Transcriptional profiling of Bacillus anthracis Sterne (34F2) during iron starvation

Abstract

Lack of available iron is one of many environmental challenges that a bacterium encounters during infection and adaptation to iron starvation is important for the pathogen to efficiently replicate within the host. Here we define the transcriptional response of B. anthracis Sterne (34F(2)) to iron depleted conditions. Genome-wide transcript analysis showed that B. anthracis undergoes considerable changes in gene expression during growth in iron-depleted media, including the regulation of known and candidate virulence factors. Two genes encoding putative internalin proteins were chosen for further study. Deletion of either gene (GBAA0552 or GBAA1340) resulted in attenuation in a murine model of infection. This attenuation was amplified in a double mutant strain. These data define the transcriptional changes induced during growth in low iron conditions and illustrate the potential of this dataset in the identification of putative virulence determinants for future study.

Figure 1. Growth of B. anthracis in iron depleted media.

B. anthracis vegetative cells were used to inoculate either IRM (black lines) or IDM (gray lines) at an initial OD₆₀₀ = 0.05. Growth was monitored over time by measuring change in OD₆₀₀. RNA was harvested at two, three, and four hours after inoculation (arrows).

Figure 2. Changes in B. anthracis gene expression induced by iron starvation over time.

Hierarchical clustering of genes identified as statistically significant by J5 test across two independent microarray experiments. Fluorescence intensities were normalized as described in the methods. Normalized intensity values from two experiments were input into Gene Pattern together and clustered using the Pearson correlation. Genes clustered into four groups, induced with a peak at three hours (A), induced at three and four hours (B), induced at four hours (C), and repressed (D). Data shown are normalized intensity values over time for samples grown in IDM only.

Figure 3. Validation of microarray data.

(A). Gene expression changes measured by Q-PCR at four hours. RNA from four experiments was tested for specific genes using real time PCR. The results of Q-PCR are represented with white bars, while corresponding values from the microarray experiments are represented with black bars. Data are presented as mean±SEM of log₂ transformed fold change where fold change is the ratio of expression in IDM to IRM. (B). Correlation analysis comparing fold change between microarray and Q-PCR for ten B. anthracis genes. Data are plotted as log₂ transformed microarray data compared to log₂ transformed Q-PCR data. The correlation coefficient (R²) between these two analyses was 0.99.

Figure 4. Categorical analysis of genes induced during iron starvation.

Pie charts represent the COG classification for the genes uniquely regulated in the conditions listed. Each pie represents 100% of the genes exhibiting significant gene expression changes following growth in iron depleted media. COG categories were identified for each differentially regulated gene based on genomic annotation (GenBank accession, NC_007530). To simplify the analysis, the COG categories were grouped into general categories as follows: Categories A, J, K, L, and RNA genes were merged into “Transcription/translation/DNA replication/RNA”; C, E, F, G, H, I, P, and Q were combined into “Transport and metabolism”; COGs N, T, and U were combined into “Cellular functions/trafficking/secretion/signal transduction”; and categories R and S were grouped with uncategorized genes and classified as “Unknown.” Other COG categories listed are reported as they were originally defined.

Figure 5. Specific analysis of metabolic changes induced in B. anthracis during iron starvation.

Each pie represents 100% of the metabolism related genes exhibiting significant gene expression changes following growth in iron depleted media (blue slice from Figure 4). Specific COG categories were identified for each gene based on genomic annotation (accession, NC_007530) and are listed as they were originally defined.

Figure 6. Attenuation of B. anthracis Δ0552, 1346 strain lacking both putative internalins.

DBA/2J mice were infected by intratracheal infection with WT (circles) or Δ0552, 1346 (squares) spores at 1.5×10⁵ spores per mouse. Mice were monitored for fourteen days, then all surviving mice were euthanized.