phoA encodes bacterial alkaline phosphatase (BAP), the founding and best-characterised member of the alkaline phosphatase family and one of the most intensively studied enzymes in molecular biology. It is a homodimeric zinc metalloenzyme carrying two Zn(2+) and one Mg(2+) per active site, which hydrolyses phosphate monoesters non-specifically through a covalent phosphoseryl intermediate on Ser102 of the mature protein. Unlike its animal relatives it is not membrane-anchored: it is synthesised with a cleavable signal peptide, exported across the inner membrane by the Sec translocase and folded in the periplasm, where it can be released by lysozyme spheroplasting. Its physiological role is phosphate scavenging under phosphate limitation, liberating inorganic phosphate from organic phosphomonoesters in the periplasm for uptake. The protein is a workhorse of methodology as much as of physiology. Its signal sequence is the canonical model substrate for studying Sec-dependent export, its crystal structure defined the family fold and metal centre, and its use as a reporter and as a dephosphorylation reagent is ubiquitous. BAP also carries a genuinely surprising side activity, oxidising phosphite to phosphate with evolution of molecular hydrogen, though two to three orders of magnitude more slowly than it hydrolyses phosphate esters.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0004035 alkaline phosphatase activity | IBA GO_REF:0000033 | ACCEPT | Summary: The core molecular function, from the unrestricted PAINT node PTN000174527 that carries this term for the whole alkaline phosphatase family. phoA is one of the seventeen seeds behind that node, so its own experimental annotations are part of what placed the assertion. Reason: Correct and ancestral. Reviewing the family (PANTHER:PTHR11596) found the catalytic serine present in all twenty-seven reviewed members with no pseudo-enzyme branch, so this term is safe family-wide and needs no clade restriction. phoA appearing in its own node's WITH/FROM list is expected rather than circular: it has direct experimental evidence for the activity, which is why the curator could use it as a descendant evidence. |
| GO:0042597 periplasmic space | IBA GO_REF:0000033 | ACCEPT | Summary: From PAINT node PTN001698225, which places periplasmic localisation at Enterobacteriaceae and is seeded from phoA itself. The enzyme is exported across the inner membrane by the Sec system and folds in the periplasm, where it acts. Reason: Correct, and the taxon restriction is the right call. The family review assessed this node SOUND specifically because confining it to Enterobacteriaceae rather than to Bacteria avoids projecting a periplasm onto the Gram-positive members of the same PANTHER subfamily, which have no periplasm in the same sense. The is_active_in qualifier is apt: this is where the enzyme works, not merely where it is found. |
| GO:0004035 alkaline phosphatase activity | IEA GO_REF:0000003 | ACCEPT | Summary: Automatic mapping from EC 3.1.3.1, duplicating the IBA and the two IDA rows. Reason: The EC assignment for phoA is about as experimentally grounded as an EC number gets; this enzyme is the reference point for the class. |
| GO:0016791 phosphatase activity | IEA GO_REF:0000002 | MODIFY | Summary: InterPro2GO mapping to the generic parent term, discarding the alkaline pH optimum that the source signature itself names. Reason: The InterPro entries behind this row are the alkaline phosphatase domain and its active site, which license GO:0004035 -- a descendant of GO:0016791 that is directly supported for this protein several times over. The same modification is proposed in all four human isozyme reviews; it is applied consistently across the family. Proposed replacements: alkaline phosphatase activity |
| GO:0042597 periplasmic space | IEA GO_REF:0000044 | ACCEPT | Summary: Automatic transfer from the UniProt subcellular location vocabulary, duplicating the IBA and IDA rows. Reason: Correct; weaker in qualifier and evidence than the rows it duplicates, but not in conflict with any of them. |
| GO:0005515 protein binding | IPI PMID:18022369 Structural basis for signal-sequence recognition by the tran... | REMOVE | Summary: From a structural study of how the SecA translocase motor recognises signal sequences. The interaction is between SecA and the phoA signal peptide, which is the canonical model substrate for Sec-dependent export. Reason: A real, directed interaction rather than a high-throughput artefact, so this is not an incorrect interaction. However, what binds SecA is the cleavable signal peptide of the precursor, not the mature enzyme, and the interaction is about getting phoA out of the cytoplasm rather than about anything phoA does once it is folded. The bare GO:0005515 term also records none of that; remove the generic binding row rather than using it as a molecular-function proxy for Sec-dependent export. Supporting Evidence: PMID:18022369 This is indeed seen with the alkaline phosphatase peptide, whose binding is inhibited by only a factor of ~4. |
| GO:0005515 protein binding | IPI PMID:19766568 Detecting folding intermediates of a protein as it passes th... | REMOVE | Summary: From a study detecting folding intermediates of a protein as it passes through the bacterial translocation channel, again using phoA as the model exported substrate. Reason: Same assessment as the other Sec-translocase rows: a genuine interaction concerning export of the precursor, not a function of the mature periplasmic enzyme. Three separate rows now record essentially the same export-substrate fact, and GO:0005515 adds no information beyond generic binding. |
| GO:0005515 protein binding | IPI PMID:19924216 Signal peptides are allosteric activators of the protein tra... | REMOVE | Summary: From the study showing that signal peptides allosterically activate the protein translocase, which used proPhoA specifically because it is a SecB-independent substrate and requires SecA as its essential receptor. Reason: The best characterised of the three translocase interactions, and still about the precursor's export rather than the enzyme's function. The paper is the clearest statement of why phoA appears in the Sec literature at all, but protein-export biology belongs on the Sec machinery rather than as a generic binding proxy on the PhoA cargo. Supporting Evidence: PMID:19924216 Like proOmpA5, proPhoA requires SecA as an essential receptor. |
| GO:0005515 protein binding | IPI PMID:23937259 Identification of protein phosphatase 1 as a regulator of th... | UNDECIDED | Summary: Cited from a paper identifying protein phosphatase 1 as a regulator of the LRRK2 phosphorylation cycle -- a human neurodegeneration study with no apparent connection to E. coli. Alkaline phosphatase is not mentioned anywhere in the cached record. The most likely explanation is that commercial alkaline phosphatase was used as a dephosphorylation reagent on LRRK2 and an automated interaction import attached it to phoA, but the cached text does not settle this. Reason: Marked undecided rather than removed because the full text is not available to me and I cannot see what the curator saw, which is the standard this project sets for experimental annotations. But the row should not be relied on: the bare protein binding term carries no functional information even if the citation is sound, and the citation cannot be verified from what is cached. Flagged in the reference review so the problem is visible to a curator who can check the full text. |
| GO:0004035 alkaline phosphatase activity | IDA PMID:8634266 Specific recognition of coiled coils by infrared spectroscop... | ACCEPT | Summary: The term is unimpeachable but the citation is almost certainly wrong. The cited paper is an infrared spectroscopy study of coiled-coil recognition in intermediate filament proteins -- desmin, neurofilament L and H -- and the string "alkaline phosphatase" does not occur anywhere in the cached record. Its only contact with phosphatases is that neurofilament tail domains were examined in phosphorylated and dephosphorylated forms, which in that era would routinely have been achieved using alkaline phosphatase as a reagent. Reason: Accepted because the assertion is beyond doubt for this protein and is independently supported by PMID:13826559, by the EC mapping and by the IBA, so nothing rests on this citation. The row is nonetheless a citation problem rather than a functional one, and it is recorded as such in the reference review. If the paper did use alkaline phosphatase as a dephosphorylation reagent, that is using the enzyme as a tool, which is thin ground for an IDA even though it technically demonstrates the activity. |
| GO:0004721 phosphoprotein phosphatase activity | IDA PMID:8634266 Specific recognition of coiled coils by infrared spectroscop... | UNDECIDED | Summary: The same coiled-coil infrared spectroscopy paper, and here it is the sole support for a substantive functional claim: that E. coli alkaline phosphatase acts as a protein phosphatase. That claim is not trivial. BAP is a broad-specificity small-molecule phosphomonoesterase, and treating a phosphoprotein with it in vitro is a standard laboratory manipulation rather than evidence of a physiological protein-phosphatase role. Reason: This is exactly the case the undecided action exists for. The cached record is abstract-only and never mentions alkaline phosphatase, so I cannot verify what was done; the project's standard forbids removing an experimental annotation on that basis. But unlike the alkaline phosphatase activity row above, this term has no independent support for phoA, so accepting it would leave a functional claim resting entirely on a citation I cannot confirm. A curator with the full text should check whether the paper contains a direct assay of phoA against a phosphoprotein or merely used it as a reagent; the two justify very different annotations, and only the first would justify this term. |
| GO:0030288 outer membrane-bounded periplasmic space | IDA PMID:14269305 RELEASE OF ALKALINE PHOSPHATASE FROM CELLS OF ESCHERICHIA CO... | ACCEPT | Summary: From the classic 1964 demonstration that alkaline phosphatase is released from E. coli by lysozyme spheroplasting -- the experiment that established the periplasm as a distinct compartment and phoA as its marker enzyme. Reason: More specific than the plain periplasmic space rows, and correctly so for a Gram-negative organism where the compartment is bounded by an outer membrane. The osmotic-shock and spheroplast release assay is direct compartment evidence rather than inference, and phoA release is still the standard positive control for periplasmic localisation. Supporting Evidence: PMID:14269305 RELEASE OF ALKALINE PHOSPHATASE FROM CELLS OF ESCHERICHIA COLI UPON LYSOZYME SPHEROPLAST FORMATION. |
| GO:0004035 alkaline phosphatase activity | IDA PMID:13826559 A fine-structure genetic and chemical study of the enzyme al... | ACCEPT | Summary: Levinthal's fine-structure genetic and chemical study of alkaline phosphatase, one of the foundational papers of bacterial molecular genetics and a genuine direct characterisation of the enzyme. Reason: The properly grounded IDA for this term, in contrast to the coiled-coil paper cited for the same term above. Combining fine-structure genetics with chemical characterisation of the enzyme, it establishes the activity in the organism where the gene lives. Supporting Evidence: PMID:13826559 A fine-structure genetic and chemical study of the enzyme alkaline phosphatase of E. coli. I. Purification and characterization of alkaline phosphatase. |
| GO:0008270 zinc ion binding | IDA PMID:14487220 Alkaline phosphatase of Escherichia coli: a zinc metalloenzy... | ACCEPT | Summary: From the 1962 paper that identified E. coli alkaline phosphatase as a zinc metalloenzyme -- the work that first established the metal dependence now known to be shared across the whole family. Reason: Core to the catalytic function rather than incidental metal association: the binuclear zinc centre polarises the substrate phosphate and activates the serine nucleophile. The family review records the metal ligands as conserved from phoA to the human isozymes, and this is the paper at the origin of that chain of evidence. Supporting Evidence: PMID:14487220 Alkaline phosphatase of Escherichia coli: a zinc metalloenzyme. |
| GO:0016311 dephosphorylation | IDA PMID:13826559 A fine-structure genetic and chemical study of the enzyme al... | KEEP AS NON CORE | Summary: The root-level biological process corresponding to the enzyme's chemistry, restating the molecular function as a process. Reason: Accurate but nearly contentless -- it would be true of any phosphatase and says nothing about what phoA dephosphorylates or why. The physiologically informative process for this gene is phosphate scavenging under phosphate limitation, which GOA does not record. Kept rather than removed because it is correct and because removing it would leave the gene with almost no process annotation. Supporting Evidence: file:ECOLI/phoA/phoA-deep-research-falcon.md Its precise role is therefore **phosphorus scavenging during limitation**, complementary to uptake systems and pathways for alternative phosphorus compounds such as glycerol phosphates and phosphonates. |
| GO:0042597 periplasmic space | IDA PMID:14269305 RELEASE OF ALKALINE PHOSPHATASE FROM CELLS OF ESCHERICHIA CO... | ACCEPT | Summary: The same lysozyme-spheroplast release experiment, annotated to the general periplasmic space term alongside the more specific outer-membrane-bounded one. Reason: Correct; a parent-and-child pair from one experiment is redundant but not wrong, and the more specific row is the one that carries the information. Supporting Evidence: PMID:14269305 RELEASE OF ALKALINE PHOSPHATASE FROM CELLS OF ESCHERICHIA COLI UPON LYSOZYME SPHEROPLAST FORMATION. |
| GO:0030613 oxidoreductase activity, acting on phosphorus or arsenic in donors | IDA PMID:15148399 A new activity for an old enzyme: Escherichia coli bacterial... | KEEP AS NON CORE | Summary: From the discovery that BAP is responsible for the phn-independent phosphite oxidation pathway in E. coli. Highly purified enzyme, confirmed as BAP by N-terminal sequencing, oxidises phosphite to phosphate with evolution of molecular hydrogen. The term is the right description of the chemistry: phosphite, a phosphorus-containing group, acts as the electron donor. Reason: Genuine and directly demonstrated, so it should not be removed, but it is firmly a side activity rather than a core function. The paper's own numbers make the point: phosphite oxidation runs at 62-242 milliunits per milligram while phosphate ester hydrolysis on the same preparation runs at 41-61 units per milligram, two to three orders of magnitude faster. This is catalytic promiscuity in an enzyme with a permissive active site, notable enough that the authors called it unprecedented, and it does not displace phosphomonoester hydrolysis as what the enzyme is for. Supporting Evidence: PMID:15148399 Highly purified BAP catalyzed Pt oxidation with specific activities of 62-242 milliunits/mg and phosphate ester hydrolysis with specific activities of 41-61 units/mg. |
| GO:0000287 magnesium ion binding | IDA PMID:3910843 Refined structure of alkaline phosphatase from Escherichia c... | ACCEPT | Summary: From the refined 2.8 A crystal structure of E. coli alkaline phosphatase, which resolved the trinuclear metal site and defined the fold that the whole family shares. Reason: Direct structural evidence for the third catalytic metal. Together with the zinc rows this completes the Zn/Zn/Mg centre that the family review records as conserved from phoA to the human isozymes. Supporting Evidence: PMID:3910843 The third site, M3, 5 A from site M2 and 7 A from site M1, is occupied by magnesium or, in the absence of magnesium, by zinc. |
| GO:0008270 zinc ion binding | IDA PMID:3910843 Refined structure of alkaline phosphatase from Escherichia c... | ACCEPT | Summary: The two catalytic zinc sites as resolved in the refined crystal structure, confirming structurally what the 1962 metal-analysis paper established chemically. Reason: Correct and independently supported. Two IDA rows for the same term from a chemical and a structural study is exactly the kind of duplication that should be kept. Supporting Evidence: PMID:3910843 Two metal sites, M1 and M2, 3.9 A apart, are occupied by zinc. |
| GO:0033748 hydrogenase (acceptor) activity | IDA PMID:15148399 A new activity for an old enzyme: Escherichia coli bacterial... | MARK AS OVER ANNOTATED | Summary: From the discovery that BAP oxidises phosphite to phosphate with evolution of molecular hydrogen, which the authors describe as a Pt-dependent, H2-evolving hydrogenase. The term is defined as catalysis of H2 + A = AH2. The observed reaction is phosphite + H2O = phosphate + H2: water is a co-substrate, and the oxidation incorporates an oxygen from it, so the reaction is not a clean reverse of the definition in which a single reduced acceptor AH2 gives up H2. Reason: The activity is real and directly demonstrated, so the row is not wrong, but the term is a loose fit for the chemistry and the companion term GO:0030613, annotated from the same experiment, fits it literally: its definition -- a phosphorus-containing group acting as the hydrogen or electron donor and reducing a hydrogen or electron acceptor -- describes phosphite donating hydride to water-derived protons exactly, and names the donor. Recording the same observation a second time under a hydrogenase term whose stoichiometry it does not match is over-annotation rather than error. It is also a side activity: the same paper measures phosphite oxidation at 62-242 milliunits per milligram against 41-61 units per milligram for phosphate ester hydrolysis. This row has been reassessed more than once during review; the present reading rests on comparing the two terms' definitions against the reaction as written in the paper. Supporting Evidence: PMID:15148399 Surprisingly, BAP catalyzes the oxidation of Pt to phosphate and molecular H2. |
| GO:0016036 cellular response to phosphate starvation | IEP PMID:2241934 From cell membrane to nucleotides: the phosphate regulon in ... | NEW | Summary: Proposed new annotation, and the one that fills the real gap in this gene's record. GOA gives phoA only GO:0016311 dephosphorylation as a biological process, which restates the chemistry and says nothing about why the cell makes the enzyme. phoA is the canonical effector of the Pho regulon: its transcription is repressed by high environmental phosphate and induced under phosphate limitation through the PhoR/PhoB two-component system, and the enzyme then scavenges phosphate from organic phosphomonoesters in the periplasm. Reason: This is the physiological purpose of the enzyme and the reason its expression is regulated at all, so its absence leaves the gene's process annotation essentially empty. The term is well precedented in this organism and for this class of gene: E. coli phoR carries GO:0016036 with IEP evidence, and appA, another secreted phosphatase effector of the same regulon, carries it with IDA. IEP is the correct code here because the evidence is the expression response -- phoA transcription tracking environmental phosphate -- rather than a direct assay of the protein. Note the annotation is about the cell's response that phoA participates in as an effector, not a claim that phoA senses phosphate; sensing is done by PhoR, and the deep research is explicit that phoA is a regulon effector rather than a sensor. Supporting Evidence: PMID:2241934 The repression of phoA (the gene encoding alkaline phosphatase) by high Pi concentrations in the medium requires the presence of an intact Pst operon (pstS, pstC, pstA, pstB and phoU) and phoR. file:ECOLI/phoA/phoA-deep-research-falcon.md PhoA is an **effector enzyme of the Pho phosphate-starvation regulon**. |
| GO:0042803 protein homodimerization activity | IDA PMID:3910843 Refined structure of alkaline phosphatase from Escherichia c... | NEW | Summary: Proposed new annotation. E. coli alkaline phosphatase is an obligate homodimer, and the refined crystal structure states it directly: the enzyme forms an isologous dimer with two reactive centres 32 A apart. UniProt records isozymes 1 and 3 as dimers of identical chains. GOA gives phoA no oligomerisation term. Reason: Proposed for consistency with the family review, which records the dimer as functionally load-bearing across the whole family rather than incidental: in the human placental paralogue the crystal structure attributes mammalian-specific allostery to the interface, in human ALPL monomeric mutants are catalytically dead and act dominant-negatively, and in human ALPI disease alleles act partly by impairing dimerization. phoA is where that quaternary arrangement is ancestral, and here the evidence is direct: the dimer is resolved crystallographically in the cited paper, so IDA is the correct code. Supporting Evidence: PMID:3910843 Alkaline phosphatase-phosphomonoester hydrolase (EC 3.1.3.1) is a metalloenzyme that forms an isologous dimer with two reactive centers 32 A apart. |
Loading supporting content…
Download this section (compressed HTML)Q: Does E. coli alkaline phosphatase have any physiological protein-phosphatase role, or is the GO:0004721 annotation an artefact of the enzyme's use as a laboratory dephosphorylation reagent? The distinction matters beyond this gene, because alkaline phosphatase is used as a reagent in thousands of papers and each is a potential source of the same annotation error.
Q: Is the phosphite-oxidising, hydrogen-evolving side activity of any physiological consequence to E. coli, or purely a laboratory observation on purified enzyme? Its specific activity is two to three orders of magnitude below the phosphatase reaction, but phosphite is a real environmental phosphorus source and the phn pathway exists to use it.
Q: Why has the periplasmic arrangement been retained across Enterobacteriaceae while the eukaryotic lineages relocated the same enzyme to the vacuole membrane and the cell surface? The compartment changed three times in this family while the chemistry did not, which makes it a good system for asking what constrains enzyme localisation independently of catalysis.
Experiment: Retrieve the full text of PMID:8634266 and determine whether alkaline phosphatase appears as an assay subject or as a dephosphorylation reagent for the neurofilament tail domains. If the latter, the two annotations drawn from it should be re-sourced -- GO:0004035 to PMID:13826559, which supports it properly, and GO:0004721 withdrawn. This is a records check rather than a laboratory experiment, and it is the single highest-value action on this gene's annotation.
Hypothesis: The GO:0004721 phosphoprotein phosphatase annotation rests on the use of alkaline phosphatase as a reagent rather than on a directed assay of phoA against a protein substrate.
Type: literature verification
Experiment: Compare growth and hydrogen evolution of phn-deleted E. coli carrying or lacking phoA on phosphite as sole phosphorus source, under phosphate starvation conditions that induce the pho regulon. This would establish whether the activity reported on purified enzyme has any physiological weight, and would settle whether a hydrogen biosynthesis process term is warranted.
Hypothesis: Phosphite oxidation by BAP is measurable in vivo under phosphate limitation and contributes to phosphorus acquisition when phosphite is the sole P source.
Type: bacterial growth and genetics
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)