Affinage

SMG5

Nonsense-mediated mRNA decay factor SMG5 · UniProt Q9UPR3

Length
1016 aa
Mass
113.9 kDa
Annotated
2026-06-10
19 papers in source corpus 16 papers cited in narrative 16 extracted findings
Cross-family judge vs UniProt: Affinage preferred faithfulness: 6/6 claims corpus-supported (100%)

Mechanistic narrative

Synthesis pass · prose summary of the discoveries below

SMG5 is a core nonsense-mediated mRNA decay (NMD) factor that couples phosphorylation of the central NMD helicase UPF1 to downstream mRNA destruction (PMID:12554878, PMID:34172724). Although its PIN domain lacks the canonical catalytic residues for RNase H-type nuclease activity, SMG5 itself has no intrinsic endonuclease function (PMID:17053788). Instead, it acts as a bifunctional scaffold. Through its 14-3-3-like domain it heterodimerizes back-to-back with SMG7, and this heterodimer recognizes the C-terminal phosphorylated Ser-Gln motifs of UPF1 in a phospho-dependent manner (PMID:23348841, PMID:25013172); SMG5 recruits protein phosphatase 2A to dephosphorylate UPF1, thereby resetting the UPF1 phosphorylation cycle (PMID:12554878, PMID:12554664). The SMG5-SMG7 module additionally licenses the endonucleolytic branch of NMD, since loss of this pathway inactivates SMG6-mediated cleavage and either SMG5 or SMG7 alone suffices to authorize SMG6 endonucleolysis (PMID:34172724). SMG5 further contributes directly to cleavage through its catalytically dead PIN domain, which docks onto the SMG6 PIN domain to form a composite active site in which a conserved SMG5 aspartate completes the catalytic tetrad, enhancing SMG6 endonuclease activity ~10-fold (PMID:41763597, PMID:41714610). In vivo, SMG5-dependent NMD is essential for clearing PTC-containing transcripts; conditional loss in mouse tissues stabilizes specific substrates and disrupts development, including a Porcn/Wnt5a axis required for craniofacial morphogenesis (PMID:40071146) and an Hnrnpl-dependent splicing program required for oligodendrocyte myelination (PMID:40930975).

Mechanistic history

Synthesis pass · year-by-year structured walk · 10 steps
  1. 2003 High

    Established SMG5's first defined molecular role: linking the NMD machinery to UPF1 dephosphorylation rather than acting as an autonomous enzyme.

    Evidence Co-purification/IP and phosphatase assays in human cells, replicated by yeast two-hybrid and IP in C. elegans

    PMID:12554664 PMID:12554878

    Open questions at the time
    • Did not determine whether PP2A recruitment is direct or scaffolded through SMG7
    • Structural basis of phospho-UPF1 recognition unresolved
  2. 2006 High

    Resolved why SMG5 is not a nuclease: its PIN domain lacks the acidic catalytic triad present in the active SMG6 PIN domain, defining SMG5 as a non-catalytic NMD factor.

    Evidence X-ray crystallography of human SMG5/SMG6 PIN domains with in vitro ssRNA assays and Drosophila dominant-negative genetics

    PMID:17053788

    Open questions at the time
    • Did not address whether the SMG5 PIN domain has any function despite lacking catalysis
    • No interaction tested between SMG5 and SMG6 PIN domains
  3. 2013 High

    Defined the structural logic of the SMG5-SMG7 module, showing 14-3-3-like heterodimerization is required for high-affinity phospho-UPF1 binding and for NMD.

    Evidence Crystal structure of C. elegans SMG5-SMG7 plus structure-based mutagenesis and human cell NMD reporters

    PMID:23348841

    Open questions at the time
    • Degradative activity localized to SMG7, leaving SMG5's contribution to decay undefined
    • Did not address SMG6 branch coupling
  4. 2014 High

    Reconstituted phospho-dependent UPF1 recognition by purified SMG5-SMG7, distinguishing it from SMG6's largely phospho-independent UPF1 binding.

    Evidence In vitro reconstitution with purified proteins, crystallography of the SMG6 14-3-3-like domain, and binding assays

    PMID:25013172

    Open questions at the time
    • Did not test functional consequences of the two recognition modes in cells
    • Did not address how SMG5-SMG7 and SMG6 binding are coordinated on UPF1
  5. 2012 Medium

    Proposed that SMG5 bridges phospho-UPF1 to decapping via PNRC2 and Dcp1a, placing SMG5 at a specific decay step.

    Evidence Co-IP, siRNA knockdown, tethering and microarray in a single lab

    PMID:23234702

    Open questions at the time
    • Directly contradicted by a later study finding no SMG5-PNRC2 interaction
    • Single-lab interaction not independently reproduced
  6. 2018 Medium

    Challenged the SMG5-PNRC2 bridging model, reassigning PNRC2-driven 5'-3' decay to a direct UPF1-PNRC2 interaction while confirming the SMG5-SMG7 requirement.

    Evidence Co-IP, tethering, siRNA knockdown and NMD reporters

    PMID:29348139

    Open questions at the time
    • Discrepancy with the earlier PNRC2 study not mechanistically resolved
    • Single-lab study
  7. 2018 Medium

    Demonstrated genetic essentiality of Smg5 for NMD in vivo, covering both SMG6-dependent and SMG6-independent decay routes.

    Evidence Drosophila genetics, NMD reporters and epistasis analysis

    PMID:29903866

    Open questions at the time
    • Molecular identity of the SMG6-independent route not defined
    • Epistasis with Smg1 not extended to mammalian systems
  8. 2021 High

    Established that the SMG5-SMG7 pathway licenses the SMG6 endonucleolytic branch, with SMG5 able to substitute for SMG7.

    Evidence siRNA depletion of SMG5/SMG7 with transcriptome-wide RNA-seq and NMD reporters in human cells

    PMID:34172724

    Open questions at the time
    • Did not define the physical basis of SMG6 licensing by SMG5
    • Did not determine redundancy mechanism between SMG5 and SMG7
  9. 2026 High

    Revealed the molecular basis of SMG6 licensing: the catalytically dead SMG5 PIN domain directly contacts the SMG6 PIN domain to form a composite active site that completes the catalytic tetrad and enhances cleavage ~10-fold.

    Evidence AlphaFold modeling, in vitro pulldowns and reconstituted endonuclease assays, compensatory mutagenesis and cellular NMD assays across C. elegans and human components, converging across three concurrent studies

    PMID:41638882 PMID:41714610 PMID:41763597

    Open questions at the time
    • Composite PIN interface predicted by AlphaFold but not experimentally solved by crystallography or cryo-EM
    • Stoichiometry and regulation of cPIN assembly in cells undefined
  10. 2025 Medium

    Connected SMG5-dependent NMD to mammalian developmental programs by identifying tissue-specific PTC-containing substrates whose stabilization drives phenotypes.

    Evidence Conditional knockout mice in neural crest and oligodendrocyte lineages with transcriptome analysis, Western blot, rescue and histology (Porcn/Wnt5a; Hnrnpl/Mag/Nfasc)

    PMID:39199410 PMID:39704269 PMID:40071146 PMID:40930975

    Open questions at the time
    • Some substrate-to-phenotype links rely on transcriptome correlation
    • c-MYC and p38 MAPK effects mechanistically inferred rather than fully reconstituted
    • Single-lab tissue models

Open questions

Synthesis pass · forward-looking unresolved questions
  • How phospho-UPF1 recognition by the SMG5-SMG7 14-3-3 module, PP2A recruitment, and assembly of the SMG5-SMG6 composite PIN are temporally coordinated on a single mRNP remains unresolved.
  • No integrated structure of UPF1-SMG5-SMG6/SMG7 on substrate
  • Order of dephosphorylation versus endonucleolytic licensing unknown

Mechanism profile

Synthesis pass · controlled-vocabulary classification · explore literature graph →
Molecular activity
GO:0060090 molecular adaptor activity 3 GO:0003723 RNA binding 2 GO:0098772 molecular function regulator activity 2
Localization
GO:0005829 cytosol 1
Pathway
R-HSA-8953854 Metabolism of RNA 3 R-HSA-74160 Gene expression (Transcription) 1
Partners
PNRC2PP2ASMG1SMG6SMG7UPF1UPF2UPF3X
Complex memberships
SMG5-SMG6 composite PIN (cPIN) heterodimerSMG5-SMG7 14-3-3-like heterodimer

Evidence

Reading pass · 16 per-paper findings extracted from the source corpus
Year Finding Method Journal Conf PMIDs
2006 Crystal structures of human SMG5 and SMG6 PIN domains reveal that SMG6 has the canonical acidic triad required for RNase H-type nuclease activity and can degrade single-stranded RNA in vitro, whereas SMG5 lacks key catalytic residues and has no intrinsic nuclease activity. An SMG6 with an inactive PIN domain acts as a dominant-negative inhibitor of NMD in Drosophila. X-ray crystallography, in vitro ssRNA degradation assay, Drosophila dominant-negative genetics The EMBO journal High 17053788
2003 Human SMG5 (hSmg5/7a) co-purifies with UPF1, UPF2, UPF3X, SMG1, and the catalytic subunit of protein phosphatase 2A (PP2A), and is required for dephosphorylation of UPF1 (but not UPF2), indicating SMG5 recruits PP2A to UPF1. SMG5 is predominantly cytoplasmic in HEK293T cells. Co-purification/co-immunoprecipitation, Western blot phosphatase assay, subcellular fractionation/Western blot RNA (New York, N.Y.) High 12554878
2003 C. elegans SMG-5 interacts with SMG-7, SMG-2 (UPF1 ortholog), and both the structural (PR65) and catalytic subunits of PP2A, as determined by immunoprecipitation and yeast two-hybrid assays, directing PP2A to dephosphorylate SMG-2. Co-immunoprecipitation, yeast two-hybrid The EMBO journal High 12554664
2013 Crystal structure of C. elegans SMG5-SMG7 complex shows their 14-3-3-like domains heterodimerize in an unusual perpendicular back-to-back orientation. Heterodimerization increases affinity of the complex for phosphorylated UPF1, and the degradative activity of the SMG5-SMG7 complex resides in SMG7. Structure-based mutations that disrupt the interaction impair NMD in human cells. X-ray crystallography, in vitro binding assays, structure-based mutagenesis, human cell NMD reporter assays Genes & development High 23348841
2014 Using purified components reconstituted in vitro, the SMG5-SMG7 14-3-3-like heterodimer recognizes a short C-terminal phosphorylated segment of UPF1 (containing the last two Ser-Gln motifs) in a phospho-dependent manner, whereas SMG6 has both a weak phospho-dependent and a dominant phospho-independent interaction with UPF1 mediated by a low-complexity region bordering its 14-3-3-like domain and the UPF1 helicase domain and C-terminal tail. Crystal structure of SMG6 14-3-3-like domain confirms a phosphoserine-binding site. In vitro reconstitution with purified proteins, crystal structure, binding assays Nucleic acids research High 25013172
2012 PNRC2 preferentially forms a complex with SMG5 (not SMG6 or SMG7); SMG5 bridges UPF1 (via phospho-dependent interaction) to PNRC2, which then connects to the decapping factor Dcp1a. Downregulation of PNRC2 abolishes the SMG5-Dcp1a interaction. Tethering assays place UPF1, SMG5, and PNRC2 at the same step in NMD. Co-immunoprecipitation, siRNA knockdown, tethering assay, microarray Nucleic acids research Medium 23234702
2018 An interaction between SMG5 and PNRC2 was not detected physically or functionally in NMD reporters; instead, UPF1 directly interacts with PNRC2 and triggers 5'-3' exonucleolytic decay in tethering assays. SMG5-SMG7 complex interaction and its functional requirement for NMD was confirmed. PNRC2 knockdown does not affect NMD reporter RNA levels. Co-immunoprecipitation, tethering assay, siRNA knockdown, NMD reporter assay RNA (New York, N.Y.) Medium 29348139
2021 Loss of the SMG5-SMG7 pathway also inactivates the SMG6 endonucleolytic branch, demonstrating that SMG5-SMG7 recruitment is required to authorize/license SMG6-mediated cleavage. Either SMG5 or SMG7 alone is sufficient to support SMG6 endonucleolysis and activate NMD, revealing SMG5 can substitute for SMG7. siRNA knockdown of SMG5/SMG7, transcriptome-wide RNA-seq, NMD reporter assays in human cells Nature communications High 34172724
2018 In Drosophila, Smg5 is genetically essential for NMD activity, required for both SMG6-dependent endonucleolytic cleavage and an additional SMG6-independent degradation mechanism. Epistasis analysis shows Smg1 becomes essential for NMD when Smg5 function is partially compromised. Drosophila genetics, NMD reporter assays, epistasis analysis Genetics Medium 29903866
2026 In C. elegans, the PIN domain of SMG-5, though catalytically inactive by canonical criteria, contains highly conserved residues essential for NMD. AlphaFold predicts a direct PIN-PIN interaction between SMG-5 and SMG-6, validated by in vitro pulldowns. Compensatory salt-bridge flip mutations confirm this interface is functionally required for SMG-6-mediated mRNA cleavage. AlphaFold structural prediction, in vitro pulldown, C. elegans genetic complementary mutagenesis, NMD reporter assays RNA (New York, N.Y.) Medium 41638882
2026 AlphaFold models predict a high-confidence SMG5-SMG6 PIN domain interface forming a composite active site, where a conserved SMG5 aspartate (D893) completes the canonical tetrad required for PIN-domain catalysis. In vitro, SMG6 alone has weak endonuclease activity that is enhanced ~10-fold by the addition of the SMG5 PIN domain. Mutations at the predicted interface, RNA-binding sites, or active site abolish this composite endonuclease activity and impair cellular NMD. AlphaFold structural prediction, in vitro reconstituted endonuclease assay, mutagenesis, cell-based NMD assay Journal of molecular biology Medium 41763597
2026 SMG5 and SMG6 interact via their PIN domains to form a composite PIN (cPIN) heterodimer with full endonuclease activity. Reconstituted SMG5-SMG6 cPIN heterodimers show high in vitro activity; SMG5 completes the SMG6 active site and substrate binding site. Mutations at their predicted interaction surfaces, RNA-binding sites, or active site attenuate or abolish cPIN activity in vitro and impair cellular NMD. Structural predictions (AlphaFold), biochemical in vitro reconstitution, mutagenesis, cell-based NMD analysis Nature communications High 41714610
2025 Conditional knockout of Smg5 in mouse craniofacial neural crest cells causes abnormal accumulation of a PTC-containing Porcn transcript, leading to reduced Porcn protein and impaired Wnt5a/JNK signaling, resulting in hypoplastic mandibles, tongue mispositioning, and cleft palate. Wnt5a addition to SMG5-deficient CNC explants ameliorates cell death. Conditional knockout mouse model, transcriptome analysis, Western blot for Porcn, Wnt5a rescue in CNC explants iScience Medium 40071146
2024 SMG5 knockout in mouse embryonic germ cells causes spermatogenesis failure (Sertoli cell-only phenotype) due to defective spermatogonial differentiation and maintenance. SMG5 loss leads to hyperactivation of the p38 MAPK signaling pathway causing widespread cell death during spermatogonial differentiation. Conditional knockout mouse model, transcriptome analysis, Western blot for p38 MAPK pathway components, histology FASEB journal Medium 39704269
2024 Smg5 knockout mESCs are viable but show delayed differentiation. SMG5 loss causes upregulation of c-MYC protein (but not c-Myc mRNA), attributed to enhanced protein synthesis, and dysregulates alternative splicing of stem cell differentiation regulators. Conditional knockout mESCs, Western blot, polysome profiling inference, transcriptome/alternative splicing analysis Biomolecules Medium 39199410
2025 Conditional knockout of Smg5 in oligodendrocyte lineage cells impairs oligodendrocyte differentiation and myelination due to failure to degrade PTC-containing Hnrnpl variant transcripts. Excess HNRNPL disrupts alternative splicing of myelin-associated genes Mag and Nfasc. Conditional knockout mouse model, RNA analysis, Western blot, electron microscopy of myelin, motor function assays The Journal of neuroscience Medium 40930975

Source papers

Stage 0 corpus · 19 papers · ranked by NIH iCite citations
Year Title Journal Citations PMID
2006 Structures of the PIN domains of SMG6 and SMG5 reveal a nuclease within the mRNA surveillance complex. The EMBO journal 170 17053788
2003 Characterization of human Smg5/7a: a protein with similarities to Caenorhabditis elegans SMG5 and SMG7 that functions in the dephosphorylation of Upf1. RNA (New York, N.Y.) 135 12554878
2003 SMG-5, required for C.elegans nonsense-mediated mRNA decay, associates with SMG-2 and protein phosphatase 2A. The EMBO journal 122 12554664
2014 Phospho-dependent and phospho-independent interactions of the helicase UPF1 with the NMD factors SMG5-SMG7 and SMG6. Nucleic acids research 90 25013172
2013 An unusual arrangement of two 14-3-3-like domains in the SMG5-SMG7 heterodimer is required for efficient nonsense-mediated mRNA decay. Genes & development 79 23348841
2021 SMG5-SMG7 authorize nonsense-mediated mRNA decay by enabling SMG6 endonucleolytic activity. Nature communications 73 34172724
2012 SMG5-PNRC2 is functionally dominant compared with SMG5-SMG7 in mammalian nonsense-mediated mRNA decay. Nucleic acids research 73 23234702
2013 The RNA helicase Ddx5/p68 binds to hUpf3 and enhances NMD of Ddx17/p72 and Smg5 mRNA. Nucleic acids research 31 23788676
2018 Dissecting the functions of SMG5, SMG7, and PNRC2 in nonsense-mediated mRNA decay of human cells. RNA (New York, N.Y.) 27 29348139
2018 Multiple Nonsense-Mediated mRNA Processes Require Smg5 in Drosophila. Genetics 14 29903866
2016 MicroRNA 433 regulates nonsense-mediated mRNA decay by targeting SMG5 mRNA. BMC molecular biology 10 27473591
2025 Fine-tuning of Wnt signaling by RNA surveillance factor Smg5 in the mouse craniofacial development. iScience 6 40071146
2024 SMG5, a component of nonsense-mediated mRNA decay, is essential for the mouse spermatogonial differentiation and maintenance. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 5 39704269
2024 RNA Surveillance Factor SMG5 Is Essential for Mouse Embryonic Stem Cell Differentiation. Biomolecules 4 39199410
2026 The PIN domain of SMG-5 functionally interacts with SMG-6 to stimulate NMD. RNA (New York, N.Y.) 3 41638882
2026 Active Site Assembly by SMG5 as a Mechanism for SMG6 Endonuclease Licencing in Nonsense-mediated mRNA Decay. Journal of molecular biology 3 41763597
2024 SMG5 Inhibition Restrains Hepatocellular Carcinoma Growth and Enhances Sorafenib Sensitivity. Molecular cancer therapeutics 3 38647536
2026 Composite SMG5-SMG6 PIN domain formation is essential for NMD. Nature communications 2 41714610
2025 Smg5 Enhances Oligodendrocyte Differentiation via Nonsense-Mediated mRNA Decay of Hnrnpl Variant Transcripts. The Journal of neuroscience : the official journal of the Society for Neuroscience 2 40930975

Missed literature

Know a paper Affinage missed for SMG5? Flag it for the maintainers and the community.

No submissions yet.