Fig 1.
MR/P operon and MrpH domain organization.
(A) The genes encoding MR/P fimbriae are organized as an operon. An invertible element (IE) in the mrp promoter controls mrp expression (inset), and MrpI is the recombinase that flips the IE. Therefore, mrpI mutants are either locked “ON” or “OFF” for MR/P fimbriae. The tip-located TDA is encoded by mrpH (yellow). (B) Depiction of linear MrpH. The predicted N-terminal domain (NTD) was used for protein crystallization. Because the C-terminal domain (CTD) has not yet been crystallized, the beginning residue shown here is an estimate.
Fig 2.
Three-dimensional crystal structure of MrpHntd.
(A) Cartoon representation (left) and topology diagram (right) of MrpH159. β-strands are labelled A-G from the N-terminus to the C-terminus. β-strands forming a β-sheet are in the same color: β-strands E, B, A and G are in green, β-strands F, C and D are in blue. Helices α1 and α2 are colored grey. Coils are colored tan. Highly conserved residues (carbon atoms in grey) identified by multiple sequence alignment (Fig 3), a non-conserved disulfide (carbons in tan), and the bound glutamic acid molecule (carbons in pink) are shown as sticks. Zn2+ is shown as a purple sphere. (B) Cartoon representation (left) and topology diagram (right) of the canonical fimbrial NTD of FimH (pdb id 1uwf) using the same coloring scheme as in (A).
Fig 3.
Multiple sequence alignment of representative MrpHntd amino acid sequences.
Aligned MrpHntd amino acid sequences from 16 different organisms, with residues numbered according to their position in the native sequence. Identical residues are highlighted with red shading and white text, and similar residues with a blue frame and red text. The top secondary-structure depiction and bottom disulfide bond indications (shown as yellow brackets) are derived from the MrpHntd structure.
Table 1.
Native data collection and refinement statistics.
Fig 4.
Close-up of the metal and ligand binding site in MrpHntd.
(A) MrpH153 ligand binding site with tartrate bound. (B) MrpH159 ligand binding site with glutamate bound. In both (A) and (B), residues interacting with the ligand or with the zinc ion are shown as sticks and labelled. Zinc is shown as a purple sphere. Water molecules are shown as red spheres. Hydrogen bonds and bonds to the zinc are shown as dotted lines with bond distances indicated next to the lines.
Table 2.
Results of TSA experiments.
Fig 5.
The top surface of MrpHntd is highly positively charged.
Side (left) and top (right) views showing the electrostatic surface of MrpHntd. A crevice extending from the ligand-binding site (tartrate displayed as a stick model) could serve as a binding site for an extended, possibly negatively charged, ligand and/or receptor. Alternatively, the positively charged surface could facilitate approach to negatively charged surfaces without being directly involved in binding. Positive surfaces are blue, neutral surfaces are white, and negative surfaces are red.
Fig 6.
Biofilm formation is dependent on MR/P fimbriae and pH.
(A) Biofilm formation in LB medium. MR/P locked-ON formed significantly more biofilm than all other tested strains. Wild-type biofilm formation trended higher than MR/P locked-OFF, but the difference was not significant. The last three columns show an mrpH mutant (ΔmrpH), either without a plasmid or complemented with pGEN-Pmrp-lux or pGEN-Pmrp-mrpH. (B-D) Biofilm formation was more pronounced at acidic pH. (B) Quantification of biofilm formation in Minimal A. Significance calculated using one-way ANOVA with Dunnett’s multiple comparisons test. (C) Representative biofilm cultures (top) and biofilm staining (bottom). (D) Growth curve analysis of P. mirabilis HI4320 L-ON and L-OFF at pH 6 and 7. (E) Growth curve wells showing floating pellicles for L-ON and planktonic growth for wild type and L-OFF.
Fig 7.
MrpH-dependent biofilm formation occurs on urinary catheter segments.
Pieces of a silicone Foley catheter were added to culture tubes containing Minimal A, pH 6. Medium was inoculated 1:100 with overnight culture of P. mirabilis, and cultures were aerated at 37C for 24 h. Catheter segments were stained with crystal violet, then removed to a new tube for extraction and quantification of stain. (A) biofilm formation by P. mirabilis HI4320 (wt), L-ON, and L-OFF strains shown in previous figures (n = 3). (B) biofilm formation by an independently constructed locked ON strain (mrpI-ON), a double mutant (mrpI-ON ΔmrpH), and the double mutant complemented with a plasmid containing either luciferase (lux) or mrpH, both under the native mrp operon promoter (Pmrp) (n = 5–6). Statistical significance was assessed versus mrpI-ON. (C) a representative example of biofilm formation by the strains shown in B.
Fig 8.
Biofilm formation requires divalent cations.
(A) EDTA (n = 4–6) and (B) TPEN (n = 6–8) inhibit biofilm formation. EtOH refers to locked-ON bacteria with ethanol vehicle added. (C) Addition of increasing concentrations of TPEN shifts P. mirabilis locked-ON from biofilm to planktonic growth. (D) Biofilm formation by P. mirabilis locked-ON does not differ significantly with chelators EGTA (Ca2+-specific) and deferoxamine (DF; iron-specific). ***P < 0.001; ****P < 0.0001 compared to L-ON by one-way ANOVA.
Fig 9.
MrpH zinc binding is essential to its function.
(A) When P. mirabilis mrpI-ON was cultured in chelexed medium ± 5 μM metal supplementation, only zinc restored biofilm formation. Significance against “No metal” columns (OD600 in orange and A570 in purple) calculated by one-way ANOVA with Dunnett’s multiple comparisons test; *P < 0.05; **P < 0.01; n = 3. (B) Biofilm formation is induced by physiologically-relevant levels of zinc. P. mirabilis mrpI-ON was cultured in chelexed Minimal A + CAA with increasing amounts of ZnSO4 for 22 h. Planktonic growth (OD600) was recorded, then biofilm was stained and quantified (A570). The zinc concentration that induces biofilm formation appears to be 50–100 nM. Significance against 0 nM ZnSO4 columns (OD600 in orange and A570 in purple) calculated by one-way ANOVA with Dunnett’s multiple comparisons test; **P < 0.01; ****P < 0.0001; n = 5.
Fig 10.
Zinc-dependent biofilm formation is reversible.
(A and B) Time course of biofilm reversal. (A) mrpI-ON was cultured overnight in chelexed Minimal A + CAA. The next day, 5 μM ZnSO4 was added, and cultures were incubated for an additional 90 min. (B) mrpI-ON was cultured overnight in chelexed Minimal A + CAA + 5 μM ZnSO4. The next day, 10 μM TPEN was added, and cultures were incubated for an additional 3h. Planktonic growth (OD600) and biofilm formation (A570) were measured in replicate cultures at each time point. n = 5–7. Statistics: one-way ANOVA with Dunnett’s multiple comparisons test, compared with T = 0. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
Fig 11.
Zinc-coordinating histidines are required for biofilm formation.
Biofilm formation by an MR/P locked-ON mutant, a locked-ON mrpH double mutant, and the double mutant complemented with wild-type or mutated mrpH expressed from plasmid pGEN-Pmrp-mrpH (+mrpH). pGEN-Pmrp-luxCDABE (EV) is a negative control plasmid and has luciferase genes under the control of the native mrp promoter. Significance compared against pGEN-Pmrp-mrpH using one-way ANOVA with Dunnett’s multiple comparisons test. Compared with the negative (EV) control, these mutants were not significantly different: H72A, H74A, H117A, and R118A; E127A P = 0.0130.
Fig 12.
(A) P. mirabilis (mrpI-ON or mrpI-ON ΔmrpH) was cultured statically for 48h, washed, then mixed with 3% guinea pig erythrocytes in 2-fold decreasing bacterial dilutions. Hemagglutination (HA) was recorded after 30 min as the last dilution that resulted in HA. Statistical significance calculated against pGEN-Pmrp-mrpH complementation strain (+ mrpH). (B) HA by afimbriate E. coli BW25113 ΔfimA, cultured as in (A), expressing mrpA-H or related mutants from a plasmid. (C) HA mediated by mrpI-ON cultured in chelexed medium without or with 5 μM ZnSO4 (-Zn or +Zn, respectively). Either 5 μM ZnSO4 or 10 μM TPEN was added (-Zn+Zn, or +Zn-Zn, respectively), and after 3 h, bacteria were assayed for HA. Representative HA reactions from the undiluted column of bacteria are shown on the bottom.
Fig 13.
Zinc-coordinating residue H72 contributes to UTI.
Bacterial recovery from female CBA/J mice at 7 dpi. P. mirabilis HI4320 mrpHΔkan (kanS) containing the empty vector pGEN-MCS (EV) was competed in a 1:1 ratio with HI4320 mrpHΩkan (kanR) containing either wild-type mrpH (pGEN-Pmrp-mrpH) or mrpH H72A (pMB-H72A). (A and B) CFU recovery of P. mirabilis from urine or organs. Bar indicates median CFU. Limit of detection = 200 CFU, indicated with a dotted line. (C and D) Competitive indices (CI) of the experiments shown in A and B, showing the ratio of recovered (output) to input vector:complemented bacteria. The dotted line at y = 0 indicates no competitive advantage for either strain. Log CI > 0 indicates that the mrpH-complemented bacteria outcompeted the vector control. ns, not significant.