Haemophilus influenzae OxyR: characterization of its regulation, regulon and role in fitness

Abstract

To prevent damage by reactive oxygen species, many bacteria have evolved rapid detection and response systems, including the OxyR regulon. The OxyR system detects reactive oxygen and coordinates the expression of numerous defensive antioxidants. In many bacterial species the coordinated OxyR-regulated response is crucial for in vivo survival. Regulation of the OxyR regulon of Haemophilus influenzae was examined in vitro, and significant variation in the regulated genes of the OxyR regulon among strains of H. influenzae was observed. Quantitative PCR studies demonstrated a role for the OxyR-regulated peroxiredoxin/glutaredoxin as a mediator of the OxyR response, and also indicated OxyR self-regulation through a negative feedback loop. Analysis of transcript levels in H. influenzae samples derived from an animal model of otitis media demonstrated that the members of the OxyR regulon were actively upregulated within the chinchilla middle ear. H. influenzae mutants lacking the oxyR gene exhibited increased sensitivity to challenge with various peroxides. The impact of mutations in oxyR was assessed in various animal models of H. influenzae disease. In paired comparisons with the corresponding wild-type strains, the oxyR mutants were unaffected in both the chinchilla model of otitis media and an infant model of bacteremia. However, in weanling rats the oxyR mutant was significantly impaired compared to the wild-type strain. In contrast, in all three animal models when infected with a mixture of equal numbers of both wild-type and mutant strains the mutant strain was significantly out competed by the wild-type strain. These findings clearly establish a crucial role for OxyR in bacterial fitness.

Figure 1. Transcriptional response of H. influenzae to the addition of H₂O₂.

Comparison of transcription of the (A) hktE and (B) pgdX genes of the H. influenzae strain 86-028NP and the isogenic oxyR mutant strain HI2285. Transcriptional status of the genes was determined in strain 86-028NP(closed circles) and strain HI2285 (open triangles) grown in sBHI throughout the experiment and in strain 86-028NP(open circles) and strain HI2285 (closed triangles) grown in sBHI to which 150 µM H₂O₂ was added at 60 minutes.

Figure 2. Growth of H. influenzae strains in sBHI supplemented with one of three oxidizing agents.

Growth of H. influenzae type b strain 10810 and the isogenic oxyR mutant HI2283 when supplemented with (A) 150 µM H₂O₂, (B) 100 µM cumene hydroperoxide and (C) 100 µM Tert-butyl peroxide. In each panel growth in sBHI is represented by closed circles for 10810 and by closed triangles for HI2283, while growth in sBHI supplemented with the specified perturbant is represented by open circles for strain 10810, and by open triangles for strain HI2283. Values are mean±SD for quintuplicate results from representative experiments. The Mann-Whitney test was used to compare growth of strain10810 and HI2283 over the entire growth period for each perturbant. For each perturbant, the growth of the mutant strain was statistically different from growth of the wild-type strain (P<0.0001 for all analyses).

Figure 3. Transcriptional response of H. influenzae to sequential additions of H₂O₂.

Percent change in transcript levels of the hktE (closed circles) and oxyR (closed triangles) upon addition of H₂O₂. H Influenzae strain 86-028NP was grown in sBHI for 60 minutes at which point 150 µM H₂O₂ was added (corresponds to time 0 in the graph above), after a further 20 minutes of growth a second addition of 150 µM H₂O₂ was made.

Figure 4. Transcriptional response of an H. influenzae pgdX mutant to the addition of H₂O₂.

Comparison of transcription of the hktE gene of the H. influenzae strain 10810 and the isogenic pgdX mutant strain HI2159. H. influenzae was grown for 60 minutes in sBHI and then supplemented with 150 µM H₂O₂. Transcript levels of hktE were determined on two samples, one taken immediately prior to H₂O₂ addition (black bars) and the second taken 4 minutes after H₂O₂ addition (grey bars). (A) hktE transcript levels of individual strains are normalized to the pre-addition transcript level of that specific strain. (B) all hktE transcript levels are normalized to the pre-addition transcript level of strain 10810.

Figure 5. Impact of the oxyR mutation in rat models of virulence and competitive fitness.

Bacteremic titers in (A) 5-day old rats and (B) 30-day old rats infected with a mixture of equal numbers of the wild-type Hib strain 10810 (open circles) and the oxyR mutant strain HI2283 (closed circles). (C) Bacteremic titers in 30-day old rats infected with a mixture of equal numbers of the wild-type Hib strain 10810 (open circles) and the oxyR mutant strain HI2283 (closed circles). Results for bacteremic titers are all means±SD from groups of 10 rats. (D) Percentage of freshly drawn blood samples from 30-day old rats infected with a mixture of equal numbers of the wild-type Hib strain 10810 and the oxyR mutant strain HI22283 containing detectable colonies of the wild-type strain (black columns) or the mutant strain (white columns). (^* P = 0.021, ^** P<0.002,).

Figure 6. Impact of the oxyR mutation in a chinchilla model of competitive fitness.

Percentage of successfully tapped ears infected with a mixture of equal numbers of the wild-type strain 86-028NP and the oxyR mutant strain HI2285 containing detectable colonies of the wild-type strain (black columns) or the mutant strain (white columns). Using Fisher’s Exact test to compare percentages of effusions with detectable wild-type or mutant strains P<0.0001 on both day 7 and day 11.

Figure 7. Impact of the pgdX mutation in rat models of competitive fitness.

Bacteremic titers in (A) 5-day old rats and (C) 30-day old rats infected with a mixture of equal numbers of the wild-type Hib strain 10810 (open circles) and the pgdX mutant strain HI2159 (closed circles). Results are means±SD from groups of 10 rats. Percentage of freshly drawn blood samples from (B) 5-day old rats and (D) 30-day old rats infected with a mixture of equal numbers of the wild-type Hib strain 10810 and the pgdX mutant strain HI2159 containing detectable colonies of the wild-type strain (black columns) or the mutant strain (white columns). (^* P<0.005, ^** P = 0.005, ^*** P<0.0001).

Figure 8. Impact of the pgdX mutation in a chinchilla model of competitive fitness.

Percentage of successfully tapped ears infected with a mixture of equal numbers of the wild-type strain 86-028NP and the pgdX mutant strain HI2334 containing detectable colonies of the wild-type strain (black columns) or the mutant strain (white columns). Using Fisher’s Exact test to compare percentages of effusions with detectable wild-type or mutant strains P = 0.0002 on day 7 and P = 0.022 on day 11.

Figure 9. Transcription of genes in the H. influenzae OxyR regulon in vivo.

Comparison of transcript levels of the individual members of the OxyR regulon during colonization of the chinchilla ear with (A) either the wild-type NTHi strain 86-028NP or the oxyR mutant strain HI2285 (ΔoxyR) and (B) with either the wild-type NTHi strain 86-028NP or the pgdX mutant strain HI2334 (ΔpgdX). Data points represent Q-PCR values of fold change in transcripts by comparison with non-stressed in vitro grown 86-028NP.

Figure 10. Schematic of the proposed regulation of the H. influenzae OxyR system.

The oxyR (pink arrow) and pgdX (green arrow) genes are divergently transcribed from the same intergenic region. Under reducing conditions, OxyR is synthesized de novo with reduced disulfides at positions 199 and 208. On exposure to elevated peroxide levels, these reduced disulfides are sequentially oxidized to create a disulfide bridge. The oxidized OxyR is able to bind to promoters upstream of members of the OxyR regulon. As a result of OxyR binding to the promoter the transcription of pgdX increases. PgdX acts to reduce the disulfides of OxyR, releasing it from the promoters and halting further transcription. In addition, binding of OxyR to the pgdX promoter temporarily sterically inhibits transcription and further de novo synthesis of OxyR. The dotted line represents the redox status of the cell, with the upper section of the figure representing reducing conditions and the lower section representing oxidizing conditions.