Development of a suite of Proteus mirabilis-derived urea-inducible promoters

Abstract

Catheter-associated urinary tract infections (CAUTIs) are a significant burden on healthcare systems, accounting for up to 40% of hospital-acquired infections globally. A prevalent CAUTI pathogen, Proteus mirabilis, is an understudied Gram-negative bacterium. One sequela of P. mirabilis CAUTI is the production of urinary stones, which complicates treatment and clearing of the infection. Stone formation is induced by the activity of urease, a nickel-metalloenzyme that is regulated by UreR in a urea-dependent manner. As urea is abundant in the urinary tract, urease genes are highly expressed during experimental UTI. We sought to leverage the urease promoter to create an expression system that would enable urea-inducible expression of genes during in vitro experiments as well as during experimental UTI. During preliminary studies, we observed unexpectedly high levels of basal expression of the urease promoter. This was somewhat dependent on the presence of regulator UreR. To further develop this expression system, we generated a series of reporter constructs to assess the impact of specific promoter elements on promoter activity in the presence and absence of urea. Elements of interest included known regulatory binding sites, alternative translational start sites, and single-nucleotide polymorphisms identified through comparative genomics. This work describes a suite of urea-inducible promoters, constructed during this study, that exhibit a variety of expression dynamics, providing a customizable platform for gene expression.IMPORTANCEUrea is an inexpensive molecule that can easily be supplied during in vitro experiments. A urea-inducible promoter would also be activated by environments where urea naturally occurs, such as in the urinary tract. Thus, the development of a urea-inducible system for selective gene expression is of great interest to the field of uropathogenesis as it would enable selective gene induction during experimental urinary tract infection. This expression system would also have important applications for recombinant protein production in biotech and manufacturing.

Keywords: Proteus mirabilis; inducible expression; promoter; urease.

Fig 1

UreR contributes to basal activity of PureD. (A) Schematic depicting elements of the urease promoter of P. mirabilis. (B) Schematic of reporter constructs pMF2 and pMF6. (C) The promoter activity of each construct was measured via Miller assay after growth to mid-logarithmic phase in buffered medium or buffered medium supplemented with 100 mM urea. Significance was determined via two-way ANOVA with Tukey’s multiple comparisons correction (*P < 0.05; **P < 0.01, ****P < 0.0001). (D) To measure the urea-responsiveness of each construct, the induction ratio (Miller units_induced/Miller units_uninduced) was calculated. pMF2 and pMF6 were assessed in a wild-type background (grey bars) or ureR mutant (white bars). Statistically significant changes in urea-responsiveness were assessed via ordinary Kruskal-Wallis with Dunn’s correction for multiple comparisons (**P < 0.01).

Fig 2

UreR binding sites play distinct roles in urease expression. (A) Schematic depicting reporter constructs pMF6-lacZYA, pMF6∆UreR1-lacZYA, pMF6∆UreR2-35-lacZYA, pMF6∆H-NS::spacer-lacZYA, and pMF6∆Reg-35-lacZYA. (B) The individual and cumulative effects of regulator binding sites on promoter activity was assessed via Miller assay after growth to mid-logarithmic phase (OD₆₀₀ = 0.5–0.9) in buffered medium or buffered medium supplemented with 100 mM urea. Data were normalized to the activity of pMF2-lacZYA cultured in buffered medium. Significance was determined via two-way ANOVA with Tukey’s multiple comparisons correction (*P < 0.05; **P < 0.01, ****P < 0.0001). (C) To measure the urea-responsiveness of each construct, the induction ratio was calculated. Statistical significance was assessed via ordinary Kruskal-Wallis with Dunn’s correction for multiple comparisons (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig 3

Alternate start codon usage impacts promoter activity. (A) Schematic of pMF6 and pMF6-derived reporter constructs driving expression from alternate translational start positions at –6, –15, and −33 using ATG (∆6, ∆15, and ∆33) or the native alternate start codon (∆6-GTG, ∆15-TTG, and ∆33-TTG). (B and C) The promoter strength of constructs depicted in (A) was assessed via Miller assay after growth to mid-logarithmic phase in buffered medium or buffered medium supplemented with 100 mM urea. Data were normalized to the activity of pMF2-lacZYA cultured in buffered medium. Statistical significance was determined using two-way ANOVA with Tukey’s multiple comparisons correction (*P < 0.05; **P < 0.01, ****P < 0.0001). (D and E) The induction ratio for each construct was quantified to compare the urea-inducibility of the reporter constructs. Significance was determined via (D) Kruskal-Wallis with Dunn’s correction for multiple comparisons (*P < 0.05, ***P < 0.001) or (E) ordinary one-way ANOVA with the Dunnett correction for multiple comparisons (****P < 0.0001). To determine whether the increased activity of pMF6∆6-GTG-lacZYA was driven by increased transcription or translation, we measured the expression of lacZ via qRT-PCR (F) and the production of LacZ via Western blotting (G). Samples were cultured in the same conditions as in (B and C). qRT-PCR data were normalized to lacZ transcript from pMF6-lacZYA cultured in buffered medium and to housekeeping gene rpoA. Expression was assessed as significantly upregulated or downregulated compared to pMF6-lacZYA using one-sample t test with a hypothetical mean of 1 (*P < 0.05; **P < 0.01). Comparative analysis was calculated using the unpaired t test (**P < 0.01). In (G), a cross-reactive protein bound by α-LacZ served as a loading control.

Fig 4

P. mirabilis SNP confers decreased response to urea. (A) Schematic depicts reporter constructs pMF6-lacZYA, pMF6-G(−16)T-lacZYA, and pMF6-T(−91)A-lacZYA. (B and C) SeqLogo graphs depict the relative abundance of nucleotides at each SNP locus in 470 P. mirabilis genomes. (D) The effect of each SNP on promoter activity was assessed via Miller assay after growth to mid-logarithmic phase (OD₆₀₀ = 0.5–0.9) in buffered medium or buffered medium supplemented with 100 mM urea. Data were normalized to the activity of pMF2-lacZYA cultured in buffered medium. Statistical significance was determined using two-way ANOVA with Tukey’s multiple comparisons correction (***P < 0.001). (E) SNPs had no effect on the urea-responsiveness of each promoter as calculated by ordinary one-way ANOVA with Dunnett’s multiple comparisons test.

Fig 5

Identification of Morganellaceae SNPs within P_ureD. The intergenic region between ureR and ureD from 15 Morganellaceae species that encode a UreR-regulated urease locus were aligned using Clustal Omega. The resulting alignment was viewed in NCBI’s MSA Viewer and colored to highlight variations from the consensus sequence in red. Gaps in coverage are depicted in white. Below the alignment schematic, the frequencies of nucleotides within the regulatory region of this bidirectional promoter are depicted using a SeqLogo graph. The HI4320 sequence is depicted underneath the SeqLogo, and loci where HI4320 differs from the consensus sequence are emphasized in underlined, bold text. The −10 and −35 sites facilitating RNA polymerase binding (white), UreR binding sites (yellow), and H-NS binding sites (blue) are also depicted. SNPs within the UreR2 binding site selected for further analysis are indicated with arrows.

Fig 6

Morganellaceae SNP enhances urea-responsive activity of P_ureD. (A) Schematic depicting reporter constructs pMF6-lacZYA, pMF6-C256T-lacZYA, and pMF6-G265A-lacZYA. (B) The effect of each SNP on promoter activity was assessed via Miller assay after growth to mid-logarithmic phase (OD₆₀₀ = 0.5–0.9) in buffered medium or buffered medium supplemented with 100 mM urea. Data were normalized to the activity of pMF2-lacZYA cultured in buffered medium. Significance was determined via two-way ANOVA with Tukey’s multiple comparisons correction (*P < 0.05). (C) SNPs had no effect on the urea-responsiveness of each promoter as calculated by ordinary one-way ANOVA with Dunnett’s multiple comparisons test.