InhA1, NprA, and HlyII as candidates for markers to differentiate pathogenic from nonpathogenic Bacillus cereus strains

Abstract

Bacillus cereus is found in food, soil, and plants, and the ability to cause food-borne diseases and opportunistic infection presumably varies among strains. Therefore, measuring harmful toxin production, in addition to the detection of the bacterium itself, may be key for food and hospital safety purposes. All previous studies have focused on the main known virulence factors, cereulide, Hbl, Nhe, and CytK. We examined whether other virulence factors may be specific to pathogenic strains. InhA1, NprA, and HlyII have been described as possibly contributing to B. cereus pathogenicity. We report the prevalence and expression profiles of these three new virulence factor genes among 57 B. cereus strains isolated from various sources, including isolates associated with gastrointestinal and nongastrointestinal diseases. Using PCR, quantitative reverse transcriptase PCR, and virulence in vivo assays, we unraveled these factors as potential markers to differentiate pathogenic from nonpathogenic strains. We show that the hlyII gene is carried only by strains with a pathogenic potential and that the expression levels of inhA1 and nprA are higher in the pathogenic than in the nonpathogenic group of strains studied. These data deliver useful information about the pathogenicity of various B. cereus strains.

FIG. 1.

Expression profiles of inhA1, nprA, and hlyII among B. cereus strains. Logarithmic representation of the expression levels of inhA1, nprA, and hlyII among 57 (54 for inhA1) B. cereus nonpathogenic (NP), food-poisoning (FP), and clinical (C) strains normalized to the mean expression value of the two reference genes rpoA and rpoB. For each strain, the mean value of two independent experiments done in duplicate is represented as a cross. For each pathogenic profile, the mean value is represented as a dot and the median value as a bold line. The × symbol represents the nprA expression level of the NVH 0391/98 strain.

FIG. 2.

Relative expression profiles. For each strain, the inhA1 (x axis) expression value was plotted against the nprA (y axis, gray dashed line) or the hlyII (dark dashed line) expression value. Alternatively, nprA expression values (x axis) were plotted against the hlyII values (y axis, plain line). The regression line was drawn, and the average ratio between expression values was calculated.

FIG. 3.

Virulence of strains to G. mellonella insects and expression correlations. Nine strains were tested for their virulence toward G. mellonella insects by injection into the insect larvae hemocoel. (A) Mortality was recorded 24 h postinfection. Error bars show standard deviations of the means of the results from two independent experiments. (B and C) A logarithmic regression line represents the correlation between levels of inhA1 (B) or nprA (C) expression of these strains and virulence. The square represents the BK strain, for which data were not included in the regression line calculation. Black diamonds represent all of the other strains that were tested for virulence.