DHTKD1 (human UniProt Q96HY7, HGNC:23537) supplies the E1 reaction of the mitochondrial oxoadipate dehydrogenase complex. Its ThDP-dependent decarboxylation and transfer to DLST-bound lipoamide form one catalytic function. DLST subsequently transfers glutaryl to CoA, and DLD reoxidizes the lipoyl group. The official product is 919 residues; no alternative-product set is supplied in the reviewed UniProt record.
| Source | Evidence used | Scope and limits |
|---|---|---|
| PMID:29191460(https://pubmed.ncbi.nlm.nih.gov/29191460/) | Recombinant human E1a activity, preference for oxoadipate over oxoglutarate, and functional assembly with E2/E3. | Verified original abstract only; the original full text was not recovered. The abstract's overall product wording must not collapse the separate E1, E2 and E3 reactions. |
| PMID:32633484(https://doi.org/10.1021/acschembio.0c00114), publisher supplement | Original manuscript and supplement: ThDP-bound human structure, isolated E1 and reconstituted-complex activity. | Supplement S7 uses human residues 25–919 with an N-terminal His tag in bacteria. S8 separates DCPIP reduction from three-enzyme NADH production. S10 describes a distinct C-terminally tagged, methylated crystallography preparation. Substrate/inhibitor poses are docking, not substrate-bound crystal observations. S12 depicts pathway convergence, not tryptophan flux. |
| PMID:32695416(https://doi.org/10.1107/S205225252000696X) | Original Methods/Results: ThDP/Mg-dependent DHTKD1 dimer and DLST association. | Human 45–919 crystallography protein is engineered. Kinetic values concern 2-oxoadipate. Tentatively fitted substrate density was unconfirmed and not deposited as bound substrate. The cryo-EM reconstruction resolves DLST core, not a complete E1/E2/E3 complex. |
| PMID:32303640(https://pubmed.ncbi.nlm.nih.gov/32303640/) | Selected original Results and Methods: human recombinant E1/E2 chemistry and G729R effects on intermediate channeling. | A disease-associated experimental variant is not a natural alternative product. Acyl transfer and complete-complex outputs are distinguished from the isolated E1 reaction. |
| PMID:23141293(https://pmc.ncbi.nlm.nih.gov/articles/PMC3516599/) | Original Results/Figure 3: human patient-fibroblast wild-type complementation and labeled-lysine conversion to 2-oxoadipate. | The normal repository cache is abstract-only; the full public original was recovered separately. This is direct lysine-pathway evidence. Introduction/Figure 1 and Discussion explain convergence with tryptophan; no tryptophan tracer experiment is claimed. |
| PMID:32160276(https://pmc.ncbi.nlm.nih.gov/articles/PMC7206849/) | Original Results/Figures 2–3 and Methods: human HEK293 knockouts, DHTKD1-knockout-controlled endogenous immunoprecipitation, proximity ligation and human recombinant hybrid assemblies. | DHTKD1 overexpression increased oxoadipate activity; OGDH increased oxoglutarate activity. Recombinant hybrid assemblies nevertheless show substrate overlap. Mouse reverse OGDH immunoprecipitation did not recover DHTKD1. These results do not establish a universal hybrid stoichiometry or a second DHTKD1 core. |
The current GO:0160166 definition is the specific E1 reaction and is a child of GO:0016624. Consequently, the broad oxidoreductase annotation is valid but merits refinement; the earlier claim that it applies only to the complete complex was incorrect. ThDP binding is retained as a mechanism-associated annotation within the same core.
The lysine and tryptophan replacements of the existing amino-acid-catabolism row describe DHTKD1's own catalytic participation. Row 8 already includes GO:0006569 as a supported replacement of the broad term, and the same pathway role is retained in the catalytic core. The separate historical NEW object is retired as duplicate coverage, with its source identity preserved. This is deduplication, not withdrawal of tryptophan-pathway convergence. No new process is proposed, no comparator-absence argument is used, and no target activity was found in the fixed GO-CAM index.
PMID:23141294(https://pmc.ncbi.nlm.nih.gov/articles/PMC3516600/) Results/Figure S3 reports endogenous DHTKD1 colocalization with MitoTracker in human HepG2 cells. The separately recovered original is more extensive than the unchanged abstract-only cache. HPA reports supported mitochondrial localization with HPA037949 in A-431, MCF7 and U2OS cells. These fluorescence observations resolve mitochondria, not the matrix.
The original MitoCoP study, PMID:34800366 combines mitochondrial fractionation, spatial proteomics, importomics and curated evidence. Its original methods and Table S1 were read. Sheet (A), row 305 identifies Q96HY7 as the majority and simplified protein ID and marks experimental identification and MitoCoP inclusion. It scores Class 1 in subtractive proteomics (85.2% mean sequence coverage), Cluster 1 in spatial proteomics, and Class 1 in importomics; sheet (B), row 278 confirms DHTKD1 in the gene list. The importomics entry is a TOMM40-depletion response, not a purified DHTKD1 import assay. The membrane-association category is Ambiguous. The matrix entry is a compiled GO annotation, so it is not used as independent direct matrix localization. A fresh download failed, but a previously authenticated copy of the same publisher supplement supplied the actual table evidence.
The source mouse ortholog A2ATU0 carries a curator-assigned matrix IDA and amino-acid/lysine IMP from PMID:28545977(https://pubmed.ncbi.nlm.nih.gov/28545977/). The abstract describes mouse Dhtkd1/Gcdh perturbation; the full localization assay was not recovered. The annotation is therefore recognized as experimental donor grounding without claiming a personally verified mouse localization experiment. It is not a transfer chain consisting solely of human-derived similarity statements.
Human Reactome supplies an explicit target-specific matrix path:
Thus the DLD- and DLST-associated rows support location through complex membership, not transfer of their catalytic activities. The more precise matrix replacements use this human curated evidence and the mouse experimental annotation together; they are not presented as finer resolution of the original human fluorescence or HTP assays.
| GOA source | Literal public interaction identifier | Interpretation |
|---|---|---|
| BioPlex 2.0, PMID:28514442 | IM-25778-3723 / EBI-21626201 |
Human PGLYRP3 Q96LB9 bait, DHTKD1 Q96HY7 prey; HEK293T anti-tag co-IP, C-terminal tandem bait tag. Positive co-complex association retained as non-core. |
| HuRI, PMID:32296183 | IM-25472-124467 / EBI-24412093; companion pooling IM-25472-13779 and array IM-25472-38686 |
Human BMAL1 O00327-8 with Q96HY7; validated Y2H plus screen records. The sequence checks identify the 625-residue partner, whose deletion at canonical position 274 leaves the bHLH DNA-binding domain intact. Refined to transcription-factor binding with assay limits. |
| BioPlex 3.0, PMID:33961781 | IM-29278-17115 / EBI-54798933 |
Human PGLYRP3 bait and DHTKD1 prey, again HEK293T anti-tag co-IP with a C-terminal tandem bait tag. Retained as non-core. |
These identifiers are searchable in IntAct; the underlying PSICQUIC target query also returns unrelated DHTKD1 pairs, which are not substituted for the three imported sources. The BioPlex papers' original clone/culture/AP-MS Methods establish tagged bait capture from lysates, not purified binary binding. PGLYRP3 is a secreted peptidoglycan-recognition and bacterial-killing protein; these assays do not show that DHTKD1 binds a signal-initiating receptor or an enzyme of an established specific class.
GO:0140297 describes binding to a DNA-binding transcription factor. The official BMAL1 O00327 record describes sequence-specific E-box recognition and transcriptional activation; its isoform-8 deletion at canonical residue 274 leaves the bHLH domain at residues 72–125 intact. Together with the positive validated pair EBI-24412093, this supports a known partner-class refinement as qualitative tagged-Y2H binding capacity. The term does not require DHTKD1 to regulate transcription or circadian timing. HuRI Results describes quadruplicate retesting and sequence confirmation, while orthogonal validation is study-wide, not established for this individual pair. Native colocalization, endogenous complex formation and affinity remain unresolved. PGLYRP3 lacks an established applicable finer class here; its non-core treatment does not rank human-cell AP-MS below Y2H.
PMID:23141294(https://pmc.ncbi.nlm.nih.gov/articles/PMC3516600/) Results/Figure 4 reports reduced ATP and NAD(H) after DHTKD1 silencing in HEK293T cells. Those pool measurements do not identify a direct reaction. Independent support for GO:0006091 comes from DHTKD1 performing the E1 step in a precursor- and redox-generating sequence: it decarboxylates 2-oxoadipate and reductively glutarylates DLST-bound lipoamide. The PMID:32633484 publisher supplement, S7–S8, separately assays isolated human E1 through DCPIP reduction and NADH production by reconstituted DHTKD1/DLST/DLD. The overall OADH event and the individual E1, DLST and DLD events above preserve the same role separation: DLST transfers glutaryl to CoA and DLD reduces NAD+. Retention is therefore grounded in a performed catalytic step, not knockdown necessity alone. No autonomous DHTKD1 NADH production or ATP synthesis is asserted. The broad annotation remains contextual to the specific catabolic core.
The project curation instructions govern the non-core retention of positive generic interactions. Supported specificity changes use MODIFY even when a specific term is already represented by another source. Original GOA identities, evidence codes and qualifiers remain intact. Shared PANTHER and pending module content were not replaced. This guide distinguishes public originals from unchanged abstract-only publication caches; cache availability is not treated as the boundary of actual original access.