Skip to main content
. 2025 May 15;11(5):001405. doi: 10.1099/mgen.0.001405

Table 3. In vitro models for safety and efficacy testing for microbiome-targeting products/therapies.

Main advantage Main disadvantage Most relevant measurable safety endpoint
2D cell culture models Ease of useLow costRelatively rapid assessmentScalableModerate/high throughputGood reproducibility Limited time in cultureLack of 3D architecture – does not mimic in vivo host–microbiome interactions Inflammatory cytokine productionCell viability
Primary human cells May retain behaviour of human cells in tissues Multiple donors needed to reflect inter-person variability
Cell lines Engineered to enable continuous passageProvide consistency across studiesEasier to source Does not necessarily retain the behaviours of human cells in tissues
3D models 3D architectureMore tissue level functionMore representative of microbe-surface interaction (skin) Higher costOften require specialized expertise and equipment ViabilityCytokine productionGene expressionStructural changes (fluorescence microscopy, electron microscopy, imaging)Can study microbial communities (viability, growth, metabolites, toxins, compositional changes of mixed populations)
3D skin equivalents ReproducibilityCommercially available Barrier properties do not resemble that of human skin
Ex vivo human skin maintained in organ culture Gold standardCloser resemblance to barrier properties of real skin maintained (better mimicking of host–microbiome interaction) Availability
Organoids Preserving organ, disease and patient-specific characteristics and enabling repeated experiments Luminal side of cells oriented towards the centre – obstacle to study host microbiome interactionNot possible to directly co-culture with strictly anaerobic bacteria because it is an aerobic model Single or multiple omics readout across many different samples
Linearized organoids Apical (luminal) side accessibleCan be co-cultured with bacteria and microbial derived metabolites introduced into the culture media Not possible to directly co-culture with strictly anaerobic bacteria because it is an aerobic modelMore expensive than normal 3D organoids Single or multiple omics readout across samples; TEER (transepithelial electrical resistance) for barrier integrity and imaging
Microfluidic devices Incorporate organoid cells with other cell types (immune cells and fibroblasts)Can replicate anaerobic conditions (gut) Infrastructure and cost needsStandardisation is challenging Multiple omics readouts from the same sample under select conditions
Organ-on-a chip Reproducible and controllable environment for both host and microbial cellsPrecise manipulation of microscale fluids, mimicking the physiological environment of the human organs Different platforms have different attributesOften difficult to choose the most adequate one Imaging, multi-omics and cytokine measurements