KDTA (At5g03770, Q8VZA5) notes

Deep research via the falcon provider failed (authentication), so the literature
summary below was compiled manually from cached publications and PubMed searches.
No -deep-research-*.md file was produced.

Identity

• Swiss-Prot Q8VZA5, "Probable 3-deoxy-D-manno-octulosonic acid transferase,
  mitochondrial" (AtKdtA), 447 aa, single-copy nuclear gene. Belongs to the
  glycosyltransferase group 1 family, GT30 subfamily (UniProt SIMILARITY, ECO:0000305);
  Pfam PF04413 Glycos_transf_N; InterPro IPR039901 Kdotransferase; PANTHER PTHR42755:SF1.
• Ortholog of E. coli WaaA/KdtA (P0AC75), the bifunctional enzyme that transfers two
  Kdo residues from CMP-Kdo onto lipid IVA (EC 2.4.99.12 and 2.4.99.13).
• Li et al. 2011 report 34% identity to E. coli KdtA
  PMID:21709257.

Function: lipid A precursor pathway (not RG-II)

• Séveno et al. 2010 (abstract only cached) tested the long-standing hypothesis that
  AtKDTA adds Kdo to the pectin rhamnogalacturonan II (RG-II). Null mutants had no
  developmental phenotype and unaltered RG-II
  PMID:20124190.
  GFP fusion targeted to mitochondria
  PMID:20124190.
  Conclusion: PMID:20124190.
• Li et al. 2011 (full text cached) identified Arabidopsis orthologs of LpxA, LpxC,
  LpxD, LpxB, LpxK and KdtA and characterised insertion mutants by LC-ESI/MS.
  The atkdta-1 (SALK_035981) null accumulates the LpxK substrate and is the only line
  in which lipid IVA (the KdtA acceptor) was detected:
  PMID:21709257;
  PMID:21709257.
  Lipid X was also 2.2-fold elevated in atkdta-1
  PMID:21709257.
• Caveat: Kdo-containing products were not detected, and the plant end product is unknown
  PMID:21709257;
  PMID:21709257.
  No in vitro enzyme assay of AtKdtA has been reported in the cached literature, nor
  complementation of an E. coli kdtA mutant (contrast AtLpxA, which complements
  PMID:22545860). Catalytic activity is therefore inferred from
  orthology plus the substrate-accumulation phenotype.
• Mutants are viable without visible phenotype
  PMID:21709257.
• Localisation: Li et al. did not themselves image KdtA; they cite Séveno
  PMID:21709257.
  Lipid X in atlpxb-1 is enriched in mitochondria
  PMID:21709257.
  Sub-mitochondrial location (outer vs inner membrane, matrix) for KdtA is not established.

Kdo in RG-II vs lipid A (disambiguation)

• Kdo is also a constituent of pectic RG-II; CMP-Kdo synthetase (CKS, At1g53000) is
  essential for pollen tube growth and is also mitochondrial
  PMID:21893514;
  PMID:21893514.
  The RG-II Kdo transferase is a different, Golgi enzyme (still unidentified as far as
  the cached literature shows); AtKDTA is not it (Séveno 2010 above). Annotations of
  KDTA to RG-II/cell-wall processes would be wrong.
• Duncan et al. 2011 (mitochondrial outer membrane proteomics) concerns KDSB/CKS
  (At1g53000), not KdtA PMID:21896887
  (note this paper mislabels the cytidylyltransferase as a "Kdo transferase"). Not used
  as evidence for KdtA.
• Guan et al. 2017: mitochondrial FAS (mtKAS/mtHD) mutants have reduced
  3-hydroxytetradecanoic acid, described as a hallmark of lipid A-like molecules
  PMID:28202596.
  Abstract only; context for a mitochondrial acyl-chain supply to the pathway.

Wei et al. 2020 RAR flag (PMID:32470107)

The RAR paper flagged plant GO:0009245 lipid A biosynthetic process annotations as
likely misannotations. For KDTA this flag is refuted: the IBA (PTN000602593, E. coli
WaaA) is supported by gene-specific mutant chemistry (lipid IVA accumulates only in
atkdta-1) and by Séveno 2010 ruling out the alternative RG-II role. Low RAR reflects
sparse plant curation of a minor, non-essential pathway, not absence.

Searches

• PubMed: "(lipid A OR Kdo transferase OR KdtA) AND (Arabidopsis OR plant) AND
  (mitochondria OR lpx)" -> 5 hits (21709257, 20124190, 21893514, 21896887, 28202596).
  No later AtKdtA-specific enzymology paper found.