acetylation, lactylation, and crotonylation in vivo, whether its E3 ubiquitin ligase activity is functionally linked to its BAF-dependent chromatin-remodeling role, and whether its C2H2 zinc finger domain contributes to DNA binding within the BAF complex.
• No quantitative in vivo comparison of acyl-mark preference
• E3 ligase function not reconstituted in a physiological BAF complex context
• C2H2 domain DNA-binding capacity not directly tested