SAMD8 (Q96LT4 / SMSr) — Function-Assignment Review OpenScientist openscientist-autonomous 3 artifacts 2026-09-20T23:37:57.212850

OpenScientist SAMD8 GO decision table (text/csv)

OpenScientist final report (text/html)

OpenScientist final report (application/pdf)

SAMD8 (Q96LT4 / SMSr) — Function-Assignment Review

Hypothesis slug: lipid-reaction-spectrum-pathways-and-compartments
Focus: function_assignment
Gene: human SAMD8 (Sphingomyelin synthase-related protein 1, SMSr), UniProt Q96LT4, 415 aa

Seed hypothesis (paraphrased)

Human SAMD8 has (a) sphingomyelin synthase or ceramide cholinephosphotransferase activity,
(b) contributes to ceramide biosynthesis, and/or (c) localizes to Golgi / plasma membrane.
Assess each component separately.

Executive Judgment

Verdict by component:

Seed component	Verdict
(a) Sphingomyelin synthase / ceramide cholinephosphotransferase activity	Refuted / over-annotated
(b) Contributes to ceramide biosynthesis (direct)	Refuted as direct catalysis; weakly/indirectly supported as regulation
(c) Localizes to Golgi / plasma membrane (native)	Refuted / paralog-and-mutant artifact
Established core: ER-resident CPE synthase, SAM-dependent ER residence	Supported

Overall: the seed's three propositions are not supported for the native full-length human protein. SAMD8/SMSr is a monofunctional ceramide phosphoethanolamine (CPE) synthase (EC 2.7.8.-) resident in the endoplasmic reticulum membrane, with its catalytic site in the ER lumen. It is not a sphingomyelin synthase; it does not synthesize ceramide (its reaction consumes ceramide to make CPE); and it is not natively a Golgi/plasma-membrane protein — those localizations belong to the paralogs SMS1/SGMS1 (Golgi) and SMS2/SGMS2 (plasma membrane) and appear for SMSr only in oligomerization-defective SAM mutants. A newer biochemical body of work adds a promiscuous glycerophospholipid hydrolase / PLC-PAP activity generating diacylglycerol, which broadens the "reaction spectrum" but still does not include SM synthesis and reinforces a regulatory rather than biosynthetic-bulk role.

Most important caveats: (1) In-vivo mouse KO shows SMSr is dispensable for ceramide levels and development, qualifying its "ceramide homeostasis regulator" role as context-specific (cultured cells). (2) The newer PLC/PAP hydrolase activities are largely in-vitro (purified enzyme / overexpression) and should be curated cautiously. (3) The current UniProt/GO record already carries the exact over-annotated IBA terms this hypothesis probes (SM synthase, ceramide cholinephosphotransferase, Golgi membrane, plasma membrane).

Evidence Matrix

#	Citation (PMID)	Evidence type	Supports/Refutes/Qualifies	Claim tested	Key finding	Context	Confidence & limitations
1	19506037 (Vacaru 2009, JCB)	Direct assay + localization + mutant	Refutes SM; Supports CPE/ER	Enzymatic identity & compartment	SMSr catalyzes CPE synthesis in the ER lumen, ~300× less product than SMS1-derived SM; blocking activity raises ER ceramide → "sensor rather than converter"	Human cells, heterologous + endogenous	High; foundational paper
2	24259670 (Tafesse 2014, JBC)	Mutant phenotype + localization	Supports ER/CPE; Qualifies ceramide role	Is SMSr an SM synthase or ceramide-maker?	"SMSr (SAMD8), an ER-resident CPE synthase," suppressor of ceramide-induced mitochondrial apoptosis; SAM domain required	Human cultured cells	High (cell culture); phenotype is loss-of-function ceramide accumulation, not ceramide synthesis
3	28120887 (Cabukusta 2017, JCB)	Localization + structural/biochemical	Supports ER; explains Golgi artifact	Native compartment & basis of ER residence	"SMSr/SAMD8 is an ER-resident CPE synthase"; ER residency via SAM-domain oligomerization; SAM mutants → partial redistribution to Golgi	Human cells; crosslinking, native PAGE	High; Golgi signal is mutant-dependent
4	27729449 (Cabukusta 2016)	Biophysical (single-molecule)	Supports ER + oligomer	Oligomeric state of ER sensor	SMSr forms SAM-dependent oligomers in the ER	HeLa, TIRF photobleaching	Medium-high; supports sensor model
5	25667419 (Bickert 2015, JLR)	Mutant phenotype (in vivo mouse)	Refutes bifunctional SM; Qualifies ceramide role	Monofunctional CPE vs SM; in-vivo ceramide role	"SMSr serves as monofunctional CPE synthase" (vs SMS2 bifunctional SM+CPE); CPE >300× lower than SM; KO did not affect ceramide levels in brain/tissues; no developmental phenotype	Mouse, multiple tissues	High; strongest in-vivo evidence; qualifies homeostasis role
6	14685263 (Huitema 2004, EMBO Rep)	Direct assay + localization (paralogs)	Competing/explains paralog over-annotation	Where do SM synthases localize?	SM synthesis transfers phosphocholine from PC; SMS1 = Golgi, SMS2 = plasma membrane	Human/mouse/C. elegans	High; establishes that Golgi/PM + SM-synthase belong to paralogs
7	33621517 (Murakami & Sakane 2021, JBC)	Direct assay (purified enzyme)	Qualifies/extends reaction spectrum	Additional activities beyond CPE	Purified SMSr hydrolyzes PE, PA, PI, PC → DAG in absence of ceramide; PAP activity ~300× > CPE activity	Purified human SMSr; COS-7 overexpression	Medium; in-vitro/overexpression, physiological relevance open
8	UniProt Q96LT4 (record; cites 33621517, 34332077, 38388831)	Database/curated	Supports ER + CPE + PLC; flags over-annotation	Curated function/location/GO	Single location: ER membrane; EC 2.7.8.-; SAM domain 12–78; 6 TM helices; PLC on PE → phosphoethanolamine + DAG, proposed SPT-activating switch; GO set contains over-annotated IBA terms	Human curated	Orientation-level; used to identify GO terms to re-examine
9	This work (computed, UniProt Q96LT4/Q86VZ5/Q8NHU3)	Structural/evolutionary (computational)	Qualifies — explains paralog over-annotation	Are SM-synthase/Golgi/PM terms paralog carry-over?	SAMD8 is the sequence outgroup (42–48% id to SGMS1/2, which are 64% to each other); shared 6-TM catalytic triad but distinct location/donor specificity	In silico pairwise alignment	Medium; NW with simple scoring, not a curated MSA; supports (not proves) carry-over hypothesis

GO Curation Implications (leads — require curator verification)

Current GO annotations on Q96LT4 (from UniProt cross-references) and recommended action:

GO ID	Aspect / Term	Current evidence	Recommendation	Rationale
GO:0002950	MF: ceramide phosphoethanolamine synthase activity	IEA:Ensembl	Retain & strengthen (upgrade to experimental, e.g. IDA from PMID 19506037/25667419)	This is the correct, primary catalytic identity
GO:0047493	MF: ceramide cholinephosphotransferase activity	IBA:GO_Central	Remove / do-not-annotate (or NOT)	= SM-synthase reaction (uses phosphocholine); SMSr is monofunctional CPE synthase (PMID 25667419)
GO:0033188	MF: sphingomyelin synthase activity	IBA:GO_Central	Remove / do-not-annotate	Paralog (SMS1/SMS2) activity; SMSr makes no SM (PMID 19506037, 25667419)
GO:0000139	CC: Golgi membrane	IBA:GO_Central	Remove / do-not-annotate	Native protein is ER; Golgi only in SAM-mutant redistribution (PMID 28120887); Golgi = SMS1 (PMID 14685263)
GO:0005886	CC: plasma membrane	IBA:GO_Central	Remove / do-not-annotate	Paralog SMS2 localization (PMID 14685263); no native SMSr PM evidence
GO:0006686	BP: sphingomyelin biosynthetic process	NAS:UniProtKB	Remove / re-examine	SMSr does not make SM
GO:0046513	BP: ceramide biosynthetic process	IDA:UniProtKB	Re-examine / generalize	SMSr's reaction consumes ceramide to make CPE; it does not biosynthesize ceramide. Prefer "ceramide metabolic process" / "sphingolipid biosynthetic process (CPE)" or a regulation term
GO:2000303	BP: regulation of ceramide biosynthetic process	IDA:UniProtKB	Retain	Consistent with sensor role (PMID 19506037) and proposed SPT-activating switch
GO:0005783 / GO:0005789	CC: ER / ER membrane	IDA:HPA / IDA:UniProtKB	Retain (core CC)	Well-supported native localization

Candidate NEW MF terms to consider (leads, in-vitro caution): phospholipase C activity (GO:0004629) and/or phosphatidate phosphatase / lipid phosphatase activity, and diacylglycerol biosynthetic process — supported by PMID 33621517 but currently in-vitro/overexpression; annotate cautiously (e.g., IDA with a comment, or hold pending in-cell confirmation).

Recommended primary MF for SAMD8 is ceramide phosphoethanolamine synthase activity (GO:0002950); core CC is endoplasmic reticulum membrane (GO:0005789); core BP is regulation of ceramide/sphingolipid homeostasis. "Protein binding" is not needed as a summary term.

Paralog Comparison (computed provenance)

Sequences and features fetched from UniProt (Q96LT4, Q86VZ5, Q8NHU3); pairwise identity by Needleman–Wunsch (match +1 / mismatch −1 / gap −2).

Protein	Acc	Length	TM helices	SAM domain	Active-site residues	Subcellular location
SAMD8/SMSr	Q96LT4	415	6	Yes (12–78)	His301, His344, Asp348	ER membrane
SGMS1/SMS1	Q86VZ5	413	6	Yes	285, 328, 332	Golgi membrane
SGMS2/SMS2	Q8NHU3	365	6	No	229, 272, 276	Plasma membrane; Golgi

Pairwise identity: SAMD8–SGMS1 = 42.4%, SAMD8–SGMS2 = 48.2%, SGMS1–SGMS2 = 63.9%.

Interpretation: The two bona fide SM synthases (SGMS1/SGMS2) are more similar to each other (64%) than either is to SAMD8 (42–48%), so SAMD8 is the divergent outgroup. All three share the 6-TM catalytic core and the conserved SMS/lipid-phosphate-phosphatase active-site triad — which is exactly why phylogenetic/automated pipelines propagate "sphingomyelin synthase activity" and "ceramide cholinephosphotransferase activity" onto SAMD8 (IBA/IEA). However, active-site conservation does not determine head-group donor specificity (PC vs PE); the experimental record shows SAMD8 is PE-specific (CPE synthase). Homology therefore cannot override direct substrate–product assays. (Correction to earlier note: the SAM domain is shared with SGMS1 — it is not unique to SAMD8 — but SAMD8's SAM-driven oligomerization is what confers its ER residence.)

Mechanistic Scope

Immediate molecular function: transfer of phosphoethanolamine from PE onto ceramide → ceramide phosphoethanolamine (CPE) + DAG (EC 2.7.8.-); additionally, promiscuous in-vitro hydrolysis of glycerophospholipids (PE/PA/PI/PC) → DAG (PLC/PAP-type). It removes/relocates phospho-head groups; it does not add phosphocholine to ceramide (i.e., no SM-synthase direction).
Cellular function: ER-membrane enzyme with lumenal catalytic site; SAM-domain homo-oligomerization anchors it in the ER and underlies a proposed ceramide-sensing function.
Downstream/indirect (not direct MF): suppression of ceramide-induced mitochondrial apoptosis, maintenance of ER-exit-site/secretory-pathway integrity, proposed activation of serine palmitoyltransferase and DAG signaling to DGKδ. These are consequences of loss-of-function or regulatory coupling, not direct catalytic outputs, and one (ceramide accumulation on loss) is not reproduced in vivo.

Conflicts and Alternatives

Paralog confusion (primary risk): SMS1/SGMS1 (Golgi SM synthase) and SMS2/SGMS2 (plasma-membrane, bifunctional SM+CPE) supply exactly the Golgi/PM and SM-synthase/cholinephosphotransferase signals seen as IBA annotations on SAMD8. These are family-tree propagations, not SAMD8-specific evidence.
Construct/mutant artifact: Golgi localization of SMSr is observed only for SAM-domain oligomerization mutants (PMID 28120887); native full-length protein is ER.
In-vitro vs in-vivo: ceramide accumulation on SMSr loss is a cultured-cell phenotype; mouse KO shows no ceramide change and no overt phenotype (PMID 25667419) — cautions against a strong "ceramide biosynthesis/homeostasis" BP call.
Reaction-spectrum ambiguity: the purified enzyme's dominant in-vitro activity is DAG generation via lipid-phosphatase/PLC chemistry, not CPE synthesis (PMID 33621517); physiological weighting of these activities is unresolved.

Knowledge Gaps

Physiological substrate/direction in vivo. Checked: in-vitro assays (CPE synthase; PAP/PLC). Gap matters because MF choice (CPE synthase vs lipid phosphatase/PLC) hinges on it. Resolve with cell-based lipidomics of SAMD8-null vs rescue, and inducible catalytic-dead rescue.
SPT-activation switch. Checked: UniProt "by similarity" note citing 34332077/38388831. Gap: whether SAMD8 directly regulates SPT in human cells. Resolve with SPT-activity/flux measurement upon SAMD8 modulation.
Any bona fide non-ER pool. Checked: HPA IDA ER; mutant-only Golgi. Gap: whether a minor native Golgi/ER-Golgi-interface pool exists. Resolve with high-resolution/APEX proximity localization of endogenous protein.
Ceramide "biosynthetic process" IDA basis. Checked: GO record. Gap: the direct experiment behind GO:0046513 IDA is unclear given SMSr consumes ceramide. Resolve by tracing the annotation's supporting figure.

Discriminating Tests

Head-group donor assay: side-by-side ceramide + PE vs ceramide + PC with purified SAMD8 and SMS2 control → confirms CPE-only vs SM (discriminates cholinephosphotransferase claim).
Endogenous localization: knock-in tag or validated antibody + organelle markers/APEX → ER vs Golgi/PM (discriminates compartment claim).
Catalytic-dead rescue lipidomics in SAMD8-KO human cells → whether ceramide/CPE/DAG/SM pools change (discriminates biosynthesis vs regulation).
Comparative paralog panel (SGMS1/SGMS2/SAMD8) in the same assay → attributes SM/Golgi/PM properties to the correct gene, exposing IBA over-propagation.
In-vivo relevance: re-examine mouse/tissue lipidomes (leveraging PMID 25667419) for CPE/DAG rather than ceramide endpoints.

Curation Leads (require curator verification)

Action: Reclassify SAMD8 MF away from SM-synthase family terms toward CPE-synthase.
Remove/flag as paralog over-annotation (IBA): GO:0033188 (SM synthase), GO:0047493 (ceramide cholinephosphotransferase), GO:0000139 (Golgi membrane), GO:0005886 (plasma membrane); re-examine GO:0006686 (SM biosynthetic process, NAS) and GO:0046513 (ceramide biosynthetic process, IDA).
Retain/strengthen: GO:0002950 (ceramide phosphoethanolamine synthase activity) — upgrade from IEA to experimental (IDA) using PMID 19506037 / 25667419; GO:0005789 (ER membrane); GO:2000303 (regulation of ceramide biosynthetic process).
Consider NEW (in-vitro caution): phospholipase C / lipid-phosphatase MF and diacylglycerol biosynthetic process from PMID 33621517.
Candidate references + verbatim snippets to verify:
PMID 25667419: "SMS-related protein (SMSr) serves as monofunctional CPE synthase"; "blocking its catalytic activity did not affect ceramide levels ... in the brain or any other tissue."
PMID 19506037: "which catalyses the synthesis of the SM analogue ceramide phosphoethanolamine (CPE) in the ER lumen."
PMID 28120887: "SMSr/SAMD8 is an ER-resident ceramide phosphoethanolamine synthase"; "their substitution causes ... a partial redistribution of the enzyme to the Golgi."
PMID 14685263: "human SMS1 is localised to the Golgi, SMS2 resides primarily at the plasma membrane."
PMID 33621517: "purified SMSr showed a DG-generating activity via hydrolysis of PE, phosphatidic acid (PA), phosphatidylinositol (PI), and phosphatidylcholine (PC) in the absence of ceramide."
Suggested curator questions: Is the GO:0046513 IDA traceable to a direct ceramide-synthesis experiment, or a regulatory observation? Are the SM-synthase/Golgi/PM IBA terms desirable to keep given monofunctional CPE + ER data? Should in-vitro PLC/PAP activity be annotated now or held?

Limitations of this review

Two of the three UniProt-cited PLC papers (PMID 34332077, 38388831) were not retrievable via the available PubMed search and are represented only through the UniProt curated summary (database-level orientation).
Programmatic UniProt fetch succeeded (Q96LT4) and provided GO/feature provenance; no local bioinformatics files were provided, so sequence/structure claims rely on UniProt features (SAM domain 12–78, 6 TM helices, ER membrane) rather than a fresh de-novo analysis.