DPF3

DPF3 (BAF45c) is a tissue-restricted, histone-reading subunit of the BAF (SWI/SNF) chromatin remodeling complex that couples chromatin recognition to lineage-specific transcriptional programs in heart, muscle, and adipose tissue (PMID:18765789, PMID:28428261). Its tandem (double) PHD finger domain directly binds both methylated and acetylated lysines on histones H3 and H4, anchoring the BAF complex to chromatin, and loss of dpf3 in vivo deregulates structural and regulatory genes to produce defective cardiac looping and disrupted muscle fibers (PMID:18765789). In brown adipocytes, DPF3 is a direct transcriptional target of EBF2 and is required for chromatin accessibility at EBF2-bound enhancers and for catecholamine-responsive brown fat gene expression and mitochondrial function (PMID:28428261). DPF3 expression is itself driven by upstream signaling inputs, being a transcriptional target of STAT5 downstream of IL-3 (PMID:24155890), and in renal cancer a germline risk allele creates an enhancer HIF-binding motif that drives allele-specific DPF3 overexpression, promoting cell growth and oncogenic chromatin/transcriptional changes (PMID:34390653, PMID:35148991). Through its short DPF3a isoform, DPF3 also acts in the cytoplasm-to-chromatin TGF-β axis: DPF3a binds SNIP1 within a SMAD4–p300 complex, relieving SNIP1-mediated repression of p300 HAT activity to increase local histone acetylation and activate cell-movement genes that drive kidney cancer migration and metastasis (PMID:35945219). Beyond its nuclear chromatin role, DPF3 has a non-canonical localization to centriolar satellites, the centrosome, spindle midzone, and midbodies, where it is required for kinetochore-microtubule attachment stability, faithful mitotic progression and genome stability, and for axoneme extension during ciliogenesis (PMID:38661008). The gene encodes a full-length and a truncated XZ isoform with restricted, developmentally distinct expression (PMID:11845289).

1. 2001 Medium

   Established the gene's isoform structure and tissue restriction, defining a full-length form bearing all d4 domains and a truncated XZ form lacking the C-terminal tandem PHD fingers — the structural basis for later isoform-specific functions.

   Evidence cDNA cloning, Northern blot and RT-PCR isoform expression across chicken and mouse

   PMID:11845289

   Open questions at the time

   • No molecular function assigned to either isoform at this stage
   • Does not establish how the PHD-less XZ isoform differs functionally
2. 2008 High

   Defined DPF3's core molecular function as a BAF complex subunit whose double PHD finger reads both methylated and acetylated histone lysines, linking chromatin recognition to cardiac and muscle gene programs in vivo.

   Evidence BAF Co-IP, methylated/acetylated histone peptide pull-downs, zebrafish morpholino knockdown with cardiac/muscle phenotypes, Mef2a promoter analysis

   PMID:18765789

   Open questions at the time

   • Does not resolve which BAF target genes are directly bound versus indirectly deregulated
   • Structural basis of dual methyl/acetyl recognition not defined
3. 2013 Medium

   Placed DPF3 downstream of cytokine signaling by showing it is a direct STAT5 transcriptional target induced by IL-3, indicating its expression is signal-responsive.

   Evidence STAT5 ChIP at the DPF3 promoter and Stat5a/Stat5b shRNA knockdown in Ba/F3 cells

   PMID:24155890

   Open questions at the time

   • Functional consequence of STAT5-driven DPF3 in hematopoietic cells not tested
   • No link to DPF3's chromatin-remodeling role in this context
4. 2017 High

   Identified DPF3 as a brown-fat-selective BAF subunit acting downstream of EBF2 to maintain enhancer accessibility, extending its chromatin role to adipose lineage programming and mitochondrial function.

   Evidence EBF2/BRG1 Co-IP, ChIP-seq, ATAC-seq and DPF3 siRNA knockdown in brown adipocytes

   PMID:28428261

   Open questions at the time

   • Whether DPF3 histone-reading is required for EBF2 enhancer accessibility not directly tested
   • In vivo brown-fat phenotype of DPF3 loss not established
5. 2019 Low

   Proposed a tumor-suppressive role in breast cancer where DPF3 loss promotes proliferation/motility via JAK2/STAT3 activation, but the mechanistic connection is unresolved.

   Evidence siRNA/overexpression in breast cancer lines with proliferation, migration assays and JAK2/STAT3 phospho-Western blots

   PMID:31076105

   Open questions at the time

   • Pathway placement rests on Western blot only without mechanistic dissection of how DPF3 regulates JAK2/STAT3
   • Opposite (oncogenic) direction reported in renal cells, leaving tissue context unresolved
6. 2022 Medium

   Resolved how a germline risk SNP drives DPF3 in renal cancer, showing the variant creates an active enhancer HIF-binding motif that supports allele-specific HIF-driven DPF3 overexpression and oncogenic growth.

   Evidence Reporter/luciferase assays, eQTL analysis, allele-specific HIF ChIP in primary renal cells, ATAC-seq and CRISPR/siRNA DPF3 deletion with proliferation assays

   PMID:34390653 PMID:35148991

   Open questions at the time

   • Direct chromatin targets mediating DPF3's pro-growth effect not fully defined
   • Single-lab functional system
7. 2022 High

   Defined an isoform-specific cytoplasmic-to-chromatin mechanism: DPF3a binds SNIP1 in a SMAD4–p300 complex to relieve repression of p300 HAT, activating TGF-β migration genes and driving metastasis.

   Evidence Co-IP of DPF3a with SNIP1/SMAD4/p300, HAT activity assay, histone-acetylation ChIP, ccRCC migration assays and in vivo metastasis model

   PMID:35945219

   Open questions at the time

   • Whether DPF3a's PHD-reading versus protein-protein interaction drives SNIP1 binding not separated
   • Generality beyond renal cancer not tested
8. 2024 High

   Revealed a non-canonical extranuclear function, placing DPF3 at centriolar satellites, centrosomes, spindle midzone and midbodies where it is required for kinetochore-microtubule stability, mitotic fidelity, and ciliogenesis.

   Evidence Immunofluorescence/live-cell imaging localization and siRNA/CRISPR loss-of-function with mitotic and axoneme-length phenotypes

   PMID:38661008

   Open questions at the time

   • Molecular partners at centrosome/midbody not identified
   • How a chromatin-remodeling subunit is targeted to centriolar satellites is unknown
   • Relationship between chromatin and mitotic/ciliary functions unresolved

• How DPF3 switches between its nuclear BAF chromatin role and its centrosomal/ciliary role, and what determines its opposing tumor-suppressive versus oncogenic outputs across tissues, remains open.
• No structural model of dual histone-mark recognition
• Centrosomal recruitment mechanism and partners unidentified
• Tissue-specific determinants of oncogenic versus suppressive DPF3 function undefined