Synthetic extremophiles via species-specific formulations improve microbial therapeutics

Abstract

Microorganisms typically used to produce food and pharmaceuticals are now being explored as medicines and agricultural supplements. However, maintaining high viability from manufacturing until use remains an important challenge, requiring sophisticated cold chains and packaging. Here we report synthetic extremophiles of industrially relevant gram-negative bacteria (Escherichia coli Nissle 1917, Ensifer meliloti), gram-positive bacteria (Lactobacillus plantarum) and yeast (Saccharomyces boulardii). We develop a high-throughput pipeline to define species-specific materials that enable survival through drying, elevated temperatures, organic solvents and ionizing radiation. Using this pipeline, we enhance the stability of E. coli Nissle 1917 by more than four orders of magnitude over commercial formulations and demonstrate its capacity to remain viable while undergoing tableting and pharmaceutical processing. We further show, in live animals and plants, that synthetic extremophiles remain functional against enteric pathogens and as nitrogen-fixing plant supplements even after exposure to elevated temperatures. This synthetic, material-based stabilization enhances our capacity to apply microorganisms in extreme environments on Earth and potentially during exploratory space travel.

Fig. 1 |. Survey of commercially available dry formulations of microbial probiotics.

We quantified the viability of a wide representative range of commercial probiotic products (Supplementary Table 1). Viable cell counts were defined as the number of CFUs assessed through dilution plating on the appropriate solid medium and quantified via programmatic image analysis (Methods). Percent viable rates relative to promised cells were computed by dividing the CFU count by the CFU counts printed on the product label. Percent viable rates relative to total cells were computed by dividing the CFU counts by the number of total countable cells determined via automated fluorescence cytometry. N = 4 or 5 independent dosage forms. Geometric means and geometric 95% confidence intervals are plotted in black over individual replicates. Broad phylogenetic classes of the component microorganisms are noted in colour and specific genera are noted in italics (Supplementary Fig. 1 for detailed make-up). neg., negative; pos., positive; Bifido., Bifidobacterium; Lacto., Lactobacillus; caps., capsules.

Fig. 2 |. Development and validation of high-throughput pipeline for dry stabilized microbial materials.

a, We developed a high-throughput pipeline (batch mixing, freeze-drying, plating and quantification) to assess the capacity of a library of materials generally recognized as safe (GRAS) to stabilize microorganisms (Methods). b, Representative dried formulations made in high-throughput pipeline and corresponding colony counting images. Scale bars, 1 cm. c, Result of materials with the top 33%, showing the superior viability of some materials relative to others after drying and storage for 24 h at room temperature; results are across four organisms and two concentrations and are ordered by descending capacity to stabilize E. coli Nissle 1917 at the 5X concentration. The viability score is a composite colony count normalized to the maximum observable growth defined as 1 (Methods). Supplementary Table 2 contains material identities and concentrations corresponding to the noted Material index material indices, and Supplementary Fig. 4 contains full results. Black boxes mark the known lyoprotectants ATCC reagent 20 (index 18) and trehalose (index 35). N = 3 independent biological replicates. d, Correlation analysis of top-performing materials for each organism. e, Verification of top materials defined in c via a more precise assay (freeze-drying in vials and individual CFU counts on plates) and the ability to preserve viability for 1 and 30 days at room temperature. Material indices marked with asterisks are at 1X concentration, while others are at 5X. Horizontal dashed line represents the limit of detection. Where noted by a vertical bar, the water sample was assayed at higher concentrations to achieve a lower limit of detection. N = 8 independently dried vials for all day 1 groups. N = 4 independently dried vials for day 30 E. coli Nissle 1917, E. meliloti and L. plantarum groups. N = 5 independently dried vials for day 30 S. boulardii groups. Geometric mean and s.d. plotted over individual replicates.

Fig. 3 |. Application of pipeline to make E. coli Nissle 1917 into a synthetic extremophile.

a, The library of materials was assayed individually (x axis) and in combination with melibiose (y axis) for each member’s capacity to stabilize E. coli Nissle 1917 at room temperature and 50 °C for 24 h. Black circles mark the mean viability score (Methods) along both dimensions, and error bars represent the s.e.m. N = 3 independent biological replicates. Coloured shapes mark material combinations with melibiose selected for further characterization. Horizontal dashed line is a reference that marks the viability score of melibiose combined with vehicle (water, blue circle). Diagonal dashed line is the identity. All concentrations were at 1X as defined in Supplementary Table 2. FOS, fructooligosaccharides. b, The relative concentrations (rel. conc.) of each of the components in the selected two-material combinations (melibiose + yeast extract or melibiose + caffeine) were varied, and the resulting effect on viability was measured after storage at 37 °C for 23 days. X, standard concentration of each material as defined in Supplementary Table 2. White letters D and E mark the selected formulations for further characterization. N = 3 independent biological replicates. c, Direct stability comparison of formulation (Form.) D (as noted in b), the parent formulation (melibiose) and capsules of the commercial E. coli Nissle 1917 product Mutaflor after storage at room temperature for one month. Mean and s.e.m. of log values plotted over individual replicates. Independent biological replicates were used for each group: Mutaflor (N = 5), melibiose (N = 4) and formulation D (N = 3). Ordinary one-way analysis of variance (ANOVA) is as follows: Mutaflor–melibiose, ****P < 0.0001; Mutaflor–formulation D, ****P < 0.0001. d, Characterization of ultra-long stability at 37 °C compared to maltodextrin (stabilizer used in Mutaflor). Geometric mean and s.d. plotted for each time point. N = 3 independently measured samples from one large material batch. Lower dashed line is the limit of detection. Upper dashed line marks the 10% viability level. e, Transmission electron micrographs of E. coli Nissle 1917 after drying in formulation D or maltodextrin and optionally exposed to high temperature. Arrowheads mark detached inner membranes.

Fig. 4 |. Synthetic extremophiles withstand harsh processing manufacturing processes.

a, Microbial tableting and coating allow simple dosing, transport and tuning of release to access a wide range of applications. b, Photos and scanning electron micrographs of synthetic extremophiles processed via industrially relevant steps. c, Stability of the E. coli Nissle 1917 synthetic extremophile (Form. D) through each process relative to parent powder. Maltodextrin is the stabilizer in the commercial product Mutaflor. The optimized coating is described in the Methods. Dotted horizontal line shows the limit of detection. Geometric mean and s.d. plotted over individual replicates. N = 3 independent biological replicates. d, Tablets of the E. coli Nissle 1917 (EcN) synthetic extremophile withstand exposure to simulated gastric fluid (SGF, in compendial disintegration apparatus) when coated with the enteric polymer EUDRAGIT S 100. Geometric mean and s.d. plotted over individual replicates. N = 6 tablets. e, Tableting allows tuning of release kinetics through the inclusion of a matrix former (HPMC). Synthetic extremophile E. coli Nissle 1917 cells carry a biosynthetic pathway (the luxCDABE operon), and the released luminescence was tracked over time. Lines connect the means. Individual replicates plotted. N = 2 tablets.

Fig. 5 |. Synthetic extremophiles remain effective microbial therapeutics despite extreme temperature and radiation exposure.

a, The E. coli Nissle 1917 synthetic extremophile (Form. D) is more effective than the commercial product Mutaflor against three enteric pathogens (S. flexneri 2a, S. enterica Infantis and S. sonnei) in vitro. Formulation D and Mutaflor were incubated at 37 °C for five days, rehydrated and mixed with the noted pathogen for 6 h in PBS. The pathogen load was quantified through differential plating on MacConkey agar and imaging (E. coli, stained pink, lactose fermentation positive (Lac+); pathogens, colourless, lactose fermentation negative (Lac−)). For each sub-panel of the plot on the right, the mean and s.e.m. of log values are plotted over individual replicates. Independent biological replicates were used for each group: S. flexneri 2a (N = 6), S. enterica Infantis (N = 4) and S. sonnei (N = 4). Dashed line is the limit of detection. Unpaired t-test, two-tailed P is as follows: S. flexneri 2a, ****P < 0.0001; S. enterica Infantis, ****P < 0.0001; S. sonnei, *P = 0.0228. b, Formulation D is also more effective than Mutaflor at reducing S. flexneri 2a in a live-animal intestinal model of infection where pathogens are directly injected into intestinal segments. Formulation D and Mutaflor were processed as in a, but instead injected in the jejunum of live Yorkshire pigs and extracted after 6 h. Each replicate was quantified as in a. Mean and s.e.m. of log values plotted over individual replicates. N = 5 independent intestinal segments. Unpaired t-test, two-tailed P = 0.0078. c, E. meliloti is a nitrogen-fixing bacterium that supplies plants with nitrogen at symbiotic root nodules (inset). Synthetic extremophiles of E. meliloti were exposed to 50 °C, hydrated and tested for functionality in a root nodulation assay with M. truncatula. Nodulated seedlings (black arrowheads) can be quantified at day 12 post-root inoculation. d, Quantification of nodulation assay (c) compared to the material viability score. Mean and s.e.m. plotted over individual replicates. N = 3 independent biological replicates for the viability score groups. Nodulation percent was calculated in independent replicate seedling pouches for each group: controls left of dashed line (N = 8); maltose, sucrose, polydextrose (N = 5); maltodextrin, soytone (N = 4). e, EcN powder exposed to ionizing radiation is plated to quantify viability (CFU ml⁻¹). Synthetic extremophile E. coli Nissle 1917 cells are robust to ionizing radiation compared to a liquid suspension of cells. Lines connect the means. Individual replicates plotted. N = 4. Dotted line is the limit of detection (LOD). Grey shading indicates the viability range of sample not exposed to radiation. Two-way ANOVA is as follows: 0.5 kGy, ****P < 0.0001; 0.1 kGy, ****P < 0.0001.