id: F7A4N8 gene_symbol: CTDSP2 product_type: PROTEIN status: DRAFT taxon: id: NCBITaxon:9796 label: Equus caballus description: >- F7A4N8 is a 174-residue product predicted from one alternative transcript (ENSECAT00000020049) of the horse CTDSP2 gene (ENSECAG00000018788, chromosome 6), which encodes the small CTD phosphatase SCP2. Its first 84 residues are essentially identical to the N-terminal non-catalytic region of human CTDSP2, after which the transcript uses an extended exon whose extra sequence encodes 90 unrelated residues and a stop codon. The product therefore lacks the entire FCP1-homology (HAD-type) phosphatase domain, including the DxDx(T/V) motif that carries the nucleophilic and proton-donor aspartates. Other transcripts of the same gene (Ensembl canonical ENSECAT00000107019, UniProt A0A9L0T5X1, 312 aa; and ENSECAT00000138494, A0A9L0R207, 252 aa) do encode the complete catalytic domain, and the full-length product groups with CTDSP2 orthologs (PANTHER PTHR12210:SF187 members in human, mouse and cow) rather than with CTDSP1 or CTDSPL. In mammals the full-length enzyme is a nuclear Mg2+-dependent phosphatase that preferentially removes phosphate from Ser5 of the RNA polymerase II CTD and is recruited by REST to silence RE-1-containing neuronal genes in non-neuronal cells. Because the truncated F7A4N8 product lacks the catalytic domain and its transcript carries a premature stop codon upstream of four exon junctions, it has no demonstrable enzymatic function, and it may be a nonsense-mediated-decay transcript or annotation artefact rather than a stable protein. existing_annotations: [] suggested_questions: - question: Is ENSECAT00000020049, the transcript encoding F7A4N8, supported by horse RNA-seq or long-read data? Its 375-bp exon extends 273 nt upstream of the canonical 102-bp exon acceptor and introduces a premature stop codon upstream of four further exon junctions, the signature of a nonsense-mediated-decay or intron-retention isoform rather than a stable protein product. experts: [] - question: Which horse CTDSP2 model best represents the real full-length product? The Ensembl canonical model (A0A9L0T5X1, 312 aa) has an N-terminal extension that does not begin with methionine and omits the 13-residue SDLLQCLQYQFYQ exon present in human CTDSP2, whereas A0A9L0R207 (252 aa) keeps that exon but omits the 19-residue exon that forms part of the catalytic fold. Neither model reproduces the human 271-residue architecture exactly. experts: [] - question: If a short F7A4N8-like protein is made, does its retained N-terminal segment (which in the SCP proteins lies outside the catalytic domain) have any function, such as competing with full-length CTDSP2 for partners, or is it simply degraded? experts: [] suggested_experiments: - hypothesis: The F7A4N8-encoding transcript (ENSECAT00000020049) is a minor, NMD-targeted isoform, and the dominant horse CTDSP2 transcript encodes a full-length catalytic protein like human CTDSP2. description: Quantify junction-spanning reads for the extended exon versus the canonical 102-bp exon acceptor in public horse RNA-seq (e.g. FAANG tissues), and compare isoform abundance in horse cells with and without NMD inhibition (cycloheximide or UPF1 knockdown). Confirm full-length ORFs by long-read (Iso-Seq) sequencing. experiment_type: transcript isoform quantification / NMD sensitivity - hypothesis: Phosphatase activity on RNA polymerase II CTD Ser5 is carried by the full-length horse CTDSP2 product and is absent from F7A4N8, which lacks the catalytic domain. description: Express the full-length horse CTDSP2 product and F7A4N8, and assay phosphate release from a Ser5-phosphorylated CTD heptapeptide substrate, with human CTDSP2 as a positive control and a catalytic-aspartate (D-to-A) mutant as a negative control. experiment_type: in vitro phosphatase assay with Ser5-phosphorylated CTD substrate references: - id: file:HORSE/CTDSP2/CTDSP2-uniprot.txt title: 'UniProtKB F7A4N8: selected horse CTDSP2 protein model' findings: - statement: F7A4N8 is an unreviewed (PE 4, predicted) 174-residue TrEMBL record imported from an Ensembl gene model, with no equine experimental characterization. record_section: [PE] supporting_text: 'PE 4: Predicted;' reference_section_type: DATABASE_ENTRY - statement: The record maps to Ensembl transcript ENSECAT00000020049 of gene ENSECAG00000018788 and to the VGNC CTDSP2 locus; the gene symbol identifies the locus, not this transcript's function. record_section: ["DR:Ensembl"] supporting_text: 'DR Ensembl; ENSECAT00000020049.4; ENSECAP00000016443.3; ENSECAG00000018788.4.' reference_section_type: DATABASE_ENTRY - statement: The only sequence features predicted on the record are a disordered, low-complexity stretch inside the transcript-specific C-terminal tail; no phosphatase domain signature is matched. record_section: ["FT:REGION"] supporting_text: /note="Disordered" reference_section_type: DATABASE_ENTRY reference_review: relevance: HIGH correctness: VERIFIED review_notes: Source record for the reviewed product. Naming is computational (Ensembl model plus VGNC symbol); it fixes the locus identity but establishes no function for this transcript product. The re-review replaced a former finding that quoted the GN line, which restates a label. - id: file:HORSE/CTDSP2/CTDSP2-bioinformatics/RESULTS.md title: CTDSP2 human-horse sequence comparison findings: - statement: A forced global alignment of F7A4N8 against human CTDSP2 pairs 170 residues; the re-review analysis shows the identity is concentrated in the first 84 residues and the rest of the pairing is alignment noise against an unrelated tail. supporting_text: Paired coverage is 62.7% of human O14595 (271 aa) and 97.7% of horse F7A4N8 (174 aa). reference_section_type: RESULTS - statement: The alignment is not a reciprocal orthology test and does not by itself license transferring the human catalytic annotation to this product. supporting_text: This measures conservation between the identified records. It is not a reciprocal orthology analysis and does not itself validate a functional annotation. reference_section_type: RESULTS reference_review: relevance: MEDIUM correctness: VERIFIED review_notes: Earlier reproducible Biopython global alignment. Its feature table correctly records that both human catalytic aspartates are unaligned in F7A4N8. The 57.1% identity over 170 paired residues overstates similarity beyond residue 84, because a global alignment with these gap penalties pairs the unrelated C-terminal tail. Superseded for gene-model and placement questions by RESULTS-rereview.md. - id: file:HORSE/CTDSP2/CTDSP2-bioinformatics/RESULTS-rereview.md title: CTDSP2 (horse) re-review gene model and phylogenetic placement findings: - statement: The horse CTDSP2 gene has three Ensembl protein-coding transcripts. F7A4N8 comes from the non-canonical ENSECAT00000020049, whose sixth exon (375 bp) extends 273 nt upstream of the 102-bp exon used by both other transcripts. supporting_text: 76298535-76298161 (375 bp) reference_section_type: RESULTS - statement: F7A4N8 shares only an 84-residue prefix with human CTDSP2 and lacks the DxDx(T/V) catalytic motif found at human position 107. supporting_text: '| horse_CTDSP2gene_F7A4N8 | 174 | 84 | 109/164 | none |' reference_section_type: RESULTS - statement: The Ensembl canonical horse product A0A9L0T5X1 (312 aa) retains the catalytic DLDET motif at position 148. supporting_text: "| horse_CTDSP2gene_A0A9L0T5X1 | 312 | 0 | 248/258 | [(148, 'DLDET')] |" reference_section_type: RESULTS - statement: Over the catalytic region, the full-length horse product is 99.4% identical to human CTDSP2 and groups with CTDSP2 orthologs, well separated from CTDSP1 (78%) and CTDSPL (81%). supporting_text: '| human_CTDSP2_O14595 | 0.994 |' reference_section_type: RESULTS reference_review: relevance: HIGH correctness: VERIFIED review_notes: Re-review analysis (compare_sequences_rereview.py), computed from cached UniProt sequences and the Ensembl REST gene model, with no hand-entered values. The comparators human, mouse and cow CTDSP2 are all PANTHER PTHR12210:SF187 members in the PANTHER 19.0 classification files; horse is not a PANTHER reference proteome, so its placement rests on this tree plus the VGNC orthology assignment. - id: file:human/CTDSP2/CTDSP2-uniprot.txt title: 'UniProtKB O14595: human CTDSP2 catalytic features' findings: - statement: Human CTDSP2 is a small CTD phosphatase that preferentially dephosphorylates Ser-5 of the POLR2A C-terminal heptapeptide repeats and negatively regulates RNA polymerase II transcription. record_section: ["CC:FUNCTION"] supporting_text: "FUNCTION: Preferentially catalyzes the dephosphorylation of 'Ser-5'" reference_section_type: DATABASE_ENTRY - statement: Human CTDSP2 is recruited by REST to RE-1-element neuronal genes, silencing them in non-neuronal cells. record_section: ["CC:FUNCTION"] supporting_text: by REST to neuronal genes that contain RE-1 elements, leading to reference_section_type: DATABASE_ENTRY - statement: The catalytic machinery sits in the FCP1-homology domain spanning human residues 97-255, outside the 84-residue N-terminal segment that F7A4N8 shares with human CTDSP2. record_section: ["FT:DOMAIN"] supporting_text: DOMAIN 97..255 reference_section_type: DATABASE_ENTRY - statement: Catalysis requires the aspartate pair Asp107 (4-aspartylphosphate intermediate) and Asp109 (proton donor), neither of which has a counterpart in F7A4N8. record_section: ["FT:ACT_SITE"] supporting_text: ACT_SITE 107 reference_section_type: DATABASE_ENTRY - statement: Human CTDSP2 is nuclear, though by similarity rather than by direct experiment. record_section: ["CC:SUBCELLULAR LOCATION"] supporting_text: 'SUBCELLULAR LOCATION: Nucleus {ECO:0000250}.' reference_section_type: DATABASE_ENTRY reference_review: relevance: HIGH correctness: VERIFIED review_notes: Curated Swiss-Prot record for the human ortholog (same PANTHER subfamily SF187 as the full-length horse product). It describes the characterized biology and gives the residue coordinates that show this truncated product lacks the catalytic domain. - id: PMID:12721286 title: A novel RNA polymerase II C-terminal domain phosphatase that preferentially dephosphorylates serine 5. full_text_unavailable: true findings: - statement: The small CTD phosphatase (SCP) family, of which CTDSP2 is a member, preferentially dephosphorylates Ser5 within the RNA polymerase II CTD consensus repeat. supporting_text: This study describes a family of small CTD phosphatases (SCPs) that preferentially catalyze the dephosphorylation of Ser5 within the consensus repeat. reference_section_type: ABSTRACT - statement: The SCPs repress transcription through their phosphatase activity. Expressing SCP1 inhibits activated transcription, whereas a phosphatase-inactive mutant enhances it, so the repressor role cannot be transferred to a product that lacks the active site. supporting_text: Expression of SCP1 inhibits activated transcription from a number of promoters, whereas a phosphatase-inactive mutant of SCP1 enhances transcription. reference_section_type: ABSTRACT reference_review: relevance: MEDIUM correctness: VERIFIED review_notes: One of the two primary sources behind the human CTDSP2 FUNCTION annotation. Abstract-only in the local cache, and the abstract foregrounds SCP1; it cannot establish by itself which assays were performed on CTDSP2/SCP2. Cited for family biology, not as equine evidence. - id: PMID:15681389 title: Small CTD phosphatases function in silencing neuronal gene expression. full_text_unavailable: true findings: - statement: REST/NRSF recruits the SCP family (SCP1 to SCP3, including CTDSP2) to RE-1-element neuronal genes to silence them in non-neuronal cells. supporting_text: We show that REST/NRSF recruits SCPs to neuronal genes that contain RE-1 elements, leading to neuronal gene silencing in non-neuronal cells. reference_section_type: ABSTRACT - statement: SCP-mediated neuronal gene silencing is conserved as far as Drosophila, consistent with the full-length horse CTDSP2 product retaining the conserved catalytic domain. supporting_text: Thus, SCP activity is an evolutionarily conserved transcriptional regulator that acts globally to silence neuronal genes. reference_section_type: ABSTRACT reference_review: relevance: MEDIUM correctness: VERIFIED review_notes: The REST/neuronal-silencing source behind the curated human record. Abstract-only in the local cache; it addresses the SCP family (SCPs 1 to 3) rather than CTDSP2 in isolation. Cited for pathway context, not as equine evidence.