thrB (PA5495, P29364) notes

Identity

UniProt P29364 (KHSE_PSEAE), 316 aa, Swiss-Prot reviewed, PE 3 (inferred from
homology). Homoserine kinase type II (HAMAP MF_00301 Homoser_kinase_2, NCBIfam
TIGR00938 thrB_alt, InterPro IPR005280, "pseudomonas-type ThrB family"); an
aminoglycoside phosphotransferase (APH, Pfam PF01636) / protein-kinase-like fold
enzyme, unrelated to the GHMP-family ThrB of E. coli and B. subtilis.
Catalytic activity in the entry is the HAMAP reaction (L-homoserine + ATP =
O-phospho-L-homoserine + ADP), step 4/5 of threonine synthesis from aspartate
[file:PSEAE/thrB/thrB-uniprot.txt "L-threonine biosynthesis; L-threonine from L-aspartate: step 4/5."].

Biology (literature)

Cloning: hom, thrB and thrC cloned and sequenced; thrB lies apart from hom-thrC
and its product resembles no known ThrB; insertional inactivation gave no
phenotype, so an isozyme was postulated
PMID:1333566.
Redundancy with ThrH: only the thrB thrH double mutant is a threonine
auxotroph PMID:10220164.
ThrH is not an ATP-dependent kinase: it is a phosphoserine phosphatase with
phosphoserine:homoserine phosphotransferase activity
PMID:14699121.
So ThrB is the ATP-dependent homoserine kinase of PAO1, inferred from family
membership plus genetics.
No direct enzyme assay of purified PA5495 was found (PubMed search
"homoserine kinase Pseudomonas aeruginosa thrB" returns only PMID:1333566,
PMID:10220164, PMID:14699121). The cited 1999 abstract assays no ThrB activity;
the full text is not available.
Consistency with genes/PSEAE/hom: that review found PseudoCAP attached
PMID:10220164 (all IDA, same date) to hom, thrB and thrH, and judged it the wrong
paper for hom. For thrB it is the right paper (it is about thrB); the issue here
is the evidence code, not the reference.

PseudoCAP IBA-seed assessment

PTN node	node term	n_seeds	leaf IBAs	seed annotation (term, evidence, PMID)	verdict	rationale
PTN001464661	GO:0009088 L-threonine biosynthetic process	1 (sole seed)	10 (10 taxa)	GO:0009088 L-threonine biosynthetic process, IDA, PMID:10220164 (PseudoCAP)	SOUND	The paper shows that thrB contributes to threonine synthesis (double thrB thrH mutant is auxotrophic), and ThrB catalyses step 4. This is genetic-interaction evidence, so IGI not IDA. Node breadth is a concern: the node also covers B. subtilis AmiN/YerI (O34640), an experimentally characterised amicoumacin kinase in an organism that has a GHMP ThrB, plus uncharacterised Synechocystis sll1119 and Deinococcus DR_0394.
PTN001464661	GO:0004413 homoserine kinase activity	1 (sole seed)	10 (10 taxa)	GO:0004413 homoserine kinase activity, IDA, PMID:10220164 (PseudoCAP)	UNVERIFIABLE	The abstract contains no direct assay of ThrB; it is a genetic study. The activity is very probably right for PA5495 (HAMAP MF_00301, gene redundancy with ThrH, which does the reaction by a different chemistry), so the IBD conclusion is plausible for true type II HSKs. But its only anchor is an IDA whose assay is not visible, and the node includes AmiN, whose characterised activity is antibiotic phosphorylation.
PTN001464650	GO:0019202 amino acid kinase activity	2	53 (34 taxa)	GO:0004413 homoserine kinase activity, IDA, PMID:10220164 (PseudoCAP); second seed human HYKK (A2RU49)	SOUND	Broad term at a node spanning type II HSK and animal hydroxylysine kinases; both seeds are amino acid kinases, and the HYKK seed supports the IBD independently of PseudoCAP. Leaves are mainly metazoan HYKK, for which the general term is correct.

Decisions summary

ACCEPT: homoserine kinase activity (IBA, IEA, IDA), L-threonine biosynthetic
process (IBA, IEA, IDA).
PSEUDOCAP: WRONG_EVIDENCE_CODE CHECK on both IDA rows (PROBABLE for the BP,
which is genetic, IGI; SUSPECTED for the MF, since the full text might contain an
extract assay).
GO_CENTRAL: MISSING_NEGATION CHECK (SUSPECTED) for B. subtilis AmiN/YerI under
PTN001464661.

PANTHER tree check (2026-11-05)

Topology fetched with projects/PSEUDOCAP/panther_tree.py node PTN001464661 --depth 3 --leaves (PTHR21064), plus two --contains queries. The node is a Eubacteria
SPECIATION node, 10 leaves, all PTHR21064:SF6.

Child node	Event / taxon	Leaves	Subfamilies	Verdict
PTN007684856	SPECIATION, Terrabacteria	3 (A9WJ27 Chloroflexus; PTN001843206 Bacillus: O34640 AmiN, Q81HY9)	SF6:3	Over-propagated; AmiN is an experimentally shown amicoumacin kinase and B. subtilis has a separate GHMP ThrB
PTN001843208	LEAF, Deinococcus radiodurans	1 (Q9RXC1, uncharacterised)	SF6	No support
PTN002326065	LEAF, Synechocystis	1 (P73341, uncharacterised)	SF6	No support
PTN001843209	LEAF, Fusobacterium nucleatum	1 (Q8RF07, named homoserine kinase)	SF6	Plausible but unverified; needs its own IBD if kept
PTN002326066	SPECIATION, Proteobacteria	4 (P29364 ThrB, Q4W557 thrB, Q89UU4 thrB, Q8PCP9)	SF6:4	Correct home for both IBDs

--contains P29364,Q4W557,Q89UU4 returns PTN002326066 (4 leaves); --contains P29364,Q8RF07 returns the root PTN001464661 itself, so there is no intermediate node
that keeps the Fusobacterium leaf without also keeping the Terrabacteria amicoumacin
kinase clade. The earlier concern is therefore confirmed by the topology and sharpened:
the defect is kept, raised SUSPECTED -> CONFIRMED, and now names PTN002326066 as the
target node. NOT annotations on the Bacillus leaves become unnecessary if the IBDs move.
The P. aeruginosa rows in this review are unaffected (P29364 is inside the target clade,
so its IBAs stand).