A phosphite (Pt)-dependent biological containment strategy, achieved by introducing a Pt-metabolic pathway and disrupting endogenous phosphate transporters, renders Escherichia coli growth strictly dependent on Pt, a compound rarely detected in natural environments, thereby preventing unintended environmental spread. In this study, we demonstrated that expression of phosphate regulon (Pho regulon) genes was markedly upregulated in a Pt-dependent E. coli strain due to the elimination of phoU, a negative regulator of the Pho regulon, along with the high-affinity phosphate transporter pstSCAB. However, further genetic modification of this strain for detailed analysis was hindered by the presence of multiple antibiotic resistance markers. To overcome this limitation, we reconstructed a Pt-dependent E. coli strain using CRISPR-Cas12a-mediated genome editing, enabling the removal of the antibiotic resistance markers and facilitating subsequent genetic manipulation. Using this strain, we disrupted the PhoBR two-component regulatory genes and found that deletion of phoBR alleviated the constitutive overexpression of Pho regulon genes and partially restored growth of the Pt-dependent strain. These findings provide mechanistic insights and technical advances for the refinement and practical application of Pt-dependent biocontainment strategy.
Artificial human microbiota can be produced in gut simulators from cryopreserved stocks. They are used for in vitro fermentation models and as alternative material for fecal microbiota transplantation therapy. However, current methods have limited information on microbial structure at the genus level and present challenges during cryopreservation. In this study, we used an edible glycerol stock of fecal batch culture instead of fresh feces to create artificial gut microbiota. Three glycerol stocks, generated through in vitro fecal fermentation with different prebiotic additives (such as fructooligosaccharide, xylan, pectin, and guar gum), were combined. Profiling via 16S rRNA gene amplicon sequencing revealed that the artificial gut microbiota derived from the combined glycerol stocks showed more amplicon sequence variants than those from a single glycerol stock. In the artificial microbiota, relative abundance values of common genera such as Bifidobacterium, Bacteroides, Prevotella, Faecalibacterium, and Escherichia were more than 10 % of those found in the original feces. Other commensal genera such as Collinsella, Anaerobutyricum hallii (formerly Eubacterium hallii) group, Anaerostipes, Blautia, Dorea, Lachnospiraceae UCG-004, and Oscillospiraceae UCG-003 were similarly maintained. Our data indicated that combining glycerol stocks of fecal cultures with different additives in a batch-type gut simulator is a useful option for producing artificial gut microbiota, the taxonomic compositions of which are comparable to those of the original feces.

