Gene 2 of Bacillus phage phi29 encodes p2 (gp2), the 575-residue family-B ("alpha-like") DNA polymerase that carries out the whole of phi29 genome replication by itself. It is the archetype of protein-primed replication: instead of an RNA or DNA primer, it forms a heterodimer with a free molecule of the phage terminal protein p3, recognises the origins at the two 5' ends of the linear 19 kb genome, and covalently attaches the initiating dAMP to the hydroxyl of Ser-232 of the priming terminal protein, a template-directed transfer onto a protein acceptor that the polymerase itself catalyses. The same enzyme then elongates that TP-dAMP product to full genome length, so the polymerase and the terminal protein are the only proteins needed to replicate the genome in vitro. Three activities reside in one chain: DNA-templated DNA polymerisation, terminal-protein deoxynucleotidylation, and a 3'-5' exodeoxyribonuclease that proofreads, housed in an independent N-terminal active site with its own metal site, between which the primer terminus travels without leaving the enzyme. Two structural features make the polymerase autonomous where other replicases need helicases and clamps: a downstream-template tunnel too narrow for duplex DNA, which couples polymerisation to strand displacement, and topological encirclement of both the downstream template and the upstream duplex, which gives processivity of tens of kilobases with no accessory protein. These properties are the basis of its wide use in multiple displacement amplification and in single-molecule nanopore sequencing.
Definition: Catalysis of the transfer of a deoxyribonucleoside monophosphate from a deoxyribonucleoside 5'-triphosphate to a hydroxyl group on an amino acid side chain of a priming protein, forming a phosphoester bond between the protein and the deoxynucleotide and releasing diphosphate. The nucleotide transferred is specified by a DNA template, and the resulting protein-dNMP adduct serves as the primer for DNA strand synthesis.
Justification: This is a catalytic activity of the phi29 DNA polymerase, distinct from its DNA-directed DNA polymerase activity in both acceptor and assay: the acceptor is a serine hydroxyl of the terminal protein rather than the 3'-OH of a nucleic acid, the reaction proceeds in the complete absence of DNA with the same linkage chemistry and the same Ser232 and YCDTD requirements, and it has its own metal preference. It is shared by the protein-primed replicases of phi29 and relatives, of PRD1 and Bam35, of adenovirus, and of linear plasmids, and it is the reaction that makes protein priming possible. GO represents the priming event only on the process side (GO:0006270 DNA replication initiation, GO:0006269 DNA replication, synthesis of primer) and represents protein nucleotidylation only as a biological process (GO:0018175) or as an adenylylation process (GO:0018117), with no molecular function term for a protein-accepting nucleotidyltransferase of this kind. Without such a term the polymerase's second catalytic activity can only be annotated as DNA polymerase activity, which misstates the acceptor, and the companion review of the terminal protein (UniProtKB:P03681) has had to propose a separate protein-primer term for the acceptor side of the same reaction. The two proposals are complementary: this one is the enzyme's activity, that one is the primer's.
Parent term: nucleotidyltransferase activity
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000166 nucleotide binding | IEA GO_REF:0000120 | ACCEPT | Summary: Electronic row from the family-B polymerase signatures and the nucleotide-binding keyword. The claim is that the protein binds nucleotides, which for this enzyme means the dNTP substrate of polymerisation and of terminal-protein deoxynucleotidylation. It is directly documented: UniProt carries curated BINDING features for a nucleotide analogue (5-methyl-UTP) at residues 254, 371, 383 and 458, and mutating Lys383 of motif B destroys dNTP use while leaving DNA affinity and the exonuclease untouched. Reason: Correct, and as specific as GO allows. The precise claim would be binding of deoxyribonucleoside triphosphates, but GO has no such term - GO:0032546 deoxyribonucleoside binding sits in the unphosphorylated nucleoside branch and GO:0032564 dATP binding is specific to the initiating nucleotide only, while elongation uses all four dNTPs. GO:0000166 is therefore kept rather than modified, and the missing term is raised in suggested_questions. The row is a mechanistic component of the polymerase activity rather than a separate core function. Supporting Evidence: PMID:9199402 mutant K383P showed no detectable capacity to incorporate dNTP substrates using either DNA or TP as primer, although its affinity for DNA was not affected file:BPPH2/2/2-uniprot.txt /ligand="5-methyl-UTP" |
| GO:0001882 nucleoside binding | IEA GO_REF:0000002 | REMOVE | Summary: InterPro2GO row from IPR014416 "DNA-directed DNA polymerase, family B, phi29-like virus". Checked against the InterPro API: that entry maps to three GO terms, and one of them is GO:0001882 nucleoside binding. A nucleoside is a base plus sugar with no phosphate, and in GO this is not a parent of nucleotide binding: GO:0000166 descends from GO:1901265 nucleoside phosphate binding, whereas GO:0001882 descends from GO:0097367 carbohydrate derivative binding and all its children are unphosphorylated nucleosides. The mapping therefore asserts something different from, not weaker than, the dNTP binding this enzyme has. Reason: Demonstrably wrong electronic mapping, which is the sanctioned use of REMOVE for an IEA row. Nothing in fifty years of work on this polymerase reports binding of a nucleoside; the substrate at the polymerisation site is a dNTP, the curated BINDING features are for a nucleotide analogue, and the Michaelis constants in the mutational literature are Km values for dNTPs. The claim the mapping was presumably meant to make, nucleotide binding, is already annotated from IPR006172 and from UniProt keyword KW-0547, so removing this row loses nothing. The defect is in the InterPro2GO mapping and propagates to every phi29-like family-B polymerase, so it is filed as a source defect against InterPro rather than treated as a problem with this entry. Supporting Evidence: PMID:9199402 the Km for dNTPs being 200-fold higher than that of the wild-type enzyme |
| GO:0003676 nucleic acid binding | IEA GO_REF:0000002 | MODIFY | Summary: InterPro2GO row at the nucleic-acid level. True, but the nucleic acid this enzyme binds is DNA in every documented case: template and primer duplex at the polymerisation site, and a free 3' single-stranded terminus at the exonuclease site. There is no evidence of RNA binding, and the enzyme actively discriminates against ribonucleotides. Reason: Correct but less specific than the evidence supports, so MODIFY rather than KEEP_AS_NON_CORE. The replacement is GO:0003677 DNA binding, an is_a descendant that this entry already carries from KW-0238, and which UniProt supports with mapped DNA-binding regions in the linker (192-229) and at the C-terminus (563-575). Proposed replacements: DNA binding Supporting Evidence: file:BPPH2/2/2-uniprot.txt Involved in DNA-binding and TP interaction file:BPPH2/2/2-uniprot.txt Interacts with DNA. |
| GO:0003677 DNA binding | IEA GO_REF:0000120 | ACCEPT | Summary: Keyword and signature row for DNA binding. Well supported and curated at residue level: the YxGG/A motif in the linker region (192-229) and the C-terminal residues 563-575 are both mapped as DNA-binding by mutagenesis, and the crystal structure shows the enzyme encircling both the downstream template strand and the upstream duplex product. Reason: Correct and kept at this level deliberately. The enzyme binds double-stranded DNA (the primer-terminus duplex and the upstream product it encircles) and single-stranded DNA (the displaced strand, the downstream template threaded through the narrow tunnel, and the 3' single-stranded terminus at the exonuclease site), so neither GO:0003690 double-stranded DNA binding nor GO:0003697 single-stranded DNA binding would be a refinement of the claim - each would be a partial restatement of it. GO:0003677 is as precise as the evidence allows. It is a component of the catalytic activities rather than a core function in its own right. Supporting Evidence: PMID:15546620 explained by its topological encirclement of both the downstream template and the upstream duplex DNA file:BPPH2/2/2-uniprot.txt The conserved motif YxGG/A located between the 3'-5' exonuclease and polymerization domains is important for DNA-binding |
| GO:0003887 DNA-directed DNA polymerase activity | IEA GO_REF:0000120 | ACCEPT | Summary: The central molecular function, arriving here electronically from the family-B signatures, EC 2.7.7.7 and KW-0239, but experimentally established for this exact protein: purified p2 with purified terminal protein replicates phi29 DNA to full length, and on primed M13 template it synthesises chains of more than 70 kb with no accessory protein, displacing the non-template strand as it goes. UniProt gives EC 2.7.7.7 with experimental evidence from PMID:3863101. Reason: Correct and maximally specific for the template-directed polymerisation activity; this is the canonical representation of the gene's core function, and the rows on the ancestors GO:0034061, GO:0016779 and GO:0016740 are modified to point here. Core. Supporting Evidence: PMID:2498321 the phi 29 DNA polymerase is able to synthesize DNA chains greater than 70 kilobase pairs PMID:3863101 the phi 29 DNA polymerase functions, not only in the formation of the p3-dAMP covalent initiation complex but also in the elongation of the latter, as the only DNA polymerase to produce full-length phi 29 DNA |
| GO:0004518 nuclease activity | IEA GO_REF:0000043 | MODIFY | Summary: Keyword row (KW-0540 Nuclease). The nuclease activity of this protein is a 3'-5' exonuclease that removes mononucleotides from a free 3' terminus of single-stranded DNA, housed in an independent N-terminal domain (residues 1-191) with its own metal site, and serving proofreading. UniProt records EC 3.1.11.- with experimental evidence from PMID:2790959. Reason: True but far less specific than the evidence supports, so MODIFY rather than KEEP_AS_NON_CORE. The replacement GO:0008310 single-stranded DNA 3'-5' DNA exonuclease activity is an is_a descendant of both this term and of GO:0004527, and its definition - sequential cleavage of mononucleotides from a free 3' terminus of a single-stranded DNA molecule - is exactly what was assayed. The directionality and the single-stranded substrate requirement are established by site-directed mutagenesis of the three conserved Exo motifs and by the mutational mapping of the single-stranded DNA ligands. Proposed replacements: single-stranded DNA 3'-5' DNA exonuclease activity Supporting Evidence: PMID:2790959 specifically inactivated the 3'----5' exonuclease activity of the enzyme file:BPPH2/2/2-uniprot.txt The N-terminus contains the 3'-5' exonuclease activity and strand displacement ability |
| GO:0004527 exonuclease activity | IEA GO_REF:0000043 | MODIFY | Summary: Keyword row (KW-0269 Exonuclease), one level below the nuclease row above and resting on the same curated activity: a 3'-5' exonuclease in the N-terminal domain that acts on a single-stranded 3' terminus and proofreads polymerisation errors by an intramolecular route, the primer terminus travelling between the two active sites without dissociating from the DNA. Reason: Same judgement as the GO:0004518 row, kept consistent: correct but general, and the specific descendant GO:0008310 single-stranded DNA 3'-5' DNA exonuclease activity states both the polarity and the substrate strandedness that the experiments establish. Modified rather than accepted because the evidence fixes the term one level finer; this is a core function of the protein, as proofreading is what makes its fidelity useful. Proposed replacements: single-stranded DNA 3'-5' DNA exonuclease activity Supporting Evidence: file:BPPH2/2/2-uniprot.txt DNA polymerase edits the polymerization errors using an intramolecular pathway as the primer terminus travels from one active site to the other without dissociation from the DNA PMID:1310035 The results obtained support the existence of independent metal binding sites for the synthetic and degradative activities of phi 29 DNA polymerase |
| GO:0006260 DNA replication | IEA GO_REF:0000120 | MODIFY | Summary: Keyword and signature row placing the protein in DNA replication. True, and this protein does not merely participate in replication, it is the replicase: it performs the initiation chemistry and the elongation that follow it, and purified p2 plus terminal protein suffice to replicate the genome. Reason: Correct but less specific than the evidence supports. Both sub-steps are separately documented for this protein and both have terms that are descendants of GO:0006260, so both are listed rather than one being picked: GO:0006270 DNA replication initiation, which the entry already carries from the CACAO rows and which the mutant-extract and TP-deoxynucleotidylation experiments establish, and GO:0006271 DNA strand elongation involved in DNA replication, which PMID:3863101 shows explicitly by separating the formation of the p3-dAMP initiation complex from its elongation to full-length genome. Note that GO:0039693 viral DNA genome replication, which this entry also carries, is not a descendant of GO:0006260 (its ancestors run through GO:0019079 viral genome replication and GO:0006259 DNA metabolic process), so it does not by itself make this row redundant. Proposed replacements: DNA replication initiation DNA strand elongation involved in DNA replication Supporting Evidence: PMID:3863101 the phi 29 DNA polymerase functions, not only in the formation of the p3-dAMP covalent initiation complex but also in the elongation of the latter, as the only DNA polymerase to produce full-length phi 29 DNA PMID:2118623 in vivo, the mutations only affect the initial steps of the replication process |
| GO:0016740 transferase activity | IEA GO_REF:0000043 | MODIFY | Summary: Keyword row (KW-0808 Transferase), the top of the transferase branch. The transfer this protein catalyses is nucleotidyl transfer onto a DNA 3'-OH or onto the serine hydroxyl of the terminal protein, both of which are several levels below this term. Reason: True but uninformative at this level, and a specific descendant is available and already annotated, so MODIFY to GO:0003887 DNA-directed DNA polymerase activity (an is_a descendant through GO:0016772 and GO:0016779 that this entry carries from EC 2.7.7.7). The second transfer reaction, deoxynucleotidylation of the terminal protein, has no GO term and is filed under proposed_new_terms, so it cannot be named as a replacement here. Proposed replacements: DNA-directed DNA polymerase activity Supporting Evidence: file:BPPH2/2/2-uniprot.txt Reaction=DNA(n) + a 2'-deoxyribonucleoside 5'-triphosphate = DNA(n+1) + diphosphate |
| GO:0016779 nucleotidyltransferase activity | IEA GO_REF:0000043 | MODIFY | Summary: Keyword row (KW-0548 Nucleotidyltransferase), the immediate parent of the DNA polymerase branch. Correct - both of this enzyme's synthetic activities are nucleotidyl transfers - but the template-directed DNA polymerase term states the main one precisely. Reason: Correct but general, handled consistently with the GO:0016740 and GO:0034061 rows: replace with the is_a descendant GO:0003887 DNA-directed DNA polymerase activity, which this entry already carries with experimental support. Worth noting that this parent is the only existing term that covers the terminal-protein deoxynucleotidylation reaction at all, which is why the proposed new molecular function term for that reaction is placed under it. Proposed replacements: DNA-directed DNA polymerase activity Supporting Evidence: file:BPPH2/2/2-uniprot.txt primer terminal protein (TP) deoxynucleotidylation, which is the formation of a covalent linkage (phosphoester) between the hydroxyl group of a specific serine residue in TP and 5'-dAMP PMID:1730646 the phi 29 DNA polymerase, in the absence of DNA, is able to catalyze the formation of a covalent complex between the phi 29 terminal protein (TP) and 5'-dAMP |
| GO:0016787 hydrolase activity | IEA GO_REF:0000043 | MODIFY | Summary: Keyword row (KW-0378 Hydrolase). Unlike the hydrolase keyword on the sister entry P03681 (terminal protein), which traces to a superseded peptidoglycan-hydrolase attribution, this one is sound: the hydrolytic activity is the 3'-5' exonuclease, and UniProt records EC 3.1.11.- for it with experimental evidence. Reason: Correct but at the top of the hydrolase branch. The specific descendant GO:0008310 single-stranded DNA 3'-5' DNA exonuclease activity is the activity actually assayed, and it is the replacement proposed for the nuclease and exonuclease rows too, so all three hydrolytic rows resolve to one term. Proposed replacements: single-stranded DNA 3'-5' DNA exonuclease activity Supporting Evidence: PMID:2790959 specifically inactivated the 3'----5' exonuclease activity of the enzyme file:BPPH2/2/2-uniprot.txt Exonuclease activity has a proofreading purpose |
| GO:0034061 DNA polymerase activity | IEA GO_REF:0000120 | MODIFY | Summary: Row from an ARBA rule plus the RHEA reaction for DNA polymerisation. The claim is right, but GO:0034061 is the template-agnostic parent, and this enzyme's polymerisation is DNA-template-directed, as the catalytic activity line and the replication of the phi29 genome itself show. Reason: Correct but one level too general: GO:0003887 DNA-directed DNA polymerase activity is an is_a child that the entry already carries. Modified rather than accepted because the template specificity is established, not assumed - the enzyme discriminates against ribonucleotides at a single tyrosine and replicates a DNA genome. The EXP row below carries the same term and is modified identically. Proposed replacements: DNA-directed DNA polymerase activity Supporting Evidence: file:BPPH2/2/2-uniprot.txt Reaction=DNA(n) + a 2'-deoxyribonucleoside 5'-triphosphate = DNA(n+1) + diphosphate |
| GO:0039693 viral DNA genome replication | IEA GO_REF:0000120 | ACCEPT | Summary: Electronic row from the phi29-like polymerase family signature and the viral DNA replication keyword, restating what the IDA row from PMID:3863101 and the CACAO IMP row from PMID:820555 assert from experiment: this protein replicates the phage's own DNA genome. Reason: Correct, maximally specific for the process, and the gene product performs the process itself rather than being required for it - it makes every phosphodiester bond of both daughter strands. Reinforces the experimental rows for the same term rather than duplicating them wrongly. Core. Supporting Evidence: PMID:3863101 A system that replicates bacteriophage phi 29 DNA with protein p3 covalently attached to the two 5' ends, using as the only proteins the phi 29 DNA polymerase and the terminal protein, is described. |
| GO:0046872 metal ion binding | IEA GO_REF:0000043 | MODIFY | Summary: Keyword row (KW-0479 Metal-binding). The metal is identified: UniProt records Mg2+ as the cofactor with four experimental references, and carries six curated Mg2+ BINDING features (145, 169, 249, 250, 456, 458), several annotated catalytic, from crystal structures 1XI1 and 2PYJ and from mutational analysis of the YxDTDS motif. Separate metal sites serve the synthetic and the degradative activities, and protein-primed initiation specifically prefers Mn2+ because of a large increase in dATP affinity. Reason: Correct but less specific than the evidence supports, so MODIFY rather than KEEP_AS_NON_CORE. GO:0000287 magnesium ion binding is an is_a descendant and is supported by two independent lines: the curated cofactor statement with its four experimental references, and six structurally mapped Mg2+ ligand residues. The Mn2+ preference of the initiation reaction is a metal-activation property measured in vitro rather than a second cofactor, so no manganese term is proposed. Proposed replacements: magnesium ion binding Supporting Evidence: file:BPPH2/2/2-uniprot.txt Name=Mg(2+); Xref=ChEBI:CHEBI:18420; PMID:1310035 the initiation of DNA polymerization using a protein as a primer, a special synthetic activity carried out by phi 29 DNA polymerase, exhibited a strong preference for Mn2+ as metal activator |
| GO:0006270 DNA replication initiation | IMP PMID:6402761 Factors involved in the initiation of phage phi 29 DNA repli... | ACCEPT | Summary: CACAO row (ECO:0006099, temperature-sensitive mutant phenotypic evidence, dated 2025-02-17). Blanco et al. 1983 dissected the in vitro requirements for forming the p3-dAMP complex, the first step of phi29 DNA replication, using extracts of Bacillus subtilis infected with phi29 mutants in genes 2, 3, 5, 6 and 17. The paper is titled for this gene and its headline result is about this gene: initiation is completely dependent on a functional gene 2 product, while the products of genes 5, 6 and 17 are dispensable for it. This is the one CACAO row in this entry whose cited paper is primarily about the gene under review. Reason: The term matches the assay exactly: the readout is formation of the initiation complex, which is the process GO:0006270 names, and the mutant-extract design is what makes IMP the right evidence code. There is no phenotype-as-process problem, because what the mutant abolishes is the chemical step itself rather than a downstream outcome. The qualifier is wrong - this protein catalyses the initiation reaction, it does not act upstream of it - and that is recorded as a source defect rather than as a reason to reject the claim; the term, reference and evidence code all stand. The cached record is abstract-only (EuropePMC returns a server error for PMC325798 and the 1983 PMC copy is scanned images), so the figure-level evidence could not be checked, and the GONUTS wiki that held the student's evidence pages is now permanently retired (gowiki.tamu.edu returns HTTP 410 and the Internet Archive has no snapshot of the protein or paper pages). The claim is independently corroborated by the in vitro reconstitution with purified proteins (PMID:3863101), by the DNA-independent TP deoxynucleotidylation reaction (PMID:1730646), and by the ts2 mutants of this gene, whose lesions affect only the initial steps of replication in vivo (PMID:2118623). The same row on the sister entry P03681 was accepted in its review. Core. Supporting Evidence: PMID:6402761 The formation of the initiation complex is completely dependent on the presence of a functional gene 2 product, in addition to protein p3 and phi29 DNA-protein p3 as template. PMID:6402761 The products of genes 5, 6 and 17 do not seem to be needed in the formation of the initiation complex. PMID:2118623 in vivo, the mutations only affect the initial steps of the replication process |
| GO:0039693 viral DNA genome replication | IMP PMID:820555 Bacillus subtilis phage phi29. Characterization of gene prod... | ACCEPT | Summary: CACAO row (ECO:0000315, mutant phenotype evidence, dated 2025-02-10) citing the 1976 Carrascosa et al. survey that mapped phi29 cistrons to their protein products and identified the cistrons whose mutants fail to synthesise phage DNA. The paper is a multi-gene survey: the same PMID is cited on CACAO rows for phi29 genes 1, 3, 6 and 17 (viral DNA genome replication) and genes 4 and 15 (viral release by cytolysis). Gene 2 is one of the DNA-synthesis cistrons. Reason: The claim is correct for this gene and the term is at the right granularity - the replication of the phage genome, not a downstream outcome such as burst size or lysis, which the same paper measures for other genes. The abstract states that mutants in six genes do not synthesise phage DNA but does not name gene 2, and the cached record is abstract-only with GONUTS now permanently retired, so the student's reading cannot be verified directly; the row is accepted under curator deference, and gene 2 is in any case beyond doubt one of those cistrons, since the companion in vitro work used gene 2 mutant extracts and showed the requirement (PMID:6402761) and the purified protein alone replicates the genome (PMID:3863101). The qualifier is wrong for the same reason as the rows above and is filed as a source defect. The identically sourced row on the sister entry P03681 was accepted in its review. Core. Supporting Evidence: PMID:820555 Mutants in six genes do not synthesize phage DNA. PMID:820555 The protein products of 10 out of the 17 cistrons detected in the genome of phage phi29 have been identified. PMID:6402761 extracts from B. subtilis infected with phi29 mutants in genes 2, 3, 5, 6 and 17, involved in DNA synthesis, have been used |
| GO:0006270 DNA replication initiation | IMP PMID:6402761 Factors involved in the initiation of phage phi 29 DNA repli... | ACCEPT | Summary: Second CACAO row for the same term and the same paper, submitted nine days before the one above and differing only in the ECO refinement: ECO:0007161 nonsense mutation phenotypic evidence here against ECO:0006099 temperature-sensitive mutant phenotypic evidence there. Both collapse to IMP in the GAF, so the pair is invisible in the annotation as distributed. Reason: Judged identically to the other GO:0006270 row, as consistency requires: right term, right paper, right evidence code, wrong qualifier (filed as a source defect). The pair is not called a duplicate submission, because phi29 has both suppressor-sensitive (nonsense) and temperature-sensitive mutant collections for this gene - four independently isolated ts2 mutants of gene 2 were cloned and sequenced in PMID:2118623 - so two rows recording two mutant classes from one paper would be legitimate. The abstract does not say which mutants Blanco et al. used, the full text is unavailable, and the GONUTS evidence pages are retired, so the question cannot be settled; the honest reading is that the two ECO refinements cannot both be checked, not that one is known to be wrong. If GO wishes to collapse them, the row to keep is whichever ECO the paper's methods support. Supporting Evidence: PMID:6402761 The formation of the initiation complex is completely dependent on the presence of a functional gene 2 product, in addition to protein p3 and phi29 DNA-protein p3 as template. PMID:2118623 The cloning and complete sequencing of gene 2 from four independently isolated temperature-sensitive mutants in the phage phi 29 DNA polymerase (ts2 mutants) is reported. |
| GO:0034061 DNA polymerase activity | EXP PMID:3863101 Replication of phage phi 29 DNA with purified terminal prote... | MODIFY | Summary: UniProt experimental row, the source of EC 2.7.7.7 in the entry. Blanco and Salas 1985 reconstituted phi29 DNA replication from purified terminal protein and purified polymerase alone, showed that all restriction fragments of the genome were labelled, and showed that the same enzyme makes both the p3-dAMP initiation complex and the full-length product. Reason: The evidence is direct and the activity is beyond doubt, but the term is the template-agnostic parent. The experiment is replication of a DNA genome from a DNA template, so GO:0003887 DNA-directed DNA polymerase activity - an is_a child that this entry already carries, and the term UniProt itself derives EC 2.7.7.7 from - is the correct level. Modified for consistency with the IEA row carrying the same term, not because the evidence is doubted; this paper is the primary support for the core polymerase function. Proposed replacements: DNA-directed DNA polymerase activity Supporting Evidence: PMID:3863101 Restriction analysis of the 32P-labeled DNA synthesized in vitro showed that all phi 29 DNA fragments were labeled. file:BPPH2/2/2-uniprot.txt Polymerase responsible for protein-primed viral DNA replication by strand displacement with high processivity and fidelity |
| GO:0039693 viral DNA genome replication | IDA PMID:3863101 Replication of phage phi 29 DNA with purified terminal prote... | ACCEPT | Summary: UniProt IDA row from the same 1985 reconstitution. This is the strongest evidence in the entry for the process: with p2 and p3 as the only proteins, about half the DNA molecules reached full genome length within a 20-minute chase, and ATP was not required at either step, so no host primase or helicase is involved. Reason: Correct, maximally specific, direct, and on the gene product that performs the process. This is the canonical row for the gene's biological process and the one the IEA and CACAO rows for the same term reinforce. It also carries the correct involved_in qualifier, which is the comparator establishing that the CACAO rows' qualifier is wrong. Core. Supporting Evidence: PMID:3863101 A system that replicates bacteriophage phi 29 DNA with protein p3 covalently attached to the two 5' ends, using as the only proteins the phi 29 DNA polymerase and the terminal protein, is described. PMID:3863101 after a 20-min chase, about half of the DNA molecules had reached apparently full-length phi 29 DNA |
| GO:0045004 DNA replication proofreading | IMP PMID:2790959 A conserved 3'->5' exonuclease active site in prokaryotic an... | NEW | Summary: Proposed new annotation. The 3'-5' exonuclease of this polymerase is a proofreading exonuclease, and GO:0045004 is defined as correction of replication errors by a DNA polymerase using a 3'-5' exonuclease activity - a definition this protein satisfies literally. GOA records the exonuclease activity (as the keyword-derived nuclease, exonuclease and hydrolase rows) but not the process the activity serves, so the proofreading function is currently invisible on the process side. Reason: The gene product performs the work of this process itself: it is the enzyme that excises the mispaired nucleotide, and it does so without releasing the primer, which is why its fidelity is a property of the single chain rather than of a complex. Site-directed mutagenesis of the predicted exonuclease active site specifically abolishes the degradative activity while leaving synthesis intact, which is separation-of-function evidence that the excision is this protein's job and makes IMP the right evidence code. UniProt states the proofreading purpose in its curated FUNCTION line. The comparator check supports the proposal rather than warning against it: GO:0045004 is annotated to the proofreading exonucleases and proofreading polymerases of other replication systems (dnaQ and polC across bacteria in QuickGO), which is the same role this protein occupies, so its absence here is a gap rather than a convention. The term is not an ancestor or descendant of any other term proposed in this review, and the activity side is separately captured by the replacement term GO:0008310. Supporting Evidence: PMID:2790959 specifically inactivated the 3'----5' exonuclease activity of the enzyme file:BPPH2/2/2-uniprot.txt Exonuclease activity has a proofreading purpose file:BPPH2/2/2-uniprot.txt DNA polymerase edits the polymerization errors using an intramolecular pathway as the primer terminus travels from one active site to the other without dissociation from the DNA |
Loading supporting content…
Download this section (compressed HTML)Q: Will InterPro repoint the GO mapping of IPR014416 (DNA-directed DNA polymerase, family B, phi29-like virus) from GO:0001882 nucleoside binding to GO:0000166 nucleotide binding? Nucleoside binding is not an ancestor of nucleotide binding in GO, so the mapping asserts binding of an unphosphorylated nucleoside, which no member of this family has been shown to do, and it propagates to every phi29-like family-B polymerase.
Q: Should GO add a binding term for deoxyribonucleoside triphosphates? For a replicative DNA polymerase the most specific statement available is GO:0000166 nucleotide binding, because GO:0032546 deoxyribonucleoside binding is in the unphosphorylated nucleoside branch and the only dNTP-level terms are single-nucleotide ones such as GO:0032564 dATP binding. The gap is what leaves the generic parent as the correct annotation here.
Q: Does GO want a molecular function term for the deoxynucleotidylation of a priming protein, as proposed here, and should it be coordinated with the protein-primer term proposed in the review of the phi29 terminal protein (UniProtKB:P03681)? The two describe the enzyme and the acceptor sides of one reaction, and both phi29 proteins are currently annotated to terms that misstate their roles in it.
Q: Should the two CACAO rows on GO:0006270 that cite PMID:6402761 and differ only in their ECO refinement (ECO:0007161 nonsense mutation and ECO:0006099 temperature-sensitive mutant, submitted nine days apart) be collapsed to one? Both appear as IMP in the GAF, so the distinction is invisible downstream, and the paper's abstract does not state which mutant class was used.
Q: Now that the GONUTS wiki is permanently retired (gowiki.tamu.edu returns HTTP 410 and the Internet Archive holds no snapshot of its protein or paper pages), is the figure-level evidence behind CACAO annotations archived anywhere? Several thousand GOA rows carry assigned_by CACAO, many of them citing abstract-only papers, and the reasoning that justified each term is no longer retrievable.
Q: Which mutant alleles did Blanco et al. 1983 use for the gene 2, 3, 5, 6 and 17 extracts? The paper is paywalled and its PMC copy is scanned images, and the answer decides both the ECO refinement question above and how strongly the initiation requirement is established for each gene.
Experiment: Build a mutant panel in the polymerisation site and in the terminal-protein interface (the YxGG/A linker, the TPR1 insertion, the YCDTD motif) and assay each protein in parallel for DNA-independent TP-dAMP formation, template-directed TP-dAMP formation and primer extension on a DNA primer-template. A mutant competent for DNA-primed synthesis but null for TP deoxynucleotidylation, together with the converse, would give the separation-of-function evidence that an IDA annotation to the proposed new molecular function term requires. Published single mutants hint at both directions but have not been assayed as one panel against the three readouts.
Hypothesis: The terminal-protein deoxynucleotidylation activity can be separated from DNA-directed polymerisation by mutation in both directions, which would establish it as an independent molecular function rather than a context of the polymerase reaction.
Type: enzyme activity assay
Experiment: Measure steady-state and single-turnover excision rates on a defined substrate series: a 3' single-stranded tail, a fully paired blunt duplex, a single mispaired 3' terminus in duplex, and a recessed 3' end, each with the wild-type enzyme and with a polymerisation-dead mutant to prevent re-extension. The strandedness requirement is inferred from mutational mapping of single-stranded DNA ligands rather than measured against a substrate panel, and it is the premise of the term this review proposes as the replacement for three keyword-derived rows.
Hypothesis: The proofreading exonuclease acts on a single-stranded 3' terminus and not on a paired terminus within duplex DNA, as GO:0008310 asserts.
Type: enzyme activity assay
Experiment: Combine quantitative PCR of origin-proximal and internal genome regions with single-molecule imaging of labelled polymerase in infected Bacillus subtilis at permissive and restrictive temperature for the ts2(24) and ts2(98) alleles, measuring initiation events per genome end and elongation rate separately. This would convert the in vivo claim behind the GO:0006270 annotations from a mutant phenotype to a measured rate, and would test whether the two ts alleles differ in vivo as they do in vitro.
Hypothesis: In vivo, the protein-primed initiation at the two genome ends is the rate-limiting step that the temperature-sensitive gene 2 lesions affect, as the 1990 analysis inferred from phenotype.
Type: imaging
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)