AIGR Gene Hypothesis Deep Research — Final Report

Blinded Sequence-Only Functional Assignment: Salmonella Typhimurium LT2 ybbO (Q7CR11)

Focus type: function_assignment · Hypothesis slug: seq-only-blinded-activity
Term under evaluation: alcohol dehydrogenase (NADP+) activity — GO:0008106
Organism: Salmonella enterica subsp. enterica serovar Typhimurium str. LT2 (NCBITaxon:99287)

Executive Judgment

Verdict: SUPPORTED (as an evidence-transfer / ISS–ISO curation lead).

Starting only from the 256-amino-acid query sequence with gene name and accession deliberately withheld, this investigation established the protein's identity, family, catalytic integrity, and most likely specific activity, and tested the seed hypothesis against competing activities (hydroxysteroid dehydrogenase, retinol dehydrogenase).

The blinded query was de-anonymized by an exact UniParc CRC64 checksum match to Salmonella Typhimurium LT2 gene ybbO (UniProt Q7CR11). The protein is a textbook classical short-chain dehydrogenase/reductase (SDR): Pfam PF00106, InterPro IPR002347, PROSITE PS00061, with the Rossmann-fold NAD(P)-binding superfamily signature. Its closest experimentally characterized relative is Escherichia coli K-12 YbbO (P0AFP4), to which it is 90.6% identical over 256 aa with 100% conservation of the catalytic tetrad and the NADP(+)-selecting basic residues. E. coli YbbO carries exactly the seed GO term — GO:0008106 "alcohol dehydrogenase (NADP+) activity" — supported by two independent direct-assay (IDA) primary papers (PMID: 25108218 and PMID: 27134112).

The seed hypothesis — "NADPH-dependent aldehyde reductase that reduces aliphatic aldehydes to primary alcohols" — is therefore supported with high confidence and can be narrowed precisely: the preferred substrate class is medium-to-long-chain aliphatic aldehydes (C6–C18, optimal C16/C18). The two competing activities were evaluated and resolved: the hydroxysteroid dehydrogenase (HSD) alternative is refuted — it is a CDD subfamily-naming artifact (cd05374), not substrate evidence — and the retinol dehydrogenase activity is real but promiscuous and non-native, since neither Salmonella nor E. coli has an endogenous retinoid pathway. Most important caveat: the Salmonella ortholog itself has not been directly assayed; all enzymatic evidence comes from the E. coli ortholog. The correct curation action is therefore an ISS/ISO evidence transfer, not an assertion of direct assay (IDA) on Q7CR11.

Key Findings

F001 — Identity: the blinded query is Salmonella LT2 YbbO (Q7CR11), 91% identical to assay-characterized E. coli YbbO

The 256-residue query sequence was identified by an exact UniParc CRC64 checksum match (checksum 22B5C5A48E6D813A → UPI000005A1DF), which uniquely maps the sequence to Salmonella Typhimurium LT2 gene ybbO (UniProtKB Q7CR11). This is a deterministic identity match, not a similarity inference.

The nearest reviewed, experimentally characterized homolog is E. coli K-12 YbbO (P0AFP4, 269 aa), which carries a 13-residue N-terminal extension relative to the query. After accounting for that offset, the ungapped identity is 90.6% over 256 aa. The cofactor-binding glycine-rich loop and catalytic region are visibly conserved (query LITGCSSGIGL…LAGCRKPDD vs E. coli LITGCSSGIGL…LAGCRKPDD). E. coli YbbO is annotated GO:0008106 "alcohol dehydrogenase (NADP+) activity" with evidence code IDA:EcoCyc (direct assay). Because the deletion of ybbO in E. coli caused a >90% reduction in long-chain fatty alcohol production (PMID: 27134112), the ortholog's role as an NADP(+)-dependent aldehyde reductase is directly established, and the query is its very close homolog.

F002 — Specific activity: broad-specificity NADPH aldehyde reductase (aliphatic C6–C18, optimal C16/C18); retinal reduction is promiscuous; no hydroxysteroid activity

Enzyme kinetics of E. coli YbbO show a wide aldehyde substrate specificity ranging from C6 to C18, with maximum affinity and catalytic efficiency for C18 and C16 chain-length aldehydes (PMID: 27134112). An independent functional screen identified YbbO among five new aldehyde reductases (ALRs) — YahK, DkgA, GldA, YbbO, and YghA — that convert aldehydes to the corresponding alcohols (PMID: 25108218). These two lines of direct evidence converge on aliphatic aldehydes as the preferred substrate class, precisely matching the seed hypothesis once narrowed to medium/long-chain aliphatic aldehydes.

Several metabolic-engineering studies show that ybbO also reduces retinal to retinol (PMID: 25726762, PMID: 34850377, PMID: 39307776, PMID: 41093029). However, in those systems retinal remains a major product (PMID: 34850377), indicating retinal is an inefficient, promiscuous, non-native substrate rather than the physiological one. Neither Salmonella nor E. coli possesses an endogenous retinoid biosynthetic pathway, so retinol dehydrogenase cannot be the gene's native function. There is no sequence or literature evidence for hydroxysteroid or dedicated retinol dehydrogenase classification.

F003 — Annotation provenance: two independent IDA papers underpin GO:0008106; the Salmonella ortholog lacks it; InterPro = classical SDR; the HSD label is a CDD artifact

QuickGO shows that E. coli YbbO (P0AFP4) is annotated GO:0008106 "alcohol dehydrogenase (NADP+) activity" with evidence IDA (ECO:0000314) from two primary references: PMID: 25108218 and PMID: 27134112. By contrast, the Salmonella ortholog Q7CR11 currently carries only GO:0016491 (oxidoreductase, evidence IBA ECO:0000318, PANTHER PTN000672121) — confirming the seed term has not yet been propagated to the query protein.

InterPro assigns Q7CR11 unambiguously to the classical SDR family:

Critically, the CDD cd05374 assignment is a broad classical-SDR cluster named after a representative member (17β-HSD-like) — it is not evidence that YbbO acts on steroid substrates. This resolves the "hydroxysteroid dehydrogenase" competing hypothesis as a naming artifact.

F004 — Catalytic integrity: 100% conservation of the SDR catalytic tetrad and NADP-binding residues justifies maximal functional transfer

Residue-level alignment of Salmonella Q7CR11 against E. coli P0AFP4 (offset +13) shows identity at every functionally critical position:

The catalytic signature region GAYAASKYA (containing the Tyr–x–x–x–Lys motif diagnostic of SDR catalysis) is identical between the two proteins. The Arg34/Lys35 pair is the hallmark of NADPH preference in the SDR family — the structural literature establishes that a conserved pair of basic residues contacting the cofactor 2′-phosphate confers NADPH over NADH selectivity (PMID: 8805511). With overall identity of 90.6% and 100% catalytic/cofactor conservation, the conditions for confident ortholog-based functional transfer (ISS/ISO) are maximally satisfied.

Mechanistic Model / Interpretation

Reaction catalyzed (native, inferred)

   R–CHO  +  NADPH  +  H⁺   ⇌   R–CH₂OH  +  NADP⁺
 (aliphatic                      (primary
  aldehyde,                       alcohol)
  C6–C18;
  optimal C16/C18)

The enzyme is a cytoplasmic, NADPH-dependent aliphatic aldehyde reductase operating in the thermodynamically favorable reductive direction (aldehyde → alcohol). Its physiological role in the native organism is best understood as aldehyde detoxification / reductive metabolism of reactive aliphatic aldehydes, consistent with E. coli YbbO's contribution to fatty alcohol production from fatty aldehyde intermediates.

SDR catalytic logic

NADPH (2'-P recognized by Arg34 / Lys35)
     │ hydride (4-pro-S)
     ▼
   Tyr146 (general acid) ── Ser132/133 (substrate positioning)
│                      │
└── Lys150 lowers Tyr pKa, binds nicotinamide ribose
     │
     Asn81 — part of proton-relay to bulk solvent
     ▼
    R–CHO carbonyl protonated → R–CH₂OH

The query possesses the complete, canonical catalytic apparatus of classical SDRs (Asn–Ser–Tyr–Lys tetrad plus the Rossmann glycine-rich cofactor loop) with no substitutions at any catalytic position. The NADP-selecting Arg/Lys pair explains the enzyme's NADPH (not NADH) cofactor preference, consistent with the seed hypothesis and with the structural determinants documented for NADPH-preferring SDRs (PMID: 8805511).

Substrate-class reasoning and disambiguation of competing hypotheses

The mechanistic conclusion is that GO:0008106 captures the primary molecular function, best specified as reduction of aliphatic aldehydes to primary alcohols; retinol dehydrogenase activity is a downstream in vitro curiosity exploited by metabolic engineers, and hydroxysteroid dehydrogenase activity is unsupported.

Evidence Matrix

GO Curation Implications

Lead curation action (requires curator verification): Add GO:0008106 "alcohol dehydrogenase (NADP+) activity" to Salmonella YbbO (Q7CR11) as an ISS/ISO evidence-transfer annotation, with with/from = UniProtKB:P0AFP4 (the directly-assayed E. coli ortholog), citing the two IDA references that anchor the ortholog term.

• The term supports a Molecular Function (MF) assignment. GO:0008106 should be retained/added as the most informative MF term; the generic GO:0016491 (oxidoreductase, IBA) currently on Q7CR11 is correct but less specific and can be refined.
• A more specific child term describing aliphatic/long-chain aldehyde reductase activity would best capture the kinetics (C6–C18, optimal C16/C18) if an appropriate GO term exists; otherwise GO:0008106 is the best-supported node.
• Biological Process: a lead-level BP annotation to aldehyde catabolism / detoxification is plausible but not directly demonstrated in Salmonella; flag as a question for the curator rather than asserting.
• Cellular Component: cytoplasm (consistent with soluble SDR); low priority, inferable but not experimentally pinned for the query.
• "protein binding" is not an appropriate recommendation here — a specific catalytic MF term is well supported.

Mechanistic Scope

The immediate molecular function being tested is a catalytic, NADPH-dependent hydride transfer that reduces the carbonyl of an aliphatic aldehyde to a primary alcohol. This is a direct gene-product activity grounded in: (1) complete conservation of the SDR catalytic tetrad and NADP-selecting residues in the query, and (2) direct in vitro kinetics plus a loss-of-function metabolic phenotype on the near-identical E. coli ortholog.

The following are explicitly downstream, pleiotropic, or context-specific and must not be conflated with the core MF:
- Fatty alcohol production (a pathway consequence observed when YbbO acts on fatty aldehyde intermediates in engineered strains).
- Retinol production (a non-native, promiscuous reduction exploited only in engineered β-carotene/retinoid strains; absent any native retinoid pathway).
- Any biocide-tolerance phenotypes (e.g., glutaraldehyde tolerance) are organism/stress-response phenomena and are not evidence that YbbO's core function is glutaraldehyde reduction.

The seed hypothesis correctly targets the direct molecular activity, and the evidence supports it at the activity level while allowing precise substrate narrowing.

Conflicts and Alternatives

1. Retinol dehydrogenase (competing activity). Multiple engineering papers report ybbO-mediated retinal → retinol conversion, and one screen even ranked ybbO highest among tested alcohol dehydrogenases for this reaction (95.6% conversion; PMID: 39307776). This is genuine catalytic promiscuity, not evidence of native function: retinal remains a major product in these systems (PMID: 34850377), and the host bacteria lack any endogenous retinoid pathway. Resolution: retinol reduction is a non-core, in-vitro/heterologous activity; do not replace GO:0008106 with a retinol dehydrogenase term.

2. Hydroxysteroid dehydrogenase (competing activity). The CDD domain hit cd05374 is labeled "17β-hydroxysteroid dehydrogenase-like, classical (c) SDRs." This is a broad cluster named after a representative, not a substrate determination. Steroid specificity in true HSDs depends on dedicated substrate-pocket residues (e.g., His221 in human 17β-HSD1; PMID: 9525918), for which there is no evidence in YbbO. Resolution: refute the HSD alternative as a database-naming artifact.

3. Paralog / family over-annotation risk. Classical SDRs are a large family and functional transfer across distant SDRs is dangerous. Here, however, the transfer is not generic family inference — it is from a 90.6%-identical ortholog with 100% catalytic conservation and two IDA papers. This is a best-case scenario for ISS/ISO transfer, which mitigates (but does not eliminate) the usual over-annotation concern.

4. Organism-specific difference. The Salmonella protein has not been assayed; a formal possibility of divergent substrate preference remains, though near-identity makes substantive functional divergence unlikely.

Limitations and Knowledge Gaps

Discriminating Tests

The following would most efficiently distinguish the supported hypothesis from its alternatives:

1. In vitro kinetic substrate panel on purified Salmonella YbbO (Q7CR11). Test a ladder of aliphatic aldehydes (C2–C18) ± NADPH/NADH. Predicted outcome if hypothesis holds: NADPH-dependent reduction with maximal efficiency near C16/C18, mirroring the E. coli ortholog. Discriminates: confirms aliphatic-aldehyde reductase and NADPH preference; upgrades evidence to IDA.
2. Steroid-substrate counter-assay. Test estradiol/androstenedione and other hydroxysteroids. Predicted outcome: negligible activity. Discriminates: falsifies the HSD alternative directly.
3. Retinal vs aliphatic-aldehyde competition kinetics. Compare catalytic efficiency (kcat/Km) for retinal vs C16 aldehyde. Predicted outcome: aliphatic aldehyde ≫ retinal. Discriminates: confirms retinol reduction is promiscuous, not core.
4. ΔybbO Salmonella phenotyping + metabolomics. Assess aldehyde accumulation and lipid-derived alcohol pools under stress. Discriminates: establishes native BP (detoxification vs biosynthetic).
5. Structure/docking. AlphaFold model + cofactor and substrate docking to verify the Asn–Ser–Tyr–Lys tetrad geometry and the NADP 2′-phosphate pocket. Discriminates: structural support for ISS transfer.

Proposed Follow-up Actions (Curation Leads)

All items are leads requiring curator verification.

Candidate annotation update
- Add GO:0008106 (alcohol dehydrogenase (NADP+) activity, MF) to Q7CR11 with evidence ISS (ECO:0000250) or ISO (ECO:0000266), with/from = UniProtKB:P0AFP4.
- Retain GO:0016491 as a parent but note GO:0008106 is the more informative term.
- Do not assign retinol dehydrogenase or hydroxysteroid dehydrogenase terms.

Candidate references to verify (exact snippets)
- PMID: 27134112: "deletion of ybbO gene, which encodes NADP(+) dependent aldehyde reductase, led to >90% reduction in long chain fatty alcohol production" and "YbbO has wide substrate specificity ranging from C6 to C18 aldehyde, with maximum affinity and efficiency for C18 and C16 chain length aldehyde, respectively."
- PMID: 25108218: "we found five new ALRs using this screening method (YahK, DkgA, GldA, YbbO, and YghA)."
- PMID: 34850377: "Escherichia coli ybbO led to increase in retinol production, but retinal remained as a major product" (to justify treating retinol activity as non-core).

Suggested curator questions
- Is an ISS/ISO transfer from a 90.6%-identical ortholog with 100% catalytic conservation acceptable under local annotation policy for adding GO:0008106?
- Does a more specific GO child term for long-chain/aliphatic aldehyde reductase activity exist and better fit the kinetics?
- Should a lead-level BP term (aldehyde catabolic process) be opened as a question pending Salmonella phenotyping?

Suggested experiments
- Purify Q7CR11; run the aliphatic-aldehyde kinetic panel and a steroid/retinal counter-assay (Discriminating Tests 1–3) to convert the ISS lead into a direct (IDA) annotation and definitively close out the HSD and retinol alternatives.

Evidence Base (Literature Summary)

• PMID: 27134112 — Identification of long chain specific aldehyde reductase and its use in enhanced fatty alcohol production in E. coli. The anchor IDA reference: ΔybbO abolishes >90% of long-chain fatty alcohol; kinetics define C6–C18 substrate range with C16/C18 optima. Supports the seed activity and provides substrate narrowing.
• PMID: 25108218 — Toward aldehyde and alkane production by removing aldehyde reductase activity in E. coli. Independent functional screen identifying YbbO as one of five aldehyde reductases. Supports GO:0008106 (second IDA reference).
• PMID: 25726762 — Selective retinol production… from metabolically engineered E. coli. Shows ybbO produces retinol via promiscuous activity; notes E. coli has no inherent retinoid pathway. Qualifies (non-native side activity).
• PMID: 34850377 — Selective production of retinol by engineered S. cerevisiae… Shows retinal remains a major product despite ybbO expression. Qualifies (retinol reduction inefficient).
• PMID: 39307776 and PMID: 41093029 — Retinoid engineering studies confirming ybbO retinal-reductase promiscuity (knockout prevents retinal loss). Qualify (competing, non-core).
• PMID: 8805511 — Crystal structure of mouse lung carbonyl reductase… Establishes that a conserved basic pair contacting the cofactor 2′-phosphate confers NADPH preference in SDRs. Supports the query's inferred NADPH selectivity (Arg34/Lys35).
• PMID: 9525918 — Unusual charge stabilization of NADP+ in 17β-HSD. Illustrates that true HSD steroid specificity relies on dedicated substrate residues, reinforcing that cd05374's "17β-HSD-like" label does not imply steroid activity. Qualifies / refutes HSD alternative.

Conclusion

The blinded sequence-only test resolves cleanly. The query is Salmonella Typhimurium LT2 YbbO (Q7CR11), a classical SDR with a fully intact catalytic tetrad (Asn81, Ser132/133, Tyr146, Lys150) and NADP-selecting basic pair (Arg34/Lys35), 90.6% identical to the assay-characterized E. coli ortholog P0AFP4. The seed hypothesis — NADPH-dependent aliphatic aldehyde reductase, GO:0008106 — is SUPPORTED and can be narrowed to medium/long-chain aliphatic aldehydes (C6–C18, optimal C16/C18). The hydroxysteroid dehydrogenase alternative is refuted as a CDD naming artifact; the retinol dehydrogenase activity is real but promiscuous and non-native. Because the Salmonella protein has not itself been assayed, the correct curation lead is an ISS/ISO transfer of GO:0008106 from UniProtKB:P0AFP4, not a direct-assay assertion.

Artifacts

• OpenScientist analysis provenance
• OpenScientist evidence matrix
• OpenScientist go decision table
• OpenScientist final report
• OpenScientist final report