P03316 — Sindbis virus structural polyprotein (p130) — curation notes

Provenance of this file

Deep research was unavailable for this gene: no deep-research API keys are
configured in this container, so just deep-research could not be run and no
-deep-research-<provider>.md file exists or should be created. These notes were
written directly from (a) the cached UniProt record P03316-uniprot.txt, (b) the
publications cached under publications/, and (c) two papers I fetched with
just fetch-pmid (PMID:12424249, PMID:1944569). The GONUTS wiki
(gowiki.tamu.edu), which normally records the CACAO student's evidence, is
offline (see projects/CACAO.md), so the three CACAO rows were judged from the
papers alone.

Identity and gene symbol

UniProt entry POLS_SINDV / P03316, "Structural polyprotein", AltName p130.
UniProt assigns no gene name — the GOA file's GENE NAME column carries the
protein name "Structural polyprotein", not a symbol — so the folder and the
review's gene_symbol both use the accession P03316. This follows the
CLAUDE.md convention for entries with no gene symbol at all ("use the accession
as the folder name"). The stub's gene_symbol: P03316 is therefore appropriate
and I have left it as seeded.

alternative_products (seeded) records the ribosomal-frameshift relationship to
P0DOK0 (the frameshifted structural polyprotein, i.e. the TF-containing product);
P0DOK0 is a separate UniProt entry and is not reviewed here.

One entry, five mature chains

This is a polyprotein: every annotation on P03316 is really an annotation on one
of its cleavage products, and which chain a row belongs to is the main thing a
reviewer has to get right here.

Processing order [file:SINDV/P03316/P03316-uniprot.txt "Capsid protein is
auto-cleaved during polyprotein translation, unmasking a signal peptide at the
N-terminus of the precursor of E3/E2"], then host signal peptidase in the ER
yields pE2, 6K and E1, and host furin cleaves pE2 to E3 + E2 in trans-Golgi
vesicles.

Capsid protein

Autoprotease. The C-terminal domain is a chymotrypsin-like serine protease
that cleaves the capsid off the nascent polyprotein; UniProt records
EC 3.4.21.90 and the reaction "Autocatalytic release of the core protein from the
N-terminus of the togavirus structural polyprotein by hydrolysis of a -Trp-|-Ser-
bond" [file:SINDV/P03316/P03316-uniprot.txt]. The domain statement is tied to a
publication: "The C-terminus functions as a protease during translation to cleave
itself from the translating structural polyprotein (PubMed:1944569)". This is the
basis for GO:0004252 and, at the right granularity, for protein autoprocessing
rather than bare proteolysis.

RNA binding and the T=4 nucleocapsid. 240 copies form a T=4 icosahedral
capsid and bind the genomic RNA [file:SINDV/P03316/P03316-uniprot.txt "Forms an
icosahedral capsid with a T=4 symmetry composed of 240 copies of the capsid
protein"]. Cryo-EM of whole virions resolves the complementary capsid/envelope
lattices PMID:3829124 — note that this 1987 35 Å reconstruction
called the core T=3; the modern consensus (and the UniProt text) is T=4 for both
shells.

Localization in infected cells. Live imaging of a tetracysteine-tagged capsid
shows diffuse cytoplasm plus discrete foci, including foci at the plasma membrane
that are candidate assembly/egress sites [PMID:23785213 "Capsid puncta were also
observed at the PM."; "This staining was distributed diffusely in the cytoplasm
as before but also showed distinct foci"]. Nuclear shuttling is asserted only
by similarity to the VEEV capsid (Q8JUX5), via an N-terminal NLS plus a
supraNES; there is no SINV-specific measurement in any source I could read.

Innate-immune evasion. BioID2 proximity labelling from the SINV capsid
identified host IRAK1, and the interaction inhibits IRAK1-dependent TLR
signalling [PMID:33673546 "the alphaviral capsid protein specifically inhibits
IRAK1-dependent signaling"; "this capsid-IRAK1 interaction is conserved across
multiple alphavirus species"]. This is the basis for both the GO:0005515 row
(partner P51617 = IRAK1, a kinase) and the GO:0039722 row.

Host-driven degradation. The capsid is a selective-autophagy (virophagy)
cargo: the host HECT E3 SMURF1 co-immunoprecipitates with it and is required for
its delivery to autophagosomes PMID:22020285. This is a host
activity directed at the capsid, not a viral function, which is why the two
GO:0044389 rows are kept but not treated as core.

Spike glycoprotein E2 (and pE2/E3)

E2 is the attachment protein. Random-insertion mutagenesis of the PE2/E1 genes
followed by selection against binding-competent virus isolated mutants mapping
near the PE2 cleavage site and to E2 residues 69–74, all defective for binding
[PMID:7684466 "Both NB1 and NB2 were defective for binding to CEF and BHK-21
cells."; "these results implicate PE2 cleavage and domains in the N-terminal
portion of E2 as important determinants of alphavirus binding and penetration"].
The receptors were later identified as VLDLR/LRP8 [file:SINDV/P03316/P03316-uniprot.txt
"Plays a role in viral attachment to target host cell, by binding to the cell
receptors VLDLR or LRP8/APOER2 (PubMed:34929721)"].

E2 reaches the plasma membrane by the secretory pathway and is surface-exposed in
infected cells PMID:875134.

6K protein

6K is a small, palmitoylated, multi-pass membrane protein present in virions at
low stoichiometry [PMID:2408229 "the 6K protein can be detected in purified
preparations of virions"; "in virions the ratio of 6K to E1 + E2 ranged from 0.08
to 0.12"].

Viroporin / channel activity. 6K permeabilizes bacterial and mammalian
membranes, and interfacial-domain mutants lose that activity
PMID:12424249. Functionally, SINV release is
sensitive to the Vpu channel blocker HMA and the Vpu channel helix complements 6K
PMID:36191050, and UniProt records the pharmacology
as curated activity regulation [file:SINDV/P03316/P03316-uniprot.txt "The channel
activity is blocked by 5-N, N-Hexamethylene amiloride."].

Caveat worth recording. Heterologous expression of 6K cRNA in Xenopus oocytes
found no intrinsic channel conductance; the currents seen were endogenous
oocyte channels upregulated downstream of ER calcium-store depletion
[PMID:17483865 "Although no intrinsic channel properties were seen, cell shrinkage
was observed within 24 h."; "6K upregulated endogenous currents"]. So the ion-channel
annotations rest on permeabilization, inhibitor pharmacology and heterologous
channel complementation, not on a recorded 6K single-channel conductance. The
oligomeric structure and conduction mechanism are explicitly open
PMID:36191050.

Localization and budding. 6K traffics with E2 through the secretory pathway
[PMID:36191050 "6K localizes to ER and Golgi"; "6K and E2 colocalize on Golgi in
the secretory pathway"] and its channel function is required to make CPV-II from
the Golgi and to deliver spikes to the plasma membrane: "in the absence of 6K,
CPV-II, which originates from the Golgi apparatus, is not detected in infected
cells, with a concomitant reduction in the glycoprotein transport to the plasma
membrane" PMID:36191050. 6K is also at the plasma membrane in oocytes
PMID:17483865.

Spike glycoprotein E1

E1 is a class II fusion protein held inactive by E2 in the mature virion. In a
receptor-free liposome system, SINV fusion is strictly low-pH dependent, needs
cholesterol plus sphingolipid in the target membrane, and is preceded by E1/E2
dissociation and formation of a trypsin-resistant E1 homotrimer
[PMID:10482600 "Fusion was optimal at pH 5.0, with a threshold at pH 6.0, and
undetectable at neutral pH, supporting a cell entry mechanism of SIN involving
fusion from within acidic endosomes."; "the E2/E1 heterodimeric envelope
glycoprotein of the virus dissociated, with formation of a trypsin-resistant E1
homotrimer, which kinetically preceded the fusion reaction, thus suggesting that
the E1 trimer represents the fusion-active conformation of the viral spike"].
E1 reaches the plasma membrane independently of pE2/E2 PMID:2806407, and is
surface-detectable in infected cells PMID:875134.

Where the spikes actually are

Both E1 and E2 sit in the lipid envelope that surrounds the nucleocapsid; the
80 trimeric spikes protrude from that bilayer, not from the capsid shell
PMID:10482600 and
PMID:3829124. This matters for the CACAO row on GO:0098029
icosahedral viral capsid, spike, whose GO definition is "A short structure
attached to an icosahedral virion capsid, and used for attachment to the host
cell" and whose only ontology placement is part_of GO:0019030 icosahedral viral capsid (checked against QuickGO, 2026-10-29). For an enveloped alphavirus the
spike is an envelope structure, so the GO term asserts the wrong compartment.
GO has no envelope-spike CC term (a QuickGO text search for "spike" returns only
GO:0098029 outside the neuroscience senses), which is why this review files one
under proposed_new_terms.

CACAO rows at a glance

Three rows carry ASSIGNED BY = CACAO:

1. GO:0019062 virion attachment to host cell / IMP / PMID:7684466 — the paper's
   whole point is binding-defective E2/PE2 insertion mutants. Right protein,
   right process, right evidence code, right granularity (the paper does not
   identify a receptor, and the row does not claim one). ACCEPT.
2. GO:0020002 host cell plasma membrane / IDA / PMID:17483865 — supported by the
   abstract's own statement, but the assay is 6K cRNA injected into Xenopus
   oocytes, not an infected host cell, and the same sentence reports the ER
   equally. Independent infected-cell data (PMID:875134, PMID:2806407,
   PMID:2408229, PMID:23785213) carry the claim. ACCEPT, with the
   heterologous-expression caveat and the note that the student took one of two
   compartments from a single sentence.
3. GO:0098029 icosahedral viral capsid, spike / IDA / PMID:10482600 and
   GO:0061025 membrane fusion / IDA / PMID:10482600 — the paper is a liposome
   fusion study of E1. GO:0061025 is the generic root of the fusion branch while
   the paper pins down the specific one (low-pH, endosome-like); GO:0098029
   puts the glycoprotein spike on the capsid. Both are term-choice problems, not
   evidence problems: MODIFY in each case.

Failure shapes for the CACAO taxonomy: one parent-level restatement
(membrane fusion where the paper establishes the endosomal fusion step), one
wrong structural compartment — a shape not yet in the taxonomy: the term names
the right supramolecular assembly (the spike) but attaches it to the wrong parent
structure (capsid rather than envelope), which only the term's ontology placement
reveals; and one heterologous-system localization read as host localization
(the oocyte row), a near-relative of the catalogued single-assay localization
problem.