Verdict: SUPPORTED. PP_1969 (UniProt Q88LG4, Pseudomonas putida KT2440) is a
bona fide MoaA ortholog, and GO:0061798 (GTP 3',8'-cyclase activity) is the
correct, appropriately specific molecular function. None of the three
characteristic TreeGrafter failure modes apply:
Single most decisive evidence: 1:1 conservation of the C-terminal
GTP-binding [4Fe-4S] cluster II ligand cysteines (PP_1969 Cys268/Cys271/Cys285 ↔
E. coli MoaA Cys257/Cys260/Cys274) within the conserved twitch motif — this
cluster directly anchors the GTP substrate and defines GTP 3',8'-cyclase catalysis.
Paralogy caveat (resolved, Iteration 2): P. putida KT2440 carries three
MoaA-type radical-SAM paralogs — Q88E69 (the gene formally named moaA), Q88K11,
and Q88LG4/PP_1969. This is a textbook failure-mode-1 trap, but it does not
overturn the call: PP_1969 is actually the paralog most similar to the
characterized E. coli MoaA (40.9% vs 37.5% for the named moaA), and all three
paralogs retain the full six-cysteine active site. PP_1969's assignment is
therefore well justified; the multiplicity reflects lineage-specific MoaA
expansion, not functional reassignment.
| Citation | Evidence type | Call | Claim tested | Key finding | Context | Confidence/limits |
|---|---|---|---|---|---|---|
| PMID 16632608 (Hänzelmann & Schindelin 2006) | Structural/biochemical (direct) | Supports | MoaA = GTP→precursor-Z enzyme using two [4Fe-4S] clusters | MoaA binds 5'-GTP at C-terminal [4Fe-4S] cluster; N-terminal cluster cleaves SAM | E. coli/S. aureus MoaA, crystal + biochem | High; reference enzyme, not PSEPK directly |
| PMID 41183211 (Pang et al. 2025) | Mechanistic (direct) | Supports | MoaA is a radical SAM GTP 3',8-cyclase | Confirms current term/mechanism nomenclature | Radical SAM GTP 3',8-cyclase | High |
| UniProt Q88LG4 xrefs | Review/database + computational | Supports | Domain architecture = MoaA | PF04055 + PF06463; IPR040064 MoaA-like; IPR010505 twitch; PANTHER SF0 | PSEPK PP_1969 | High |
| This report: NW alignment vs P30745 | Computational (evolutionary) | Supports | Ortholog with intact active site | 40.9% id; 6/6 cluster cysteines conserved & spaced | PP_1969 vs E. coli MoaA | High; heuristic aligner, but motifs unambiguous |
| This report: paralog scan (UniProt, Iter 2) | Computational (evolutionary) | Qualifies/supports | Is PP_1969 the MoaA or a diverged paralog? | 3 MoaA paralogs in KT2440; PP_1969 is CLOSEST to E. coli MoaA; all 3 keep 6/6 cys | PSEPK | High |
| This report: AlphaFold AF-Q88LG4 v6 (Iter 2) | Computational (structural) | Supports | Are both [4Fe-4S] sites geometrically intact? | SG–SG 5.8–6.9 Å per triad; pLDDT 86–95; 21.4 Å inter-cluster cleft | PP_1969 model | High for geometry; predicted (no metals in model) |
| UniProt Q88LG4 GO:0061799 (IEA:TreeGrafter) | Database propagation | Competing/qualifies | Does PP_1969 also do the MoaC step? | No MoaC (PF01967) domain present → sibling-term error | PSEPK PP_1969 | High |
Pairwise alignment PP_1969 (Q88LG4) vs characterized E. coli MoaA (P30745):
40.0% identity (126/315 aligned positions).
| Cluster | Role | E. coli MoaA | PP_1969 | Status |
|---|---|---|---|---|
| N-term radical-SAM, Cx₃Cx₂C | SAM reductive cleavage → 5'-dAdo• | Cys24 / Cys28 / Cys31 | Cys36 / Cys40 / Cys43 | Conserved, correct spacing |
| C-term cluster II (GTP-binding) | Anchors 5'-GTP substrate | Cys257 / Cys260 / Cys274 | Cys268 / Cys271 / Cys285 | Conserved, correct spacing |
C-terminal motif alignment (twitch domain):
E.coli P30745: D F C A T C N R L R V S S I G K L H L C L (257,260,...,274)
PP_1969 Q88LG4: P F C R T C S R L R L S S T G W L H G C L (268,271,...,285)
Conserved GGEP motif present (…ITGGEPLV…, ~res 86). All catalytic/metal-binding
residues are intact — no evidence of pseudo-enzyme degeneracy.
3D geometry (AlphaFold AF-Q88LG4-F1 v6, mean pLDDT 90):
| Cluster | Cys triad | SG–SG distances (Å) | pLDDT |
|---|---|---|---|
| N-term radical-SAM | 36 / 40 / 43 | 5.9, 6.9, 6.0 | 86–95 |
| C-term GTP-binding | 268 / 271 / 285 | 6.4, 5.8, 6.3 | 88–92 |
Both triads show SG–SG spacing consistent with 3-Cys coordination of a [4Fe-4S]
cluster (one unique Fe left open for SAM cleavage / GTP binding respectively). The
two cluster centroids sit 21.4 Å apart, flanking the central substrate cleft —
the canonical MoaA arrangement. (Metals are not present in the AF model; geometry
is inferred from the apo cysteine positions.)
Paralog placement (UniProt PANTHER PTHR22960, taxon 160488):
| Accession | Gene | Length | % id to E. coli MoaA | 6-Cys active site |
|---|---|---|---|---|
| Q88LG4 (PP_1969, target) | — | 337 | 40.9% | 6/6 intact |
| Q88E69 | moaA | 334 | 37.5% | 6/6 intact |
| Q88K11 | — | 322 | 36.3% | 6/6 intact |
PP_1969 and Q88K11 are a recent duplication (67.0% mutual identity). PP_1969 is
the closest paralog to the characterized cyclase → not a diverged outlier.
Twitch / GTP-binding region conservation (Iteration 3). Aligning the C-terminal
substrate-binding region (E. coli cols ~248–282) shows the cluster-II motif intact
in all three paralogs:
E.coli : ...DFCATCNRLRVSSIGKLHLCL...
Q88E69 : ...NFCATCNRVRLTVEGRLLLCL...
Q88K11 : ...PFCRSCSRLRLSSTGRLYGCL...
PP_1969: ...PFCRTCSRLRLSSTGWLHGCL...
PP_1969 vs E. coli identity is balanced across domains — N-terminal radical-SAM
41%, C-terminal twitch 39% — i.e. the GTP-recognition domain is not selectively
eroded. This rules out the subtler form of failure-mode 1 (clusters kept but
substrate specificity re-tuned).
PP_1969 directly has the stated function. GO:0061798 is supported and
specific; the only curation fix is removing the co-propagated MoaC sibling term
GO:0061799.