Genome-wide transcriptional analysis of the phosphate starvation stimulon of Bacillus subtilis

Abstract

Bacillus subtilis responds to phosphate starvation stress by inducing the PhoP and SigB regulons. While the PhoP regulon provides a specific response to phosphate starvation stress, maximizing the acquisition of phosphate (P(i)) from the environment and reducing the cellular requirement for this essential nutrient, the SigB regulon provides nonspecific resistance to stress by protecting essential cellular components, such as DNA and membranes. We have characterized the phosphate starvation stress response of B. subtilis at a genome-wide level using DNA macroarrays. A combination of outlier and cluster analyses identified putative new members of the PhoP regulon, namely, yfkN (2',3' cyclic nucleotide 2'-phosphodiesterase), yurI (RNase), yjdB (unknown), and vpr (extracellular serine protease). YurI is thought to be responsible for the nonspecific degradation of RNA, while the activity of YfkN on various nucleotide phosphates suggests that it could act on substrates liberated by YurI, which produces 3' or 5' phosphoribonucleotides. The putative new PhoP regulon members are either known or predicted to be secreted and are likely to be important for the recovery of inorganic phosphate from a variety of organic sources of phosphate in the environment.

FIG. 1.

Representative growth curves and sampling points of various B. subtilis strains grown in LPM. OD₆₀₀ values (•), APase production (▪), concentration of P_i in the medium (▴), and RNA isolation time points are shown for the wild-type strain 168 (A) and phoR-null (B) and sigB-null (C) mutant strains. Isolated RNA was used for the Northern blots and macroarray experiments. The hybridization for the Northern blot was carried out with DIG-labeled RNA probes specific for the phoA gene. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to phoA-specific riboprobes. For the DNA array analysis, total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex.). Normalization and quantification were performed as described in Materials and Methods.

FIG. 1.

Representative growth curves and sampling points of various B. subtilis strains grown in LPM. OD₆₀₀ values (•), APase production (▪), concentration of P_i in the medium (▴), and RNA isolation time points are shown for the wild-type strain 168 (A) and phoR-null (B) and sigB-null (C) mutant strains. Isolated RNA was used for the Northern blots and macroarray experiments. The hybridization for the Northern blot was carried out with DIG-labeled RNA probes specific for the phoA gene. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to phoA-specific riboprobes. For the DNA array analysis, total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex.). Normalization and quantification were performed as described in Materials and Methods.

FIG. 2.

Northern blot analyses and transcriptional profiles for the phoA gene. (A) RNA was isolated from wild-type B. subtilis strain 168 (Bs 168) and sigB-null and phoR-null mutants. Bacteria were grown in LPM, and samples were taken 2 h before (T₋₂) and 0 h (T₀) and 5 h (T₅) after entry into the stationary growth phase, which was provoked by phosphate starvation. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to a phoA-specific riboprobe. Transcript size was determined by comparison with DIG-labeled RNA size markers (Roche Diagnostics, Mannheim, Germany). (B) RNA was isolated from wild-type B. subtilis (168) (♦) and sigB-null (▴) and phoR-null (•) mutants grown in LPM. Samples were taken before, during, and after entry into the stationary growth phase, which was provoked by phosphate starvation. Total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex.). Normalization and quantification were performed as described in Materials and Methods.

FIG. 3.

Growth and reporter activity of B. subtilis yfkN-lacZ fusion mutants grown in LPM. (A) OD₆₀₀ values of lacZ fusion yfkN-lacZ (♦), ΔsigB yfkN-lacZ (▴), and ΔphoR yfkN-lacZ (•) mutants are shown with closed symbols. APase activities of yfkN-lacZ (⋄), ΔsigB yfkN-lacZ (▵), and ΔphoR yfkN-lacZ (○) strains are shown with open symbols. PNP, p-nitrophenyl. (B) Specific β-galactosidase activities of yfkN-lacZ (♦), ΔsigB yfkN-lacZ (▴), and ΔphoR yfkN-lacZ (•) strains are shown with closed symbols.

FIG. 4.

Northern blot analyses and transcriptional profiles for the yfkN gene. (A) RNA was isolated from wild-type B. subtilis strain 168 (Bs 168) and sigB- and phoR-null mutants. Bacteria were grown in LPM, and samples were taken 2 h before (T₋₂) and 0 h (T₀) and 5 h (T₅) after entry into the stationary growth phase, which was provoked by phosphate starvation. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to yfkN-specific riboprobes. Transcript size was determined by comparison with DIG-labeled RNA size markers (Roche Diagnostics, Mannheim, Germany). The arrows labeled 16S rRNA and 23S rRNA indicate the locations of these rRNA species that are known to trap smaller RNA species (1). (B) RNA was isolated from wild-type B. subtilis (168) (♦) and sigB-null (▴), and phoR-null (•) mutants grown in LPM. Samples were taken before, during, and after entry into the stationary growth phase, which was provoked by phosphate starvation. Total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex). Normalization and quantification were performed as described in Materials and Methods.

FIG. 5.

Northern blot analyses and transcriptional profiles for the yjdB gene. (A) RNA was isolated from wild-type B. subtilis strain 168 (Bs 168) and sigB-null and phoR-null mutants. Bacteria were grown in LPM, and samples were taken 2 h before (T₋₂) and 0 h (T₀) and 5 h (T₅) after entry into the stationary growth phase, which was provoked by phosphate starvation. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to yjdB-specific riboprobes. Transcript size was determined by comparison with DIG-labeled RNA size markers (Roche Diagnostics, Mannheim, Germany). (B) RNA was isolated from wild-type B. subtilis (168) (♦) and sigB-null (▴) and phoR-null (•) mutants grown in LPM. Samples were taken before, during, and after entry into the stationary growth phase, which was provoked by phosphate starvation. Total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex.). Normalization and quantification were performed as described in Materials and Methods.

FIG. 6.

Growth and reporter activity of B. subtilis yurI-lacZ fusion mutants grown in LPM. (A) OD₆₀₀ values of lacZ fusion mutant yurI-lacZ (♦), ΔsigB yurL-lacZ (▴), and ΔphoR yurL-lacZ (•) strains are shown with closed symbols. APase activities of yurI-lacZ (⋄), ΔsigB yurL-lacZ (▵), and ΔphoR yurL-lacZ (○) strains are shown with open symbols. PNP, p-nitrophenyl. (B) Specific β-galactosidase activities of yurI-lacZ (♦), ΔsigB yurL-lacZ (▴), and ΔphoR yurL-lacZ (•) strains are shown.

FIG. 7.

Northern blot analyses and transcriptional profiles for the yurI gene. (A) RNA was isolated from wild-type B. subtilis strain 168 (Bs 168) and sigB- and phoR-null mutants. Bacteria were grown in LPM, and samples were taken 2 h before (T₋₂) and 0 h (T₀) and 5 h (T₅) after entry into the stationary growth phase, which was provoked by phosphate starvation. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to yurI-specific riboprobes. Transcript size was determined by comparison with DIG-labeled RNA size markers (Roche Diagnostics, Mannheim, Germany). (B) RNA was isolated from wild-type B. subtilis (168) (♦) and sigB-null (▴) and phoR-null (•) mutants grown in LPM. Samples were taken before, during, and after entry into the stationary growth phase, which was provoked by phosphate starvation. Total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex.). Normalization and quantification were performed as described in Materials and Methods.

FIG. 8.

Northern blot analyses and transcriptional profiles for the vpr gene. (A) RNA was isolated from wild-type B. subtilis strain 168 (Bs 168) and sigB-null and phoR-null mutants. Bacteria were grown in LPM, and samples were taken 2 h before (T₋₂) and 0 h (T₀) and 5 h (T₅) after entry into the stationary growth phase, which was provoked by phosphate starvation. Five micrograms of RNA was applied per lane; after the filters were capillary blotted, they were hybridized to vpr-specific riboprobes. Transcript size was determined by comparison with DIG-labeled RNA size markers (Roche Diagnostics, Mannheim, Germany). (B) RNA was isolated from wild-type B. subtilis (168) (♦) and sigB-null (▴) and phoR-null (•) mutants grown in LPM. Samples were taken before, during, and after entry into the stationary growth phase, which was provoked by phosphate starvation. Total RNA was isolated and used as the template for reverse transcriptase incorporating radiolabeled [³³P]dATP, which was hybridized to a whole-genome macroarray (Sigma-Genosys, The Woodlands, Tex.). Normalization and quantification were performed as described in Materials and Methods.

FIG. 9.

Pho stimulon of B. subtilis. Model for the activation of the PhoP and SigB (σ^B) regulons in response to phosphate starvation (energy) stress. The activation of PhoP via phosphorylation by PhoR leads to the induction of genes involved in the recovery and acquisition of P_i (28). The energy-sensing pathway of the SigB regulon is mediated via the Per-Arnt-Sim (PAS) domain of the RsbP phosphatase and RsbQ. Activated RsbP removes the serine phosphate (P) from RsbV∼P, which in turn sequestrates anti-σ^B factor RsbW. Released from its anti-σ factor, SigB is now free to interact with the core RNA polymerase to induce the nonspecific general stress genes.