Contribution of YjbIH to Virulence Factor Expression and Host Colonization in Staphylococcus aureus

Abstract

To persist within the host and cause disease, Staphylococcus aureus relies on its ability to precisely fine-tune virulence factor expression in response to rapidly changing environments. During an unbiased transposon mutant screen, we observed that disruption of a two-gene operon, yjbIH, resulted in decreased levels of pigmentation and aureolysin (Aur) activity relative to the wild-type strain. Further analyses revealed that YjbH, a predicted thioredoxin-like oxidoreductase, is predominantly responsible for the observed yjbIH mutant phenotypes, though a minor role exists for the putative truncated hemoglobin YjbI. These differences were due to significantly decreased expression of crtOPQMN and aur Previous studies found that YjbH targets the disulfide- and oxidative stress-responsive regulator Spx for degradation by ClpXP. The absence of yjbH or yjbI resulted in altered sensitivities to nitrosative and oxidative stress and iron deprivation. Additionally, aconitase activity was altered in the yjbH and yjbI mutant strains. Decreased levels of pigmentation and aureolysin (Aur) activity in the yjbH mutant were found to be Spx dependent. Lastly, we used a murine sepsis model to determine the effect of the yjbIH deletion on pathogenesis and found that the mutant was better able to colonize the kidneys and spleens during an acute infection than the wild-type strain. These studies identified changes in pigmentation and protease activity in response to YjbIH and are the first to have shown a role for these proteins during infection.

Keywords: Spx; Staphylococcus aureus; YjbH; YjbI; staphyloxanthin.

FIG 1

Staphyloxanthin production and Aur activity are reduced in the yjbIH and yjbH mutants. (A) Carotenoid pigment assay with representative photos of colony colors. Values represent averages of results from two independent experiments. Error bars represent standard errors of the means (SEM) (n = 6). *, P < 0.05; **, P < 0.01 (one-way ANOVA with Holm-Sidak multiple-comparison test). WT, wild type. (B) Protease assay using 1.0% skim milk agar plates. Per panel, images are representative of results from >3 independent experiments and of cultures grown on the same plate and adjusted for contrast similarly.

FIG 2

β-Galactosidase activities of strains containing (A and C) crtOPQMN or (B and D) aur promoter-lacZ fusions. For panels A and B, samples were taken over the course of growth until the early stationary phase. For panels C and D, cells were harvested after 8 h of growth. Data are representative of results from ≥3 independent experiments. Error bars represent SEM (n = 3). **, P < 0.01; *, P < 0.05 (Student's t test). For panel A, the sigB mutant results were statistically significantly different from the WT strain results at all time points.

FIG 3

Sensitivity of mutants to NO^· stress. (A) Strains were grown in LBGT medium with (empty symbols) or without (closed symbols) 10 mM DETA-NO. Data represent averages (n = 3) with SEM. (B) The area under the curve was calculated for each strain. **, P < 0.01; *, P < 0.05 (one-way ANOVA and Holm-Sidak multiple-comparison test).

FIG 4

Altered sensitivities of strains to ROS. Serial dilutions of the WT and yjbI, yjbH, and yjbIH mutant strains left untreated (A) or exposed to 1 M hydrogen peroxide (B) or 25 mM superoxide-producing methyl viologen (C). “+con” indicates a mutation in either katA (A and B) or sodM (C) as a control.

FIG 5

Aconitase activity and sensitivity to Fe chelation. (A and B) Levels of aconitase activity in yjbI, yjbH, and yjbIH mutant strains compared to the wild-type strain grown under conditions of (A) high aeration and (B) low aeration. (C) Aconitase activity of acnA combination mutant strains with acnA expressed from a nonnative promoter under conditions of high aeration. Data represent averages (n = 3) and standard deviations (SD). *, P < 0.05 compared to WT (one-way ANOVA with Tukey’s multiple-comparison test). (D) Sensitivity of the mutant strains exposed to a 1 mM concentration of the metal chelator 2,2,-dipyridyl.

FIG 6

Relationship of yjbIH phenotypes to Spx and ClpP. (A) Colony pigment of strains grown overnight on TSA. (B) Aur activity assay using 1% skim milk agar. (C) Serial dilutions of indicated strains spotted onto TSA containing 0.2 mM diamide. Per panel, images are of cultures grown on the same plate and were adjusted for contrast similarly. “spx+” indicates complementation of spx in the yjbIH spx double mutant.

FIG 7

Relationship of YjbIH and σ^B. The images show results of Aur activity assay with 1% skim milk agar (left panels) and colony pigment of strains grown overnight on TSA (right panels). JLB174 was the sigB strain used. All images represent cultures grown on the same plates and adjusted for contrast similarly.

FIG 8

sigB, yjbI, and yjbH mutant phenotypes in response to stress. (A and B) Serial dilutions of the WT and yjbH, sigB (JMB2745), and yjbH sigB mutant strains were exposed to either (A) 1 mM 2,2-dipyridyl or (B) 25 mM methyl viologen. Pictures are representative of results from multiple experiments and include fhuA and sodM positive controls (+con) for panels A and B, respectively. (C) Aconitase activity under conditions of high aeration in yjbI and yjbH mutants with and without sigB. Data represent averages (n = 3) and SD. *, P < 0.01 compared to WT; ns, no significant difference (one-way ANOVA with Tukey’s multiple-comparison test).

FIG 9

Relationship of the yjbIH phenotypes to Spx and σ^B. (A) Colony pigment of strains grown overnight on TSA. (B) Aur activity assay using 1% skim milk agar.

FIG 10

The yjbIH mutant strain showed increased virulence in a murine sepsis model. (A) For the low dose, mice were infected with 4 × 10⁶ CFU of WT (n = 4) or yjbIH (n = 5) cells. (B) For the high dose, mice were infected with 1 × 10⁷ CFU of WT (n = 14) or yjbIH (n = 13) cells. For percent weight loss determinations, the area under the curve (AUC) was calculated for the WT (n = 12) and yjbIH (n = 11) strains. One mutant-infected mouse succumbed to infection on day 5 and was excluded from the analysis. Error bars represent SD. *, P ≤ 0.01 (Mann-Whitney test). (C) Mice were infected with 1 × 10⁷ CFU of WT or yjbIH, yjbIH complement, yjbI, or yjbH cells. For panel C, n = 10, but 3 WT-, 2 yjbIH-, and 3 yjbIH complement-infected mice were at or below the level of detection (LOD; 100 CFU/ml), and while symbols are not shown, they were included in the analysis.

FIG 11

Model of YjbIH, Spx, and σ^B interactions. (A) Known interactions of YjbH, Spx, ClpXP, and RNAP. (B) Model to account for the Aur and pigment phenotypes described here. Briefly, both Spx and σ^B regulate the production of aur and crt, but σ^B is a stronger regulator of both (denoted by thicker lines), making the sigB mutant phenotype dominant over an spx mutant phenotype. In this model, the absence of YjbH leads to increased levels of Spx, which suppresses the expression of aur and crtOPQMN. In the absence of sigB alone, Aur levels increase and crtOPQMN levels decrease. However, in the absence of YjbH and σ^B, Spx levels increase to shut down aur expression. This model accounts for the sigB mutant phenotypes being dominant over the spx mutation and how the sigB mutation can be dominant over the yjbH mutation in pigmentation but not Aur activity. However, this model is likely simplistic and it is not yet known whether Spx works directly through σ^B in some phenotypes (if other regulators are involved) or if YjbH interacts with or influences other systems, which is likely.