endosymbiosis

Deep research

Generated by just research-trait; source: research/traits/ecology/endosymbiosis-deep-research-falcon.md

# Curation report: microbial endosymbiosis

## Trait record and scope

- **Trait label:** endosymbiosis
- **Trait identifier:** `traitmech:000045`
- **Category:** ECOLOGY
- **Term kind:** CLASS
- **Mapping status:** REVIEWED
- **Parent:** `traitmech:000040`
- **Operational definition:** a microbial lifestyle/localization trait in which a microorganism resides within the cells or tissues of a host.

For causal-graph curation, the defining phenotype should be **host-internal residence**, preferably demonstrated at cellular resolution. Residence may be directly in host cytoplasm, in bacteriocytes, or inside a host-derived membrane compartment such as a symbiosome. The trait does **not** imply mutualism, obligacy, vertical transmission, nutritional provisioning, or genome reduction; these are frequent but separable properties. For example, *Symbiodolus* occurs intracellularly throughout host development and is concentrated in ovaries, but its secretion systems, nutritional benefits, and entry mechanisms remain largely inferred rather than causally demonstrated (wierz2024intracellularsymbiontsymbiodolus pages 1-2).

### Boundary cases

1. **Ectosymbiosis is excluded**, even when host and microbe exhibit strong genomic or metabolic coupling. Surface attachment to gill epithelial cells is not intracellular residence.
2. **Tissue-associated microbiota are not automatically endosymbionts.** Localization inside host tissue but outside host cells should be represented separately if the ontology distinguishes tissue-level endosymbiosis from intracellular endosymbiosis.
3. **Intracellular pathogens are a semantic boundary.** The supplied definition does not require mutual benefit, so persistent intracellular parasitic associations may satisfy the localization phenotype. Curators should avoid silently restricting the trait to mutualists.
4. **Organelles are evolutionary products of ancient endosymbiosis**, but mitochondria and plastids should not normally be annotated as extant microbial endosymbionts.
5. **Genome reduction is neither necessary nor sufficient.** It commonly follows obligate intracellular association, but reduced genomes can also occur in tightly coupled ectosymbionts.
6. **Rhizobia are stage-dependent.** Free-living rhizobia are not endosymbiotic; bacteroids enclosed within plant-derived symbiosomes are.

## Current mechanistic model

The most defensible general model is not a single universal pathway. Endosymbiosis is an emergent state produced by several modules:

1. **Entry and intracellular accommodation:** host-cell invasion or uptake, followed by residence in cytoplasm or a host-derived compartment.
2. **Immune modulation and population control:** localized immune tolerance, antimicrobial peptides, pattern-recognition proteins, metal limitation, autophagy, apoptosis, or lysosomal digestion.
3. **Bidirectional exchange:** hosts provide carbon, nitrogen precursors, iron, oxygen, or other substrates; symbionts provide amino acids, vitamins, fixed nitrogen, or chemosynthetic carbon.
4. **Transmission and persistence:** vertical, horizontal, or mixed transmission maintains the association.
5. **Evolutionary dependency:** isolation, bottlenecks, relaxed selection, and loss of repair/recombination functions promote genome degeneration and metabolic complementation.

Comparative analysis of 34 independently degenerating *Sodalis* lineages illustrates that this evolutionary module is partly deterministic and partly contingent: amino-acid-biosynthesis genes were preferentially lost, host-beneficial B-vitamin pathways retained, and redundant respiratory-chain and DNA-repair functions lost stochastically (boyd2024stochasticitydeterminismand pages 1-2). The authors identify host-cell isolation, transovarial bottlenecks, and repair/recombination loss as processes expected to reduce effective selection and increase mutational load; these are mechanistic explanations, not individually proven universal causes (boyd2024stochasticitydeterminismand pages 1-2).

## Candidate graph nodes

Identifiers below are conservative suggestions. Taxon-specific proteins are left label-only when a stable accession was not verified.

### Trait, localization, and host structures

| Candidate node | Suggested grounding | Curation note |
|---|---|---|
| endosymbiosis | `traitmech:000045` | Target trait; quote verbatim in YAML. |
| host cell | `GO:0005623` | General intracellular context. |
| cytoplasm | `GO:0005737` | Appropriate for cytoplasmic bacteriocytes such as cockroach–*Blattabacterium*. |
| membrane | `GO:0016020` | Generic only; prefer “symbiosome membrane” label where the specialized structure is intended. |
| symbiosome | label-only candidate | Host-derived membrane plus symbiosome space and bacteroid; ontology mapping should be verified locally. |
| symbiosome membrane | label-only candidate | Interface for host–rhizobium nutrient exchange. |
| bacteriocyte | label-only candidate | Specialized symbiont-bearing host cell. |
| bacteriome | label-only candidate | Organ composed of bacteriocytes in many insects. |
| root nodule | `GO:0042129` | Verify local ontology treatment before use. |
| intracellular anatomical structure | `GO:0005622` | Broad fallback; less informative than bacteriocyte/symbiosome. |

### Biological processes and modules

| Candidate node | Suggested grounding | Evidence relevance |
|---|---|---|
| nitrogen fixation | `GO:0009399` | Bacteroid conversion of N₂ to ammonia. |
| transmembrane transport | `GO:0055085` | Nutrient movement across host or symbiosome membranes. |

Showing the first 60 of 288 lines of findings; the linked file also carries the run's front matter and the prompt it was given — read the full report.