Existing Annotations Review

GO Term	Evidence	Action	Reason
GO:0004615 phosphomannomutase activity	IBA GO_REF:0000033	ACCEPT	Summary: Phylogenetically inferred molecular function that exactly matches the experimentally established catalytic activity of PMM2 (EC 5.4.2.8, Man6P <-> Man1P). This is the core molecular function of the gene product. Reason: Core, well-supported molecular function; concordant with experimental (EXP/TAS) annotations and UniProt catalytic-activity curation. Supporting Evidence: file:human/PMM2/PMM2-uniprot.txt Reaction=alpha-D-mannose 1-phosphate = D-mannose 6-phosphate
GO:0006013 mannose metabolic process	IBA GO_REF:0000033	ACCEPT	Summary: Correct, parent-level biological-process annotation. PMM2 acts on mannose phosphosugars, so participation in mannose metabolism is accurate, though the more informative process term is GDP-mannose biosynthetic process. Reason: Accurate process annotation directly describing the metabolite class the enzyme acts on; retained as core (parent of the GDP-mannose biosynthetic term).
GO:0005829 cytosol	IBA GO_REF:0000033	ACCEPT	Summary: PMM2 is a soluble cytosolic enzyme; the Man6P<->Man1P reaction occurs in the cytosol. Consistent with UniProt (Cytoplasm), HPA IDA, and Reactome. Reason: Core localization, supported by direct (IDA) and multiple inferred annotations. Supporting Evidence: file:human/PMM2/PMM2-uniprot.txt SUBCELLULAR LOCATION: Cytoplasm
GO:0006487 protein N-linked glycosylation	IBA GO_REF:0000033	KEEP AS NON CORE	Summary: PMM2 only supplies GDP-mannose, a nucleotide-sugar precursor used many steps upstream of N-glycosylation; it is not a glycosyltransferase and does not act on protein substrates. The 'involved_in' relationship is defensible because biallelic PMM2 loss causes protein hypoglycosylation (the basis of PMM2-CDG), but this is a downstream consequence (substrate guilt-by-association), not the enzyme's molecular role. Keep as a non-core process annotation. Independently corroborated in a yeast SEC53/PMM2 disease model, where PMM2 mutations are described as affecting protein N-linked glycosylation (PMID:36214454). Reason: Downstream pathway annotation reflecting precursor supply, not the direct catalytic function; retained because loss-of-function disrupts N-glycosylation, but it is not a core function of this metabolic enzyme. Supporting Evidence: PMID:36214454 mutations in the phosphomannomutase gene PMM2, which affect protein N-linked
GO:0004615 phosphomannomutase activity	IEA GO_REF:0000120	ACCEPT	Summary: Electronic assertion of the core catalytic activity, consistent with the EC 5.4.2.8 / Rhea mapping and InterPro PMM family signature. Redundant with the experimental and IBA annotations of the same term. Reason: Correct core molecular function via independent electronic methods.
GO:0005737 cytoplasm	IEA GO_REF:0000120	MODIFY	Summary: Correct but less specific parent of cytosol. The cytosol (GO:0005829) annotation is the more informative and directly supported localization. Reason: Generalizes the supported cytosol localization; replace with the more specific term. Proposed replacements: cytosol
GO:0006013 mannose metabolic process	IEA GO_REF:0000117	ACCEPT	Summary: Electronic duplicate of the IBA mannose metabolic process annotation; correct. Reason: Accurate process annotation via ARBA model; concordant with the IBA annotation.
GO:0009298 GDP-mannose biosynthetic process	IEA GO_REF:0000120	ACCEPT	Summary: Core biological process: the PMM step (Man6P->Man1P) is step 2/2 in the UniPathway route from fructose 6-phosphate to Man1P feeding GDP-mannose synthesis. Concordant with experimental, TAS, and IMP annotations. Functionally corroborated by metabolic tracer work showing PMM2 deficiency depletes GDP-mannose (PMID:37257447) and by structural work noting that Man-1-P formation is a pivotal step in GDP-mannose and dolichol-phosphate-mannose biosynthesis (PMID:40572562). Reason: Core process directly describing the metabolic pathway the enzyme commits to. Supporting Evidence: file:human/PMM2/PMM2-uniprot.txt GDP-alpha-D-mannose PMID:37257447 caused by PMM2 deficiency, presents with depleted GDP-mannose and abnormal PMID:40572562 The formation of Man-1-P is a pivotal step in the biosynthesis of GDP-mannose
GO:0005515 protein binding	IPI PMID:25416956 A proteome-scale map of the human interactome network.	MARK AS OVER ANNOTATED	Summary: Generic 'protein binding' from a proteome-scale binary (Y2H) interactome map. The interaction was experimentally detected (IPI), so it is real, but the generic GO:0005515 term carries no specific functional meaning for a cytosolic metabolic enzyme and does not identify an adapter/scaffold function. This is an uninformative over-annotation, not an incorrect one. Reason: Experimentally detected but uninformative high-throughput 'protein binding' annotation; the interaction is real but the generic term provides no functional information about PMM2.
GO:0005515 protein binding	IPI PMID:26871637 Widespread Expansion of Protein Interaction Capabilities by ...	MARK AS OVER ANNOTATED	Summary: Generic 'protein binding' from a high-throughput alternative-splicing interactome study. The interaction was experimentally detected (IPI) and is real, but as above the generic term is uninformative for this metabolic enzyme and not a core function. Reason: Experimentally detected but uninformative high-throughput 'protein binding' annotation; real interaction but no specific function established.
GO:0005515 protein binding	IPI PMID:32296183 A reference map of the human binary protein interactome.	MARK AS OVER ANNOTATED	Summary: Generic 'protein binding' from the HuRI reference binary interactome. The recorded partners (e.g. ACY3, MEOX2, SGK2 isoform) were experimentally detected (IPI) and are real interactions, but they are not connected to PMM2's catalytic role and the generic term provides no functional insight. Uninformative over-annotation rather than incorrect. Reason: Experimentally detected but uninformative high-throughput 'protein binding' annotation; real interactions but no specific function established for a cytosolic phosphomannomutase.
GO:0005829 cytosol	IEA GO_REF:0000120	ACCEPT	Summary: Electronic duplicate of the well-supported cytosol localization. Correct. Reason: Core localization confirmed by multiple independent lines of evidence.
GO:0043025 neuronal cell body	IEA GO_REF:0000107	MARK AS OVER ANNOTATED	Summary: Single Ensembl-orthology electronic annotation transferred from mouse. PMM2 is a ubiquitously expressed soluble cytosolic housekeeping enzyme (HPA: low tissue specificity); detection in a neuronal cell body reflects where the cytosol of that cell is, not a neuron-specific function or a distinct localization. Over-annotation for a soluble metabolic enzyme. Reason: Orthology-transferred, cell-type-specific component annotation that adds no functional information beyond the general cytosolic localization and risks implying a neuron-restricted role.
GO:0009298 GDP-mannose biosynthetic process	TAS Reactome:R-HSA-446205	ACCEPT	Summary: Reactome-curated involvement in GDP-mannose synthesis, matching the enzyme's committed pathway step. Core process annotation. Reason: Author-curated (TAS) support for the core GDP-mannose biosynthetic process role. Supporting Evidence: Reactome:R-HSA-446205 GDP-mannose is the mannose donor for the first 5 mannose addition reactions
GO:0004615 phosphomannomutase activity	EXP PMID:16540464 The X-ray crystal structures of human alpha-phosphomannomuta...	ACCEPT	Summary: Experimental support for phosphomannomutase activity. Although the abstract foregrounds the PMM1 crystal structures, the study explicitly compares both isozymes and shows alpha-PMM1 and alpha-PMM2 have a conserved active site and similar kinetic properties; UniProt assigns the EC 5.4.2.8 catalytic-activity ECO:0000269 evidence to this paper for PMM2. This is the primary experimental anchor of the core molecular function. Reason: Direct experimental evidence for the core catalytic activity; the defining function of the gene product. Supporting Evidence: PMID:16540464 are shown to have a conserved active-site structure and to display similar kinetic properties
GO:0004615 phosphomannomutase activity	TAS Reactome:R-HSA-3781926	ACCEPT	Summary: Reactome-curated phosphomannomutase activity (Man6P->Man1P) for PMM2. Concordant with the experimental and IBA/IEA annotations of the same term. Reason: Author-curated support for the core molecular function. Supporting Evidence: Reactome:R-HSA-3781926 PMM2) catalyses the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P)
GO:0005829 cytosol	IDA GO_REF:0000052	ACCEPT	Summary: Direct immunofluorescence localization (HPA) to the cytosol. Strongest evidence for the core cytosolic localization. Reason: Direct experimental localization supporting the core cytosolic component annotation.
GO:0009298 GDP-mannose biosynthetic process	IMP PMID:9525984 Carbohydrate-deficient glycoprotein syndrome type Ib. Phosph...	ACCEPT	Summary: The GO term (GDP-mannose biosynthetic process) is biologically correct for PMM2 and is independently and strongly supported (EXP PMID:16540464; TAS PMID:9140401 and Reactome; IEA pathway). However, the cited reference PMID:9525984 is about phosphomannose ISOMERASE (PMI/MPI) deficiency causing CDG type Ib and mannose therapy - a different enzyme (F6P<->Man6P) and a different gene - and does not study PMM2's mutase step. The annotation is therefore kept (the term is right and is a core process) but the citation appears mis-attributed; this is recorded in the reference review rather than removing an experimental-coded annotation. Reason: Core process annotation; term is correct and well supported by other evidence. The specific PMID:9525984 citation is flagged as likely mis-attributed (PMI/CDG-Ib paper) in reference_review, but the GO term itself is retained.
GO:0005829 cytosol	TAS Reactome:R-HSA-3781926	ACCEPT	Summary: Reactome-curated cytosolic localization. Concordant with IDA/IBA/IEA evidence. Reason: Author-curated support for the core cytosolic localization.
GO:0005829 cytosol	TAS Reactome:R-HSA-446201	ACCEPT	Summary: Reactome-curated cytosolic localization (PMM1,2 isomerise Man6P to Man1P). Concordant with all other localization evidence. Reason: Author-curated support for the core cytosolic localization. Supporting Evidence: Reactome:R-HSA-446201 Cytosolic phosphomannomutases 1 and 2 (PMM1 and PMM2) catalyse the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P)
GO:0004615 phosphomannomutase activity	TAS PMID:9140401 Mutations in PMM2, a phosphomannomutase gene on chromosome 1...	ACCEPT	Summary: The original cloning paper identifies PMM2 as a phosphomannomutase whose deficiency underlies CDG1. Author-stated (TAS) support for the core catalytic activity. Reason: Foundational TAS support for the core molecular function. Supporting Evidence: PMID:9140401 phosphomannomutase (PMM)8, an enzyme necessary for
GO:0009101 glycoprotein biosynthetic process	TAS PMID:9140401 Mutations in PMM2, a phosphomannomutase gene on chromosome 1...	KEEP AS NON CORE	Summary: Broad downstream process. PMM2 contributes to glycoprotein biosynthesis only by supplying GDP-mannose; it is not itself a glycosyltransferase. This is the most general of the downstream-glycosylation annotations and reflects the disease phenotype (CDG) rather than the enzyme's molecular role. Keep as non-core. Reason: General downstream-pathway annotation by precursor supply; true at the organismal/disease level but not a core function of this metabolic enzyme.
GO:0009298 GDP-mannose biosynthetic process	TAS PMID:9140401 Mutations in PMM2, a phosphomannomutase gene on chromosome 1...	ACCEPT	Summary: Author-stated (TAS) support for the core GDP-mannose biosynthetic process role, from the gene's foundational characterization paper. Reason: Core process annotation supported by the original characterization of PMM2. Supporting Evidence: PMID:9140401 phosphomannomutase (PMM)8, an enzyme necessary for

Core Functions

Cytosolic, Mg2+-dependent phosphomannomutase that reversibly isomerizes D-mannose 6-phosphate to alpha-D-mannose 1-phosphate, the committed second step supplying mannose 1-phosphate for GDP-mannose (and dolichol-phosphate-mannose) biosynthesis. This nucleotide-sugar precursor supply is the prerequisite for protein N-linked glycosylation, but the enzyme itself acts only on phosphosugar metabolites. The active enzyme is an obligate homodimer; catalytic-site variants (e.g. p.Arg141His) impair activity while dimer-interface variants (e.g. p.Phe119Leu) cause loss of function by disrupting homodimer formation.

Molecular Function:

phosphomannomutase activity

Directly Involved In:

GDP-mannose biosynthetic process mannose metabolic process

Cellular Locations:

cytosol

Substrates:

D-mannopyranose 6-phosphate(2-) alpha-D-mannose 1-phosphate(2-)

Supporting Evidence:

PMID:16540464
are shown to have a conserved active-site structure and to display similar kinetic properties
file:human/PMM2/PMM2-uniprot.txt
Reaction=alpha-D-mannose 1-phosphate = D-mannose 6-phosphate
Reactome:R-HSA-446205
GDP-mannose is the mannose donor for the first 5 mannose addition reactions
PMID:40572562
Structurally, PMM2 functions as an obligate homodimer

References

GO_REF:0000033

Annotation inferences using phylogenetic trees

GO_REF:0000052

Gene Ontology annotation based on curation of immunofluorescence data

GO_REF:0000107

Automatic transfer of experimentally verified manual GO annotation data to orthologs using Ensembl Compara

GO_REF:0000117

Electronic Gene Ontology annotations created by ARBA machine learning models

GO_REF:0000120

Combined Automated Annotation using Multiple IEA Methods

PMID:16540464

The X-ray crystal structures of human alpha-phosphomannomutase 1 reveal the structural basis of congenital disorder of glycosylation type 1a.

The two isozymes alpha-PMM1 and alpha-PMM2 share a conserved active-site structure and similar kinetic properties; analysis of mutation sites explains the genotype-phenotype relationship of CDG-1a.

"The two isozymes, alpha-PMM1 and alpha-PMM2, are shown to have a conserved active-site structure and to display similar kinetic properties."

PMID:25416956

A proteome-scale map of the human interactome network.

PMID:26871637

Widespread Expansion of Protein Interaction Capabilities by Alternative Splicing.

PMID:32296183

A reference map of the human binary protein interactome.

PMID:36214454

Evolutionary rescue of phosphomannomutase deficiency in yeast models of human disease.

The most common cause of human CDG are mutations in PMM2 that impair protein N-linked glycosylation; the yeast SEC53 gene is a homolog of human PMM2, and disease-associated alleles (including F126L, equivalent to human p.Phe119Leu) cause a slow-growth/glycosylation phenotype that can be evolutionarily rescued.

"mutations in the phosphomannomutase gene PMM2, which affect protein N-linked"

PMID:40572562

In Silico Analysis of Phosphomannomutase-2 Dimer Interface Stability and Heterodimerization with Phosphomannomutase-1.

PMM2 functions as an obligate homodimer; the catalytic p.Arg141His variant impairs activity whereas the interface variant p.Phe119Leu disrupts homodimer formation, leading to destabilization/degradation. Man-1-P formation is a pivotal step in GDP-mannose and dolichol-phosphate-mannose biosynthesis. PMM1/PMM2 heterodimers are predicted to be structurally plausible but PMM1 does not compensate for PMM2 deficiency in vivo.

"Structurally, PMM2 functions as an obligate homodimer"

PMID:37257447

Tracer metabolomics reveals the role of aldose reductase in glycosylation.

PMM2-CDG (caused by PMM2 deficiency) presents with depleted GDP-mannose and abnormal glycosylation; PMM2 itself is not directly involved in polyol metabolism. Aldose-reductase inhibition (epalrestat) diverts glucose flux toward sugar nucleotide synthesis, increasing GDP-mannose and improving glycosylation.

"Considering that the PMM2 enzyme is not"

PMID:9140401

Mutations in PMM2, a phosphomannomutase gene on chromosome 16p13, in carbohydrate-deficient glycoprotein type I syndrome (Jaeken syndrome).

PMM2 on 16p13 encodes a phosphomannomutase necessary for GDP-mannose synthesis; missense mutations cause CDG type I (Jaeken syndrome), establishing PMM deficiency as the basis of the disease.

"phosphomannomutase (PMM)8, an enzyme necessary for"

PMID:9525984

Carbohydrate-deficient glycoprotein syndrome type Ib. Phosphomannose isomerase deficiency and mannose therapy.

Describes phosphomannose ISOMERASE (PMI/MPI) deficiency causing CDG type Ib and mannose therapy - a different enzyme and gene than PMM2.

"Phosphomannose isomerase (PMI) deficiency is the cause of a new type of carbohydrate-deficient glycoprotein syndrome"

Reactome:R-HSA-3781926

Defective PMM2 does not isomerise Man6P to Man1P

PMM2 catalyses the cytosolic isomerization of Man6P to Man1P; Man1P is the precursor of GDP-mannose and dolichol-phosphate-mannose required for N-glycosylation. PMM2 mutations cause PMM2-CDG (CDG-1a).

"catalyses the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P) in the cytosol of cells"

Reactome:R-HSA-446201

PMM1,2 isomerise Man6P to Man1P

Cytosolic PMM1 and PMM2 catalyze the Man6P->Man1P isomerization; PMM2 mutations cause Jaeken syndrome (CDG-Ia).

"Cytosolic phosphomannomutases 1 and 2 (PMM1 and PMM2) catalyse the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P)"

Reactome:R-HSA-446205

Synthesis of GDP-mannose

GDP-mannose, the mannose donor for early N-glycan and Dol-P-Man synthesis, is made from fructose 6-phosphate and GTP in three steps (PMM2 catalyzes the mutase step).

"GDP-mannose is the mannose donor for the first 5 mannose addition reactions"

Suggested Experiments

Experiment: Perform affinity-purification mass spectrometry (AP-MS) on endogenously tagged PMM2 in human cells under native conditions and compare to the binary Y2H hits; test whether candidate partners (ACY3, MEOX2, SGK2) co-purify and whether their knockdown alters PMM2 activity or GDP-mannose levels.

Hypothesis: PMM2's only biologically relevant function is to supply Man1P/GDP-mannose; its reported binary protein interactions are not functionally meaningful.

Type: AP-MS / interaction validation

Experiment: Reconstitute patient PMM2 missense variants in PMM2-null cells and measure enzyme kinetics, steady-state GDP-mannose and Dol-P-Man pools, and lipid-linked oligosaccharide profiles; correlate with clinical severity to test the flux-limitation model.

Hypothesis: CDG-Ia phenotype severity tracks quantitatively with residual cytosolic GDP-mannose flux through the PMM2 step rather than with any non-catalytic property.

Type: variant reconstitution / metabolic flux analysis

Deep Research

Falcon

(PMM2-deep-research-falcon.md)

Gene Research for GO Annotation Review Falcon

Gene Research for GO Annotation Review

Target

Gene symbol: PMM2
Organism: Homo sapiens

UniProt Context

=== UNIPROT METADATA ===
UniProt ID: O15305
Entry Name: PMM2_HUMAN
Gene Name: PMM2
Protein Name: Phosphomannomutase 2
EC Number: 5.4.2.8
Organism: Homo sapiens (Human)
NCBI Taxonomy ID: 9606
Function: Involved in the synthesis of the GDP-mannose and dolichol- phosphate-mannose required for a number of critical mannosyl transfer reactions.
Subcellular Location: Cytoplasm.
Subunit: Homodimer.
======================

Research Objective

Prepare a comprehensive literature-backed research report for GO annotation review of the target gene.

Focus on:

The core biochemical activity of the gene product and any substrate specificity.
The activation or maturation mechanism of the protein, including proteolytic processing if relevant.
The best-supported subcellular localization and complexes where the active protein functions.
The strongest experimentally supported biological processes in the target organism, or in very close orthologs when transfer is justified.
Which annotations appear to reflect core function versus context-specific, pleiotropic, or over-extended roles.
Any evidence bearing on apoptosis, developmental cell death, neuronal roles, inflammatory signaling, pyroptosis, synaptic remodeling, protein processing, and localization to cytoplasm, cytosol, nucleus, or signaling complexes.

Output Requirements

Structure the report as:

Executive summary
Molecular function
Biological process roles
Cellular localization and complexes
Annotation-risk assessment
Key literature

Use citations for all substantive claims.

PMM2 Gene Research Report for GO Annotation Review

Executive Summary

Human PMM2 (phosphomannomutase 2, UniProt O15305) encodes a cytoplasmic homodimeric enzyme (EC 5.4.2.8) that catalyzes the reversible interconversion of mannose-6-phosphate and mannose-1-phosphate, a critical step in GDP-mannose biosynthesis for protein N-glycosylation (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, garapati2024acomplementc4–derived pages 1-2). The most strongly supported molecular function and biological process annotations center on phosphomannomutase activity, GDP-mannose precursor supply, and support of N-linked glycosylation (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mele2025insilicoanalysis pages 1-2, garapati2024acomplementc4–derived pages 1-2). Cellular component evidence strongly supports cytoplasmic/cytosolic localization and obligate homodimer formation (mele2025insilicoanalysis pages 1-2, mangione2025targetedmetabolomicevaluation pages 1-2, edmondson2025novelmousemodel pages 4-7). PMM2 deficiency (PMM2-CDG) causes severe neurodevelopmental phenotypes, particularly cerebellar hypoplasia and ataxia, highlighting heightened dependency of developing cerebellum on intact N-glycosylation (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7). Secondary effects of PMM2 deficiency include ER stress, mitochondrial dysfunction, altered autophagy, and inflammatory signaling abnormalities, but these represent pleiotropic downstream consequences rather than direct PMM2 functions (ligezka2023interplayofimpaired pages 1-2, pascoal2025immunopathologyinpmm2cdg pages 1-3). No convincing evidence supports direct PMM2 roles in apoptosis execution, pyroptosis, or canonical developmental cell death programs as distinct from general neurodevelopmental pathology (ligezka2023interplayofimpaired pages 1-2, pradeep2023glycosylationandbehavioral pages 1-2).

1. Molecular Function

1.1 Core Biochemical Activity and Substrate Specificity

PMM2 catalyzes the reversible conversion of mannose-6-phosphate (Man-6-P) to mannose-1-phosphate (Man-1-P), classified as EC 5.4.2.8 (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2). This phosphomannomutase activity represents the core, best-supported molecular function. Multiple independent studies confirm Man-6-P as the canonical substrate and Man-1-P as the product (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, ligezka2023interplayofimpaired pages 1-2, edmondson2025novelmousemodel pages 4-7). The reaction proceeds through a "ping-pong" catalytic mechanism typical of phosphosugar mutases, involving a phosphoenzyme/phosphohistidine intermediate (mele2025insilicoanalysis pages 1-2).

The physiological significance of this activity is production of Man-1-P as the immediate precursor for GDP-mannose synthesis (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, garapati2024acomplementc4–derived pages 1-2). GDP-mannose serves as an essential sugar donor for N-linked glycosylation in the endoplasmic reticulum, where it is used in formation and elongation of lipid-linked oligosaccharides and dolichol-phosphate-mannose (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mele2025insilicoanalysis pages 1-2, ligezka2023interplayofimpaired pages 1-2). PMM2 deficiency depletes both Man-1-P and GDP-mannose pools, causing broad protein hypoglycosylation (mangione2025targetedmetabolomicevaluation pages 1-2, radenkovic2023tracermetabolomicsreveals pages 1-4, garapati2024acomplementc4–derived pages 1-2).

Biochemical measurements in mouse brain and patient-derived cells confirm cytosolic PMM enzyme activity as the relevant functional pool (edmondson2025novelmousemodel pages 4-7). While the reaction is chemically reversible, the physiological direction emphasizes generation of Man-1-P for downstream glycosylation precursor synthesis (radenkovic2023tracermetabolomicsreveals pages 1-4, ligezka2023interplayofimpaired pages 1-2, edmondson2025novelmousemodel pages 4-7).

Feature	PMM2 finding	Evidence/notes	Citation
Gene/protein	PMM2 encodes phosphomannomutase 2, a phosphosugar mutase central to glycosylation precursor metabolism	Human PMM2 is repeatedly described as the enzyme defective in PMM2-CDG and necessary for N-linked protein glycosylation precursor supply	(edmondson2025novelmousemodel pages 1-4, garapati2024acomplementc4–derived pages 1-2)
Enzyme classification	EC 5.4.2.8	PMM2 is identified as phosphomannomutase, catalyzing intramolecular transfer of phosphate between C6 and C1 positions of mannose phosphate	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2)
Core enzymatic activity	Reversible interconversion of mannose-6-phosphate (Man-6-P) and mannose-1-phosphate (Man-1-P)	Multiple sources state that PMM2 catalyzes Man-6-P ↔ Man-1-P conversion; this is the core, best-supported molecular function	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, edmondson2025novelmousemodel pages 4-7)
Canonical substrate	Mannose-6-phosphate	PMM2 activity assays and disease models are framed around conversion of Man-6-P to Man-1-P; deficiency causes Man-1-P depletion	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, ligezka2023interplayofimpaired pages 1-2, edmondson2025novelmousemodel pages 4-7)
Canonical product	Mannose-1-phosphate	Man-1-P is explicitly described as the immediate product of PMM2 catalysis and as depleted in PMM2 deficiency	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, ligezka2023interplayofimpaired pages 1-2, garapati2024acomplementc4–derived pages 1-2)
Biochemical reaction significance	Produces the precursor for GDP-mannose synthesis	Man-1-P is converted onward to GDP-mannose, linking PMM2 directly to glycan donor production	(vignogna2022evolutionaryrescueof pages 1-2, budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2)
Pathway role	Required upstream of GDP-mannose and dolichol-phosphate-mannose biosynthesis	Recent structural/review sources note that formation of Man-1-P is pivotal for GDP-mannose and dolichol-phosphate-mannose, both sugar donors used in lipid-linked oligosaccharide formation	(mele2025insilicoanalysis pages 1-2, ligezka2023interplayofimpaired pages 1-2)
Glycosylation relevance	Supports N-linked glycosylation by supplying mannose donor pools	PMM2 deficiency lowers GDP-mannose availability and causes abnormal N-glycosylation/hypoglycosylation across models and patient samples	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, radenkovic2023tracermetabolomicsreveals pages 1-4, garapati2024acomplementc4–derived pages 1-2)
Quaternary structure	Obligate homodimer	Yeast/human disease literature and structural analysis describe PMM2 as functioning as a homodimer, with dimer formation integral to native structure	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, tran2020secondgenerationpharmacologicalchaperones pages 1-3)
Structural requirement for activity	Dimerization is important, and disruption of the dimer interface contributes to loss of function	Interface variants such as p.Phe119Leu are associated with dimerization defects; recent structural modeling emphasizes interface stability as a determinant of residual activity	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2)
Disease-illustrating alleles	p.Arg141His affects catalytic function; p.Phe119Leu impairs dimerization/interface stability	These common pathogenic alleles illustrate separable catalytic-site and dimer-interface failure modes	(mele2025insilicoanalysis pages 1-2, edmondson2025novelmousemodel pages 4-7)
Catalytic mechanism class	Shared mutase “ping-pong” phosphotransfer mechanism	A recent mechanistic comparison places PMM2 among enzymes sharing a communal ping-pong catalytic mechanism	(mele2025insilicoanalysis pages 1-2)
Catalytic intermediate	Phosphoenzyme/phosphohistidine intermediate characteristic of phosphosugar mutases	PMM2 belongs to the phosphomutase family and is discussed in the context of phosphotransfer mutase chemistry; this supports annotation to phosphomannomutase activity rather than broader signaling roles	(mele2025insilicoanalysis pages 1-2)
Reaction directionality in vivo	Reversible chemically, but physiologically important for generating Man-1-P for downstream glycosylation precursor synthesis	Mouse and human studies emphasize the biological consequence of reduced cytosolic PMM2 activity as failure to sustain Man-1-P and GDP-mannose pools	(edmondson2025novelmousemodel pages 4-7, ligezka2023interplayofimpaired pages 1-2, radenkovic2023tracermetabolomicsreveals pages 1-4)
Cellular biochemical context	Cytosolic enzyme activity measured in bulk brain/cell lysates; supplies cytosol-derived sugar nucleotides for ER/Golgi glycosylation pathways	Recent mouse work explicitly refers to cytosolic PMM2 enzyme activity, and broader glycosylation studies describe activated sugar donors as recruited from the cytosol into ER/Golgi glycosylation pathways	(edmondson2025novelmousemodel pages 4-7, mangione2025targetedmetabolomicevaluation pages 1-2)
Key metabolic phenotype when deficient	Decreased mannose-1-phosphate and GDP-mannose; broad hypoglycosylation	PMM2-deficient fibroblasts and PBMCs show depleted Man-1-P/GDP-mannose with altered glycoprotein/glycopeptide profiles	(radenkovic2023tracermetabolomicsreveals pages 1-4, mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2)
Best-supported core GO-relevant function	Phosphomannomutase activity driving mannose phosphate interconversion for glycosylation precursor biosynthesis	The strongest evidence supports concise core annotations around phosphomannomutase activity, GDP-mannose biosynthetic contribution, and protein N-glycosylation support	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, garapati2024acomplementc4–derived pages 1-2)

Table: This table summarizes the best-supported core molecular functions of human PMM2 for GO annotation review. It emphasizes the enzyme’s phosphomannomutase activity, obligate homodimeric structure, catalytic mechanism class, and role in generating glycosylation precursors.

1.2 Protein Activation and Maturation Mechanisms

No evidence for proteolytic processing was identified. PMM2 maturation appears to depend primarily on proper protein folding and quaternary structure assembly rather than post-translational cleavage events (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2).

1.3 Quaternary Structure: Obligate Homodimer

PMM2 functions as an obligate homodimer, with dimerization considered essential for enzymatic activity (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, tran2020secondgenerationpharmacologicalchaperones pages 1-3). Structural evidence from both human and yeast homolog studies confirms the homodimeric nature of active PMM2 (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2).

The importance of dimerization is illustrated by disease-causing mutations. The common pathogenic variant p.Phe119Leu (equivalent to yeast F126L) specifically disrupts the dimer interface, causing loss of function through impaired dimerization rather than catalytic-site damage (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2). In contrast, the p.Arg141His variant affects catalytic function directly (mele2025insilicoanalysis pages 1-2, edmondson2025novelmousemodel pages 4-7). Recent structural modeling emphasizes that interface stability is a critical determinant of residual enzymatic activity, with interface mutations predicted to reduce dimer stability and contribute to disease-associated variant effects (mele2025insilicoanalysis pages 1-2).

In silico modeling has suggested that PMM2/PMM1 heterodimers may be structurally plausible and energetically viable, though slightly less stable than PMM2 homodimers (mele2025insilicoanalysis pages 1-2). However, this remains predictive, and PMM1 does not compensate for PMM2 deficiency in vivo, indicating heterodimerization is not physiologically significant (mele2025insilicoanalysis pages 1-2).

2. Biological Process Roles

2.1 Core Function: GDP-Mannose Biosynthesis and N-Glycosylation

The most strongly supported biological process role is PMM2's direct contribution to GDP-mannose biosynthetic processes and protein N-linked glycosylation (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, garapati2024acomplementc4–derived pages 1-2). By catalyzing the conversion of mannose-6-phosphate to mannose-1-phosphate, PMM2 provides the immediate precursor used to generate GDP-mannose (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2). This positions PMM2 upstream of dolichol-phosphate-mannose synthesis and lipid-linked oligosaccharide precursor assembly required for N-glycosylation (mele2025insilicoanalysis pages 1-2, ligezka2023interplayofimpaired pages 1-2, garapati2024acomplementc4–derived pages 1-2).

PMM2 deficiency reduces availability of mannose donors, causing broad hypoglycosylation and altered N-glycan maturation in patient cells and sera (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, radenkovic2023tracermetabolomicsreveals pages 1-4, garapati2024acomplementc4–derived pages 1-2). Glycoproteomic analyses reveal that PMM2-deficient fibroblasts shift toward truncated and high-mannose glycan species, with decreased complex-type N-glycans (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2). Therapeutic interventions that increase GDP-mannose levels, such as liposome-encapsulated mannose-1-phosphate or aldose reductase inhibition, improve glycosylation in PMM2-deficient cells, confirming the causal role of PMM2 in maintaining glycosylation homeostasis (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, radenkovic2023tracermetabolomicsreveals pages 1-4).

These annotations (GDP-mannose biosynthetic process, protein N-glycosylation) represent core, direct roles appropriate for GO annotation (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, garapati2024acomplementc4–derived pages 1-2).

2.2 Neurodevelopmental and Cerebellar Roles

PMM2 deficiency causes severe neurodevelopmental phenotypes, particularly cerebellar hypoplasia and ataxia, in both human patients and mouse models (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7, pradeep2023glycosylationandbehavioral pages 1-2). A novel mouse model pairing a floxed Pmm2 allele with a catalytically inactive R137H knock-in allele revealed striking cell-type and developmental-stage specificity (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7).

Post-mitotic neuronal (Snap25-Cre) or astrocytic (Gfap-Cre) loss of PMM2 produced mice indistinguishable from controls across broad neurological assessments (edmondson2025novelmousemodel pages 4-7). In contrast, embryonic neural precursor-specific deletion (Nestin-Cre) caused cerebellar hypoplasia, ataxia, seizures, and early lethality (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7). Multi-omics profiling revealed widespread molecular disturbances throughout the brain, with cerebellum showing the most pronounced disruption (edmondson2025novelmousemodel pages 1-4). Glycoproteomic alterations identified in mouse models were corroborated in PMM2-CDG patient post-mortem cerebellar tissue, implicating impaired synaptic transmission as a key pathogenic mechanism (edmondson2025novelmousemodel pages 1-4).

These findings highlight heightened dependency of the developing cerebellum on intact PMM2-dependent N-glycosylation and reveal a neurodevelopmental origin of PMM2-CDG brain pathology (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7). However, these represent organism-level phenotypic consequences of defective glycosylation rather than direct gene-specific developmental functions. Such annotations should be considered context-dependent and phenotype-linked, not replacing core glycosylation annotations (edmondson2025novelmousemodel pages 1-4, parrado2022dissectingthetranscriptional pages 1-2, pradeep2023glycosylationandbehavioral pages 1-2).

2.3 Secondary and Pleiotropic Effects: ER Stress, Mitochondrial Dysfunction, Autophagy, and Inflammation

PMM2 deficiency causes multiple downstream cellular consequences that are better understood as pleiotropic disease effects rather than core PMM2 biological processes.

ER Stress and Unfolded Protein Response: Abnormal protein glycosylation in PMM2-deficient cells promotes accumulation of misfolded proteins and activation of ER stress/unfolded protein response (UPR) programs (ligezka2023interplayofimpaired pages 1-2, parrado2022dissectingthetranscriptional pages 1-2). This represents a secondary consequence of impaired glycosylation, not a direct PMM2 function.

Mitochondrial Dysfunction: PMM2-CDG patient-derived fibroblasts exhibit secondary mitochondrial dysfunction, including decreased maximal and ATP-linked respiration, reduced complex I function of the electron transport chain, and altered energy metabolism (mangione2025targetedmetabolomicevaluation pages 1-2, ligezka2023interplayofimpaired pages 1-2). Metabolomic profiling revealed lower ATP, GTP, and UTP levels, abnormal ATP/ADP and NAD+/NADH ratios, and increased oxidative stress markers (mangione2025targetedmetabolomicevaluation pages 1-2). While these findings are consistent and reproducible, they represent pleiotropic downstream effects rather than core PMM2 function (ligezka2023interplayofimpaired pages 1-2).

Autophagy and Mitophagy Alterations: PMM2-CDG fibroblasts show altered autophagy, marked by increased abundance of the autophagosome marker LC3-II and changes in abundance/glycosylation of autophagy and mitophagy pathway proteins (ligezka2023interplayofimpaired pages 1-2). Interestingly, serum sorbitol levels (a disease severity biomarker) and CDG severity scores showed inverse correlation with LC3-II abundance, suggesting autophagy may modulate disease severity (ligezka2023interplayofimpaired pages 1-2). However, this represents adaptive or pathophysiologic cellular response, not direct PMM2 regulation of autophagy.

Inflammatory Signaling: Defective glycosylation in PMM2-CDG fibroblasts alters TNFR1 structure and receptor shedding, leading to blunted downstream cytokine (IL-6, CCL5) and kinase (ERK1/2, p38, JNK2) responses after TNF-α stimulation (pascoal2025immunopathologyinpmm2cdg pages 1-3). Transcriptomic analysis revealed dysregulation of immune and signaling pathways, particularly in cells bearing R141H heterozygous variants (pascoal2025immunopathologyinpmm2cdg pages 1-3). These findings point to TNFR1 signaling dysregulation as a contributor to immune dysfunction in PMM2-CDG, but represent context-specific immunopathology rather than core PMM2 function (pascoal2025immunopathologyinpmm2cdg pages 1-3).

Metabolic Rewiring: PMM2 deficiency alters intracellular glucose flux, increasing polyol production while depleting GDP-mannose (radenkovic2023tracermetabolomicsreveals pages 1-4). Aldose reductase inhibition redirects glucose flux away from polyol production toward sugar nucleotide synthesis, improving glycosylation (radenkovic2023tracermetabolomicsreveals pages 1-4). This represents compensatory/pathophysiologic metabolism secondary to the core glycosylation defect.

Biological process / role	Summary of PMM2 involvement	Strength for GO review	Direct core role or downstream effect?	Key supporting systems	GO annotation implication	Citation
GDP-mannose biosynthetic process	PMM2 catalyzes mannose-6-phosphate to mannose-1-phosphate, providing the immediate precursor used to generate GDP-mannose	Very strong	Direct core role	Human biochemical/disease literature; patient-derived fibroblasts; mouse biochemical assays	Strongly supports core process annotation to GDP-mannose precursor generation / mannose phosphate interconversion	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, garapati2024acomplementc4–derived pages 1-2, edmondson2025novelmousemodel pages 4-7)
Dolichol-phosphate-mannose and lipid-linked oligosaccharide precursor supply	By sustaining mannose-1-phosphate and GDP-mannose pools, PMM2 supports synthesis of dolichol-phosphate-mannose and mannose incorporation into lipid-linked oligosaccharides required for N-glycosylation	Very strong	Direct core role	Pathway-focused reviews; PMM2-CDG metabolic studies	Appropriate as a core upstream glycosylation-pathway annotation	(mele2025insilicoanalysis pages 1-2, ligezka2023interplayofimpaired pages 1-2, garapati2024acomplementc4–derived pages 1-2)
Protein N-linked glycosylation	PMM2 deficiency reduces availability of mannose donors, causing broad hypoglycosylation and altered N-glycan maturation in cells and patients	Very strong	Direct core role at pathway-support level	Patient sera glycoproteomics; fibroblast glycoproteomics; metabolomics; therapy rescue studies	Strongly supported biological-process annotation; this is the principal organism-level consequence of PMM2 activity	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, radenkovic2023tracermetabolomicsreveals pages 1-4, garapati2024acomplementc4–derived pages 1-2)
Global protein glycosylation homeostasis	PMM2 activity is required to maintain normal glycoform abundance across many proteins, with deficiency shifting glycoproteins toward truncated/high-mannose species	Strong	Mostly direct, but broad/system-level	Human serum N-glycoproteomics; fibroblast treatment studies	Acceptable if framed broadly around protein glycosylation, but less specific than N-glycosylation/GDP-mannose biosynthesis	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2)
Nervous system development / neurodevelopment	PMM2 deficiency disrupts brain development, especially when severe reduction occurs in embryonic neural precursors; post-mitotic neuronal or astrocytic loss alone is insufficient in mice	Strong for disease biology, weaker as direct gene-specific developmental role	Predominantly downstream of defective glycosylation	Conditional mouse models; patient phenotypes; transcriptional profiling	Use caution: suitable only as a higher-level phenotype-linked process, not as the primary core function	(edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7, parrado2022dissectingthetranscriptional pages 1-2, pradeep2023glycosylationandbehavioral pages 1-2)
Cerebellar development	Developing cerebellum shows heightened dependence on intact PMM2-dependent N-glycosylation; embryonic neural precursor loss causes cerebellar hypoplasia and ataxia in mouse, matching human disease	Strong for organismal phenotype	Downstream consequence of core glycosylation defect	Conditional mouse model; patient imaging/clinical studies; reviews	Potentially supportable with strong phenotype context, but should not replace core glycosylation annotations	(edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7, pradeep2023glycosylationandbehavioral pages 1-2)
Synaptic transmission / synaptic function	Multi-omics in PMM2-deficient mouse brain implicated impaired synaptic transmission as a pathogenic mechanism, especially in cerebellum	Moderate	Downstream consequence	Mouse brain multi-omics; corroboration with patient post-mortem cerebellar tissue	Too context-dependent for core annotation; consider only with explicit experimental support standards	(edmondson2025novelmousemodel pages 1-4)
Response to endoplasmic reticulum stress / unfolded protein response	PMM2 deficiency causes abnormal protein glycosylation that promotes accumulation of misfolded proteins and activation of ER stress/UPR programs	Moderate to strong	Secondary downstream effect	Patient-derived fibroblasts; PMM2-deficient lymphoblastoid cells; review literature	Better treated as indirect/pathophysiologic effect, not core PMM2 biological process	(ligezka2023interplayofimpaired pages 1-2, parrado2022dissectingthetranscriptional pages 1-2)
Autophagy / mitophagy regulation	PMM2-CDG fibroblasts show altered autophagy markers and abundance/glycosylation of autophagy/mitophagy pathway proteins	Moderate	Secondary downstream effect	Patient fibroblasts	Annotation risk is high unless directly demonstrated in target organism with specific mechanistic evidence	(ligezka2023interplayofimpaired pages 1-2)
Mitochondrial function / cellular bioenergetics	PMM2 deficiency is associated with secondary mitochondrial dysfunction, including reduced maximal and ATP-linked respiration and altered complex I function	Moderate	Secondary downstream effect	Patient fibroblasts; metabolomic profiling	Should be considered pleiotropic disease consequence, not core function	(ligezka2023interplayofimpaired pages 1-2, mangione2025targetedmetabolomicevaluation pages 1-2)
Inflammatory signaling via TNF/TNFR1 pathway	Defective glycosylation in PMM2-CDG fibroblasts alters TNFR1 structure/shedding and blunts downstream cytokine and kinase responses after TNF stimulation	Moderate but context-specific	Secondary downstream effect	Patient skin fibroblasts; transcriptomics/glycomics/immune assays	High-risk annotation for core GO; better viewed as context-specific immunopathology	(pascoal2025immunopathologyinpmm2cdg pages 1-3)
Cellular metabolic rewiring / polyol metabolism	PMM2 deficiency alters intracellular glucose flux, increases polyols, and depletes GDP-mannose; aldose reductase inhibition can redirect flux toward sugar nucleotide synthesis	Moderate	Secondary downstream effect linked to core metabolic defect	Patient fibroblasts; zebrafish models; metabolomics; clinical observations	Do not treat as core PMM2 process; this is compensatory/pathophysiologic metabolism	(radenkovic2023tracermetabolomicsreveals pages 1-4)
Broad transcriptional effects on development, neuronal differentiation, cell junctions, and motility	PMM2-deficient cells show widespread transcriptomic changes, including genes linked to neuronal differentiation and stress responses	Moderate but nonspecific	Indirect downstream effect	PMM2-deficient B-lymphoblastoid cells	Not appropriate for direct GO function/process transfer without more targeted mechanistic evidence	(parrado2022dissectingthetranscriptional pages 1-2)
Apoptosis / programmed cell death	Available PMM2 literature in the retrieved set supports ER stress and developmental pathology, but does not provide strong direct evidence that PMM2 itself is a dedicated apoptosis regulator; some cited CDG model work reported apoptosis not clearly elevated in PMM2 mutant cell lines	Weak / insufficient for direct PMM2 role	Likely indirect where present	Cellular models; reviews	Avoid direct apoptosis annotation based on current evidence set	(ligezka2023interplayofimpaired pages 1-2, pradeep2023glycosylationandbehavioral pages 1-2)
Pyroptosis	No convincing evidence found in the retrieved PMM2-specific literature	Insufficient	Not established	None in retrieved PMM2-specific sources	Do not annotate	(ligezka2023interplayofimpaired pages 1-2, pascoal2025immunopathologyinpmm2cdg pages 1-3)
Developmental cell death	Mouse developmental brain pathology supports developmental dependence on PMM2, but not a specific direct role in executing developmental cell death programs	Weak to moderate	Indirect if present	Embryonic neural precursor mouse models	Avoid specific developmental cell death annotation absent direct mechanistic data	(edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7)

Table: This table distinguishes PMM2’s strongly supported core biological processes from downstream, pleiotropic, or context-specific effects observed in disease models and patients. It is useful for GO annotation review because it separates direct glycosylation-related functions from secondary neurodevelopmental, stress, immune, and metabolic consequences.

2.4 Evidence Assessment for Specific Processes

Apoptosis and Developmental Cell Death: Available evidence does not support PMM2 as a dedicated apoptosis regulator. While ER stress and developmental pathology occur in PMM2 deficiency, cellular model studies found that apoptosis levels in PMM2-deficient cell lines were comparable to controls (ligezka2023interplayofimpaired pages 1-2). The developmental brain pathology in mouse models reflects developmental dependence on PMM2 for glycosylation, not a specific direct role in executing developmental cell death programs (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7). Direct apoptosis annotation should be avoided based on current evidence (ligezka2023interplayofimpaired pages 1-2, pradeep2023glycosylationandbehavioral pages 1-2).

Pyroptosis: No convincing evidence for PMM2 involvement in pyroptosis was found in the retrieved literature (ligezka2023interplayofimpaired pages 1-2, pascoal2025immunopathologyinpmm2cdg pages 1-3).

Synaptic Function: Multi-omics profiling implicated impaired synaptic transmission as a pathogenic mechanism in PMM2-deficient cerebellum (edmondson2025novelmousemodel pages 1-4). However, this reflects downstream consequences of defective glycosylation of synaptic proteins rather than direct PMM2 regulation of synaptic processes.

3. Cellular Localization and Complexes

3.1 Subcellular Localization: Cytoplasm/Cytosol

PMM2 is best supported as a cytoplasmic/cytosolic enzyme (mangione2025targetedmetabolomicevaluation pages 1-2, edmondson2025novelmousemodel pages 4-7). Recent mouse studies explicitly measured cytosolic PMM2 enzyme activity in brain tissue (edmondson2025novelmousemodel pages 4-7). Glycosylation pathway studies describe sugar-nucleotide precursor generation occurring in the cytosol, with activated sugar donors subsequently transported to ER and Golgi for glycosylation reactions (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2).

PMM2's functional metabolic context positions it in the cytosol where it generates mannose-1-phosphate for GDP-mannose production (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2). This supports annotating PMM2 to the cytosolic side of glycosylation precursor metabolism, not to ER or Golgi lumen (mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2).

No strong evidence supports PMM2 as a nuclear protein in normal physiology. The mechanistic and biochemical evidence centers on cytosolic enzyme activity and glycosylation precursor supply (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2, edmondson2025novelmousemodel pages 4-7). Similarly, PMM2 supports ER/Golgi glycosylation indirectly through precursor supply but is not itself an ER/Golgi-resident enzyme (mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2).

3.2 Protein Complexes: Homodimer

The primary complex state is the PMM2 homodimer (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, tran2020secondgenerationpharmacologicalchaperones pages 1-3). This obligate homodimer represents the active enzymatic form, with dimer interface integrity being a key determinant of residual activity (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2). Strong support exists for annotation to protein homodimerization activity/complex, but not to large signaling complexes (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2).

Evidence for PMM2 participation in other protein complexes is limited. A PMM2-TRIM28 interaction was reported in colorectal cancer cells, affecting TRIM28 nuclear translocation and KIFC3 transcriptional regulation (peng2026pmm2interactswith pages 1-2). However, this represents a context-specific cancer mechanism rather than core PMM2 biology and is not appropriate for core cellular component annotation (peng2026pmm2interactswith pages 1-2).

Little support exists for placing PMM2 in canonical inflammatory/signaling complexes. TNFR1-pathway abnormalities are explained as consequences of defective glycosylation of other proteins, not evidence that PMM2 is a stable component of TNFR1 signaling complexes (pascoal2025immunopathologyinpmm2cdg pages 1-3).

Category	Finding	Evidence summary	GO cellular component implication	Citation
Primary subcellular localization	PMM2 is best supported as a cytoplasmic/cytosolic enzyme	Recent mouse work explicitly refers to reduced cytosolic PMM2 enzyme activity in brain after conditional disruption; broader glycosylation studies place sugar-nucleotide precursor generation in the cytosol before use in ER/Golgi glycosylation pathways (edmondson2025novelmousemodel pages 4-7, mangione2025targetedmetabolomicevaluation pages 1-2)	Strong support for cytoplasm/cytosol annotation	(edmondson2025novelmousemodel pages 4-7, mangione2025targetedmetabolomicevaluation pages 1-2)
Functional metabolic context	PMM2 acts in the cytosol to generate mannose-1-phosphate for GDP-mannose production, with downstream use of sugar donors in ER/Golgi glycosylation	PMM2 converts mannose-6-phosphate to mannose-1-phosphate; activated sugar donors are described as recruited from the cytosol and then used in ER/Golgi glycosylation pathways (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2)	Supports annotating PMM2 to the cytosolic side of glycosylation precursor metabolism, not to ER/Golgi lumen	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2)
Primary complex state	PMM2 functions as an obligate homodimer	Multiple sources describe PMM2 as a homodimer/obligate homodimer; dimerization is considered important for enzymatic activity and loss-of-function variants can disrupt the interface (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, tran2020secondgenerationpharmacologicalchaperones pages 1-3)	Strong support for annotation to protein homodimerization activity/complex, but not to large signaling complexes	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, tran2020secondgenerationpharmacologicalchaperones pages 1-3)
Dimerization evidence strength	Dimer interface integrity is a key determinant of residual activity	Common pathogenic variants illustrate separation of catalytic defects and dimerization/interface defects, especially p.Phe119Leu; recent structural modeling emphasizes interface stability (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2)	Strengthens confidence in PMM2 homodimer as the active form	(vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2)
PMM1 interaction evidence	PMM2/PMM1 heterodimerization is structurally plausible but not established as a validated physiological complex	In silico modeling suggested PMM2/PMM1 heterodimers are energetically viable, but this remains predictive; PMM1 does not compensate for PMM2 deficiency in vivo (mele2025insilicoanalysis pages 1-2)	Do not annotate PMM2 to a PMM1-containing complex without stronger experimental evidence	(mele2025insilicoanalysis pages 1-2)
Nuclear localization evidence	No strong evidence in the retrieved core PMM2 literature supports PMM2 as a bona fide nuclear protein in normal physiology	The strongest mechanistic and biochemical evidence centers on cytosolic enzyme activity and glycosylation precursor supply; no direct normal nuclear localization evidence was identified in the high-confidence PMM2 disease/metabolism studies reviewed (edmondson2025novelmousemodel pages 4-7, budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2)	Do not prioritize nucleus annotation for core GO review	(edmondson2025novelmousemodel pages 4-7, budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2)
ER/Golgi localization evidence	PMM2 supports ER/Golgi glycosylation indirectly but is not itself shown to be an ER/Golgi-resident enzyme	The pathway places PMM2 upstream in cytosolic precursor generation, while glycan assembly/processing occurs in ER and Golgi (mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2)	Avoid annotating PMM2 to ER lumen, Golgi lumen, or Golgi membrane based solely on pathway membership	(mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2)
Signaling complex evidence	Little support for placing PMM2 in canonical inflammatory/signaling complexes as a core function	TNFR1-pathway abnormalities in PMM2-CDG are explained as consequences of defective glycosylation of other proteins, not evidence that PMM2 is a stable component of TNFR1 signaling complexes (pascoal2025immunopathologyinpmm2cdg pages 1-3)	Avoid annotation to TNFR1/signaling complexes	(pascoal2025immunopathologyinpmm2cdg pages 1-3)
TRIM28 interaction	PMM2-TRIM28 interaction has been reported in colorectal cancer cells, with associated effects on TRIM28 nuclear translocation	This interaction comes from a 2026 cancer study outside the requested 2023-2024 priority window and reflects a disease-context mechanism rather than established core PMM2 biology (peng2026pmm2interactswith pages 1-2)	At most context-specific, not appropriate for core cellular component annotation in a GO review centered on canonical function	(peng2026pmm2interactswith pages 1-2)
Cytoplasm vs cytosol terminology	Available evidence supports both “cytoplasm” and more specifically “cytosol/cytosolic activity”	Disease and biochemical studies use cytosolic language for enzyme activity, while broader pathway descriptions refer to cytoplasmic production of sugar donors (edmondson2025novelmousemodel pages 4-7, mangione2025targetedmetabolomicevaluation pages 1-2)	Cytosol may be the most specific supported term; cytoplasm is also defensible	(edmondson2025novelmousemodel pages 4-7, mangione2025targetedmetabolomicevaluation pages 1-2)
Overall annotation confidence	Best-supported localization/complex model is cytosolic PMM2 homodimer supplying glycosylation precursors	This integrates biochemical function, quaternary structure, and pathway localization from human, mouse, and disease studies (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, mangione2025targetedmetabolomicevaluation pages 1-2, edmondson2025novelmousemodel pages 4-7)	Core GO cellular component annotations should prioritize cytosol/cytoplasm and homodimer, while excluding overextended nucleus, ER/Golgi lumen, and signaling-complex assignments	(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, mangione2025targetedmetabolomicevaluation pages 1-2, edmondson2025novelmousemodel pages 4-7)

Table: This table summarizes the strongest evidence for PMM2 subcellular localization and complex membership relevant to GO review. It distinguishes well-supported cytosolic homodimer localization from weaker or context-specific claims such as nuclear, ER/Golgi, or signaling-complex associations.

4. Annotation-Risk Assessment

4.1 Core Functions (Strongly Supported, Low Risk)

Phosphomannomutase activity (EC 5.4.2.8): Very strong biochemical and genetic evidence (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, edmondson2025novelmousemodel pages 4-7)
GDP-mannose biosynthetic process: Direct upstream role supported by pathway analysis, metabolomics, and rescue studies (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, radenkovic2023tracermetabolomicsreveals pages 1-4, garapati2024acomplementc4–derived pages 1-2)
Protein N-linked glycosylation: Extensive glycoproteomic evidence in human and mouse (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, mangione2025targetedmetabolomicevaluation pages 1-2, radenkovic2023tracermetabolomicsreveals pages 1-4, garapati2024acomplementc4–derived pages 1-2)
Cytoplasm/cytosol localization: Consistent with biochemical function and pathway context (mangione2025targetedmetabolomicevaluation pages 1-2, edmondson2025novelmousemodel pages 4-7)
Protein homodimerization: Strong structural and functional evidence (vignogna2022evolutionaryrescueof pages 1-2, mele2025insilicoanalysis pages 1-2, tran2020secondgenerationpharmacologicalchaperones pages 1-3)

4.2 Context-Dependent Annotations (Moderate Risk)

Nervous system development/cerebellar development: Strong phenotypic evidence but downstream of glycosylation defect; suitable only as higher-level phenotype-linked process with appropriate context (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7, pradeep2023glycosylationandbehavioral pages 1-2)
Synaptic transmission: Implicated by multi-omics but indirect; requires explicit experimental support standards (edmondson2025novelmousemodel pages 1-4)

4.3 Pleiotropic/Secondary Effects (High Annotation Risk)

ER stress/UPR: Secondary consequence of impaired glycosylation, not core function (ligezka2023interplayofimpaired pages 1-2, parrado2022dissectingthetranscriptional pages 1-2)
Mitochondrial function/bioenergetics: Secondary downstream effect observed in disease (mangione2025targetedmetabolomicevaluation pages 1-2, ligezka2023interplayofimpaired pages 1-2)
Autophagy/mitophagy: Altered in disease but indirect (ligezka2023interplayofimpaired pages 1-2)
Inflammatory signaling (TNF pathway): Context-specific immunopathology (pascoal2025immunopathologyinpmm2cdg pages 1-3)
Metabolic rewiring/polyol metabolism: Compensatory/pathophysiologic (radenkovic2023tracermetabolomicsreveals pages 1-4)

4.4 Insufficient Evidence (Do Not Annotate)

Apoptosis: No convincing direct evidence (ligezka2023interplayofimpaired pages 1-2, pradeep2023glycosylationandbehavioral pages 1-2)
Pyroptosis: No evidence found (ligezka2023interplayofimpaired pages 1-2, pascoal2025immunopathologyinpmm2cdg pages 1-3)
Developmental cell death (as distinct process): Weak mechanistic evidence (edmondson2025novelmousemodel pages 1-4, edmondson2025novelmousemodel pages 4-7)
Nuclear localization: Not supported in normal physiology (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3, garapati2024acomplementc4–derived pages 1-2, edmondson2025novelmousemodel pages 4-7)
ER/Golgi lumen localization: Upstream cytosolic enzyme (mangione2025targetedmetabolomicevaluation pages 1-2, garapati2024acomplementc4–derived pages 1-2)
PMM1 heterodimerization: Structurally plausible but not physiologically established (mele2025insilicoanalysis pages 1-2)

5. Key Literature

5.1 Recent Reviews and Clinical Studies (2022-2025)

Mele et al. (2025): Structural analysis of PMM2 dimer interface stability and heterodimerization potential (mele2025insilicoanalysis pages 1-2)
Edmondson et al. (2025): Novel mouse model revealing neurodevelopmental origin of PMM2-CDG brain pathology (edmondson2025novelmousemodel pages 1-4)
Medico et al. (2025): PMM2-CDG endocrine manifestations and N-glycosylation role (first paper retrieved, 2025 publication)
Mangione et al. (2025): Targeted metabolomics of PMM2-CDG PBMCs revealing GDP-mannose depletion and metabolic changes (mangione2025targetedmetabolomicevaluation pages 1-2)
Pascoal et al. (2025): Immunopathology in PMM2-CDG and TNF-TNFR1 signaling disruption (pascoal2025immunopathologyinpmm2cdg pages 1-3)

5.2 Biochemical and Therapeutic Studies (2022-2024)

Budhraja et al. (2024): Liposome-encapsulated mannose-1-phosphate therapy improving N-glycosylation (budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3)
Garapati et al. (2024): Complement C4-derived glycopeptide biomarker for PMM2-CDG (garapati2024acomplementc4–derived pages 1-2)
Radenkovic et al. (2023): Aldose reductase role in glycosylation and metabolic rewiring in PMM2-CDG (radenkovic2023tracermetabolomicsreveals pages 1-4)
Ligezka et al. (2023): Interplay of impaired cellular bioenergetics and autophagy in PMM2-CDG (ligezka2023interplayofimpaired pages 1-2)

5.3 Mechanistic and Model Studies (2020-2022)

Vignogna et al. (2022): Evolutionary rescue of phosphomannomutase deficiency in yeast models (vignogna2022evolutionaryrescueof pages 1-2)
Parrado et al. (2022): Transcriptional program of PMM2-deficient B-lymphoblastoid cells (parrado2022dissectingthetranscriptional pages 1-2)
Tran et al. (2020): Second-generation pharmacological chaperones for PMM2-CDG (tran2020secondgenerationpharmacologicalchaperones pages 1-3)
Sosicka et al. (2021): Chemical therapies for congenital disorders of glycosylation (sosicka2021chemicaltherapiesfor pages 1-3)

This comprehensive review provides evidence-based guidance for GO annotation of human PMM2, clearly distinguishing core molecular functions and biological processes from context-specific, pleiotropic, or over-extended roles.

References

(budhraja2024liposomeencapsulatedmannose1phosphatetherapy pages 1-3): Rohit Budhraja, Silvia Radenkovic, Anu Jain, Irena J.J. Muffels, Moulay Hicham Alaoui Ismaili, Tamas Kozicz, Akhilesh Pandey, and Eva Morava. Liposome-encapsulated mannose-1-phosphate therapy improves global n-glycosylation in different congenital disorders of glycosylation. Molecular Genetics and Metabolism, 142:108487, Jun 2024. URL: https://doi.org/10.1016/j.ymgme.2024.108487, doi:10.1016/j.ymgme.2024.108487. This article has 19 citations and is from a peer-reviewed journal.
(vignogna2022evolutionaryrescueof pages 1-2): Ryan C Vignogna, Mariateresa Allocca, Maria Monticelli, Joy W Norris, Richard Steet, Ethan O Perlstein, Giuseppina Andreotti, and Gregory I Lang. Evolutionary rescue of phosphomannomutase deficiency in yeast models of human disease. eLife, Oct 2022. URL: https://doi.org/10.7554/elife.79346, doi:10.7554/elife.79346. This article has 16 citations and is from a domain leading peer-reviewed journal.
(mele2025insilicoanalysis pages 1-2): Bruno Hay Mele, Jessica Bovenzi, Giuseppina Andreotti, Maria Vittoria Cubellis, and Maria Monticelli. In silico analysis of phosphomannomutase-2 dimer interface stability and heterodimerization with phosphomannomutase-1. Jun 2025. URL: https://doi.org/10.3390/molecules30122599, doi:10.3390/molecules30122599. This article has 2 citations.
(garapati2024acomplementc4–derived pages 1-2): Kishore Garapati, Rohit Budhraja, Mayank Saraswat, Jinyong Kim, Neha Joshi, Gunveen S. Sachdeva, Anu Jain, Anna N. Ligezka, Silvia Radenkovic, Madan Gopal Ramarajan, Savita Udainiya, Kimiyo Raymond, Miao He, Christina Lam, Austin Larson, Andrew C. Edmondson, Kyriakie Sarafoglou, Nicholas B. Larson, Hudson H. Freeze, Matthew J. Schultz, Tamas Kozicz, Eva Morava, and Akhilesh Pandey. A complement c4–derived glycopeptide is a biomarker for pmm2-cdg. JCI Insight, Apr 2024. URL: https://doi.org/10.1172/jci.insight.172509, doi:10.1172/jci.insight.172509. This article has 11 citations and is from a domain leading peer-reviewed journal.
(mangione2025targetedmetabolomicevaluation pages 1-2): Renata Mangione, Lara Cirnigliaro, Miriam Wissam Saab, Fabio Pettinato, Alessandro Barbato, Alfio Distefano, Enrico La Spina, Giuseppe Lazzarino, Giovanni Li Volti, Lucia Longhitano, Daniele Tibullo, Alessandra Pittalà, Cesarina Giallongo, Valentina Di Pietro, Giovanni Tabbi, Salvatore Antonio Longo, Andrea Graziani, Barbara Tavazzi, Angela Maria Amorini, Giacomo Lazzarino, and Rita Barone. Targeted metabolomic evaluation of peripheral blood mononucleated cells from patients with pmm2-cdg. Scientific Reports, May 2025. URL: https://doi.org/10.1038/s41598-025-98846-8, doi:10.1038/s41598-025-98846-8. This article has 3 citations and is from a peer-reviewed journal.
(edmondson2025novelmousemodel pages 4-7): Andrew C. Edmondson, Rohit Budhraja, Zijie Xia, Ashley Melendez-Perez, Cadmus Cai, Silvia Radenkovic, Ashley M. Collins, Emily J. Shiplett, Sophie F. Hill, Ala Somarowthu, Johanna Dam, Ling-Lin Pai, Mariarita Santi, Seonhee Kim, Miao He, Ethan M. Goldberg, Tamas Kozicz, Eva Morava, Akhilesh Pandey, and Zhaolan Zhou. Novel mouse model reveals neurodevelopmental origin of pmm2-cdg brain pathology. bioRxiv, Jun 2025. URL: https://doi.org/10.1101/2025.06.01.657261, doi:10.1101/2025.06.01.657261. This article has 2 citations.
(edmondson2025novelmousemodel pages 1-4): Andrew C. Edmondson, Rohit Budhraja, Zijie Xia, Ashley Melendez-Perez, Cadmus Cai, Silvia Radenkovic, Ashley M. Collins, Emily J. Shiplett, Sophie F. Hill, Ala Somarowthu, Johanna Dam, Ling-Lin Pai, Mariarita Santi, Seonhee Kim, Miao He, Ethan M. Goldberg, Tamas Kozicz, Eva Morava, Akhilesh Pandey, and Zhaolan Zhou. Novel mouse model reveals neurodevelopmental origin of pmm2-cdg brain pathology. bioRxiv, Jun 2025. URL: https://doi.org/10.1101/2025.06.01.657261, doi:10.1101/2025.06.01.657261. This article has 2 citations.
(ligezka2023interplayofimpaired pages 1-2): Anna N. Ligezka, Rohit Budhraja, Yurika Nishiyama, Fabienne C. Fiesel, Graeme Preston, Andrew Edmondson, Wasantha Ranatunga, Johan L. K. Van Hove, Jens O. Watzlawik, Wolfdieter Springer, Akhilesh Pandey, Eva Morava, and Tamas Kozicz. Interplay of impaired cellular bioenergetics and autophagy in pmm2-cdg. Genes, 14:1585, Aug 2023. URL: https://doi.org/10.3390/genes14081585, doi:10.3390/genes14081585. This article has 21 citations.
(pascoal2025immunopathologyinpmm2cdg pages 1-3): Carlota Pascoal, Pedro Granjo, Rebeka Kodríková, Marta Falcão, Ana C. Santos, Inês Teodoro, Zuzana Pakanová, Marek Nemčovič, Jan Mucha, Margarida Castro-Caldas, Ana R. Grosso, Vanessa dos Reis Ferreira, and Paula A. Videira. Immunopathology in pmm2-cdg: defective glycosylation impact in the tnfα -tnfr1 signalling pathway. Frontiers in Immunology, Sep 2025. URL: https://doi.org/10.3389/fimmu.2025.1655354, doi:10.3389/fimmu.2025.1655354. This article has 1 citations and is from a peer-reviewed journal.
(pradeep2023glycosylationandbehavioral pages 1-2): Prajitha Pradeep, Hyeyeon Kang, and Boyoung Lee. Glycosylation and behavioral symptoms in neurological disorders. Translational Psychiatry, May 2023. URL: https://doi.org/10.1038/s41398-023-02446-x, doi:10.1038/s41398-023-02446-x. This article has 116 citations and is from a peer-reviewed journal.
(radenkovic2023tracermetabolomicsreveals pages 1-4): Silvia Radenkovic, Anna N. Ligezka, Sneha S. Mokashi, Karen Driesen, Lynn Dukes-Rimsky, Graeme Preston, Luckio F. Owuocha, Leila Sabbagh, Jehan Mousa, Christina Lam, Andrew Edmondson, Austin Larson, Matthew Schultz, Pieter Vermeersch, David Cassiman, Peter Witters, Lesa J. Beamer, Tamas Kozicz, Heather Flanagan-Steet, Bart Ghesquière, and Eva Morava. Tracer metabolomics reveals the role of aldose reductase in glycosylation. Cell Reports Medicine, 4:101056, Jun 2023. URL: https://doi.org/10.1016/j.xcrm.2023.101056, doi:10.1016/j.xcrm.2023.101056. This article has 33 citations and is from a peer-reviewed journal.
(tran2020secondgenerationpharmacologicalchaperones pages 1-3): My Lan Tran, Yves Génisson, Stéphanie Ballereau, and Cécile Dehoux. Second-generation pharmacological chaperones: beyond inhibitors. Molecules, 25:3145, Jul 2020. URL: https://doi.org/10.3390/molecules25143145, doi:10.3390/molecules25143145. This article has 72 citations.
(parrado2022dissectingthetranscriptional pages 1-2): Antonio Parrado, Gonzalo Rubio, Mercedes Serrano, María Eugenia De la Morena-Barrio, Salvador Ibáñez-Micó, Natalia Ruiz-Lafuente, Reinhard Schwartz-Albiez, Ana Esteve-Solé, Laia Alsina, Javier Corral, and Trinidad Hernández-Caselles. Dissecting the transcriptional program of phosphomannomutase 2-deficient cells: lymphoblastoide b cell lines as a valuable model for congenital disorders of glycosylation studies. Glycobiology, 32(2):84-100, Aug 2022. URL: https://doi.org/10.1093/glycob/cwab087, doi:10.1093/glycob/cwab087. This article has 8 citations and is from a peer-reviewed journal.
(peng2026pmm2interactswith pages 1-2): Zheng Peng, Bing Ma, Zhou Song, Yunshan Zhao, Yang Yang, Yong Liu, Chenggang Li, and Yong Zhang. Pmm2 interacts with trim28 to recruit e2f4 and promote kifc3-mediated tumor glycolysis and colorectal cancer progression. Oncogene, 45:1145-1160, Mar 2026. URL: https://doi.org/10.1038/s41388-026-03707-x, doi:10.1038/s41388-026-03707-x. This article has 0 citations and is from a domain leading peer-reviewed journal.
(sosicka2021chemicaltherapiesfor pages 1-3): Paulina Sosicka, Bobby G. Ng, and Hudson H. Freeze. Chemical therapies for congenital disorders of glycosylation. ACS chemical biology, 17:2962-2971, Nov 2021. URL: https://doi.org/10.1021/acschembio.1c00601, doi:10.1021/acschembio.1c00601. This article has 32 citations and is from a domain leading peer-reviewed journal.

📚 Additional Documentation

Notes

(PMM2-notes.md)

PMM2 (phosphomannomutase 2, O15305) — research notes

Summary

PMM2 is a cytosolic phosphomannomutase (EC 5.4.2.8) that catalyzes the reversible
isomerization of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P), the
second of two steps converting fructose 6-phosphate to Man1P en route to GDP-mannose
(UniPathway UPA00126, UER00424; step 2/2). Man1P is the precursor of GDP-mannose and
dolichol-phosphate-mannose, the activated mannose donors used in N-linked glycosylation
(LLO/dolichol pathway), O-mannosylation, C-mannosylation, and GPI-anchor synthesis.
Biallelic loss-of-function variants cause PMM2-CDG (CDG-Ia / Jaeken syndrome), the most
common congenital disorder of glycosylation.

Core molecular function and pathway

EC 5.4.2.8, Rhea RHEA:11140: alpha-D-mannose 1-phosphate = D-mannose 6-phosphate
[file:human/PMM2/PMM2-uniprot.txt "Reaction=alpha-D-mannose 1-phosphate = D-mannose 6-phosphate"].
Catalytic activity and kinetics directly measured: KM=16 uM for alpha-D-mannose
1-phosphate; KM=13.5 uM for alpha-D-glucose 1-phosphate
[file:human/PMM2/PMM2-uniprot.txt "KM=16 uM for alpha-D-mannose 1-phosphate"].
Note PMM also acts on glucose-1,6-bisphosphate-type substrates; the physiologically
relevant reaction is the mannose phosphomutase reaction supplying GDP-mannose.
Pathway placement: "Nucleotide-sugar biosynthesis; GDP-alpha-D-mannose biosynthesis;
alpha-D-mannose 1-phosphate from D-fructose 6-phosphate: step 2/2"
[file:human/PMM2/PMM2-uniprot.txt].
HAD (haloacid dehalogenase) superfamily / eukaryotic PMM family; two-domain cap+core
architecture; Asp nucleophile (Asp12 in mature human numbering / Asp19 in PMM1) and
Mg2+ cofactor PMID:16540464. UniProt: ACT_SITE 12 (nucleophile), ACT_SITE
14 (proton donor/acceptor), multiple Mg2+ binding residues, requires metal ion (KW
Magnesium, Metal-binding; GO:0046872 metal ion binding IEA-UniProtKB-KW).
Homodimer [file:human/PMM2/PMM2-uniprot.txt "SUBUNIT: Homodimer"].
Cytoplasmic/cytosolic localization [file:human/PMM2/PMM2-uniprot.txt
"SUBCELLULAR LOCATION: Cytoplasm"]; HPA IDA localizes to cytosol (GO_REF:0000052).

Structural / mechanistic literature

PMID:16540464 (Silvaggi et al. 2006, JBC): X-ray structures of human alpha-PMM1 with
and without bound substrate; HADSF cap/core domains; "The two isozymes, alpha-PMM1 and
alpha-PMM2, are shown to have a conserved active-site structure and to display similar
kinetic properties." Provides the EXP support for GO:0004615 phosphomannomutase activity
for PMM2 (also the structural basis of CDG-1a genotype-phenotype). Abstract foregrounds
PMM1, but PMM2 is explicitly assayed/compared; this is the ECO:0000269/EXP Reactome
annotation for PMM2 phosphomannomutase activity. Do NOT remove as "wrong gene."

Disease / CDG biology

PMID:9140401 (Matthijs et al. 1997, Nat Genet): cloned PMM2 on 16p13; 11 missense
mutations in 16 CDG1 patients with documented PMM deficiency; "Our results give
conclusive support to the biochemical finding that the phosphomannomutase deficiency is
the basis for CDG1." Establishes PMM2 = enzyme, role in GDP-mannose synthesis and
glycoprotein biosynthesis. This is the TAS source for GO:0004615, GO:0009298,
GO:0009101 PMID:9140401.
PMM2-CDG (CDG-Ia, MIM:212065): autosomal recessive multisystem disorder; severe
encephalopathy, axial hypotonia, abnormal eye movements, psychomotor retardation,
peripheral neuropathy, cerebellar hypoplasia, retinitis pigmentosa, inverted nipples,
abnormal fat distribution, coagulopathy [file:human/PMM2/PMM2-uniprot.txt DISEASE].
Underglycosylation of serum glycoproteins (hypoglycosylated transferrin) is the
diagnostic hallmark.
R141H is the most common allele, never homozygous (homozygosity lethal) — supports a
hypomorphic/balanced-genotype model [reactome/R-HSA-3781926.md "The R141H mutation is
never observed in the homozygous state"; file:human/PMM2/PMM2-uniprot.txt VARIANT 141
"homozygosis of this mutation is incompatible with life"].
Reactome R-HSA-446201 ("PMM1,2 isomerise Man6P to Man1P"), R-HSA-3781926 ("Defective
PMM2 does not isomerise Man6P to Man1P"), R-HSA-446205 ("Synthesis of GDP-mannose"),
R-HSA-4043911 ("Defective PMM2 causes PMM2-CDG").

PMID:9525984 — CAUTION: this is about CDG-Ib / PMI (MPI), NOT PMM2

The FlyBase IMP annotation of GDP-mannose biosynthetic process (GO:0009298) to PMM2
cites PMID:9525984. But that paper is about phosphomannose ISOMERASE (PMI/MPI) deficiency
causing CDG type Ib and mannose therapy — a DIFFERENT enzyme and a different gene
PMID:9525984. The cached publication is full-text-available (abstract+full text both
present and concordant — they describe PMI, not PMM2). PMM2 catalyzes the Man6P<->Man1P
mutase step, whereas PMI catalyzes F6P<->Man6P (the isomerase step). The IMP evidence in
this paper does not concern PMM2's catalytic step. The GO TERM (GDP-mannose biosynthetic
process) is still biologically correct for PMM2, and the function is independently and
strongly supported (EXP PMID:16540464, TAS PMID:9140401, IEA pathway). So the right
action is ACCEPT the term but flag the citation as MISCITED/likely wrong reference in
reference_review, rather than REMOVE the annotation. (Per CLAUDE.md: do not REMOVE an
experimental annotation merely on cached-text grounds; here we keep the term and document
the citation problem.)

Protein interactions (IPI / GO:0005515 protein binding)

Three IPI "protein binding" annotations from high-throughput binary interactome maps:
PMID:25416956 (Rolland et al. 2014, proteome-scale interactome), PMID:26871637 (Yang et
al. 2016, alternative-splicing interactome expansion), PMID:32296183 (Luck et al. 2020,
HuRI reference binary interactome). Partners listed in GOA: ACY3 (Q96HD9), MEOX2 (Q6FHY5),
SGK2 isoform (Q9HBY8-2). UniProt INTERACTION lists ACY3 (NbExp=9), MEOX2, SGK2.
These are generic Y2H/binary-map hits with no demonstrated functional/biological meaning
for a cytosolic metabolic enzyme. Per curation guidance, "protein binding" (GO:0005515)
is uninformative and should not be treated as a core molecular function. Mark as
over-annotated / non-core; no specific adapter or scaffold function is established.

Localization annotations

cytosol (GO:0005829): supported by HPA IDA (GO_REF:0000052), IBA, IEA, Reactome TAS, and
UniProt "Cytoplasm". Core. cytoplasm (GO:0005737) IEA is a correct but less precise parent.
neuronal cell body (GO:0043025): single Ensembl-orthology IEA (GO_REF:0000107) transferred
from mouse Q9Z2M7. PMM2 is a ubiquitously expressed cytosolic housekeeping enzyme (HPA:
"Low tissue specificity"). No evidence of a neuron-cell-body-specific function; this is a
by-product of where the enzyme happens to be detectable in a neuron and is an
over-annotation at the gene/function level. Keep as non-core at most; better treated as
over-annotated for a soluble cytosolic enzyme.

Process annotations: core vs downstream (substrate guilt-by-association)

CORE process: GDP-mannose biosynthetic process (GO:0009298) and mannose metabolic
process (GO:0006013) — these directly describe what the enzyme product (Man1P) feeds and
what the enzyme acts on. PMM2's reaction is the committed mutase step.
DOWNSTREAM processes: protein N-linked glycosylation (GO:0006487, IBA) and glycoprotein
biosynthetic process (GO:0009101, TAS PMID:9140401). PMM2 only SUPPLIES a precursor
(GDP-mannose) many steps upstream; it is not itself a glycosyltransferase and does not
act on protein substrates. By the GO "supplies precursor for" / substrate-guilt-by-
association pattern these are true at the whole-organism level (loss causes
hypoglycosylation) but are downstream consequences, not the molecular role of the enzyme.
Treat as KEEP_AS_NON_CORE (involved_in is defensible because biallelic loss disrupts
N-glycosylation, the basis of the disease) rather than core MF/BP. glycoprotein
biosynthetic process is the most general of these and the most over-annotated.

FlyBase orthology process terms (in UniProt DR but not in GOA core list)

GO:0061728 (GDP-mannose biosynthetic process from mannose) and GO:0061729 (GDP-D-mannose
biosynthetic process from fructose-6-phosphate), IMP from FlyBase — more specific children
of GO:0009298 that precisely describe the two GDP-mannose biosynthetic routes the PMM step
participates in. These are good candidate refinements but are not in the supplied GOA TSV
rows; noted for proposed refinement.

Conclusions for core_functions

molecular_function: GO:0004615 phosphomannomutase activity (enables); metal ion binding
(GO:0046872) is a cofactor requirement, not a standalone core function.
directly_involved_in: GO:0009298 GDP-mannose biosynthetic process; GO:0006013 mannose
metabolic process.
locations: GO:0005829 cytosol.
substrates: D-mannose 6-phosphate / alpha-D-mannose 1-phosphate (CHEBI:58735 / CHEBI:58409).

Falcon integration (2026-06-21)

Integrated findings from PMM2-deep-research-falcon.md (FutureHouse Falcon, 25 citations) into
the already-complete review. Conservative enrichment only — no action values were flipped.

References added (resolved to PMID, fetched into cache, with reference_review)

PMID:36214454 — Vignogna et al. 2022, eLife, "Evolutionary rescue of phosphomannomutase
deficiency in yeast models of human disease." Yeast SEC53/PMM2 model; corroborates that PMM2
loss-of-function impairs N-linked glycosylation and that p.Phe119Leu/F126L acts via a
dimer/stability defect PMID:36214454. relevance HIGH→set MEDIUM (model organism), correctness VERIFIED.
PMID:40572562 — Hay Mele et al. 2025, Molecules, "In Silico Analysis of Phosphomannomutase-2
Dimer Interface Stability and Heterodimerization with Phosphomannomutase-1." Supports the obligate
homodimer quaternary structure and the catalytic (R141H) vs interface (F119L) defect distinction
PMID:40572562; Man-1-P pivotal for
GDP-mannose/Dol-P-Man synthesis. relevance HIGH, correctness VERIFIED.
PMID:37257447 — Radenkovic et al. 2023, Cell Reports Medicine, "Tracer metabolomics reveals
the role of aldose reductase in glycosylation." Confirms PMM2 deficiency depletes GDP-mannose
PMID:37257447 and
explicitly states PMM2 is not directly involved in polyol metabolism. relevance MEDIUM,
correctness VERIFIED.

Annotation summaries enriched (no action changes)

GO:0006487 protein N-linked glycosylation (KEEP_AS_NON_CORE): added Vignogna 2022 corroboration
verbatim supported_by.
GO:0009298 GDP-mannose biosynthetic process (IEA, ACCEPT): added Radenkovic 2023 (GDP-mannose
depletion) and Mele 2025 (Man-1-P pivotal step) verbatim supported_by.
core_functions: description now notes the obligate-homodimer active form and the R141H (catalytic)
vs F119L (interface) defect distinction; added Mele 2025 to core_functions supported_by.

Falcon claims NOT added as citations / rejected (with reason)

Aldose-reductase / polyol "metabolic rewiring" as a PMM2 function — rejected as a PMM2
function. Radenkovic 2023 itself states PMM2 is not directly involved in polyol metabolism; this
is a downstream/compensatory effect. Used only to support the core GDP-mannose role and to justify
NOT annotating polyol metabolism.
PMM1/PMM2 heterodimer (Mele 2025) — computational only; PMM1 does not compensate for PMM2 in
vivo. Does NOT justify a PMM1-containing-complex annotation. Noted in reference_review only.
Edmondson et al. 2025 mouse model (neurodevelopmental/cerebellar/synaptic) — NOT added as a
citation. It is a bioRxiv preprint (doi:10.1101/2025.06.01.657261) without a PubMed PMID, so it is
not fetchable per the citation rule; findings are downstream organism-level phenotypes, not core
PMM2 function. Recorded here in notes only.
TNFR1/inflammatory signaling (Pascoal 2025), ER stress/UPR, mitochondrial dysfunction,
autophagy (Ligezka 2023, Parrado 2022), PMM2-TRIM28 colorectal-cancer interaction (Peng 2026)
— all pleiotropic/context-specific downstream consequences per Falcon's own risk assessment; not
core function, not added. Consistent with the existing review's scope.
Budhraja 2024, Garapati 2024, Mangione 2025, Tran 2020, Sosicka 2021, Pradeep 2023, Medico 2025
— biomarker/therapy/metabolomics/general-chaperone papers reinforcing already-captured core
conclusions; not added to avoid citation bloat (Tran 2020 = PMID:32660097 is a general chaperone
review, only MEDIUM relevance and not PMM2-specific).

Validation

uv run ai-gene-review validate genes/human/PMM2/PMM2-ai-review.yaml --terms → "Valid (with 1
warnings)"; the sole warning is advisory (no annotation cites the deep-research .md file). All
added supporting_text strings are verbatim substrings of the fetched cached publications; all new
reference titles match the fetched records. No ❌ ERROR.

📄 View Raw YAML

id: O15305
gene_symbol: PMM2
product_type: PROTEIN
status: DRAFT
taxon:
  id: NCBITaxon:9606
  label: Homo sapiens
description: >-
  PMM2 encodes phosphomannomutase 2 (EC 5.4.2.8), a cytosolic, Mg2+-dependent enzyme of
  the eukaryotic phosphomannomutase family within the haloacid dehalogenase (HAD)
  superfamily. It catalyzes the reversible isomerization of mannose 6-phosphate to
  mannose 1-phosphate, the committed second step in the conversion of fructose 6-phosphate
  to GDP-mannose. The mannose 1-phosphate produced is the precursor of GDP-mannose and
  dolichol-phosphate-mannose, the activated mannose donors used by the N-linked
  glycosylation (lipid-linked oligosaccharide/dolichol) pathway and other mannosyl-transfer
  reactions. PMM2 functions as a homodimer using an aspartate nucleophile and divalent
  magnesium for phosphoryl transfer, and is a soluble, ubiquitously expressed housekeeping
  enzyme. Because it supplies a key nucleotide-sugar precursor rather than acting directly
  on glycoprotein substrates, its loss disrupts glycosylation indirectly. Biallelic
  loss-of-function variants cause PMM2-CDG (congenital disorder of glycosylation type Ia,
  Jaeken syndrome), the most common congenital disorder of glycosylation, characterized by
  underglycosylation of serum glycoproteins and a multisystem, predominantly neurological,
  phenotype.
alternative_products:
  - name: "1"
    id: O15305-1
  - name: "2"
    id: O15305-2
    sequence_note: VSP_056228, VSP_056229
existing_annotations:
  - term:
      id: GO:0004615
      label: phosphomannomutase activity
    evidence_type: IBA
    original_reference_id: GO_REF:0000033
    qualifier: enables
    review:
      summary: >-
        Phylogenetically inferred molecular function that exactly matches the
        experimentally established catalytic activity of PMM2 (EC 5.4.2.8,
        Man6P <-> Man1P). This is the core molecular function of the gene product.
      action: ACCEPT
      reason: >-
        Core, well-supported molecular function; concordant with experimental
        (EXP/TAS) annotations and UniProt catalytic-activity curation.
      supported_by:
        - reference_id: file:human/PMM2/PMM2-uniprot.txt
          supporting_text: "Reaction=alpha-D-mannose 1-phosphate = D-mannose 6-phosphate"
  - term:
      id: GO:0006013
      label: mannose metabolic process
    evidence_type: IBA
    original_reference_id: GO_REF:0000033
    qualifier: involved_in
    review:
      summary: >-
        Correct, parent-level biological-process annotation. PMM2 acts on mannose
        phosphosugars, so participation in mannose metabolism is accurate, though
        the more informative process term is GDP-mannose biosynthetic process.
      action: ACCEPT
      reason: >-
        Accurate process annotation directly describing the metabolite class the
        enzyme acts on; retained as core (parent of the GDP-mannose biosynthetic term).
  - term:
      id: GO:0005829
      label: cytosol
    evidence_type: IBA
    original_reference_id: GO_REF:0000033
    qualifier: is_active_in
    review:
      summary: >-
        PMM2 is a soluble cytosolic enzyme; the Man6P<->Man1P reaction occurs in the
        cytosol. Consistent with UniProt (Cytoplasm), HPA IDA, and Reactome.
      action: ACCEPT
      reason: >-
        Core localization, supported by direct (IDA) and multiple inferred annotations.
      supported_by:
        - reference_id: file:human/PMM2/PMM2-uniprot.txt
          supporting_text: "SUBCELLULAR LOCATION: Cytoplasm"
  - term:
      id: GO:0006487
      label: protein N-linked glycosylation
    evidence_type: IBA
    original_reference_id: GO_REF:0000033
    qualifier: involved_in
    review:
      summary: >-
        PMM2 only supplies GDP-mannose, a nucleotide-sugar precursor used many steps
        upstream of N-glycosylation; it is not a glycosyltransferase and does not act on
        protein substrates. The 'involved_in' relationship is defensible because biallelic
        PMM2 loss causes protein hypoglycosylation (the basis of PMM2-CDG), but this is a
        downstream consequence (substrate guilt-by-association), not the enzyme's molecular
        role. Keep as a non-core process annotation. Independently corroborated in a yeast
        SEC53/PMM2 disease model, where PMM2 mutations are described as affecting protein
        N-linked glycosylation (PMID:36214454).
      action: KEEP_AS_NON_CORE
      reason: >-
        Downstream pathway annotation reflecting precursor supply, not the direct catalytic
        function; retained because loss-of-function disrupts N-glycosylation, but it is not
        a core function of this metabolic enzyme.
      supported_by:
        - reference_id: PMID:36214454
          supporting_text: "mutations in the phosphomannomutase gene PMM2, which affect protein N-linked"
  - term:
      id: GO:0004615
      label: phosphomannomutase activity
    evidence_type: IEA
    original_reference_id: GO_REF:0000120
    qualifier: enables
    review:
      summary: >-
        Electronic assertion of the core catalytic activity, consistent with the EC
        5.4.2.8 / Rhea mapping and InterPro PMM family signature. Redundant with the
        experimental and IBA annotations of the same term.
      action: ACCEPT
      reason: >-
        Correct core molecular function via independent electronic methods.
  - term:
      id: GO:0005737
      label: cytoplasm
    evidence_type: IEA
    original_reference_id: GO_REF:0000120
    qualifier: located_in
    review:
      summary: >-
        Correct but less specific parent of cytosol. The cytosol (GO:0005829) annotation
        is the more informative and directly supported localization.
      action: MODIFY
      reason: >-
        Generalizes the supported cytosol localization; replace with the more specific term.
      proposed_replacement_terms:
        - id: GO:0005829
          label: cytosol
  - term:
      id: GO:0006013
      label: mannose metabolic process
    evidence_type: IEA
    original_reference_id: GO_REF:0000117
    qualifier: involved_in
    review:
      summary: >-
        Electronic duplicate of the IBA mannose metabolic process annotation; correct.
      action: ACCEPT
      reason: >-
        Accurate process annotation via ARBA model; concordant with the IBA annotation.
  - term:
      id: GO:0009298
      label: GDP-mannose biosynthetic process
    evidence_type: IEA
    original_reference_id: GO_REF:0000120
    qualifier: involved_in
    review:
      summary: >-
        Core biological process: the PMM step (Man6P->Man1P) is step 2/2 in the
        UniPathway route from fructose 6-phosphate to Man1P feeding GDP-mannose synthesis.
        Concordant with experimental, TAS, and IMP annotations. Functionally corroborated
        by metabolic tracer work showing PMM2 deficiency depletes GDP-mannose (PMID:37257447)
        and by structural work noting that Man-1-P formation is a pivotal step in GDP-mannose
        and dolichol-phosphate-mannose biosynthesis (PMID:40572562).
      action: ACCEPT
      reason: >-
        Core process directly describing the metabolic pathway the enzyme commits to.
      supported_by:
        - reference_id: file:human/PMM2/PMM2-uniprot.txt
          supporting_text: "GDP-alpha-D-mannose"
        - reference_id: PMID:37257447
          supporting_text: "caused by PMM2 deficiency, presents with depleted GDP-mannose and abnormal"
        - reference_id: PMID:40572562
          supporting_text: "The formation of Man-1-P is a pivotal step in the biosynthesis of GDP-mannose"
  - term:
      id: GO:0005515
      label: protein binding
    evidence_type: IPI
    original_reference_id: PMID:25416956
    qualifier: enables
    review:
      summary: >-
        Generic 'protein binding' from a proteome-scale binary (Y2H) interactome map.
        The interaction was experimentally detected (IPI), so it is real, but the generic
        GO:0005515 term carries no specific functional meaning for a cytosolic metabolic
        enzyme and does not identify an adapter/scaffold function. This is an
        uninformative over-annotation, not an incorrect one.
      action: MARK_AS_OVER_ANNOTATED
      reason: >-
        Experimentally detected but uninformative high-throughput 'protein binding'
        annotation; the interaction is real but the generic term provides no functional
        information about PMM2.
  - term:
      id: GO:0005515
      label: protein binding
    evidence_type: IPI
    original_reference_id: PMID:26871637
    qualifier: enables
    review:
      summary: >-
        Generic 'protein binding' from a high-throughput alternative-splicing interactome
        study. The interaction was experimentally detected (IPI) and is real, but as above
        the generic term is uninformative for this metabolic enzyme and not a core function.
      action: MARK_AS_OVER_ANNOTATED
      reason: >-
        Experimentally detected but uninformative high-throughput 'protein binding'
        annotation; real interaction but no specific function established.
  - term:
      id: GO:0005515
      label: protein binding
    evidence_type: IPI
    original_reference_id: PMID:32296183
    qualifier: enables
    review:
      summary: >-
        Generic 'protein binding' from the HuRI reference binary interactome. The recorded
        partners (e.g. ACY3, MEOX2, SGK2 isoform) were experimentally detected (IPI) and are
        real interactions, but they are not connected to PMM2's catalytic role and the generic
        term provides no functional insight. Uninformative over-annotation rather than incorrect.
      action: MARK_AS_OVER_ANNOTATED
      reason: >-
        Experimentally detected but uninformative high-throughput 'protein binding'
        annotation; real interactions but no specific function established for a cytosolic
        phosphomannomutase.
  - term:
      id: GO:0005829
      label: cytosol
    evidence_type: IEA
    original_reference_id: GO_REF:0000120
    qualifier: is_active_in
    review:
      summary: >-
        Electronic duplicate of the well-supported cytosol localization. Correct.
      action: ACCEPT
      reason: >-
        Core localization confirmed by multiple independent lines of evidence.
  - term:
      id: GO:0043025
      label: neuronal cell body
    evidence_type: IEA
    original_reference_id: GO_REF:0000107
    qualifier: located_in
    review:
      summary: >-
        Single Ensembl-orthology electronic annotation transferred from mouse. PMM2 is a
        ubiquitously expressed soluble cytosolic housekeeping enzyme (HPA: low tissue
        specificity); detection in a neuronal cell body reflects where the cytosol of that
        cell is, not a neuron-specific function or a distinct localization. Over-annotation
        for a soluble metabolic enzyme.
      action: MARK_AS_OVER_ANNOTATED
      reason: >-
        Orthology-transferred, cell-type-specific component annotation that adds no
        functional information beyond the general cytosolic localization and risks implying
        a neuron-restricted role.
  - term:
      id: GO:0009298
      label: GDP-mannose biosynthetic process
    evidence_type: TAS
    original_reference_id: Reactome:R-HSA-446205
    qualifier: involved_in
    review:
      summary: >-
        Reactome-curated involvement in GDP-mannose synthesis, matching the enzyme's
        committed pathway step. Core process annotation.
      action: ACCEPT
      reason: >-
        Author-curated (TAS) support for the core GDP-mannose biosynthetic process role.
      supported_by:
        - reference_id: Reactome:R-HSA-446205
          supporting_text: GDP-mannose is the mannose donor for the first 5 mannose addition reactions
  - term:
      id: GO:0004615
      label: phosphomannomutase activity
    evidence_type: EXP
    original_reference_id: PMID:16540464
    qualifier: enables
    review:
      summary: >-
        Experimental support for phosphomannomutase activity. Although the abstract
        foregrounds the PMM1 crystal structures, the study explicitly compares both
        isozymes and shows alpha-PMM1 and alpha-PMM2 have a conserved active site and
        similar kinetic properties; UniProt assigns the EC 5.4.2.8 catalytic-activity
        ECO:0000269 evidence to this paper for PMM2. This is the primary experimental
        anchor of the core molecular function.
      action: ACCEPT
      reason: >-
        Direct experimental evidence for the core catalytic activity; the defining function
        of the gene product.
      supported_by:
        - reference_id: PMID:16540464
          supporting_text: are shown to have a conserved active-site structure and to display similar kinetic properties
  - term:
      id: GO:0004615
      label: phosphomannomutase activity
    evidence_type: TAS
    original_reference_id: Reactome:R-HSA-3781926
    qualifier: enables
    review:
      summary: >-
        Reactome-curated phosphomannomutase activity (Man6P->Man1P) for PMM2. Concordant
        with the experimental and IBA/IEA annotations of the same term.
      action: ACCEPT
      reason: >-
        Author-curated support for the core molecular function.
      supported_by:
        - reference_id: Reactome:R-HSA-3781926
          supporting_text: PMM2) catalyses the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P)
  - term:
      id: GO:0005829
      label: cytosol
    evidence_type: IDA
    original_reference_id: GO_REF:0000052
    qualifier: located_in
    review:
      summary: >-
        Direct immunofluorescence localization (HPA) to the cytosol. Strongest evidence
        for the core cytosolic localization.
      action: ACCEPT
      reason: >-
        Direct experimental localization supporting the core cytosolic component annotation.
  - term:
      id: GO:0009298
      label: GDP-mannose biosynthetic process
    evidence_type: IMP
    original_reference_id: PMID:9525984
    qualifier: involved_in
    review:
      summary: >-
        The GO term (GDP-mannose biosynthetic process) is biologically correct for PMM2 and
        is independently and strongly supported (EXP PMID:16540464; TAS PMID:9140401 and
        Reactome; IEA pathway). However, the cited reference PMID:9525984 is about
        phosphomannose ISOMERASE (PMI/MPI) deficiency causing CDG type Ib and mannose
        therapy - a different enzyme (F6P<->Man6P) and a different gene - and does not study
        PMM2's mutase step. The annotation is therefore kept (the term is right and is a core
        process) but the citation appears mis-attributed; this is recorded in the reference
        review rather than removing an experimental-coded annotation.
      action: ACCEPT
      reason: >-
        Core process annotation; term is correct and well supported by other evidence. The
        specific PMID:9525984 citation is flagged as likely mis-attributed (PMI/CDG-Ib paper)
        in reference_review, but the GO term itself is retained.
  - term:
      id: GO:0005829
      label: cytosol
    evidence_type: TAS
    original_reference_id: Reactome:R-HSA-3781926
    qualifier: located_in
    review:
      summary: >-
        Reactome-curated cytosolic localization. Concordant with IDA/IBA/IEA evidence.
      action: ACCEPT
      reason: >-
        Author-curated support for the core cytosolic localization.
  - term:
      id: GO:0005829
      label: cytosol
    evidence_type: TAS
    original_reference_id: Reactome:R-HSA-446201
    qualifier: located_in
    review:
      summary: >-
        Reactome-curated cytosolic localization (PMM1,2 isomerise Man6P to Man1P).
        Concordant with all other localization evidence.
      action: ACCEPT
      reason: >-
        Author-curated support for the core cytosolic localization.
      supported_by:
        - reference_id: Reactome:R-HSA-446201
          supporting_text: Cytosolic phosphomannomutases 1 and 2 (PMM1 and PMM2) catalyse the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P)
  - term:
      id: GO:0004615
      label: phosphomannomutase activity
    evidence_type: TAS
    original_reference_id: PMID:9140401
    qualifier: enables
    review:
      summary: >-
        The original cloning paper identifies PMM2 as a phosphomannomutase whose deficiency
        underlies CDG1. Author-stated (TAS) support for the core catalytic activity.
      action: ACCEPT
      reason: >-
        Foundational TAS support for the core molecular function.
      supported_by:
        - reference_id: PMID:9140401
          supporting_text: "phosphomannomutase (PMM)8, an enzyme necessary for"
  - term:
      id: GO:0009101
      label: glycoprotein biosynthetic process
    evidence_type: TAS
    original_reference_id: PMID:9140401
    qualifier: involved_in
    review:
      summary: >-
        Broad downstream process. PMM2 contributes to glycoprotein biosynthesis only by
        supplying GDP-mannose; it is not itself a glycosyltransferase. This is the most
        general of the downstream-glycosylation annotations and reflects the disease
        phenotype (CDG) rather than the enzyme's molecular role. Keep as non-core.
      action: KEEP_AS_NON_CORE
      reason: >-
        General downstream-pathway annotation by precursor supply; true at the
        organismal/disease level but not a core function of this metabolic enzyme.
  - term:
      id: GO:0009298
      label: GDP-mannose biosynthetic process
    evidence_type: TAS
    original_reference_id: PMID:9140401
    qualifier: involved_in
    review:
      summary: >-
        Author-stated (TAS) support for the core GDP-mannose biosynthetic process role,
        from the gene's foundational characterization paper.
      action: ACCEPT
      reason: >-
        Core process annotation supported by the original characterization of PMM2.
      supported_by:
        - reference_id: PMID:9140401
          supporting_text: "phosphomannomutase (PMM)8, an enzyme necessary for"
core_functions:
  - description: >-
      Cytosolic, Mg2+-dependent phosphomannomutase that reversibly isomerizes
      D-mannose 6-phosphate to alpha-D-mannose 1-phosphate, the committed second step
      supplying mannose 1-phosphate for GDP-mannose (and dolichol-phosphate-mannose)
      biosynthesis. This nucleotide-sugar precursor supply is the prerequisite for
      protein N-linked glycosylation, but the enzyme itself acts only on phosphosugar
      metabolites. The active enzyme is an obligate homodimer; catalytic-site variants
      (e.g. p.Arg141His) impair activity while dimer-interface variants (e.g. p.Phe119Leu)
      cause loss of function by disrupting homodimer formation.
    molecular_function:
      id: GO:0004615
      label: phosphomannomutase activity
    directly_involved_in:
      - id: GO:0009298
        label: GDP-mannose biosynthetic process
      - id: GO:0006013
        label: mannose metabolic process
    locations:
      - id: GO:0005829
        label: cytosol
    substrates:
      - id: CHEBI:58735
        label: D-mannopyranose 6-phosphate(2-)
      - id: CHEBI:58409
        label: alpha-D-mannose 1-phosphate(2-)
    supported_by:
      - reference_id: PMID:16540464
        supporting_text: are shown to have a conserved active-site structure and to display similar kinetic properties
      - reference_id: file:human/PMM2/PMM2-uniprot.txt
        supporting_text: "Reaction=alpha-D-mannose 1-phosphate = D-mannose 6-phosphate"
      - reference_id: Reactome:R-HSA-446205
        supporting_text: GDP-mannose is the mannose donor for the first 5 mannose addition reactions
      - reference_id: PMID:40572562
        supporting_text: "Structurally, PMM2 functions as an obligate homodimer"
proposed_new_terms:
  - proposed_name: phosphomannomutase activity involved in GDP-mannose biosynthetic process
    proposed_definition: >-
      Any phosphomannomutase activity (EC 5.4.2.8; interconversion of D-mannose
      6-phosphate and alpha-D-mannose 1-phosphate) that is a step in the GDP-mannose
      biosynthetic process. Captures the precursor-supply role of phosphomannomutase 2
      (PMM2) and its orthologs within nucleotide-sugar biosynthesis, distinguishing the
      committed mutase step from downstream mannosyl-transfer functions.
    justification: >-
      Existing annotations conflate the direct molecular function (phosphomannomutase
      activity) with downstream processes (N-linked glycosylation, glycoprotein
      biosynthesis). A function-in-process term would let curators record PMM2's role with
      the correct granularity and avoid substrate guilt-by-association over-annotation.
    proposed_parent:
      id: GO:0004615
      label: phosphomannomutase activity
suggested_questions:
  - question: >-
      Beyond supplying GDP-mannose, does PMM2 have any moonlighting or regulatory role
      (e.g. metabolite sensing, complex membership) that would justify any of its
      high-throughput binary protein-interaction hits?
  - question: >-
      Does PMM2 contribute to the cytosolic pool balance between mannose-1-phosphate and
      glucose-1-phosphate handling in vivo, given its measurable activity on glucose
      1-phosphate, and does this affect interpretation of residual-activity CDG genotypes?
suggested_experiments:
  - hypothesis: >-
      PMM2's only biologically relevant function is to supply Man1P/GDP-mannose; its
      reported binary protein interactions are not functionally meaningful.
    description: >-
      Perform affinity-purification mass spectrometry (AP-MS) on endogenously tagged PMM2
      in human cells under native conditions and compare to the binary Y2H hits; test
      whether candidate partners (ACY3, MEOX2, SGK2) co-purify and whether their knockdown
      alters PMM2 activity or GDP-mannose levels.
    experiment_type: AP-MS / interaction validation
  - hypothesis: >-
      CDG-Ia phenotype severity tracks quantitatively with residual cytosolic GDP-mannose
      flux through the PMM2 step rather than with any non-catalytic property.
    description: >-
      Reconstitute patient PMM2 missense variants in PMM2-null cells and measure enzyme
      kinetics, steady-state GDP-mannose and Dol-P-Man pools, and lipid-linked
      oligosaccharide profiles; correlate with clinical severity to test the
      flux-limitation model.
    experiment_type: variant reconstitution / metabolic flux analysis
references:
  - id: GO_REF:0000033
    title: Annotation inferences using phylogenetic trees
    findings: []
  - id: GO_REF:0000052
    title: Gene Ontology annotation based on curation of immunofluorescence data
    findings: []
  - id: GO_REF:0000107
    title: Automatic transfer of experimentally verified manual GO annotation data to
      orthologs using Ensembl Compara
    findings: []
  - id: GO_REF:0000117
    title: Electronic Gene Ontology annotations created by ARBA machine learning models
    findings: []
  - id: GO_REF:0000120
    title: Combined Automated Annotation using Multiple IEA Methods
    findings: []
  - id: PMID:16540464
    title: The X-ray crystal structures of human alpha-phosphomannomutase 1 reveal the
      structural basis of congenital disorder of glycosylation type 1a.
    findings:
      - statement: >-
          The two isozymes alpha-PMM1 and alpha-PMM2 share a conserved active-site
          structure and similar kinetic properties; analysis of mutation sites explains
          the genotype-phenotype relationship of CDG-1a.
        supporting_text: The two isozymes, alpha-PMM1 and alpha-PMM2, are shown to have a conserved active-site structure and to display similar kinetic properties.
        reference_section_type: ABSTRACT
    reference_review:
      relevance: HIGH
      correctness: VERIFIED
      review_notes: >-
        PubMed-verified structural/kinetic study. Abstract foregrounds PMM1 crystal
        structures but explicitly compares both isozymes and underpins the EC 5.4.2.8
        ECO:0000269 evidence UniProt assigns to PMM2; per guidelines not removed as
        'wrong gene'.
  - id: PMID:25416956
    title: A proteome-scale map of the human interactome network.
    findings: []
    reference_review:
      relevance: LOW
      correctness: LOW_QUALITY
      review_notes: >-
        High-throughput binary interactome screen; supports only a generic, uninformative
        'protein binding' annotation with no demonstrated functional relevance to PMM2.
  - id: PMID:26871637
    title: Widespread Expansion of Protein Interaction Capabilities by Alternative Splicing.
    findings: []
    reference_review:
      relevance: LOW
      correctness: LOW_QUALITY
      review_notes: >-
        High-throughput interactome screen; supports only a generic 'protein binding'
        annotation, not a core function.
  - id: PMID:32296183
    title: A reference map of the human binary protein interactome.
    findings: []
    reference_review:
      relevance: LOW
      correctness: LOW_QUALITY
      review_notes: >-
        HuRI reference binary interactome; generic 'protein binding' hits with no
        functional meaning for a cytosolic metabolic enzyme.
  - id: PMID:36214454
    title: Evolutionary rescue of phosphomannomutase deficiency in yeast models of
      human disease.
    findings:
      - statement: >-
          The most common cause of human CDG are mutations in PMM2 that impair protein
          N-linked glycosylation; the yeast SEC53 gene is a homolog of human PMM2, and
          disease-associated alleles (including F126L, equivalent to human p.Phe119Leu)
          cause a slow-growth/glycosylation phenotype that can be evolutionarily rescued.
        supporting_text: "mutations in the phosphomannomutase gene PMM2, which affect protein N-linked"
        reference_section_type: ABSTRACT
    reference_review:
      relevance: MEDIUM
      correctness: VERIFIED
      review_notes: >-
        PubMed-verified (PMID:36214454, eLife 2022). Yeast SEC53/PMM2 model; corroborates
        that PMM2 loss-of-function impairs N-linked glycosylation and that the common
        p.Phe119Leu/F126L allele acts via dimer/stability defects. Added by Falcon
        integration to strengthen the N-glycosylation (non-core) and homodimer evidence.
  - id: PMID:40572562
    title: In Silico Analysis of Phosphomannomutase-2 Dimer Interface Stability and
      Heterodimerization with Phosphomannomutase-1.
    findings:
      - statement: >-
          PMM2 functions as an obligate homodimer; the catalytic p.Arg141His variant
          impairs activity whereas the interface variant p.Phe119Leu disrupts homodimer
          formation, leading to destabilization/degradation. Man-1-P formation is a pivotal
          step in GDP-mannose and dolichol-phosphate-mannose biosynthesis. PMM1/PMM2
          heterodimers are predicted to be structurally plausible but PMM1 does not
          compensate for PMM2 deficiency in vivo.
        supporting_text: "Structurally, PMM2 functions as an obligate homodimer"
        reference_section_type: INTRODUCTION
    reference_review:
      relevance: HIGH
      correctness: VERIFIED
      review_notes: >-
        PubMed-verified (PMID:40572562, Molecules 2025). In silico/structural study;
        supports the homodimer quaternary structure asserted in core_functions and the
        catalytic- vs interface-defect distinction (R141H vs F119L). Predicted PMM1/PMM2
        heterodimer is computational only and PMM1 does not compensate in vivo, so it does
        NOT justify a PMM1-containing-complex annotation. Added by Falcon integration.
  - id: PMID:37257447
    title: Tracer metabolomics reveals the role of aldose reductase in glycosylation.
    findings:
      - statement: >-
          PMM2-CDG (caused by PMM2 deficiency) presents with depleted GDP-mannose and
          abnormal glycosylation; PMM2 itself is not directly involved in polyol metabolism.
          Aldose-reductase inhibition (epalrestat) diverts glucose flux toward sugar
          nucleotide synthesis, increasing GDP-mannose and improving glycosylation.
        supporting_text: "Considering that the PMM2 enzyme is not"
        reference_section_type: ABSTRACT
    reference_review:
      relevance: MEDIUM
      correctness: VERIFIED
      review_notes: >-
        PubMed-verified (PMID:37257447, Cell Rep Med 2023). Confirms PMM2 deficiency
        depletes GDP-mannose (supports the core GDP-mannose biosynthetic process role) and
        explicitly states PMM2 is NOT directly involved in polyol metabolism, supporting the
        decision NOT to annotate polyol/metabolic-rewiring as a PMM2 function. The
        aldose-reductase axis is a downstream/compensatory metabolic effect, not a PMM2
        molecular function. Added by Falcon integration.
  - id: PMID:9140401
    title: Mutations in PMM2, a phosphomannomutase gene on chromosome 16p13, in carbohydrate-deficient
      glycoprotein type I syndrome (Jaeken syndrome).
    findings:
      - statement: >-
          PMM2 on 16p13 encodes a phosphomannomutase necessary for GDP-mannose synthesis;
          missense mutations cause CDG type I (Jaeken syndrome), establishing PMM deficiency
          as the basis of the disease.
        supporting_text: phosphomannomutase (PMM)8, an enzyme necessary for
        reference_section_type: ABSTRACT
    reference_review:
      relevance: HIGH
      correctness: VERIFIED
      review_notes: >-
        Foundational PMM2 cloning/disease paper; PubMed-verified. Anchors the gene's
        catalytic identity and disease association.
  - id: PMID:9525984
    title: Carbohydrate-deficient glycoprotein syndrome type Ib. Phosphomannose isomerase
      deficiency and mannose therapy.
    findings:
      - statement: >-
          Describes phosphomannose ISOMERASE (PMI/MPI) deficiency causing CDG type Ib and
          mannose therapy - a different enzyme and gene than PMM2.
        supporting_text: Phosphomannose isomerase (PMI) deficiency is the cause of a new type of carbohydrate-deficient glycoprotein syndrome
        reference_section_type: ABSTRACT
    reference_review:
      relevance: LOW
      correctness: MISCITED
      review_notes: >-
        Cited as the IMP source for PMM2 GDP-mannose biosynthetic process, but the paper is
        about phosphomannose isomerase (PMI/MPI) deficiency / CDG-Ib and mannose therapy -
        a different enzyme catalyzing F6P<->Man6P and a different gene (MPI). The GO term is
        still correct for PMM2 and is strongly supported by other references, so the
        annotation was kept; the citation itself appears mis-attributed. Recommendation:
        FlyBase should re-examine and correct this IMP annotation, since its supporting
        evidence derives from a different enzyme (PMI/MPI, CDG-Ib) and does not assay the
        PMM2 mutase step.
  - id: Reactome:R-HSA-3781926
    title: Defective PMM2 does not isomerise Man6P to Man1P
    findings:
      - statement: >-
          PMM2 catalyses the cytosolic isomerization of Man6P to Man1P; Man1P is the
          precursor of GDP-mannose and dolichol-phosphate-mannose required for
          N-glycosylation. PMM2 mutations cause PMM2-CDG (CDG-1a).
        supporting_text: catalyses the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P) in the cytosol of cells
        reference_section_type: DATABASE_ENTRY
    reference_review:
      relevance: HIGH
      correctness: VERIFIED
      review_notes: Reactome-curated reaction; consistent with UniProt and primary literature.
  - id: Reactome:R-HSA-446201
    title: PMM1,2 isomerise Man6P to Man1P
    findings:
      - statement: >-
          Cytosolic PMM1 and PMM2 catalyze the Man6P->Man1P isomerization; PMM2 mutations
          cause Jaeken syndrome (CDG-Ia).
        supporting_text: Cytosolic phosphomannomutases 1 and 2 (PMM1 and PMM2) catalyse the isomerisation of mannose 6-phosphate (Man6P) to mannose 1-phosphate (Man1P)
        reference_section_type: DATABASE_ENTRY
    reference_review:
      relevance: HIGH
      correctness: VERIFIED
      review_notes: Reactome-curated; supports cytosolic localization and catalytic activity.
  - id: Reactome:R-HSA-446205
    title: Synthesis of GDP-mannose
    findings:
      - statement: >-
          GDP-mannose, the mannose donor for early N-glycan and Dol-P-Man synthesis, is made
          from fructose 6-phosphate and GTP in three steps (PMM2 catalyzes the mutase step).
        supporting_text: GDP-mannose is the mannose donor for the first 5 mannose addition reactions
        reference_section_type: DATABASE_ENTRY
    reference_review:
      relevance: HIGH
      correctness: VERIFIED
      review_notes: Reactome-curated pathway context for the GDP-mannose biosynthetic process annotation.

PMM2

Gene Description

Proposed New Ontology Terms

phosphomannomutase activity involved in GDP-mannose biosynthetic process

Existing Annotations Review

Core Functions

References

Suggested Questions for Experts

Suggested Experiments

Deep Research

Falcon

Gene Research for GO Annotation Review

Target

UniProt Context

Research Objective

Output Requirements

PMM2 Gene Research Report for GO Annotation Review

Executive Summary

1. Molecular Function

1.1 Core Biochemical Activity and Substrate Specificity

1.2 Protein Activation and Maturation Mechanisms

1.3 Quaternary Structure: Obligate Homodimer

2. Biological Process Roles

2.1 Core Function: GDP-Mannose Biosynthesis and N-Glycosylation

2.2 Neurodevelopmental and Cerebellar Roles

2.3 Secondary and Pleiotropic Effects: ER Stress, Mitochondrial Dysfunction, Autophagy, and Inflammation

2.4 Evidence Assessment for Specific Processes

3. Cellular Localization and Complexes

3.1 Subcellular Localization: Cytoplasm/Cytosol

3.2 Protein Complexes: Homodimer

4. Annotation-Risk Assessment

4.1 Core Functions (Strongly Supported, Low Risk)

4.2 Context-Dependent Annotations (Moderate Risk)

4.3 Pleiotropic/Secondary Effects (High Annotation Risk)

4.4 Insufficient Evidence (Do Not Annotate)

5. Key Literature

5.1 Recent Reviews and Clinical Studies (2022-2025)

5.2 Biochemical and Therapeutic Studies (2022-2024)

5.3 Mechanistic and Model Studies (2020-2022)

📚 Additional Documentation

Notes

PMM2 (phosphomannomutase 2, O15305) — research notes

Summary

Core molecular function and pathway

Structural / mechanistic literature

Disease / CDG biology

PMID:9525984 — CAUTION: this is about CDG-Ib / PMI (MPI), NOT PMM2

Protein interactions (IPI / GO:0005515 protein binding)

Localization annotations

Process annotations: core vs downstream (substrate guilt-by-association)

FlyBase orthology process terms (in UniProt DR but not in GOA core list)

Conclusions for core_functions

Falcon integration (2026-06-21)

References added (resolved to PMID, fetched into cache, with reference_review)

Annotation summaries enriched (no action changes)

Falcon claims NOT added as citations / rejected (with reason)

Validation

📄 View Raw YAML