This manual synthesis describes selected original experiments and their limits. It distinguishes natural human products from historical constructs and separates hydrolysis, glycosyl transfer and downstream cellular effects. It is not a provider-generated report.
PMID:32144204, original article, Results, Figure 9E and the GlcCer/GalCer hydrolase Methods describe a separate hydrolysis assay. The cache contains only the abstract.
GalCer hydrolysis occurred in the presence of rGBA1, the HEK293T cell homogenates, and rGALC.
Human GBA2-expressing homogenates produced C6-NBD-ceramide from C6-NBD-GalCer with NB-DGJ inhibition. At pH 5.8 over 30 minutes, activity was about sevenfold below GlcCer-analog cleavage in the same preparation, not a comparison with GALC. No free-galactose measurement or physiological clearance is established. Images and supplementary material were not inspected.
PMID:17080196, original JCI article, selected enzyme-activity Results, Figure 6/7 captions and selected cloning/enzyme-assay Methods, distinguishes human and mouse cDNAs expressed in COS cells. Both increased fluorescent GlcCer cleavage over mock transfection. The authentic radiolabeled-lipid conversion panel used mouse GBA2 in HEK293 cells; a human host does not make that enzyme human. Knockout tissue accumulation and infertility are mouse outcomes. Bacterial fragments described in antibody production are immunogens, not purified active full-length human enzyme. The normal cache is abstract-only, while these selected publisher sections were accessible. Images and earlier protocols cited in Methods were not inspected, and none of these constructs was equated with a modern natural isoform by sequence.
PMID:23250757(https://pmc.ncbi.nlm.nih.gov/articles/PMC3561557/) uses tagged mouse NM_172692 constructs in HEK293 cells and primary mouse neurons. Selected protease-protection, carbonate-extraction and localization results support peripheral membrane association with cytosolic termini, including ER/Golgi pools. Membrane reconstitution used a fluorogenic substrate. This is ortholog and construct evidence; it does not establish the topology of every human pool or equal activity of all human products.
PMID:30308956(https://pmc.ncbi.nlm.nih.gov/articles/PMC6213336/), Results 2.1 and Methods 4.2–4.3, compares three D594H patient lymphoblastoid lines with four controls. Fluorogenic-substrate activity is reduced in lysates and at the surface of living cells, with substrate-entry and released-enzyme controls; GlcCer accumulation is measured separately. This supports a contextual human surface-associated enzymatic pool, not direct transmembrane mapping. Its immunoblot concerns GBA1, not GBA2 abundance. The complete main text and textual captions were read, without images, supplements or raw tables.
The single GlcCer-catabolic core accommodates these differently supported membrane pools. Its ER/Golgi locations rest on mouse ortholog experiments, whereas its plasma-membrane location has human patient-cell activity and an explicit human Reactome catalyst. An unresolved universal topology does not imply absence from any of those observed pools.
PMID:26724485(https://pmc.ncbi.nlm.nih.gov/articles/PMC4766994/) uses mouse NM_172692.3 expressed in COS-7 cells for GBA2 experiments. Selected Results/Methods distinguish reversible transfer between GlcCer and cholesterol from GlcChol hydrolysis. The main text explicitly reports the latter; its supplementary quantitative hydrolysis table was not inspected. Reduced GlcChol in knockout mice supports, but does not by itself prove, an in-vivo synthetic contribution. Proposed cholesterol transport/buffering is not a measured transport mechanism. The negative donor result under these conditions is distinct from later positive GalCer experiments.
PMID:25803043(https://pmc.ncbi.nlm.nih.gov/articles/PMC4372435/) tests mouse Gba2 loss and inhibition. Selected actin-turnover and microtubule Results distinguish an increased F/G-actin ratio and longer microtubule persistence from unchanged growth rate. These are downstream lipid-dependent regulatory phenotypes, not direct actin/tubulin binding or polymerase activities. The decisive short excerpts appear once in the review; their paired electronic contexts retain reference links.
PMID:33361282(https://pmc.ncbi.nlm.nih.gov/articles/PMC7903134/) is retained for its negative GBA2 transxylosylation result under the tested conditions. It does not transfer GBA1 substrate breadth to GBA2.
The reviewed beta-glucosidase class preserves verified glucoside substrate breadth; the primary GlcCer-specific annotation remains separate. The official is_a path from GO:0008422 to GO:0004553 supports the broad hydrolase refinement. The inspected current ontology does not supply an is_a path from GO:0004348 to GO:0008422; biochemical substrate inclusion and a currently asserted ontology edge are different claims.
The sterol-glycoside/RHEA:58264 questions remain questions about existing term definitions and mappings. The observed reverse transfer to ceramide does not establish an additional physiological biosynthetic process, so no NEW process is introduced merely to fill a pathway description.