PA5458 notes

Identity

UniProt Q9HTB2 (unreviewed, PE 4), 380 aa, locus PA5458, no gene name.
Pfam PF01757 Acyl_transf_3 (residues 7-341); InterPro IPR002656; PANTHER
PTHR23028:SF53. Nine Phobius TM helices (7-26 ... 291-312), so a polytopic
integral membrane protein, in a Gram-negative bacterium the cytoplasmic
(inner) membrane.
No deep-research provider available; literature research done manually
(PubMed, cached full text of PMID:23341552).

Literature

Only one paper studies PA5458: Hao et al. 2013 mBio, PMID:23341552 (full text cached).

PA5458 is in the pa5455-pa5459 cluster adjacent to the CPA (common polysaccharide
antigen, D-rhamnan O antigen) locus pa5447-pa5454; the 5-gene cluster is an operon
and is required for CPA biosynthesis PMID:23341552.
Domain: PMID:23341552
In-frame deletion: CPA still made but altered. PMID:23341552; PMID:23341552;
weaker MAb N1F10 binding to the mutant polymer PMID:23341552. Complementation partly restored it.
Overexpression: PMID:23341552.
Authors' interpretation: PMID:23341552.
No chemical analysis of acetylation, no enzyme assay, no localisation experiment.
Acetyltransferase activity is a domain-based prediction.

Assessment: PA5458 is very probably the membrane-bound O-acyltransferase that
decorates CPA (probably O-acetylation), i.e. a modification step of O-antigen
biosynthesis. The mutant phenotype (altered polymer structure, not merely amount)
fits the gene acting on the polymer itself, so involvement in O antigen
biosynthesis is acceptable at the gene level. The IDA evidence code on the
PseudoCAP O-antigen row is wrong: the evidence is mutant (IMP) and overexpression
analysis, not a direct assay.

Family context

Acyltransferase-3 (AT3, PF01757) proteins are polytopic inner-membrane
acyltransferases that O-acylate cell-surface glycans, with different substrates in
different lineages: S. aureus OatA O-acetylates peptidoglycan muramic acid
PMID:15661003, and
Salmonella OafA acetylates O antigen. Both sit in the same PAINT node as PA5458
(see bioinformatics).

PAINT node composition

See PA5458-bioinformatics/RESULTS.md.

PTN000546841 (GO:0009103 LPS biosynthetic process; 31 leaves): 15 leaves are in
organisms with no LPS (M. tuberculosis x5, S. coelicolor x4, S. aureus x2 incl. OatA,
L. monocytogenes OatA, S. pneumoniae spr1868, B. subtilis YrhL, B. cereus).
PTN000546715 (GO:0000271 polysaccharide biosynthetic process; 113 leaves): 109
eukaryotic, 97 nematode. Nematode AT3 proteins have no known polysaccharide role.

Comparator check: S. aureus OatA (Q2FV54) and L. monocytogenes OatA (Q8Y7I6), whose
function is peptidoglycan O-acetylation, carry only IBA GO:0009103 for process in
QuickGO (queried 2026-10-28): the propagated term is wrong for them.

PseudoCAP IBA-seed assessment

PTN node	node term	n_seeds	leaf IBAs	seed annotation (term, evidence, PMID)	verdict	rationale
PTN000546841	GO:0009103 lipopolysaccharide biosynthetic process	1	31	GO:0009103 IMP PMID:23341552 (plus GO:0009243 IDA, same paper)	WRONG	The seed is sound for PA5458 itself (the deletion alters CPA structure), but one Gram-negative O-antigen seed cannot justify LPS biosynthesis at a node that also contains Gram-positive peptidoglycan O-acetyltransferases (OatA of S. aureus and L. monocytogenes) and mycobacterial/streptomycete members: 15 of 31 leaves are in organisms that make no LPS. The node term should be withdrawn or moved to a Gram-negative O-antigen subclade (with OafA).
PTN000546715	GO:0000271 polysaccharide biosynthetic process	1	113	GO:0009103 IMP PMID:23341552 (plus GO:0009243 IDA)	WRONG	Generalising one bacterial O-antigen modification phenotype to a deep node whose leaves are 109/113 eukaryotic (97 nematode, 9 oomycete) AT3 proteins of unknown function. The seed shows modification of an already made polymer, not synthesis; no evidence for polysaccharide synthesis in the propagated clade.

Seed rows themselves: GO:0009103 IMP is acceptable but under-specific (the gene
already carries GO:0009243); GO:0009243 is the right term but its IDA code should be
IMP.