CPT2

Summary: Conserved mitochondrial localization with a resolved ancestral source.

Reason: The actual PTHR22589 tree places human P23786 beneath PTN000501944, which carries the mitochondrial IBD. The exact 658-residue target alignment and descendant evidence agree with human mitochondrial imaging, MitoCoP detection and inner-membrane enzyme characterization. The two other mitochondrial-loss records are off the target path; neither donor count nor the target itself occurring among descendants invalidates this IBA. Retain the sound broad IBA location without asserting that every cellular pool is mitochondrial.

Summary: The ancestral palmitoyltransferase activity is retained by human CPT2.

Reason: PTN000501944 is an actual target-path IBD for GO:0004095. Human CPT2 assays demonstrate palmitoylcarnitine formation from palmitoyl-CoA and carnitine; the reversible chemistry also supports physiological regeneration of matrix acyl-CoA. Current mouse and rat experimental records and human target evidence provide the descendant grounding; an off-path loss does not exclude human CPT2.

Summary: CPT2 performs the acyl-transfer step feeding mitochondrial fatty acid beta-oxidation.

Reason: The target descends from the actual fatty-acid-beta-oxidation IBD at PTN000501944. CPT2 supplies matrix acyl-CoA through its own chemistry rather than merely being a consumed substrate. Human enzyme evidence and mouse loss-of-function flux experiments support participation. The fly donor PMID:25588812 has an Editorial Expression of Concern (PMID:32737422): stocks and construct maps were not retained for repeat work. This limits that evidence without establishing a false result or erasing the independent human and mammalian grounding. Current rat expression records are context evidence, not purified-enzyme assays.

Summary: The combined electronic mapping supports the established palmitoyltransferase reaction.

Reason: The current source tuple combines ARBA00087538, mouse P52825/ENSMUSP00000030345, RHEA:12661 and EC:2.3.1.21. Current ARBA branch 1 matches all three target FunFams. The mouse record includes an experimentally supported CPT2 activity, and the literal UniProt reaction plus human assays establish the same palmitoyl-transfer chemistry. The historical row omitted WITH/FROM; the complete current tuple is documented separately rather than inserted into it.

Summary: CPT2 is associated with the mitochondrial inner membrane on its matrix side.

Reason: The current UniProt vocabulary source is SL-0168, with peripheral-membrane and matrix-side qualifiers in the curated protein record. Human purification/cloning and the exact MitoCoP target row support this location; the latter classifies CPT2 as peripheral. No transmembrane pore or obligatory transmembrane helix is inferred.

Summary: Human CPT2 has measured octanoyltransferase activity.

Reason: The current electronic source RHEA:17177 is the octanoyl-CoA/carnitine reaction, also explicitly recorded for human CPT2 with PMID:20538056 in UniProt and in the current GO definition. The genuine human-ORF yeast assay measures C8 acylcarnitine formation. This is one substrate activity of the same enzyme, not a separate isoform or second core.

Summary: Refine transport participation to the enzymatic carnitine-shuttle step.

Reason: Current ARBA00088749 branch 1 matches the actual target three-FunFam combination. GO:0006853 is a typed child of GO:0015909 and explicitly includes transferring acyl groups to and from carnitine. CPT2 performs that chemistry, while SLC25A20 carries the exchange across the membrane. The refinement specifies its transport-pathway contribution; being an enzyme does not contradict the broader process term.

Proposed replacements: carnitine shuttle

Summary: Refine the broad acyltransferase mapping to carnitine O-acyltransferase.

Reason: IPR000542 genuinely maps to broad acyltransferase activity. Human CPT2 transfers several acyl groups to carnitine, so GO:0016406 is a supported class refinement. It preserves the demonstrated substrate breadth instead of representing the enzyme only by its C16 activity; palmitoyl and octanoyl activities already have separate assertions.

Proposed replacements: carnitine O-acyltransferase activity

Summary: Retain the positive BioPlex CPT2–MCUR1 association as secondary binding.

Reason: The exact PMID:28514442 IntAct record EBI-21858544 reports C-terminally tagged MCUR1/Q96AQ8 as bait and CPT2/P23786 as prey in HEK293T anti-tag co-immunoprecipitation/AP-MS. Both reference sequence checksums reconstruct. This is positive co-complex association, without a resolved contact surface or evidence that CPT2 regulates the calcium uniporter. MCUR1 is a uniporter regulator, not itself the calcium channel; no warranted partner-class binding refinement is established here. Retain generic binding under the standing campaign policy.

Summary: Preserve the retired historical HuRI aggregate and its positive binding scope.

Reason: The old source row lacks partner identities and is ambiguous between two current HuRI assertions. It is retained as retired provenance, with the literal CYSRT1 and OTX1 current source objects appended separately. Actual positive human-protein two-hybrid records support binding; a blanket overannotation judgment based only on the generic label is not justified. The OTX1-specific refinement belongs to that separate current row, not to every partner in this old aggregate.

Summary: Retain the positive 2021 CPT2–MCUR1 co-association at its measured scope.

Reason: The actual PMID:33961781 record EBI-54762176 is a HEK293T anti-tag co-immunoprecipitation association with MCUR1 bait and CPT2 prey, not a target experiment in both screened cell lines. It preserves a C-terminal tandem-tag feature. The selected paper text supplies the AP-MS study context; target details come from the literal curated record. No direct-contact, endogenous calcium-regulatory function or channel-binding class is inferred for CPT2. Generic binding remains non-core.

Summary: Matrix-side catalysis supports the matrix localization transfer.

Reason: The current mouse P52825 source has an IMP matrix annotation tied to PMID:28916721, with the Compara protein identifier also retained in the current tuple. Human CPT2 acts on the matrix side of the inner membrane and supplies matrix acyl-CoA. This supports the location without claiming a freely soluble-only pool or deriving biology from the ordinary relationship qualifier. The donor conditional muscle experiment does not itself provide a purified human subcompartment image.

Summary: CPT2 catalyzes the terminal acyl-transfer step of the carnitine shuttle.

Reason: The mouse donor has the exact GO:0006853 IMP record with PMID:28916721. Its conditional muscle loss reduces long-chain fatty acid oxidation, while the target human activity directly supplies the acyl-CoA-forming reaction. The transfer agrees with target enzyme chemistry and the actual Reactome catalyst path, not merely a generic requirement for energy metabolism.

Summary: A source-specific mouse adipose phenotype supports contextual thermogenesis.

Reason: PMID:25578732 directly studies adipose-specific Cpt2 deletion in mice: cold-challenged animals become hypothermic and the normal brown-fat thermogenic transcriptional response is impaired. CPT2 contributes through fatty-acid utilization and its metabolic context; it is not assigned kinase or transcription-factor activity. This supports a contextual thermogenic contribution, while a corresponding human adipose perturbation assay was not read. Keep the existing ortholog transfer as non-core.

Summary: The curated shuttle contains an explicit CPT2 catalytic event.

Reason: The actual R-HSA-200425 path reaches reaction R-HSA-200410, catalyst 201039 and human CPT2 entity 200994/reference P23786. It converts acylcarnitine and CoA to acyl-CoA and carnitine at the inner membrane. This is a target-performed step, not inference from mere pathway co-membership.

Summary: The UniPathway assignment agrees with the enzyme’s contribution to fatty acid oxidation.

Reason: UPA00659 is explicitly present in the current target UniProt pathway record; the official GO methods page states that UniPathway itself is no longer maintained. CPT2 performs acyl transfer to regenerate the matrix acyl-CoA substrates used by beta-oxidation. Its contribution is biochemical, not a claim that CPT2 catalyzes the later oxidation, hydration, dehydrogenation or thiolysis reactions.

Summary: Human immunofluorescence supports a mitochondrial CPT2 pool.

Reason: Current HPA data for antibody HPA028214 report supported mitochondrial staining in A-431, U-251 MG and U2OS cells, alongside additional nuclear/nucleolar locations depending on cell line. The actual localization metadata support the existing broad mitochondrial IDA. No original image-pixel examination or exclusive mitochondrial localization is claimed; inner-membrane precision comes from separate evidence.

Summary: The human enzyme transfers multiple acyl groups to carnitine.

Reason: Actual PMID:20538056 Methods use sequence-verified complete human CPT2 ORF in yeast homogenates, with endogenous CAT2 deleted and empty-vector controls. The measured direction is acyl-CoA plus carnitine to acylcarnitine; targeting and incomplete processing in yeast are explicit. C8 through C18 activity and lower C20 activity support the broad carnitine O-acyltransferase term. A strict C18 cutoff or a purified mature human enzyme preparation is not asserted.

Summary: The disease-variant study supports CPT2 carnitine acyltransferase activity.

Reason: The complete available PMID:7711730 abstract explicitly reports COS-cell transfection, depressed CPT II catalytic activity for P50H/D553N variants, and protein-stability analysis. Full original Methods/Results were not recovered, so exact assay substrates and construct processing are not claimed. Retain the curated experimental activity with deference to its positive source and independent human substrate profiling.

Summary: Human patient-derived myocytes corroborate loss of the palmitoyltransferase step.

Reason: The complete PMID:24780397 abstract reports more C16 palmitoylcarnitine in CPT II-deficient iPSC-derived myocytes, exacerbation by heat stress and a bezafibrate response. These are cellular acylcarnitine-profile measurements rather than an isolated enzyme-turnover assay. Full original Methods were not recovered. The existing experimental activity is nevertheless clearly correct from independent human CPT2 assays; retain it without misdescribing the patient-cell evidence.

Summary: Direct C8 acylcarnitine formation supports octanoyltransferase activity.

Reason: The actual human-CPT2 yeast-homogenate experiment in PMID:20538056 includes octanoyl-CoA plus carnitine and measures the corresponding acylcarnitine by ESI-MS/MS against empty-vector controls. This activity belongs to the same broadly selective enzyme as the C16 reaction, with no isoform-specific claim.

Summary: The exact human CPT2 protein group is detected in MitoCoP.

Reason: The authenticated original Table S1 contains CPT2/P23786 in sheet A row 265 and sheet B row 240. The record has MitoCoP=1 and positive subtractive, spatial and importomics classifications; its membrane classification is peripheral. The protein group includes additional sequence entries, so the record is not used to assign an unlisted isoform. Actual study Methods and target/header cells support this HTP mitochondrial localization.

Summary: The directly measured acyl-transfer chemistry is long-chain fatty-acid metabolism.

Reason: Human CPT2 transfers long-chain acyl groups, including the measured C14/C16/C18 substrates and lower C20 activity. This is actual chemical transformation of the defined long-chain substrate class. Retain the broad process at the paper’s bidirectional metabolic scope; the assay does not by itself measure whole-cell beta-oxidation flux.

Summary: Human CPT2 directly catalyzes palmitoyl transfer to carnitine.

Reason: PMID:20538056 uses complete human CPT2 expressed in yeast homogenates and quantifies palmitoylcarnitine formation from C16 acyl-CoA and L-carnitine, with empty-vector controls. This establishes the GO:0004095 reaction in its assayed direction. Physiological matrix acyl-CoA regeneration is the reverse reaction, not a different active site.

Summary: CPT2’s biochemical step contributes to the fatty-acid-beta-oxidation pathway.

Reason: The selected original measures acylcarnitine formation and discusses reverse physiological acyl-CoA regeneration; it is not a direct beta-oxidation-flux experiment. CPT2 itself performs the substrate-supplying acyl-transfer step, which is included in the existing pathway-level participation. Independent mouse muscle and adipose experiments corroborate the flux role. Retain the curated process while keeping the original assay distinct from a downstream oxidation measurement.

Summary: CPT2 directly interconverts free carnitine and acylcarnitines.

Reason: The substrate/product measurements in PMID:20538056 demonstrate chemical transformation of carnitine-containing metabolites. The broad carnitine-metabolic-process term is accurate for this bidirectional chemistry; it does not assign carnitine biosynthesis or a carnitine transporter pore to CPT2.

Summary: Refine experimentally established acyl transfer to its carnitine acceptor class.

Reason: The same original substrate panel specifically transfers multiple acyl groups to L-carnitine. GO:0016406 preserves that directly demonstrated breadth and is more precise than general acyltransferase activity; the C16 and C8 reactions already have separate source assertions.

Proposed replacements: carnitine O-acyltransferase activity

Summary: Retain thermogenesis as a conditional physiological contribution.

Reason: PMID:25578732 directly studies adipose-specific Cpt2 deletion in mice: cold-challenged animals become hypothermic and the normal brown-fat thermogenic transcriptional response is impaired. CPT2 contributes through fatty-acid utilization and its metabolic context; it is not assigned kinase or transcription-factor activity. This supports a contextual thermogenic contribution, while a corresponding human adipose perturbation assay was not read. Keep the existing ortholog transfer as non-core.

Summary: The expression event’s output is the inner-membrane CPT2 product.

Reason: The actual R-HSA-1989773 output is human CPT2 entity 200994, residues 26–658, assigned to the mitochondrial inner membrane and connected to reference P23786. This supports the product’s location. The event describes production of CPT2 and does not establish CPT2 as a transcriptional regulator or gene-expression catalyst.

Summary: The CPT2 catalytic entity is assigned to the mitochondrial inner membrane.

Reason: R-HSA-200410 reaches CPT2 through its actual catalyst activity, with human entity 200994 and reference P23786. This is the acylcarnitine-to-acyl-CoA reaction at the inner membrane, consistent with human characterization and the peripheral matrix-facing topology.

Summary: The original human liver study identifies the palmitoyltransferase product.

Reason: PMID:1988962 links purified human liver enzyme peptides to the 658-residue CPT2 cDNA and describes the palmitoyl-CoA/L-carnitine enzyme. It does not itself provide the later recombinant substrate panel. The sequence correction PMID:1961767 changes two residues and retains the 658-residue product; neither correction is a retraction of enzyme identity. Direct activity is independently supported by PMID:20538056.

Summary: The human liver enzyme is identified as an inner-membrane protein.

Reason: The original cloning/peptide work explicitly describes the human liver inner-membrane enzyme and its 25-residue targeting leader. Later human mitochondrial proteomics and the curated topology corroborate this location. The sequence corrigendum is accounted for without changing the original source identifier.

Summary: Retain the positive CPT2–CYSRT1 binary interaction as non-core.

Reason: Three literal HuRI records identify human P23786 and A8MQ03 in two-hybrid array, prey-pooling and validated two-hybrid formats in yeast. The reference-product sequences reproduce both ROGIDs, and fusion metadata are retained separately. The positive interaction is not invalidated by expected cellular locations. A defensible more specific partner-class function is not established for this pair, so generic binding remains non-core rather than overannotated.

Summary: Refine the positive OTX1 interaction to its demonstrated transcription-factor class.

Reason: The actual HuRI records EBI-22741336, EBI-22733631 and EBI-24264064 identify the same positive human CPT2–OTX1 pair in yeast assays. Human OTX1/P32242 has direct sequence-specific promoter-binding and RNA-polymerase-II reporter activity in PMID:18849347. GO:0061629 therefore describes the assayed partner class, without claiming that CPT2 binds DNA, regulates transcription, or contacts OTX1 in a native nuclear compartment. The original OTX1 reporter required double normalization; detailed OTX2/CRX-only mutant or competition controls are not attributed to OTX1. This secondary interaction is not a second catalytic core. This refinement combines evidence for the binding event with independent evidence identifying the bound protein class. The term does not require CPT2 to modulate transcription or to contact the DNA-binding domain of OTX1. Yeast fusion assays leave endogenous encounter and physiological relevance unresolved, but compartment expectations alone do not contradict the reported interaction. The more general MCUR1 and CYSRT1 association reviews lack this same demonstrated partner-class refinement.

Proposed replacements: RNA polymerase II-specific DNA-binding transcription factor binding

Summary: Mitochondrial localization is sound despite an unresolved current ARBA route.

Reason: The current annotation cites ARBA00026962. Exhaustive condition evaluation of its 380 current branches against the actual target domains, FunFams, family and taxonomic lineage finds no matching branch. This is a limitation in reconstructing the current rule premise, not proof that an earlier rule made a false assignment. Human imaging, original enzyme characterization and exact MitoCoP detection independently establish mitochondrial localization; retain the biological assertion and mark the rule source unresolved.