PHF19

PHF19 (PCL3) is a Polycomb-like accessory subunit of PRC2 that couples recognition of active-chromatin marks to deposition of the repressive H3K27me3 modification, thereby controlling the balance between proliferation and differentiation across stem cells and multiple cancers (PMID:23104054, PMID:23160351, PMID:31383640). Its Tudor domain selectively binds H3K36me2/me3, an interaction required for full PRC2 methyltransferase activity and for recruiting both PRC2 and the H3K36me3 demethylase NO66 to target genes during differentiation (PMID:23104054, PMID:23160351). Genome-wide, PHF19 promotes H3K27me3 spreading from CpG islands and the organization of broad H3K27me3 domains, and its loss de-represses PRC2 target genes including cell-cycle inhibitors and interferon-JAK-STAT genes (PMID:31383640, PMID:32821835); in triple-negative breast cancer cells it drives micron-scale nuclear PRC2 clusters through an intrinsically disordered region that stabilizes local PRC2 occupancy and reinforces H3K27me3 macro-domains (PMID:41066230). PHF19 also directly represses individual target promoters, repressing SIAH1 to stabilize β-catenin and regulating the cell-cycle inhibitor p21 to promote proliferation (PMID:30323224, PMID:33996816). Its activity is antagonized or redirected by related PCL proteins MTF2/PCL2, whose chromatin recruitment increases upon PHF19 loss [PMID:32155117, PMID:bio_10.1101_2025.05.15.654236], and it is regulated through RNA-based mechanisms: WTAP-deposited m6A on the PHF19 3'-UTR stabilizes its transcript (PMID:40038518), while the long isoform PHF19L is recruited to m6A-modified nascent RNA via YTHDC1 to form a liquid condensate that sequesters EZH2 away from chromatin and activates otherwise-silenced genes (PMID:41129231).

1. 2012 High

   Establishing how a Polycomb-like protein reads active chromatin, the Tudor domain of PHF19 was shown to directly and selectively bind H3K36me3, defining the molecular basis for targeting PRC2 to active-mark chromatin.

   Evidence NMR spectroscopy and biochemical histone-binding assays, replicated across two concurrent papers

   PMID:23104054 PMID:23160351 PMID:23228662

   Open questions at the time

   • Does not establish in vivo genome-wide consequences of the binding
   • PHD-domain function not resolved
2. 2012 High

   Linking the histone-reading activity to enzymatic output, Tudor binding to H3K36me2/me3 was shown to be required for full PRC2 methyltransferase activity, connecting mark recognition to catalysis.

   Evidence In vitro reconstituted methyltransferase assays with Phf19 mutants plus NMR and mutagenesis

   PMID:23104054

   Open questions at the time

   • Quantitative contribution relative to other PRC2 cofactors not defined
3. 2012 High

   Defining the developmental function, PHF19 was shown to recruit PRC2 and the H3K36me3 demethylase NO66 to stem-cell genes during differentiation, providing a model for converting active to repressive chromatin.

   Evidence ChIP, knockdown-rescue in mouse ESCs and co-immunoprecipitation

   PMID:23104054 PMID:23160351

   Open questions at the time

   • NO66 association rests on Co-IP without reciprocal structural validation
   • Order of NO66 vs PRC2 recruitment not resolved
4. 2012 Medium

   Connecting PHF19 to cancer signaling, Akt was shown to activate PHF19 expression in melanoma where PHF19 silencing reduces proliferation but increases transendothelial migration, hinting at context-dependent roles.

   Evidence siRNA knockdown, proliferation and transendothelial migration assays, ChIP at the PHF19 promoter

   PMID:22487681

   Open questions at the time

   • Mechanism linking PHF19 to migration not defined
   • Single lab, single cancer type
5. 2015 Medium

   Placing PHF19 in an upstream signaling cascade, G9a was shown to upregulate PCL3/PHF19 to increase PRC2 recruitment and H3K27 methylation at the E-cadherin promoter, positioning PHF19 as a downstream effector controlling an EMT-relevant gene.

   Evidence ChIP, knockdown, overexpression and Western blot in pancreatic cancer cells

   PMID:26688070

   Open questions at the time

   • Direct vs indirect transcriptional control of PHF19 by G9a unresolved
   • Generality beyond E-cadherin not tested
6. 2018 Medium

   Identifying a direct target-gene mechanism, PHF19 was shown to bind the SIAH1 promoter and repress it to stabilize β-catenin in glioma, establishing a β-catenin-dependent route to proliferation and migration.

   Evidence siRNA knockdown, ChIP, Wnt/β-catenin inhibitor epistasis, proliferation/migration assays; separate Co-IP confirming PRC2 (EZH2/EED/SUZ12) assembly

   PMID:30131250 PMID:30323224

   Open questions at the time

   • Whether SIAH1 repression is PRC2/H3K27me3-dependent not shown
   • Single-lab tumor models
7. 2019 High

   Demonstrating genome-scale chromatin function, PHF19 was shown to drive H3K27me3 spreading from CpG islands and organization of broad repressive domains in multiple myeloma, with loss de-repressing cell-cycle and interferon-JAK-STAT genes; a parallel study showed miR-155/Ship1/Akt signaling promotes Phf19 to restrain T-cell senescence in a histone-binding-dependent manner.

   Evidence ChIP-seq, RNA-seq, shRNA, xenograft in MM; miRNA gain/loss, ChIP and T-cell assays

   PMID:31089138 PMID:31383640

   Open questions at the time

   • Mechanism of domain spreading not biochemically reconstituted
   • Direct vs secondary targets of de-repression not fully separated
8. 2020 High

   Revealing functional interplay among PCL proteins, PHF19 depletion in prostate cancer was shown to increase MTF2/PCL2 recruitment and genome-wide gain of PRC2/H3K27me3, while in HSCs Phf19 loss redistributed H3K27me3 and enhanced quiescence, indicating PHF19 shapes where, not simply how much, H3K27me3 is deposited.

   Evidence Co-IP, ChIP-seq, RNA-seq and functional assays in prostate cancer; conditional knockout, H3K27me3 ChIP-seq and reconstitution assays in mouse HSCs

   PMID:32155117 PMID:32821835

   Open questions at the time

   • Molecular basis of MTF2/PHF19 competition not structurally defined
   • Tissue-specificity of redistribution outcomes unexplained
9. 2021 Medium

   Extending the proliferation-differentiation axis, PHF19 was shown to directly regulate p21 in CML with MTF2 partially compensating to direct erythroid differentiation, and a short PHF19 isoform was found to bind β-TrCP to suppress Gli1 ubiquitination and activate Hedgehog signaling in HCC, revealing isoform- and PRC2-independent functions.

   Evidence siRNA, ChIP, differentiation assays in CML; Co-IP, ubiquitination assay, xenograft and in vivo KO in HCC

   PMID:33996816 PMID:34129846

   Open questions at the time

   • β-TrCP/Gli1 mechanism rests on single-lab Co-IP and ubiquitination assays
   • Isoform expression patterns across tissues not mapped
10. 2025 High

    Defining higher-order and RNA-coupled regulation, studies showed PHF19 forms IDR-dependent micron-scale nuclear PRC2 clusters reinforcing H3K27me3 macro-domains and TNBC motility, that WTAP-mediated 3'-UTR m6A stabilizes PHF19 mRNA in t(8;21) AML, and that PHF19L forms a YTHDC1 condensate sequestering EZH2 from chromatin to activate genes in prostate cancer.

    Evidence Spatial proteomics, high-resolution imaging and IDR mutagenesis (TNBC); m6A-seq, RNA stability, ChIP-seq, ATAC-seq, in vivo AML models; Co-IP, condensate assays, ChIP-seq, RNA-seq (prostate)

    PMID:40038518 PMID:41066230 PMID:41129231

    Open questions at the time

    • Whether clustering and condensate functions coexist in the same cells unknown
    • Switch between repressive PRC2-recruiting and EZH2-sequestering activities not mechanistically unified
11. 2025 Medium

    Probing opposing PCL functions in development, separation-of-function mutants in human pluripotent stem cells indicated PHF19 antagonizes MTF2-PRC2.1-mediated repression at developmental genes during cardiomyocyte differentiation.

    Evidence CRISPR separation-of-function mutants in hPSCs, ChIP-seq and differentiation assays (preprint)

    PMID:bio_10.1101_2025.05.15.654236

    Open questions at the time

    • Preprint, single lab, not yet peer-reviewed
    • Reconciliation with PHF19's repressive roles in other contexts unresolved

• How PHF19 switches between recruiting PRC2 to deposit H3K27me3 and sequestering EZH2 to activate genes—across isoforms, condensates, and competing PCL proteins—remains unresolved.
• No unified model integrating repressive and activating modes
• Determinants of isoform choice and condensate formation in vivo unknown
• Structural basis of PHF19 vs MTF2 chromatin partitioning not defined