All articles
Start with the Scientific Question Before Choosing the Strain
A reference strain is often the right starting point for standardizing an assay, but its name alone does not guarantee that the model is relevant. In Pseudomonas aeruginosa, for example, PAO1 and PA14 share a highly conserved core genome but differ in accessory regions and regulatory circuits. Some PAO1 sublineages also display variations in mexT or mexF that affect expression of the mexEF-oprN efflux operon and several associated phenotypes. When studying this pathway, simply specifying “PAO1” is therefore not enough: the sublineage, genotype and phenotype must be consistent with the mechanism under investigation. [1–3]
The case of Staphylococcus aureus Newman D2C calls for the same level of caution: mutations affecting two major virulence regulators, agrA and saeR, have made this sublineage markedly attenuated. In an infection model, its use may therefore give the impression that a therapeutic candidate is particularly effective when the challenge strain is simply less virulent. [4]
These examples are not anecdotal exceptions: they underpin our first working principle. We first define the phenotype that the model needs to express and then verify that the selected strain actually allows it to be observed. Depending on the project, this may involve documenting its origin, genotype, virulence level, resistance profile, or parental strain. Once this biological identity has been established, a second question becomes critical: what environment should the isolate come from?
Choose an Appropriate Origin and Level of Diversity
A human clinical isolate, a veterinary strain and an environmental isolate do not tell the same story. They have been exposed to different niches and selective pressures. Their origin can therefore reveal where a resistance mechanism has persisted, how a microorganism has adapted, or which comparisons are needed to investigate its circulation.
Work carried out at Besançon University Hospital on Aspergillus fumigatus provides a concrete illustration. Comparing azole-resistant clinical and environmental isolates, including strains recovered from tulip pots originating from imported bulbs, revealed a hospital reservoir that would have remained undetected in a study restricted to patient isolates. [5,6]
Similarly, the geographical diversity of the Borrelia burgdorferi sensu lato complex shows that a model frequently encountered in North America is not automatically representative of European or Asian issues, where other species and clinical tropisms play a major role. [7,8]
Origin is therefore not a secondary metadata field, but an experimental variable. In an antimicrobial resistance study, for example, we may favor a panel combining human, veterinary and environmental isolates when this diversity is required to explore the distribution of a phenotype or guide investigations into its transferability. For virulence or diagnostic studies, the sampling site, clinical context and geographical area may be equally important.
This reasoning enables the model to be sized appropriately. A single strain may be sufficient to develop a method; a pair of contrasting strains may help isolate a mechanism; and a broader panel becomes essential for testing the robustness or representativeness of a result. Each comparison is therefore chosen to contribute to the expected scientific demonstration. Once these criteria have been defined, the question becomes operational: where can the right biological resources be found?
Mobilize the Right Biological Resource
Smaltis already maintains several hundred biological resources within its own collections, including reference strains, characterized isolates, mutants and reporter strains. This resource base allows us to rapidly mobilize suitable models for projects involving antimicrobial resistance, virulence, diagnostics, microbiota or bioproduction.
However, this internal collection does not limit scientific choice. If the required strain is not already available, our network of hospital, veterinary and biological resource partners enables us to identify isolates that meet the criteria defined with the client: species, origin, sampling site, clinical context, genotype or resistance profile. Their availability naturally remains subject to collection accessibility, applicable authorizations and the rights associated with the samples.
Our collaboration with CEESA illustrates this capability in the veterinary field. Smaltis manages and stores collections generated through CEESA’s European animal health activities. Its Secretary General highlights both the reliability of the storage process, the rigor of traceability, and the value of being able to “process or analyze the strains directly at Smaltis.” The biological resource can therefore be identified, secured and then used within the same scientific environment.
Access to a broad network increases the possibilities, but it does not resolve every scientific question. In some situations, no natural strain can unambiguously link a phenotype to a specific gene or allow two conditions to be compared against a controlled genetic background. In such cases, the model must no longer simply be selected: it must be constructed.
Build the Model When It Does Not Exist
For a mechanistic demonstration, an isogenic pair — a parental strain and its mutant — is often more informative than two natural isolates whose genomes differ at many loci. This approach makes it possible to directly investigate the role of an efflux system, a porin, a beta-lactamase, a virulence factor, a quorum-sensing pathway, or a function involved in adhesion and biofilm formation.
Our mutant catalogue already includes single and combined deletions, enzymatic variants, and strains expressing different fluorescent markers. We can therefore select an existing model, pair it with its parental strain, or incorporate it into a strain panel designed specifically for the project.
When the required model does not exist, our team specialized in genetic engineering can build it to order through deletion, insertion, point mutation or complementation. The strategy and control steps are defined according to the function being studied and the intended future use of the strain. Our article on bacterial mutant construction describes this approach in greater detail.
Whether sourced from a collection or created specifically for the project, however, a strain only retains long-term value if its identity and history remain fully controlled.
Preserve the Scientific Value of the Strain
The initial choice is not sufficient if the strain drifts, its passage history is incomplete, or its storage conditions alter its properties. Traceability, identity, purity and viability checks, together with controlled revival procedures, therefore extend the scientific reasoning initiated at the beginning of the project. These measures make it possible to reproduce a study, compare multiple batches, and resume a program over time.
Our biobanking and microbial resource management services combine storage at -80 °C, sample tracking and, depending on requirements, quality controls, revival, production or additional analyses. Collections therefore remain secure and directly available for the next stages of the project.