DHCR7 = 7-dehydrocholesterol reductase (7-DHC reductase; Delta7-sterol reductase; EC 1.3.1.21),
the ER-membrane enzyme catalysing the FINAL step of the Kandutsch-Russell arm of cholesterol
biosynthesis: NADPH-dependent reduction of the C7-C8 double bond of 7-dehydrocholesterol (7-DHC)
to cholesterol.
7-DHC is also the precursor of vitamin D3, so DHCR7 partitions 7-DHC between cholesterol and
vitamin D synthesis [PMID:25637936 "DHCR7 is important for both cholesterol and vitamin D
synthesis"; "as 7DHC can be converted into cholesterol or vitamin D, DHCR7 represents a switch
between these vital molecules in skin cells exposed to UVB"].
ER membrane, multi-pass. PMID:9878250. UniProt SUBCELLULAR LOCATION: Endoplasmic reticulum membrane; Multi-pass membrane
protein. HPA IDA -> ER. IDA PMID:9878250 also assigns nuclear outer membrane (contiguous with
ER/nuclear envelope) and ER. 7 predicted TM helices. Note GO:0005640 nuclear outer membrane is
part of ER envelope continuum — plausible but non-core.
IMP PMID:20615952: DHCR7 + EBP (D8D7I) together reconstitute cholesterol-5,6-epoxide hydrolase
(ChEH) activity of the microsomal antiestrogen binding site (AEBS). This is a genuine, curated,
experimentally-supported second activity of DHCR7 as part of a heterodimeric AEBS complex — NOT
a spurious wrong-branch IEA. UniProt records it (AltName "Cholesterol-5,6-epoxide hydrolase
subunit DHCR7", EC 3.3.2.11) with IMP-level evidence. It is a moonlighting/complex activity, not
the core reductase function; KEEP_AS_NON_CORE (do not REMOVE an experimental annotation). The
task framing warned of an "incorrect ChEH IEA" but this specific annotation is IMP/experimental
and its full text supports it — treat accordingly.
Two 2024 Nature papers (IMP): DHCR7 is pro-ferroptotic because it consumes the anti-ferroptotic
metabolite 7-DHC. PMID:38297129. PMID:38297130. Real, experimentally
supported regulatory role (via depleting a radical-trapping sterol); non-core relative to the
biosynthetic function. KEEP_AS_NON_CORE.
SLOS (Smith-Lemli-Opitz syndrome) — autosomal recessive; elevated 7-DHC, low cholesterol;
microcephaly/2-3 toe syndactyly/developmental delay. Many SLOS missense variants in UniProt;
loss of 7-DHC reductase activity (e.g., T93M, N274K, L306R mutagenesis). Confirms enzyme is the
cause. Background for the enzymatic core function.
Human DHCR7 is a 475-residue membrane-bound reductase that uses NADPH to reduce the C7–C8 bond of 7-dehydrocholesterol. The human cDNA was tested in yeast microsomes with mock controls and cholesterol product identification; the heterologous host does not make this a yeast-enzyme experiment (PMID:9465114(https://pubmed.ncbi.nlm.nih.gov/9465114/)). The official record supplies one sequence, not a set of experimentally equivalent natural short products.
DHCR7 also contributes with EBP to microsomal cholesterol-epoxide hydrolysis. Human coexpression and MCF-7 knockdown support a joint activity; the kinetic differences suggest a contribution to substrate binding, without proving autonomous epoxide hydrolysis by DHCR7. The contribution-aware second core preserves the positive result while the original single-protein catalytic assertion is marked as over-annotated (PMID:20615952(https://pubmed.ncbi.nlm.nih.gov/20615952/)). The linked 2010 correction fixes online inhibitor-table formatting and does not revise the reconstitution experiment.
Reciprocal DHCR24 association supports enzyme binding; DHCR24-dependent enhancement of DHCR7 activity is distinct from a claim that DHCR7 activates DHCR24. The 2015 correction changes a Figure 1 chemical label only (PMID:25637936(https://pubmed.ncbi.nlm.nih.gov/25637936/)). Human HeLa TMEM147 affinity purification also identifies DHCR7, but neither that association nor TMEM147 translocon membership establishes a separate DHCR7 adaptor or chaperone function (PMID:32694168(https://pubmed.ncbi.nlm.nih.gov/32694168/)).
DHCR7 consumption of protective 7-dehydrocholesterol increases ferroptosis susceptibility in the tested systems. This is a supported metabolic regulatory role, not an additional death-execution core (PMID:38297129(https://pubmed.ncbi.nlm.nih.gov/38297129/); PMID:38297130(https://pubmed.ncbi.nlm.nih.gov/38297130/)). The complete 2026 correction clarifies resazurin-background values and reports unchanged effects after two normalizations. Its supplemental plots were not independently assessed. The first paper was read from selected author-proof passages, with final Methods access limited.
ER membrane is the supported refinement for family-derived membrane or ER assertions. The broader NK-like-cell membrane-proteomics annotation keeps its original scope, because independent human fractionation also finds DHCR7 in non-ER membrane preparations; neither result is an exclusive localization claim (PMID:19946888(https://pubmed.ncbi.nlm.nih.gov/19946888/); PMID:38297130(https://pubmed.ncbi.nlm.nih.gov/38297130/)). The separately curated nuclear outer membrane claim remains unresolved because the full localization experiment was unavailable; the accessible abstract explicitly establishes ER localization (PMID:9878250(https://pubmed.ncbi.nlm.nih.gov/9878250/)).
NADP(H) binding is part of the intrinsic reductase mechanism. The ISS donor MaSR1 is a bacterial sterol reductase with an NADPH-bound structure, not a DHCR7 ortholog; its substrate specificity is not transferred (PMID:25307054(https://pubmed.ncbi.nlm.nih.gov/25307054/)). The PAINT target sequence and lineage were reconstructed. Two literal IBD records were found; the current cholesterol-biosynthesis node is on the lineage but its literal IBD was absent from the inspected release. This explicit reconstruction limit does not negate direct human cholesterol synthesis.
The source evidence guide supplies public original URLs, Methods/Results/caption locators, short verbatim anchors, correction scopes and exact electronic-source paths. It is authored navigation, not independent experimental evidence. Generic positive association is retained where no justified narrower class is established, under the scoped project instruction.
The existing source files and earlier notes above are unchanged. The genuine default Falcon attempt failed authentication and its configured perplexity-lite fallback failed for insufficient quota; no provider-named report was authored manually. This augmentation therefore uses the explicitly documented manual primary-source and database assessment.
This entry supersedes the earlier Annotation dispositions (summary) and the broad-membrane KEEP_AS_NON_CORE interpretation in Source-specific augmentation. Those passages remain as the historical record. The current YAML contains the operative decisions: both GO:0016020 membrane rows are MODIFY to GO:0005789 endoplasmic reticulum membrane. Independent human localization, microsomal assays and the membrane-bound protein architecture support this refinement (PMID:9878250(https://pubmed.ncbi.nlm.nih.gov/9878250/); PMID:9465114(https://pubmed.ncbi.nlm.nih.gov/9465114/)). The DHCR7-specific YTS table and sampled membrane fraction in PMID:19946888(https://pubmed.ncbi.nlm.nih.gov/19946888/) remain unverified. Refining the gene-level claim does not turn that study into an ER-localization experiment or imply exclusive ER localization. The specific nuclear outer membrane claim remains UNDECIDED.
The epoxide-hydrolase decision retains the experimentally supported subunit contribution. UniProt Q9UBM7 assigns EC 3.3.2.11 and the ChEH-subunit name, but its FUNCTION section leaves the individual components' precise roles unresolved. The original coexpression and knockdown results support joint activity; autonomous DHCR7 hydrolysis remains unestablished (PMID:20615952(https://pmc.ncbi.nlm.nih.gov/articles/PMC2922168/), Figure 4 and the following Discussion). Both core entries are unchanged.
The source evidence guide now shows the actual conjunctive FunFam/taxon conditions of the dated ARBA rules and distinguishes reference identity from target-table or whole-paper verification. No new quotation, source annotation, product or biological claim is added.