propionic acid fermentation

Deep research

Generated by just research-trait; source: research/traits/metabolism/propionic_acid_fermentation-deep-research-falcon.md

# Curation report: propionic acid fermentation

## Trait record and scope

- **Trait label:** propionic acid fermentation
- **Trait identifier:** `traitmech:000029`
- **Category / kind / status:** METABOLISM / CLASS / REVIEWED
- **Parent:** `METPO:1002005`
- **Synonym:** propionate fermentation

### Recommended scope

This trait should represent an organism’s experimentally demonstrable capacity for **net fermentative production of propionate from sugars or lactate**, canonically in dairy propionibacteria through the Wood–Werkman, or methylmalonyl-CoA, pathway. The classical overall lactate stoichiometry is:

**3 lactate → 2 propionate + acetate + CO₂ + H₂O.**

The pathway couples pyruvate reduction to oxidation elsewhere in metabolism, consumes glycolysis-derived NADH, conserves energy as ATP, and commonly produces acetate and CO₂ with propionate. Theoretical Wood–Werkman fermentation from glucose has been estimated at up to 0.7 g propionate/g glucose (1.71 mol/mol) with supplemented reducing equivalents; the commonly cited pathway-level product ratio is approximately 2:1 propionate:acetate and estimated energy yield is 4 ATP/glucose. These are model/theoretical values rather than universal assay thresholds. (bucher2021propionicacidbacteria pages 3-5, gonzalezgarcia2017microbialpropionicacid pages 8-10, gonzalezgarcia2017microbialpropionicacid pages 3-5)

The best-supported exemplar is *Propionibacterium freudenreichii*. Native propionibacteria are regarded as strong biological-production candidates because propionate is a major fermentation product and the Wood–Werkman route is energetically efficient, although slow growth, acid inhibition, coproduct formation, and downstream recovery remain barriers. (gonzalezgarcia2017microbialpropionicacid pages 1-3)

### Boundaries and nearby traits

1. **Propionate presence alone is insufficient.** Propionate may arise through the acrylate, succinate/sodium-pumping, or 1,2-propanediol pathways, through amino-acid catabolism, or through engineered aerobic biosynthesis. Those routes should not automatically instantiate this narrowly defined Wood–Werkman-centered trait. (gonzalezgarcia2017microbialpropionicacid pages 6-8, gonzalezgarcia2017microbialpropionicacid pages 1-3, gonzalezgarcia2017microbialpropionicacid pages 5-6)
2. **Net production must be distinguished from consumption.** Under oxygen exposure, *P. freudenreichii* can consume previously formed propionate and acetate. Thus detection of pathway enzymes does not prove net fermentation under every condition. (loivamaa2024aerobicadaptationand pages 6-9, dank2021propionibacteriumfreudenreichiithrives pages 3-4)
3. **Gut Bacteroidia are a boundary case.** They frequently use a succinate route and show species-specific bicarbonate dependence. Their phenotype may merit a broader “fermentative propionate production” parent or a separate mechanistic child rather than direct assignment to this canonical class. (doring2024propionateproductionby pages 1-2, doring2024propionateproductionby pages 4-5)
4. **Engineered aerobic propionate production is out of scope.** An engineered *Pseudomonas taiwanensis* strain accumulated 2.8 g/L propionate aerobically using introduced succinyl-CoA-catabolic genes and an acyl-CoA hydrolase. This is propionate biosynthesis, but not classical propionic acid fermentation. (neves2024expandingpseudomonastaiwanensis pages 1-2)
5. **Assay recommendation:** require substrate depletion plus net propionate accumulation under anoxic or explicitly fermentative conditions. Acetate and CO₂ strengthen identification but should not be mandatory because ratios vary with strain, medium, redox state, and cofactors. (bucher2021propionicacidbacteria pages 3-5, doring2024propionateproductionby pages 4-5)

## Candidate nodes

Ontology identifiers below are restricted to identifiers directly supported by the retrieved literature. Chemical and gene identifiers should remain **label-only until independently checked against current ChEBI, Rhea, KEGG, MetaCyc, UniProt, and NCBI Taxonomy releases**; this avoids inventing or misassigning CURIEs.

### Trait and pathway modules

| Candidate node | Type | Grounding recommendation |
|---|---|---|
| propionic acid fermentation | Trait class | `traitmech:000029` |
| metabolism parent | Trait class | `METPO:1002005` |
| Wood–Werkman cycle | Pathway/module | Label-only pending MetaCyc/KEGG verification |
| glycolysis | Pathway/module | Label-only pending pathway-database verification |
| fermentative redox balancing | Biological process | Label-only; pathway-level interpretation |
| ATP generation during fermentation | Biological process | Label-only; do not assign a reaction-level identifier without verification |

### Organisms and taxonomic exemplars

- *Propionibacterium freudenreichii*: canonical dairy propionibacterium and principal exemplar.
- *Acidipropionibacterium acidipropionici*: native producer used in current high-density bioprocess research.
- Other dairy propionibacteria, including *Acidipropionibacterium jensenii* and *A. thoenii*: plausible taxon-specific instances, not universal mechanistic proxies.
- *Bacteroides propionicifaciens* and *Bacteroides graminisolvens*: succinate-route boundary exemplars.
- *Clostridium propionicum*, *Megasphaera elsdenii*, and *Prevotella ruminicola*: acrylate-route boundary exemplars.
- *Veillonella*, *Propionigenium*, and *Selenomonas*: succinate/sodium-coupled route boundary taxa. (gonzalezgarcia2017microbialpropionicacid pages 6-8, gonzalezgarcia2017microbialpropionicacid pages 5-6)

### Chemicals and metabolic roles

**Inputs/electron donors:** glucose and other fermentable sugars; lactate; glycerol in process-specific cofermentations.

**Central intermediates:** pyruvate, oxaloacetate, malate, fumarate, succinate, succinyl-CoA, methylmalonyl-CoA, propionyl-CoA.

**Outputs:** propionate/propionic acid; acetate/acetic acid; carbon dioxide; water. Succinate and lactate may remain as strain- and process-dependent coproducts.

Showing the first 60 of 247 lines of findings; the linked file also carries the run's front matter and the prompt it was given — read the full report.