SON

SON is a nuclear speckle scaffold protein and RNA splicing cofactor that organizes the subnuclear compartmentalization of splicing factors and promotes efficient cotranscriptional splicing of introns with weak splice sites in genes governing the cell cycle, brain development, centriole biogenesis, and kidney morphogenesis (PMID:33095160, PMID:21504830, PMID:20053686, PMID:27545680, PMID:34406792). Together with SRRM2, SON forms the organizing core of nuclear speckles; its depletion disassembles these bodies and disperses snRNPs, SR proteins, and casein kinase 1 family members (PMID:33095160, PMID:27681590). Independent of its splicing role, SON binds DNA near transcription start sites, recruits menin to inhibit MLL complex assembly and H3K4 trimethylation, and thereby represses transcription — a function antagonized by alternatively spliced short SON isoforms enriched in acute myeloid leukemia (PMID:26990989). Heterozygous loss-of-function SON mutations cause a syndromic neurodevelopmental disorder characterized by intellectual disability, brain malformations, and congenital anomalies of the kidney and urinary tract, attributable to haploinsufficient mis-splicing of target transcripts (PMID:27545680, PMID:31005274).

1. 2001 High

   Establishing that SON is a sequence-specific DNA-binding protein resolved its initial molecular activity: it binds a GA-rich consensus element upstream of the HBV core promoter and represses transcription, defining SON as a transcriptional repressor.

   Evidence Expression cloning, EMSA with blocking antibody, PCR-assisted binding-site selection, and reporter assays in hepatoma cells

   PMID:11306577

   Open questions at the time

   • Endogenous genomic targets were unknown
   • Mechanism of repression (cofactor recruitment) was not addressed
2. 2008 High

   Identification of SON as an interaction partner of the AML1-ETO NHR4 domain linked it to leukemogenesis and showed that SON binding is required for AML1-ETO-induced growth arrest, positioning SON at the intersection of proliferation control and leukemia.

   Evidence Co-immunoprecipitation, NHR4 mutagenesis, retroviral bone-marrow transplant leukemia model, and siRNA knockdown

   PMID:18952841

   Open questions at the time

   • Whether SON's role in leukemogenesis acts through its splicing or transcriptional function was unclear
   • SON's broader contribution to normal hematopoiesis was unknown
3. 2010 High

   Demonstrating that SON depletion disorganizes nuclear speckle architecture and arrests cells at metaphase established SON as a structural organizer of splicing factor compartments and linked this to cell-cycle progression.

   Evidence RNAi in HeLa cells with immunofluorescence of snRNP and SR protein markers

   PMID:20053686

   Open questions at the time

   • Whether SON acts enzymatically or as a scaffold was unresolved
   • Direct splicing targets had not been identified
4. 2011 High

   Defining SON as an RNA splicing cofactor that promotes SR-protein interaction with RNA Pol II and spliceosome components at weak splice sites in cell-cycle genes provided the core mechanistic model for its splicing function.

   Evidence RNAi, RT-PCR splicing assays, co-IP of SR proteins with RNA Pol II, live-cell imaging of mitotic defects

   PMID:21504830

   Open questions at the time

   • Genome-wide splicing target landscape was incomplete
   • Whether DNA-binding and splicing roles are coupled or independent was unknown
5. 2016 High

   Three concurrent advances resolved SON's chromatin-level mechanism, its disease relevance, and tissue-specific splicing targets: SON binds DNA at TSSs to block MLL complex assembly via menin, heterozygous SON mutations cause a neurodevelopmental syndrome through mis-splicing of brain-development genes, and SON haploinsufficiency disrupts CAKUT gene splicing in the kidney.

   Evidence ChIP-seq and RNA-seq in hematopoietic cells (transcriptional repression); patient RNA-seq, zebrafish morpholinos, and RT-PCR (neurodevelopment); SON knockdown in kidney cells and patient fibroblasts (CAKUT)

   PMID:26990989 PMID:27545680 PMID:31005274

   Open questions at the time

   • Structural basis for dual DNA-binding and RNA-splicing activities was unknown
   • How short SON isoforms are regulated in AML was not fully defined
   • Whether splicing and transcriptional targets overlap was unresolved
6. 2020 High

   Multiple studies extended SON's functional reach: SON and SRRM2 were shown to form the minimal organizing core of nuclear speckles via their intrinsically disordered regions; Son knockdown in mouse cortex caused defective neuronal migration and reduced dendritic spine density (rescued by human SON); and SON was required for macrophage autophagy and interferon response through splicing of GBF1.

   Evidence Co-depletion and IDR-deletion microscopy (speckle organization); in utero electroporation with shRNA and human SON rescue (neuronal migration); RNAi and GBF1 chemical inhibition phenocopy in macrophages

   PMID:32448361 PMID:32484234 PMID:33095160

   Open questions at the time

   • Phase-separation biophysics of SON-SRRM2 condensates were not characterized
   • In vivo brain phenotype was from knockdown, not genetic knockout
   • Macrophage findings relied on a single pathogen infection model
7. 2021 High

   SON was shown to drive oncogenic splicing in glioblastoma by maintaining PTBP1 expression and forming a complex with hnRNP A2B1 to antagonize RBFOX2-mediated exon inclusion, and separately to control centriole assembly and ciliogenesis through splicing of CEP131.

   Evidence RNA-seq and Co-IP of SON-hnRNP A2B1 with orthotopic GBM xenografts (glioblastoma); whole-genome mRNA-seq, EM of centrioles, and ciliogenesis assays (centriole/cilia)

   PMID:34406792 PMID:34548489

   Open questions at the time

   • Whether SON-hnRNP A2B1 complex exists outside GBM was not tested
   • Whether ciliogenesis defects contribute to the neurodevelopmental phenotype was unknown
8. 2023 High

   Identification of SON as a critical m6A target in hematopoietic stem cells resolved how epitranscriptomic regulation controls HSC self-renewal: m6A depletion paradoxically increases Son mRNA but depletes SON protein; SON reintroduction rescues HSC symmetric commitment and suppresses an inflammatory gene program including CCL5 while maintaining MYC expression.

   Evidence m6A profiling, conditional SON knockout and overexpression in mouse HSCs, transplantation assays, RNA-seq, human HSC validation

   PMID:38065069

   Open questions at the time

   • Mechanism by which m6A promotes SON translation was not defined
   • Whether SON's anti-inflammatory role in HSCs is through splicing, transcription, or both was not dissected

• Major open questions include the structural basis for SON's bifunctional DNA-binding/splicing-cofactor activities, the biophysical rules governing SON-SRRM2 phase separation in nuclear speckle assembly, and whether SON's transcriptional repression and splicing functions are coupled or independently regulated in different cell types.
• No high-resolution structure of any SON domain
• Phase-separation properties not biophysically characterized
• Cell-type-specific separation of transcriptional vs. splicing roles not achieved