Scientists Use CRISPR to Strip Antibiotic Resistance From a Widely Used Probiotic

One of the world’s most widely consumed probiotic bacteria just got a genetic safety upgrade, and the tool that made it possible was already hiding inside the microbe itself. In a study published in Microbial Biotechnology, researchers report that they used the native CRISPR-Cas machinery of Bifidobacterium animalis subsp. lactis BLC01 to disable tetW, a tetracycline resistance gene carried by most strains of this subspecies, without introducing any foreign DNA into the final organism. The resulting derivative, named BLC01-2F3G10, lost its tetracycline resistance entirely while retaining every probiotic trait the team measured, from acid and bile tolerance to adhesion to human intestinal cells. The work offers a proof of concept for a safe-by-design strategy that could reshape how next-generation probiotics are engineered and regulated.

The concern driving the research is well documented. Bifidobacteria are among the earliest colonizers of the human gastrointestinal tract and dominate the gut microbiota of infants and healthy adults, and B. animalis subsp. lactis is the most commonly used probiotic in foods and supplements, with documented benefits ranging from improved colonic barrier function and mitigation of antimicrobial treatment side effects to enhanced oral health, relief of infant colic and cholesterol-lowering activity. Yet phylogenomic surveys show that tetW, which encodes a ribosomal protection protein that blocks tetracycline from binding the bacterial ribosome, is widely distributed across strains of this subspecies. The gene is frequently flanked by mobile genetic elements, raising the possibility of horizontal gene transfer to other members of the gut microbiota. Comparative analyses of tetW loci from human intestinal Bifidobacterium strains have found 98 to 100 percent identity within a 2.1 kilobase core region, and conserved sequences matching tetW, tetO and tetS have been detected in commensal and pathogenic bacteria spanning the Arcanobacterium, Streptococcus, Corynebacterium, Campylobacter and Listeria genera. The tetW locus also has a GC content of 51.9 percent, considerably lower than the 60.5 percent GC content of the BLC01 genome, a signature consistent with horizontal acquisition.

Under European Food Safety Authority guidelines, any antimicrobial resistance gene is considered a hazard and may preclude Qualified Presumption of Safety status unless its intrinsic nature is demonstrated. Recent metagenomic analyses of commercial probiotic products have detected more than 70 distinct resistance genes, including hybrid tet(W/N/W) variants often linked to integrative conjugative elements. Against this backdrop, the research team set out to eliminate the resistance trait at its source. Their approach falls under the umbrella of New Genomic Techniques, which recent EU regulation defines as a diverse group of methods capable of producing organisms with modifications equivalent to those obtainable by conventional breeding or with more complex changes. EFSA has stated that applying these techniques to microorganisms does not pose novel hazards compared with established genomic techniques or conventional mutagenesis, with respect to the technique itself, and the agency has outlined comparative risk assessment strategies based on substantial equivalence between the parental strain and the edited derivative.

The technical centerpiece of the study is the exploitation of BLC01’s own immune system. Bioinformatic analysis identified an endogenous Type I-U CRISPR-Cas system consisting of a CRISPR array with 19 spacers interspersed with a conserved 36-nucleotide direct repeat, located immediately downstream of the cas operon. Spacer analysis against viral sequence databases revealed a conserved 5′-CAC-3′ protospacer-adjacent motif. The team designed a 33-nucleotide spacer targeting the 5′ region of tetW and cloned it into a synthetic mini-CRISPR array on the pAM1 shuttle vector, complete with the native leader sequence, two direct repeats and a rho-independent transcription terminator. When expressed, this construct mimics native CRISPR activity, producing a guide RNA that directs the endogenous Cascade-Cas3 complex to the tetW locus. A two-kilobase repair template carried on the same plasmid then steered homology-directed repair, introducing seven nucleotide substitutions that create three consecutive premature stop codons at positions 62 to 64 of the TetW protein.

The editing worked with striking efficiency. Of 96 individual clones screened, two displayed a tetracycline-sensitive phenotype. The researchers cured one mutant of the editing plasmid and subjected the resulting clone, BLC01-2F3G10, to whole-genome sequencing using both Illumina and Oxford Nanopore platforms. Comparison with the wild-type genome confirmed the intended mutations in tetW and revealed only two additional changes: a single cytosine deletion in a non-coding region and an adenine-to-cytosine substitution in the lgt gene, which encodes a phosphatidylglycerol-prolipoprotein diacylglyceryl transferase. That substitution changes a threonine to a proline at the boundary of an alpha helix, but three-dimensional structural modeling showed the overall conformation of the protein remained unaltered, suggesting the mutation is functionally neutral. Crucially, because the editing plasmid was removed, the final strain carries no exogenous DNA, a feature that substantially strengthens its biosafety and regulatory profile.

The functional consequences were unambiguous. The minimum inhibitory concentration of tetracycline for the parental strain was 32 micrograms per milliliter, well above the EFSA microbiological cut-off of 8 micrograms per milliliter. After tetW inactivation, the MIC dropped to 1 microgram per milliliter, comfortably below the threshold. Growth kinetics confirmed the loss of resistance: at a sublethal tetracycline concentration of 0.5 micrograms per milliliter, the edited strain showed a marked delay in exponential growth, and at a lethal concentration of 1 microgram per milliliter it failed to grow beyond an optical density of roughly 0.1 over 24 hours, while the wild type retained partial growth. For the seven other antimicrobials tested, including ampicillin, gentamicin, kanamycin, streptomycin, erythromycin, clindamycin and chloramphenicol, only minor variations in MIC values were observed, and all remained at or below EFSA cut-offs, indicating the edit specifically abolished tetracycline resistance without altering susceptibility to unrelated drug classes.

Equally important, the edit left the probiotic machinery intact. Both strains survived three hours of exposure to pH 2, 3 and 4 with no significant differences between them, tolerated bile salt concentrations up to 2 percent, and showed similar growth kinetics under osmotic stress at sodium chloride concentrations of 2.5 to 3.5 percent. Both produced exopolysaccharides when grown on glucose, fructose, sucrose or lactose as sole carbon sources, with no differences in colony morphology. Auto-aggregation was strong in both strains, reaching 79.75 percent for the wild type and 83.30 percent for the edited derivative after four hours, and co-aggregation with Escherichia coli ATCC 25922 and Salmonella enterica UC3605 was similarly low and variable in both. Under the INFOGEST static in vitro digestion protocol, which simulates oral, gastric and intestinal phases, both strains maintained bacterial loads of approximately 8 log CFU per milliliter throughout the experiment.

Adhesion assays reinforced the picture of functional equivalence. Using Caco-2 and HT-29 human intestinal epithelial cell lines, the team found that both bifidobacterial strains adhered at levels comparable to Lacticaseibacillus rhamnosus ATCC 53103, the gold-standard positive control. On Caco-2 cells, the edited strain actually showed the highest adhesion of any strain tested, reaching 90.56 percent, while the negative control, Lactobacillus delbrueckii subsp. lactis DSM 2072, managed only 6.63 percent. On HT-29 monolayers quantified by real-time PCR, BLC01 and BLC01-2F3G10 exhibited adhesion levels 10.45 and 8.74 times higher than the positive control, respectively, with no statistically significant difference between them. These results indicate that tetW inactivation did not disturb the surface-associated proteins and envelope components that mediate host interaction, colonization and immunomodulatory effects.

A final and critical question was stability. Resistance genes can sometimes revert or be regained under selective pressure, so the team passaged the edited strain for five consecutive days in medium containing tetracycline at 0.1 and 0.5 micrograms per milliliter, concentrations above the reported minimal selective concentration of 0.01 micrograms per milliliter but below the mutant’s MIC. After 122 generations, no revertant colonies capable of growing at the 8 micrograms per milliliter cut-off were detected. By eliminating a mobile, widely conserved resistance gene without introducing new determinants, the edited strain reduces the theoretical risk of horizontal gene transfer within the gut resistome, a concern underscored by metagenomic evidence linking probiotic-associated tetracycline resistance to mobile elements. The authors argue that minimal, well-characterized edits of this kind, which abolish resistance without deleting large genomic regions, are particularly attractive from a regulatory standpoint because they reduce the likelihood of unintended effects and simplify molecular characterization. While in vivo studies will be needed to confirm the strain’s behavior in the complex intestinal ecosystem, the study demonstrates that endogenous CRISPR editing can serve as a precision safety tool, providing a generalizable framework for developing next-generation probiotics that are both effective and aligned with evolving regulatory and societal expectations.

Subject of Research: Removal of the tetracycline resistance gene tetW from Bifidobacterium animalis subsp. lactis using its endogenous CRISPR-Cas system under a safe-by-design framework

Article Title: Endogenous CRISPR‐Based Removal of Tetracycline Resistance in Bifidobacterium animalis subsp. lactis Through a Safe‐by‐Design Approach

Article References: Endogenous CRISPR‐Based Removal of Tetracycline Resistance in Bifidobacterium animalis subsp. lactis Through a Safe‐by‐Design Approach. (n.d.). https://doi.org/10.1111/1751-7915.70443

Image Credits: AI Generated

DOI: 10.1111/1751-7915.70443

Keywords: CRISPR, probiotics, Bifidobacterium animalis subsp. lactis, antimicrobial resistance, tetW, tetracycline, genome editing, safe-by-design, New Genomic Techniques, horizontal gene transfer, EFSA, gut microbiota