ALPP encodes placental alkaline phosphatase (PLAP, the Regan isozyme), one of the three tissue-specific human alkaline phosphatases whose genes cluster on chromosome 2q34-q37 and which share 90-98% sequence similarity with each other while being only about 57% identical to tissue-nonspecific ALPL. Like its paralogues it is a GPI-anchored homodimeric metalloenzyme displayed on the outer leaflet of the plasma membrane, with each active site containing two Zn(2+) and one Mg(2+) and catalysis proceeding through a covalent phosphoseryl intermediate on the conserved active-site serine; a further structural Ca(2+) site and a mammalian-specific crown domain sit outside the catalytic centre. PLAP hydrolyses phosphate monoesters at alkaline pH with broad in vitro specificity, and two properties distinguish it biochemically from the other isozymes: exceptional heat stability, which is the classical basis for identifying it in serum, and uncompetitive inhibition by L-amino acids such as L-leucine and L-phenylalanine, attributable to Glu429 near the active site entrance. Expression is confined almost entirely to the syncytiotrophoblast of the placenta, where transcript levels rise several hundred-fold from first trimester to term, and the protein is highly polymorphic in the population. Despite decades of study its physiological substrate and its specific role in placental biology remain unresolved, so the well-defined pyrophosphatase and mineralization functions of ALPL should not be transferred to it. It is clinically prominent as an oncofetal marker: ectopic expression occurs in seminoma and other germ cell tumours, which is why it is under investigation as a solid-tumour immunotherapy target.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0004035 alkaline phosphatase activity | IBA GO_REF:0000033 | ACCEPT | Summary: The core molecular function, placed by PAINT at a deep node of the alkaline phosphatase family. The donor list spans bacterial, fungal and vertebrate members including E. coli PhoA (P00634), all four human isozymes and rodent orthologues, which is the expected shape for an ancient enzymatic activity. Reason: Correct and independently supported for the human protein by direct assay (PMID:1939159) and by the 1.8 A crystal structure of PLAP (PMID:11124260). The IBA is also at the right level of specificity: it names the chemistry without committing to a physiological substrate, which matters here because unlike ALPL, ALPP has no established physiological substrate. Supporting Evidence: PMID:11124260 The crystal structure, determined at 1.8-A resolution, reveals that during evolution, only the overall features of the enzyme have been conserved with respect to Escherichia coli. |
| GO:0005886 plasma membrane | IBA GO_REF:0000033 | ACCEPT | Summary: PLAP is GPI-anchored to the outer leaflet of the plasma membrane, with the catalytic domain facing the extracellular space, so the plasma membrane is genuinely where the enzyme acts and is_active_in is the correct qualifier. UniProt records the GPI anchor at Asp506 with experimental evidence (PMID:2153284). Reason: Ecto-orientation is shared across the whole family and is what makes these enzymes ectophosphatases. The narrower node used for this IBA (PTN000904735, seeded from human and mouse ALPL together with the rat intestinal enzymes Alpi and Akp3) reflects the animal GPI-anchored membrane form specifically, and its seeds span two branches of the family, which is the right placement. |
| GO:0004035 alkaline phosphatase activity | IEA GO_REF:0000003 | ACCEPT | Summary: Automatic mapping from EC 3.1.3.1, duplicating the IBA and the experimental rows. Reason: The EC assignment for ALPP is itself experimentally grounded rather than predicted, so the mapping is sound. |
| GO:0005886 plasma membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Automatic transfer from the UniProt subcellular location vocabulary (SL-0039, cell membrane), which for ALPP rests on experimental demonstration of the GPI anchor. Reason: Correct, though weaker in both qualifier and evidence than the IBA row it duplicates. |
| GO:0016791 phosphatase activity | IEA GO_REF:0000002 | MODIFY | Summary: InterPro2GO mapping from IPR001952 and IPR018299 to the generic parent term, discarding the alkaline pH optimum that the signature itself names. Reason: The InterPro entries that generated this row are specifically the alkaline phosphatase domain and its active site, so they license GO:0004035, a descendant of GO:0016791 that is already directly supported for this protein. Annotating only to the parent adds nothing. Proposed replacements: alkaline phosphatase activity |
| GO:0005515 protein binding | IPI PMID:25416956 A proteome-scale map of the human interactome network. | MARK AS OVER ANNOTATED | Summary: Two hits from the Rolland et al. proteome-scale binary interactome map: KRTAP5-9 (P26371) and KRTAP4-12 (Q9BQ66). Both partners are keratin-associated proteins, a small cysteine-rich family that is among the most frequently recovered and least biologically interpretable classes of partner in yeast two-hybrid screens. Reason: GO:0005515 with no functional interpretation is uninformative by the project's own curation standard, and here there is a specific reason for caution rather than a general one: both recovered partners are KRTAPs, cysteine-rich hair-matrix proteins with no plausible route to encountering a GPI-anchored ectoenzyme of the syncytiotrophoblast. The screen is a legitimate high-throughput resource and the rows are not fabricated, but they should not be read as evidence that PLAP has a protein-binding function. Supporting Evidence: PMID:25416956 A proteome-scale map of the human interactome network. |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | MARK AS OVER ANNOTATED | Summary: Roughly thirty partners from the HuRI reference binary interactome. The list is dominated by keratin-associated proteins (Q5T5A8, Q5T5B0, Q5T752, Q5T753, Q5T754, Q5T871, Q5TA76, Q5TA77, Q5TA81, Q5TCM9, P26371 and others), with the remainder mostly coiled-coil and zinc-finger proteins. Reason: Same assessment as the other interactome row, and more strongly: a partner list of thirty proteins in which one sticky family predominates is the classic signature of assay-driven rather than biology-driven recovery. Nothing in the set suggests a real complex or an adapter role for PLAP, and no follow-up study has validated any of these interactions. If any partner is later confirmed, it should be annotated with a specific term rather than with the uninformative parent. Supporting Evidence: PMID:32296183 A reference map of the human binary protein interactome. |
| GO:0004035 alkaline phosphatase activity | IDA PMID:1939159 Mutation of a single amino acid converts germ cell alkaline ... | ACCEPT | Summary: Direct measurement of alkaline phosphatase activity for PLAP, GCAP, three site-directed GCAP mutants and six GCAP/PLAP chimeras expressed in COS-1 cells. The study established that a single residue at position 429 accounts for the classical differential sensitivity of PLAP and GCAP to L-leucine, EDTA and heat. Reason: Direct assay of the human enzyme, and unusually informative for isozyme annotation: it shows that PLAP and GCAP differ by only seven residues yet are biochemically distinguishable, and pins the difference to Glu429 in PLAP. This is the experimental basis for treating ALPP and ALPG as distinct entities rather than interchangeable ones. Supporting Evidence: PMID:1939159 We report that the differential reactivity of PLAP and GCAP depends critically on a single amino acid at position 429. GCAP with Gly-429 is strongly inhibited by L-leucine, EDTA, and heat, whereas PLAP with Glu-429 is resistant. |
| GO:0000287 magnesium ion binding | ISS GO_REF:0000024 | ACCEPT | Summary: Curator sequence-similarity transfer from rat intestinal AP (P15693) of the catalytic magnesium site. The claim is correct, but the evidence code understates what is available: the 1.8 A crystal structure of human PLAP itself resolves the metal ensemble, and UniProt cites PubMed:11124260 directly for the Mg(2+) coordinating residues at positions 64, 177 and 333. Reason: Magnesium is one of the three catalytic metals in every alkaline phosphatase active site and is required for activity, so this is a genuine molecular function rather than incidental metal association. A curator revisiting this row could upgrade it from ISS to IDA on the strength of the human structure; the term itself needs no change. Supporting Evidence: PMID:11124260 The location of the active site at the bottom of a large valley flanked by an interfacial crown-shaped domain and a domain containing an extra metal ion on the other side suggest that the substrate of PLAP could be a specific phosphorylated protein. |
| GO:0008270 zinc ion binding | ISS GO_REF:0000024 | ACCEPT | Summary: Sequence-similarity transfer of the two catalytic zinc sites. As with magnesium, UniProt annotates the coordinating residues (64, 114, 338, 342, 379, 380, 454) with direct evidence from the human PLAP structure rather than by similarity. Reason: The binuclear zinc centre performs the chemistry, polarising the substrate phosphate and activating the serine nucleophile, so this is core to the catalytic function. The transfer is correct and the underlying human structural evidence is stronger than the ISS code implies. |
| GO:0009986 cell surface | IDA PMID:15907827 Effect of glycosylphosphatidylinositol specific phospholipas... | KEEP AS NON CORE | Summary: Cell-surface localisation from a study of GPI-specific phospholipase D expression in chronic myeloid leukaemia. The cached record is abstract-only and the abstract discusses GPI-PLD, CD55 and CD59 without mentioning alkaline phosphatase; PLAP is a standard GPI-PLD substrate, so it was very plausibly used in the assays reported in the full text, but that cannot be confirmed from what is cached here. Reason: The localisation claim is certainly true of PLAP on independent grounds -- it is a GPI-anchored ectoprotein and its plasma membrane residence is annotated several times over -- so there is no case for removing it, and the curator will have seen the full text. It is kept as non-core because GO:0009986 is a less precise statement than the plasma membrane rows already present, not because the evidence is doubted. See the reference review for the citation caveat. |
| GO:0004035 alkaline phosphatase activity | IDA PMID:2133555 Two alkaline phosphatase genes are expressed during early de... | ACCEPT | Summary: The term is right but the citation is a poor fit for it. The cited paper is a mouse preimplantation-embryo study that used reverse transcription PCR to show which of the mouse alkaline phosphatase genes are transcribed; it measured transcripts, in mouse, and its subject gene is the mouse embryonic AP rather than human ALPP. Reason: Accepted because the assertion itself -- that ALPP enables alkaline phosphatase activity -- is beyond doubt and is directly supported by PMID:1939159, PMID:11124260 and the EC assignment, and because the full text is not available to me and the curator may have drawn on data the abstract does not describe. But the abstract states its organism and method explicitly, so the mismatch is worth recording rather than passing over: a transcript survey in mouse embryos is not a direct assay of human PLAP activity, and this row should not be counted as independent experimental support. Flagged in the reference review accordingly. Supporting Evidence: PMID:2133555 Oligonucleotide primers were constructed from the three sequences and used in the reverse transcription-polymerase chain reaction technique to establish that two of the three AP isozymes are transcribed during preimplantation development. |
| GO:0004035 alkaline phosphatase activity | NAS PMID:3042787 Nucleotide sequence of the human placental alkaline phosphat... | ACCEPT | Summary: Non-traceable author statement from the paper that sequenced the human placental ALP gene, which identifies it as encoding "the classic heat-stable placental alkaline phosphatase" and situates it among the three chromosome-2 ALP genes. Reason: A weak evidence code attached to a correct and, in this case, foundational claim: this is the paper that defined the gene. The activity is independently established by direct assay, so nothing rests on the NAS alone. The paper is also useful for the isozyme-discrimination problem that pervades this family, since it explicitly distinguishes placental ALP-1 from placental ALP-2 (now ALPG) and from intestinal ALP. Supporting Evidence: PMID:3042787 One of these genes (the placental ALP-1) encodes the classic heat-stable placental alkaline phosphatase. |
| GO:0009897 external side of plasma membrane | IC PMID:2153284 Selectivity of the cleavage/attachment site of phosphatidyli... | NEW | Summary: Proposed new annotation. Every location row in GOA says "plasma membrane" or "cell surface" without recording which face of the membrane the enzyme occupies. PLAP was the model protein in which GPI anchoring was worked out: near-saturation mutagenesis showed the C-terminal hydrophobic peptide is removed and the glycolipid attached at Asp-484 of the mature protein (Asp506 in precursor numbering), and that substitutions blocking that processing trap the protein in the endoplasmic reticulum with very low activity. That paper establishes the anchor; it does not itself observe which leaflet the anchored protein faces. Reason: Coded IC rather than IDA because the exoplasmic orientation is inferred by the curator from the demonstrated GPI anchor, not observed in the cited experiment: a GPI anchor is attached in the ER lumen and places the whole mature protein on the exoplasmic leaflet, so a GPI-anchored enzyme at the plasma membrane is on its external face. The inference is secure -- it is the same logic by which serum PLAP exists at all and by which it works as a tumour marker -- but it is an inference. The sibling proposals are direct and keep IDA: ALPI by cell-surface flow cytometry of a GPI-signal truncation (PMID:29567797) and ALPG by PI-PLC release of the enzyme from intact cells (PMID:2162249). GO:0009897 is a descendant of GO:0005886 and conflicts with no existing row. The more specific term for GPI-anchored topology, GO:0031362 anchored component of external side of plasma membrane, is obsolete (GO obsoleted it because it represents protein topology rather than a cellular component), so GO:0009897 is the correct current term. Supporting Evidence: PMID:2153284 The COOH-terminal domain of pre-pro-PLAP provides a signal for processing during which a largely hydrophobic 29-residue COOH-terminal peptide is removed, and the PI-G moiety is added to the newly exposed Asp-484 terminus. PMID:2153284 The bulk of the PLAP protein produced by category II mutants and some produced by category I mutants were sequestered within the cell, apparently in the endoplasmic reticulum (ER). |
| GO:0005509 calcium ion binding | ISS GO_REF:0000024 | NEW | Summary: Proposed new annotation. In addition to the catalytic Zn/Zn/Mg centre, mammalian alkaline phosphatases carry a structural calcium site that has no counterpart in the bacterial enzyme; UniProt annotates four calcium-coordinating residues in ALPP (238, 291, 292, 307) by similarity to ALPL, where the identity of the metal was established by synchrotron X-ray fluorescence. The PLAP structure paper itself describes the enzyme as having "a domain containing an extra metal ion" alongside the crown domain. Reason: GOA gives ALPP zinc and magnesium but not calcium, which leaves the mammalian-specific part of its metal complement unrecorded. The evidence is ISS rather than IDA because the metal's identity was determined in the paralogue rather than in ALPP, which is the honest code here. Note that this site is structural: as UniProt states for ALPL, calcium binding does not influence phosphatase activity, so this is a real but non-catalytic molecular function. Supporting Evidence: PMID:11124260 The location of the active site at the bottom of a large valley flanked by an interfacial crown-shaped domain and a domain containing an extra metal ion on the other side suggest that the substrate of PLAP could be a specific phosphorylated protein. |
| GO:0042803 protein homodimerization activity | IDA PMID:11124260 Crystal structure of alkaline phosphatase from human placent... | NEW | Summary: Proposed new annotation. PLAP is an obligate homodimer; the crystal structure resolves an unusually intimate interface in which a long N-terminal alpha-helix from one monomer wraps around the other and a residue from one subunit reaches into the active site of its partner. UniProt records the homodimer with four independent experimental references. Reason: This is the one binding term in this family worth asserting, because the dimer is a functional rather than an incidental unit: the structure attributes the allosteric behaviour specific to mammalian alkaline phosphatases to the quality of that interface, and in the ALPL paralogue monomeric mutants are catalytically dead. Recording it makes the family's quaternary requirement explicit and gives the two uninformative GO:0005515 rows something concrete to be contrasted with. Supporting Evidence: PMID:11124260 Allostery is probably favored by the quality of the dimer interface, by a long N-terminal alpha-helix from one monomer that embraces the other one, and similarly by the exchange of a residue from one monomer in the active site of the other. |
Loading supporting content…
Download this section (compressed HTML)Q: What does placental alkaline phosphatase actually dephosphorylate in the placenta? The enzyme is one of the most strongly induced transcripts at term and reaches very high abundance on the syncytiotrophoblast surface, yet no physiological substrate has been demonstrated. The crystal structure suggests a phosphorylated protein rather than a small molecule, on the basis of the active site sitting at the bottom of a valley flanked by the crown domain.
Q: Is there a phenotype associated with loss of ALPP? The gene is highly polymorphic and null or near-null alleles are recognised, which would make human genetics the most direct route to a function, but no clear ALPP-deficiency phenotype has been described. Does redundancy with ALPG, or with soluble enzymes, mask it?
Q: What accounts for the almost complete restriction of ALPP to placenta despite roughly 90 percent exonic identity with ALPI and ALPG, and what drives its ectopic reactivation in seminoma and other tumours? Understanding the switching between the three chromosome-2 genes would clarify how much of the isozyme divergence is regulatory rather than catalytic.
Q: Does PLAP have any physiologically meaningful protein partner? The two interactome-derived GO:0005515 rows are dominated by keratin-associated proteins and look assay-driven, leaving the question genuinely open rather than answered.
Experiment: Compare the extracellular phosphoproteome and phosphometabolome of term placental explants or trophoblast organoids with and without a selective ALPP inhibitor, or after ALPP knockout, using a heat step to separate PLAP activity from other alkaline phosphatases. Substrates whose phosphorylation state shifts specifically on ALPP loss would be the first direct candidates, and the design controls for the isozyme-confusion problem that has made most published ALP substrate work uninformative about ALPP specifically.
Hypothesis: PLAP's physiological substrate is a phosphorylated protein or peptide presented at the syncytiotrophoblast surface, rather than a small-molecule phosphomonoester.
Type: comparative phosphoproteomics with genetic and pharmacological perturbation
Experiment: Extend the chimera and point-mutant panel of PMID:1939159 by profiling purified PLAP, GCAP and the position-429 variants against a broad phosphomonoester substrate library under matched conditions. If Glu429 alters substrate preference and not merely L-leucine and heat sensitivity, that would give the first functional rationale for maintaining two nearly identical genes.
Hypothesis: The seven residues that differ between PLAP and GCAP, and Glu429 in particular, tune substrate preference rather than only inhibitor sensitivity.
Type: comparative enzyme kinetics on engineered variants
Experiment: Test the reported KRTAP partners by co-immunoprecipitation and proximity labelling from term placental tissue and trophoblast cell lines, where both proteins would have to be co-expressed and topologically accessible for the interaction to be real. A negative result would justify demoting the two GO:0005515 rows outright rather than merely flagging them as over-annotated.
Hypothesis: The keratin-associated protein interactions reported for ALPP in binary interactome screens are assay artefacts and do not occur in trophoblast.
Type: interaction validation in a physiological cell type
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)