Pseudomonas aeruginosa lasR mutant fitness in microoxia is supported by an Anr-regulated oxygen-binding hemerythrin

Abstract

Pseudomonas aeruginosa strains with loss-of-function mutations in the transcription factor LasR are frequently encountered in the clinic and the environment. Among the characteristics common to LasR-defective (LasR-) strains is increased activity of the transcription factor Anr, relative to their LasR+ counterparts, in low-oxygen conditions. One of the Anr-regulated genes found to be highly induced in LasR- strains was PA14_42860 (PA1673), which we named mhr for microoxic hemerythrin. Purified P. aeruginosa Mhr protein contained the predicted di-iron center and bound molecular oxygen with an apparent K_d of ∼1 µM. Both Anr and Mhr were necessary for fitness in lasR+ and lasR mutant strains in colony biofilms grown in microoxic conditions, and the effects were more striking in the lasR mutant. Among genes in the Anr regulon, mhr was most closely coregulated with the Anr-controlled high-affinity cytochrome c oxidase genes. In the absence of high-affinity cytochrome c oxidases, deletion of mhr no longer caused a fitness disadvantage, suggesting that Mhr works in concert with microoxic respiration. We demonstrate that Anr and Mhr contribute to LasR- strain fitness even in biofilms grown in normoxic conditions. Furthermore, metabolomics data indicate that, in a lasR mutant, expression of Anr-regulated mhr leads to differences in metabolism in cells grown on lysogeny broth or artificial sputum medium. We propose that increased Anr activity leads to higher levels of the oxygen-binding protein Mhr, which confers an advantage to lasR mutants in microoxic conditions.

Keywords: Anr; Pseudomonas aeruginosa; hemerythrin; lasR; microoxic growth.

Fig. 1.

Increased fitness of the ∆lasR mutant in microoxic conditions is dependent on anr. (A) Assay scheme for the competition of P. aeruginosa strains (yellow culture) against a P. aeruginosa strain tagged with a constitutively expressed lacZ (blue culture). (B) Fitness of wild type (WT), ∆anr, ∆anr + anr, ∆lasR, ∆lasR∆anr, and ∆lasR∆anr + anr relative to the lacZ-labeled WT on tryptone agar at 0.2% O₂ at 37 °C for 48 h. Using a one-way ANOVA with multiple comparisons test, a–b, a–c, and b–c are significantly different; P < 0.0001. (C) Relative fitness of ∆lasR and ∆lasR + lasR using the assay conditions described for assays in B. **P < 0.01 by unpaired t test.

Fig. 2.

mhr transcription and Mhr protein levels are Anr-regulated and higher in the lasR mutant strains. (A and B) β-Galactosidase activity in strains bearing an mhr-lacZ promoter fusion. ****P < 0.0001 by one-way ANOVA, multiple comparisons test; ns, not significant. (A) PA14 wild-type (WT), ∆anr, and WT bearing a promoter fusion variant in which the Anr-box was mutated as described in Materials and Methods (mut. Anr-box). (B) β-Galactosidase activity in WT, ∆lasR, and ∆lasR∆anr containing the mhr-lacZ promoter fusion. (C) Western blot using a rabbit α-Mhr polyclonal antibody. Lanes from left to right: WT, ∆anr, ∆lasR, ∆lasR∆anr, and 12.5 ng Mhr that was purified from E. coli. Total protein stained with REVERT, used for normalization, is shown. (D–F) Quantification of Mhr by Western blot for pairs of lasR mutant and lasR+ strains, normalized to total protein using the REVERT stain (Licor). Each point within a group represents band intensity data from a separate experiment on a different day. *P < 0.05, **P < 0.01 by ratio paired t test. (D) Mhr levels in PA14 WT and ∆lasR. (E) Mhr levels in the natively LasR− keratitis isolate 262K and a derivative in which its lasR allele was replaced with that from strain PA14. (F) Mhr levels for a LasR+ chronic CF isolate and its genetically related LasR− partner recovered from the same patient. Cells were grown as colony biofilms in 0.2% oxygen for 16 h on tryptone agar for all experiments in this figure.

Fig. 3.

Mhr is an oxygen-binding hemerythrin. (A) Structural model of P. aeruginosa Mhr based on the crystal structure for M. capsulatus hemerythrin protein (Methylococcus Hr) (Protein Data Bank ID 4XPX) (36). The region around the di-iron center was zoomed in to display the key oxygen-binding residues. (B) Electronic absorption spectra of deoxy-Mhr and oxy-Mhr in a 20-mM Tris⋅HCl buffer at pH 8. The broad peak around 488 nm is characteristic of the oxy-Mhr form. (C) Plot of the absorbance changes versus the concentration of free (unbound) oxygen. Fits of the absorbance changes versus the concentration of total oxygen in SI Appendix, Fig. S3, have yielded an apparent K_d value of 0.74 μM.

Fig. 4.

Fitness of the ∆lasR mutant in microoxic and normoxic conditions is dependent on mhr. (A) Microoxic fitness of the wild type (WT), ∆lasR, ∆mhr, and ∆lasR∆mhr relative to the lacZ-labeled WT under 0.2% O₂. (B) Microoxic fitness of ∆mhr carrying the empty vector pMQ70 (EV) or a pMQ70-mhr (mhr) against the lacZ-labeled WT carrying pMQ70. (C) Microoxic fitness of the ∆lasR∆mhr + EV or ∆lasR∆mhr + mhr. In B and C, 0.2% arabinose and carbenicillin (300 μg/mL) was added to the agar medium. (D) Normoxic (21% O₂) fitness of WT, ∆lasR, ∆anr, and ∆lasR∆anr. (E) Normoxic fitness of ∆lasR, ∆mhr, and ∆lasR∆mhr. All assays were performed grown on tryptone agar at 37 °C for 48 h. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.00001 by one-way ANOVA, multiple comparisons (A, D, and E) or t test (B and C). ns, not significant.

Fig. 5.

mhr is coregulated with ccoNOPQ-2, the ∆mhr microoxic fitness phenotype requires high-affinity cytochrome c oxidases, and Mhr affects cell metabolism. (A) Similarity in expression pattern for genes within the Anr regulon in a publicly available data compendium determined using a machine-learning–derived model for gene expression, eADAGE. Similarities of expression pattern for Anr-regulated genes (ovals) are depicted as edge length. High-affinity cytochrome C oxidases subunits ccoN2, ccoO2, and ccoP2 are shown in blue, and denitrification structural genes are shown in green. (B) Microoxic (0.2% O₂) fitness comparison for ∆cco1∆cco2 (∆cco) and ∆cco1∆cco2∆mhr (∆cco∆mhr) was determined by competition against the constitutively lacZ-labeled ∆cco1∆cco2 (∆cco att::lacZ). Assay was performed on tryptone agar at 37 °C for 4 d. (C) Metabolomics analysis of 16-h colony biofilms on either LB agar or ASM agar under 21% oxygen. Plotted log₂-transformed fold difference (∆lasR/∆lasR∆mhr) met the following criteria: 1) P < 0.05 by false discovery rate analysis and 2) log₂-transformed fold difference magnitude at least 0.5 on LB or ASM. Plotted metabolites were grouped by pathway: (a) energy generation and (b) sulfur metabolism. Each point represents the average of five replicates. See Dataset S1 for additional information.