AIGR Deep Research Report — pvdD (Q88F79, PP_4219), Pseudomonas putida KT2440

Focus: Function-assignment hypothesis — "PvdD participates in monocarboxylic acid biosynthesis (GO:0072330)"

Organism: Pseudomonas putida KT2440 (NCBITaxon:160488) · Gene: pvdD · UniProt: Q88F79 · Locus: PP_4219 · Focus type: function_assignment

Summary

The seed hypothesis proposes that P. putida KT2440 PvdD should be curated under GO:0072330 "monocarboxylic acid biosynthetic process" because, in building its native pyoverdine, it produces molecules that could fall within the broad monocarboxylic-acid chemical class. Importantly, the hypothesis asks the reviewer to do the honest chemistry — to count the free carboxyl groups of the actual KT2440 products (PVD G4R and G4R A, the succinamide/cyclized and succinate/linear variants) rather than inferring the count from amino-acid names — and to consider that "being a pyoverdine rather than an enterobactin synthetase does not alone exclude this broad chemical class."

Performing exactly that structural accounting refutes the hypothesis. The native KT2440 pyoverdine is a polycarboxylic chromopeptide siderophore carrying 2–3 free carboxylic acid groups (Asp β-COOH, threo-β-hydroxy-Asp β-COOH, and — in the succinate variant only — the free succinate carboxyl). A molecule with two or three free carboxyls is, by the term's own definition, not a monocarboxylic acid. The chemical-class defense the hypothesis invites therefore fails on chemistry, not on taxonomy. Separately, PvdD's direct molecular activity is nonribosomal peptide synthesis — it is the terminal, chain-releasing NRPS subunit — and it does not itself synthesize the succinyl side chain (the fragment that most resembles a small monocarboxylic acid); that acyl group is installed by separate maturation enzymes.

The annotation itself is also weak and unstable. All nine current GO annotations on Q88F79 are IEA (electronic, zero experimental). GO:0072330 specifically originates from UniProtKB-ARBA association-rule automation and is absent from the live GOA/QuickGO extract, where the corresponding broad automated process term is instead GO:0044550 "secondary metabolite biosynthetic process." GO:0072330 co-occurs with a cluster of clearly erroneous enterobactin-branch TreeGrafter mismaps — a textbook signature of NRPS-family paralog over-annotation. The correct, biologically supported terms are GO:0002049 "pyoverdine biosynthetic process" and GO:0019290 "siderophore biosynthetic process" (BP), with GO:0031177 "phosphopantetheine binding" retained (MF). Verdict: REFUTED / OVER-ANNOTATED.

Key Findings

Finding 1 — The native KT2440 pyoverdine is a polycarboxylic chromopeptide siderophore, not a monocarboxylic acid

The hypothesis is right to demand a genuine free-carboxyl count from the actual structures. When that accounting is performed, it defeats the hypothesis rather than supporting it.

Structural mass-spectrometry studies define the real KT2440 products — the pyoverdines PVD G4R and G4R A — as a three-part architecture: (i) a dihydroxyquinoline-type chromophore, (ii) a heptapeptide chain Asp–Orn–OHAsp–Dab–Gly–Ser–cOHOrn, and (iii) a C-4 diacid side chain attached to the chromophore (succinic acid or its monoamide/"succinamide" derivative; a malamide/malic variant is reported for the related P. taiwanensis strain, not KT2440). PMID: 25895945 states that the pyoverdine "consists of a dihydroxyquinoline-type chromophore, a peptide chain of variable length and conformation, and a side chain composed of a dicarboxylic acid or its monoamide derivative," and that "the three PVDs of P. putida possess the same peptide chain of seven amino acids, Asp-Orn-OHAsp-Dab-Gly-Ser-cOHOrn, with a cyclicized portion present in two of the PVDs."

Free-carboxyl accounting:

Thus the molecule carries 2–3 free carboxylic acid groups — it is polycarboxylic. Its iron-binding triad is catecholate (chromophore) + hydroxamate (cOHOrn) + α-hydroxy-carboxylate (OHAsp), a classic mixed-ligand siderophore rather than a simple organic acid. Product identity is independently confirmed by PMID: 28631237: "In Pseudomonas sp. putida KT2[2]40 samples, the known pyoverdines G4R and G4R A were readily confirmed."

Because GO:0072330 is defined by a product class carrying exactly one carboxyl group, a molecule with two or three free carboxyls is excluded by the term's own definition. The hypothesis's chemical-class reasoning, applied rigorously, lands on refutation.

Finding 2 — GO:0072330 on Q88F79 is automated ARBA over-annotation; the direct function is the terminal NRPS module of pyoverdine biosynthesis

The UniProt Q88F79 entry describes a 3,470-amino-acid multi-modular protein with Pfam architecture Condensation (PF00668) + AMP-binding/adenylation (PF00501) + PP-binding/thiolation-PCP (PF00550) + a second AMP-binding module + Thioesterase (PF00975) — the canonical signature of a terminal, chain-releasing NRPS subunit that adds the final residues and releases/cyclizes the mature peptide.

The direct functional evidence comes from the well-characterized P. aeruginosa ortholog. PMID: 25015884 establishes that "the Pseudomonas aeruginosa nonribosomal peptide synthetase (NRPS) PvdD contains two modules that each incorporate an l-threonine residue at the C-terminal end of pyoverdine." PMID: 26268580 reinforces the mechanism: "Non-ribosomal peptide synthetase (NRPS) enzymes govern the assembly of amino acids and related monomers into peptide-like natural products." PvdD's immediate biochemical acts are therefore ATP-dependent amino-acid adenylation, phosphopantetheine-mediated thiolation, condensation (peptide-bond formation), and thioesterase-catalyzed release/macrocyclization — nonribosomal peptide synthesis, not the biosynthesis of a free monocarboxylic acid.

Within the IEA-only GO panel, GO:0072330 is an IEA:UniProtKB-ARBA call. The remaining machine annotations include an enterobactin cluster that is simply wrong for a pyoverdine synthetase — GO:0009366 "enterobactin synthetase complex," GO:0047527 "2,3-dihydroxybenzoate–serine ligase activity," and GO:0009239 "enterobactin biosynthetic process" — all TreeGrafter mismaps arising because both enterobactin and pyoverdine systems are NRPS-family. Only GO:0031177 (phosphopantetheine binding) and the generic amino-acid-activation concept are appropriate; GO:0072330 is not.

Finding 3 — GO definitions and live QuickGO provenance confirm GO:0072330 is inapt automated annotation; all Q88F79 evidence is IEA

To make the decision auditable, term definitions and the live annotation set were pulled directly from EBI QuickGO.

• GO:0072330 "monocarboxylic acid biosynthetic process" = "the chemical reactions and pathways resulting in the formation of monocarboxylic acids, any organic acid containing one carboxyl (–COOH) group." The KT2440 pyoverdine has 2–3 free carboxyls → fails the single-COOH definition. Reinforced by PMID: 25895945: the side chain is "a dicarboxylic acid or its monoamide derivative."
• GO:0002049 "pyoverdine biosynthetic process" = "formation of the siderochrome pyoverdine" → exact fit.
• GO:0019290 "siderophore biosynthetic process" → fits as the parent process.
• GO:0043041 ("amino acid activation for nonribosomal peptide biosynthetic process") is now OBSOLETE — cannot be retained verbatim.

A live QuickGO extract for Q88F79 returned 9 annotations, ALL IEA (ECO:0000501 UniProt or ECO:0007826 TreeGrafter), zero experimental:

Notably, GO:0072330 appears in the UniProtKB entry (ARBA prediction) but is NOT in the current GOA/QuickGO extract, where the broad automated BP is instead GO:0044550. This instability — the term flickering between resources across releases — is itself diagnostic of a low-confidence, non-curated automated call that should not be promoted into a curated review.

Mechanistic Model / Interpretation

The refutation rests on separating three levels: the enzyme's direct reaction, the product's chemical class, and the automated annotation provenance. All three point away from GO:0072330.

   DIRECT MOLECULAR FUNCTION (what PvdD actually catalyzes)
   ┌─────────────────────────────────────────────────────────────┐
   │  PvdD = terminal NRPS subunit (3,470 aa)                      │
   │  [C] – [A] – [T/PCP] – [A] – [T/PCP] – [TE]                   │
   │   condensation · adenylation · thiolation · thioesterase      │
   │  → activates aa, forms peptide bonds, releases C-terminus     │
   │  MF: GO:0031177 (Ppant binding); NRPS / amino-acid activation │
   └─────────────────────────────────────────────────────────────┘
              │  assembles
              ▼
   PRODUCT (native KT2440 pyoverdine: PVD G4R / G4R A)
   ┌─────────────────────────────────────────────────────────────┐
   │  Chromophore (dihydroxyquinoline, catecholate Fe ligand)      │
   │      │ C1-carboxyl = AMIDE to peptide (not free acid)         │
   │      │ side chain = succinate / succinamide / malamide        │
   │  Peptide: Asp–Orn–OHAsp–Dab–Gly–Ser–cOHOrn                    │
   │      Asp β-COOH ........ FREE acid (1)                        │
   │      OHAsp β-COOH ...... FREE acid + α-OH-acid Fe ligand (1)  │
   │      cOHOrn ............ CYCLIC HYDROXAMATE (not acid)        │
   │  Side-chain COOH (succinate variant only) .... FREE acid (1)  │
   │  ───────────────────────────────────────────────            │
   │  Σ FREE carboxyls = 2–3  →  POLYCARBOXYLIC                    │
   └─────────────────────────────────────────────────────────────┘
              │  therefore
              ▼
   CHEMICAL-CLASS TEST against GO:0072330
   ┌─────────────────────────────────────────────────────────────┐
   │  GO:0072330 = "one carboxyl (-COOH) group" (monocarboxylic)  │
   │  Product carries 2–3 free carboxyls  →  DOES NOT MATCH        │
   │  Correct BP: GO:0002049 pyoverdine biosynthetic process       │
   │              GO:0019290 siderophore biosynthetic process      │
   └─────────────────────────────────────────────────────────────┘

The key logical trap the seed sets — "do not infer the count solely from amino-acid names" and "the broad chemical class is not excluded just because it's a pyoverdine" — is answered by doing the accounting properly. The molecule is polycarboxylic, so it fails the monocarboxylic definition. The broad-class defense would apply only if the product had exactly one free carboxyl; it does not. And even under the most permissive reading, PvdD does not synthesize the succinyl acyl group (installed by separate acylation/tailoring machinery), so attributing free-acid production to PvdD's catalytic activity is doubly unsupported. PvdD's thioesterase-mediated macrocyclization (to cyclo-OHOrn) in fact removes a free carboxyl by lactamization rather than creating a monocarboxylic acid.

Evidence Base (Evidence Matrix)

How the literature supports or challenges the findings

• Structural characterization of multiple pyoverdines... PMID: 25895945 — Primary structural anchor defining the KT2440 scaffold and the dicarboxylic/monoamide side chain. Challenges the monocarboxylic classification by demonstrating multiple carboxyl groups.
• Structural characterization of pyoverdines produced by P. putida KT2440... PMID: 28631237 — Independently confirms G4R/G4R A identity. Supports the product identity used in the accounting.
• Biosynthesis of novel Pyoverdines by domain substitution in an NRPS of P. aeruginosa. PMID: 25015884 — Direct functional characterization of PvdD as a two-module terminal NRPS in the ortholog. Supports the true molecular function.
• Portability of the thiolation domain in recombinant pyoverdine NRPS. PMID: 26268580 — Establishes NRPS mechanism/engineering for pyoverdine synthetases. Supports the NRPS classification.
• Siderophore-mediated iron acquisition in P. entomophila L48 and P. putida KT2440. PMID: 19459056 — Physiological context; KT2440 makes a single characterized pyoverdine and no second siderophore. Orients the siderophore biology.

GO Curation Implications (leads — require curator verification)

• GO:0072330 (monocarboxylic acid biosynthetic process, BP): Do not curate / recommend removal of reliance on it. IEA-ARBA only, chemically inapt (product is polycarboxylic), unstable across releases, and not PvdD's direct activity. Not suitable as a positive curation statement.
• GO:0009366 / GO:0047527 / GO:0009239 (enterobactin complex / DHB-Ser ligase / enterobactin biosynthesis): Remove — wrong pathway/product (pyoverdine ≠ enterobactin); NRPS-family paralog mismap.
• GO:0044550 (secondary metabolite biosynthetic process, BP): present in live GOA (IEA:UniProt); acceptable as a broad parent but should be superseded by the specific siderophore/pyoverdine terms.
• GO:0043041 (BP): now OBSOLETE — do not retain verbatim; replace with a current NRPS/pyoverdine term.
• Retain / prefer instead:
• MF: GO:0031177 phosphopantetheine binding (retain); consider adenylation/ligase and NRPS thioesterase activity terms supported by domain content (GO:0003824 "catalytic activity" is too generic).
• BP: add the specific GO:0002049 pyoverdine biosynthetic process and/or GO:0019290 siderophore biosynthetic process.
• CC: cytoplasm/cytosol (GO:0005737/GO:0005829) plausible; remove enterobactin synthetase complex (GO:0009366).
• Avoid "protein binding" as an endpoint; specific NRPS/siderophore terms are supported.

Mechanistic Scope

Direct activity (what PvdD does): ATP-dependent amino-acid adenylation → phosphopantetheine-mediated thiolation → condensation (peptide-bond formation) → thioesterase-catalyzed release/macrocyclization of the C-terminal pyoverdine peptide. This is captured by NRPS/peptide-biosynthesis MF concepts and by GO:0002049 / GO:0019290 at the process level; GO:0031177 is directly supported by the PP-binding/PCP domains.

Out of scope (downstream / other enzymes): chromophore formation, N-acylation with the succinate/malate side chain, side-chain amidation, and iron chelation. The free carboxyls in the finished siderophore are properties of the assembled multi-enzyme product and of separate building-block and tailoring enzymes — not a monocarboxylic-acid product of PvdD. Iron acquisition and growth/competition phenotypes are downstream physiological consequences, not PvdD's molecular function.

Conflicts and Alternatives

1. Paralog confusion: the enterobactin annotations and the ARBA monocarboxylic-acid rule both derive from the generic NRPS/AMP-binding domain, not from PvdD-specific evidence.
2. Strain-specificity: the malic-acid side chain and the Orn-Asp-OHAsn-Thr… sequence emphasized by PMID: 28631237 pertain to P. taiwanensis VLB120; the KT2440 backbone is Asp-Orn-OHAsp-Dab-Gly-Ser-cOHOrn (PMID: 25895945). Do not conflate.
3. Product-variant ambiguity: succinate vs. succinamide (monoamide) vs. malamide variants shift the exact free-carboxyl count (2 vs. 3), but every variant is polycarboxylic (≥2 free –COOH), so the monocarboxylic classification fails regardless.
4. Charitable reading: even if the succinate side chain is formally a monocarboxylic acyl group, that acyl group is not synthesized by PvdD, so the term still fails for PvdD specifically.
5. The seed's "subsequent" reference: PMID:31451546 (arginine biosynthesis modulating pyoverdine production under oxidative-stress adaptation) is a physiology/genetics paper — it provides no pyoverdine structural/carboxyl data and does not support a monocarboxylic-acid function for PvdD.
6. Annotation instability: GO:0072330 (UniProtKB-ARBA) vs GO:0044550 (live GOA) for the same protein shows the broad automated BP is version-dependent and should not drive curation.

Limitations and Knowledge Gaps

Discriminating Tests

1. Curator provenance check (fastest, decisive): Confirm in QuickGO/GOA that GO:0072330 is IEA:UniProtKB-ARBA and absent from the curated GOA extract for Q88F79. This alone justifies not promoting it.
2. Structural free-COOH audit: For each KT2440 variant (G4R, G4R A, succinamide, malamide), enumerate free carboxyls from published structures; enumerate protonation-state–resolved carboxyls of apo- vs ferri-pyoverdine. Expected: ≥2 in all cases → fails monocarboxylic definition.
3. Domain-architecture verification: Re-run InterPro/Pfam on Q88F79 to confirm C-A-T-A-T-TE terminal-NRPS architecture and PP-binding domains (supporting GO:0031177), and the absence of enterobactin-specific domains that would justify the mismapped terms.
4. Ortholog cross-check: Align Q88F79 with P. aeruginosa PvdD to confirm orthology and transfer the two-module terminal-NRPS function (PMID: 25015884) as ISS/IBA.
5. Mutant experiment (definitive in-organism): Delete pvdD in KT2440 and confirm loss/truncation of PVD G4R/G4R A by LC–HR-MS/MS, establishing GO:0002049 with experimental evidence, plus complementation to show PvdD builds the C-terminal peptide (not the acyl side chain).

Proposed Follow-up Actions / Curation Leads (verify before applying)

• Action: Reject the hypothesis as a basis to assert GO:0072330; flag GO:0072330 and all enterobactin terms as IEA over-annotations.
• Candidate references (with snippets to verify):
• PMID: 25895945: "side chain composed of a dicarboxylic acid or its monoamide derivative"; "the same peptide chain of seven amino acids, **Asp-Orn-OHAsp-Dab-Gly-Ser-cOHOrn." → polycarboxylic structure.
• PMID: 28631237: "the known pyoverdines G4R and G4R A were readily confirmed." → product identity.
• PMID: 25015884: "PvdD contains two modules that each incorporate an l-threonine residue at the C-terminal end of pyoverdine." → true molecular function.
• PMID: 26268580: "NRPS enzymes govern the assembly of amino acids... into peptide-like natural products." → NRPS classification.
• Candidate replacement / new GO terms: GO:0002049 (pyoverdine biosynthetic process), GO:0019290 (siderophore biosynthetic process); retain GO:0031177 (phosphopantetheine binding). Note GO:0043041 is obsolete.
• Suggested curator questions: Does any source provide experimental (EXP/IDA/IMP) support for GO:0072330 on PvdD? Should broad chemical-class BP terms be suppressed when a specific pathway term exists?
• Suggested experiments: ΔpvdD deletion + LC–HR-MS/MS pyoverdine readout for in-organism experimental evidence; InterPro/Pfam re-scan to document domain support for GO:0031177 and absence of enterobactin domains.

Conclusion

The hypothesis that PvdD (Q88F79) should be curated under GO:0072330 "monocarboxylic acid biosynthetic process" is refuted. The native KT2440 pyoverdine is a polycarboxylic chromopeptide siderophore (2–3 free –COOH), which fails the term's single-carboxyl definition; the annotation is unsupported, unstable IEA:UniProtKB-ARBA automation co-occurring with erroneous enterobactin mismaps; and PvdD's direct function is nonribosomal peptide synthesis (terminal NRPS module of the pyoverdine synthetase). The curation should instead use GO:0002049 (pyoverdine biosynthetic process) and GO:0019290 (siderophore biosynthetic process), retain GO:0031177 (phosphopantetheine binding), and remove the erroneous enterobactin terms (GO:0009366, GO:0047527, GO:0009239).

Artifacts

• OpenScientist final report
• OpenScientist final report