antibiotic resistance

traitmech:000088 · CLASS · REVIEWED

A physiological capacity to grow in the presence of antibiotic concentrations that inhibit susceptible cells, mediated by efflux, target modification, drug inactivation, or reduced permeability.

Mechanisms of antibiotic resistance

Evidence-backed causal sketch linking resistance determinants (efflux, target modification, drug inactivation) to growth under antibiotic exposure.

Mechanisms of antibiotic resistance Interactive directed graph showing evidence-backed causal relationships for antibiotic resistance.

Edge evidence

  • drug efflux pump confers antibiotic resistance METPO:2007700

    Drug efflux is one determinant enabling resistance.

    • DOI:10.1038/nrmicro3380 Blair et al. review efflux among resistance mechanisms.
  • antibiotic resistance participates in response to antibiotic biolink:participates_in

    Resistance is part of the cellular response to antibiotics.

    • DOI:10.1038/s41579-022-00820-y Updated review of antibiotic-resistance mechanisms.
  • drug efflux pump decreases intracellular antibiotic concentration RO:0002212

    Efflux pump overexpression lowers intracellular antibiotic concentration, producing resistance phenotypes.

    • DOI:10.3390/pharmaceutics16020170 Efflux pumps reduce intracellular antibiotic concentrations; upregulation under antibiotic exposure produces resistance phenotypes (broad across Gram-positive/negative).
  • outer membrane porin decreases antibiotic influx RO:0002212

    Porin loss/modification reduces antibiotic influx, a major resistance mechanism.

    • DOI:10.3390/pharmaceutics16020170 Reduced permeability/porin loss and modification of outer membrane porin channels listed as major resistance mechanisms.
  • beta-lactamase hydrolyzes beta-lactam antibiotic METPO:2007808

    Beta-lactamases hydrolyze the beta-lactam ring, inactivating the drug.

    • DOI:10.1038/s43856-024-00591-y Drug hydrolysis by beta-lactamases; beta-lactamases hydrolyze the beta-lactam ring. Strong canonical mechanism.
  • QRDR mutation in DNA gyrase/topoisomerase IV confers resistance to fluoroquinolone

    QRDR mutations in DNA gyrase/topoisomerase IV confer fluoroquinolone resistance (class-level edge).

    • DOI:10.3389/fphar.2024.1444781 QRDR mutations in DNA gyrase and topoisomerase IV cause quinolone/fluoroquinolone resistance.
  • 23S rRNA methylation decreases binding of macrolide antibiotic

    23S rRNA methylation modifies the ribosomal target, decreasing macrolide binding and conferring resistance.

    • DOI:10.3389/fphar.2024.1444781 Methylation of 23S rRNA confers linezolid, chloramphenicol, clindamycin, and macrolide resistance.

Provenance

Source
METPO (2025-11-25)
Definition source
DOI:10.1038/nrmicro3380

Parent traits (1)

Synonyms (1)

  • antimicrobial resistance RELATED_SYNONYM · DOI:10.1038/nrmicro3380

kg-microbe context

Matched 1 kg-microbe node via parent_proxy.

  • METPO:1000059 [-2.682, -2.070, -3.656, -0.652, …]

512-dim DeepWalkSkipGramEnsmallen embedding from kg-microbe (2026-04-25).

Nearest neighbors in embedding space

Top-8 cosine-similar METPO traits from the 2026-04-25 deepwalk (512-D).

Deep research

Generated by just research-trait; source: research/traits/physiology/antibiotic_resistance-deep-research-falcon.md

Unreviewed literature output — not curated TraitMech content Ontology identifiers suggested below have not been resolved against their ontologies, and some are known to be wrong. Check any CURIE against the source before using it.
# TraitMech curation report: antibiotic resistance

## Record and scope summary

- **Trait:** antibiotic resistance
- **Identifier:** **`traitmech:000088`**
- **Category / kind / status:** PHYSIOLOGY / CLASS / REVIEWED
- **Parent:** `METPO:1000059`
- **Recommended operational meaning:** a microbial cell or population can **replicate at an antibiotic concentration that inhibits a defined susceptible comparator**, usually observed as an increased minimum inhibitory concentration (MIC) or categorical nonsusceptibility under a specified susceptibility-testing standard. MIC is the lowest concentration preventing visible replication; therefore, resistance is fundamentally a concentration-, medium-, incubation-, species-, and breakpoint-dependent phenotype rather than an unconditional property of a gene. (maeda2024laboratoryevolutionof pages 1-2)

The supplied definition—“a physiological capacity to grow in the presence of antibiotic concentrations that inhibit susceptible cells, mediated by efflux, target modification, drug inactivation, or reduced permeability”—is consistent with current mechanistic understanding. Canonical mechanisms additionally include **target replacement/protection** and **cell-envelope remodeling**. Resistance may be intrinsic or acquired by mutation or horizontal gene transfer. (maeda2024laboratoryevolutionof pages 1-2, zhu2022clinicalperspectiveof pages 4-5)

### Boundaries and nearby traits

1. **Tolerance is not resistance.** Tolerant populations survive longer exposure without necessarily having an increased MIC; time-kill measures are more appropriate than growth inhibition alone. (maeda2024laboratoryevolutionof pages 1-2, maeda2024laboratoryevolutionof pages 12-13)
2. **Persistence is not resistance.** Persistence is survival of a minority, typically non-growing or slow-growing subpopulation, without a stable population-wide MIC increase. It should be represented separately unless the graph explicitly models an evolutionary route from persistence to inherited resistance. (maeda2024laboratoryevolutionof pages 1-2, maeda2024laboratoryevolutionof pages 12-13)
3. **Heteroresistance is adjacent but distinct.** An isogenic population contains a resistant minority while the majority remains susceptible; routine MIC testing can report the isolate as susceptible. Population analysis profiling is the reference detection approach. Consequently, heteroresistance should not automatically instantiate the population-wide trait. (xu2025epidemiologymechanismsand pages 1-2)
4. **Multidrug resistance is a classification**, requiring resistance across multiple antimicrobial categories; it is not a separate molecular mechanism. A single resistance determinant can be narrow-spectrum or pleiotropic. (zhu2022clinicalperspectiveof pages 2-4)
5. **Biofilm-associated recalcitrance** can combine diffusion effects, altered physiology, tolerance, persistence, and inherited resistance. “Biofilm formation → antibiotic resistance” is too broad for unqualified curation.
6. **Antimicrobial resistance** is broader than antibiotic resistance because it includes antiviral, antifungal, and antiparasitic resistance. This graph should remain bacterial-antibiotic focused unless TraitMech intentionally uses the synonym broadly.

## Candidate nodes grouped by type

Identifiers below are deliberately conservative. Where an exact stable CURIE was not verified from the retrieved evidence, the node is left **label-only** rather than assigned a potentially incorrect identifier.

### Trait and assay nodes

- `traitmech:000088` — antibiotic resistance
- `METPO:1000059` — supplied parent trait
- minimum inhibitory concentration (MIC), label-only
- elevated MIC, label-only
- susceptible comparator, label-only
- antibiotic susceptibility testing, label-only
- clinical breakpoint, label-only
- population analysis profiling, label-only

### Environmental and experimental factors

- antibiotic exposure / selection pressure
- repeated antibiotic exposure
- subinhibitory antibiotic concentration
- inhibitory antibiotic concentration
- growth medium, inoculum, incubation time and temperature
- antibiotic combination or sequential/alternating treatment
- membrane permeabilizer
- metabolic inhibitor

Repeated exposure selects adaptive variants, whereas alternating drugs can slow evolution where collateral-sensitivity trade-offs apply. Laboratory evolution coupled to whole-genome sequencing and phenotyping is a current implementation for identifying such paths. (maeda2024laboratoryevolutionof pages 1-2, maeda2024laboratoryevolutionof pages 12-13, maeda2024laboratoryevolutionof pages 6-7)

### Genes, proteins, enzymes, transporters and complexes

- **Drug inactivation:** β-lactamases; ESBLs; AmpC; carbapenemases; NDM/VIM/IMP metallo-β-lactamases; OXA enzymes; aminoglycoside-modifying enzymes
- **Target replacement/remodeling:** `mecA`, `mecC`, PBP2a; `vanA`, `vanB`, `vanM`, VanHAX; D-Ala-D-Lac ligase
- **Target mutation/modification:** `gyrA`, `gyrB`, `parC`, `parE`; `erm`-family 23S-rRNA methyltransferases; 16S-rRNA methylases; `cfr`
- **Efflux:** AcrAB–TolC; AcrB; RND, MFS, MATE, SMR and ABC transporter families; `tetA`; `msrA`
- **Permeability:** OmpF; OprD/OprD2; porin loss or downregulation
- **Envelope remodeling:** `mcr-1`, `mcr-3`, `mcr-9`; MCR phosphoethanolamine transferases; EptA; PmrAB and PhoPQ regulatory systems
- **Evolution/dissemination:** plasmid, transposon and other mobile genetic element

### Chemicals, structures, cellular locations and processes

Showing the first 60 of 224 lines of findings; the linked file also carries the run's front matter and the prompt it was given — read the full report.

Curation history

  1. · PROPOSED_FROM_RESEARCH · claude

    Proposed candidate PHYSIOLOGY trait (antibiotic resistance) from literature research. Distinct from persister-cell tolerance (non-genetic).

  2. · CURATED_CAUSAL_GRAPH · claude

    Added evidence-backed causal graph (resistance mechanisms) with GO node grounding and RO/biolink predicate groundings; promoted PROPOSED to REVIEWED.

  3. · GROUND_CAUSAL_NODES · claude

    Grounded 1 causal-node grounding field(s) via mappings/node_grounding.tsv (UniProtKB:A3RNR9×1).

  4. · ENRICH_CAUSAL_GRAPH · claude

    Added 5 evidence-backed generic edges (9 new nodes) from the deep-research report.

  5. · GROUND_CAUSAL_PREDICATES · claude

    Grounded 2 causal-edge predicate_id field(s) via mappings/predicate_grounding.tsv (RO:0002212×2).

  6. · GROUND_CAUSAL_NODES · claude

    Grounded 2 causal-node grounding field(s) via mappings/node_grounding.tsv (CHEBI:27933×1, CHEBI:25105×1).

  7. · GROUND_CAUSAL_PREDICATES · claude

    Grounded 1 causal-edge predicate_id field(s) via mappings/predicate_grounding.tsv (METPO:2000013×1).

  8. · GROUND_CAUSAL_NODES · claude

    Grounded 2 causal-node grounding field(s) via mappings/node_grounding.tsv (UniProtKB:A0A024EG85×1, UniProtKB:A0A059L9H0×1).

  9. · RETRACT_DEAD_UNIPROT_GROUNDINGS · claude

    Retracted 3 UniProtKB grounding(s) whose accessions are deleted from UniProt; nodes demoted to label-only pending re-grounding (docs/GROUNDING_POLICY.md)

  10. · GROUND_CAUSAL_NODES · claude

    Grounded 1 causal-node grounding field(s) via mappings/node_grounding.tsv (GO:0008800×1).

  11. · MIGRATE_ENABLES_TRAIT_EDGES · claude

    Migrated 1 causal edge(s) off enables/RO:0002327 with a TRAIT object (1 to confers), issue 302. RO:0002327 has range 'biological process or activity', which a trait (a disposition) cannot satisfy, so the previous form entailed trait is-a BiologicalProcessOrActivity. The replacements are proposed in proposals/metpo_traitmech_v8 and are placeholder ids until METPO mints them.

  12. · MIGRATE_MICROBE_DOMAIN_EDGES_PART2 · claude

    Re-grounded 1 causal edge(s) off microbe-domain METPO predicates onto their causal-graph counterparts (1 to hydrolyzes), issue 301 part 2. The previous predicates are transitively rdfs:subPropertyOf METPO:2000001, whose rdfs:domain is METPO:1000525 (microbe), so a causal-graph subject entailed that the subject IS a microbe; CausalNodeTypeEnum has no organism member, so no such edge could ever satisfy the domain. Each replacement is a 1:1 mirror of its source predicate that changes only the domain, so the claim each edge makes is unchanged and directions are unchanged. The replacements are proposed in proposals/metpo_traitmech_v9 and are placeholder ids until METPO mints them.