A global investigation of the Bacillus subtilis iron-sparing response identifies major changes in metabolism

Abstract

The Bacillus subtilis ferric uptake regulator (Fur) protein is the major sensor of cellular iron status. When iron is limiting for growth, derepression of the Fur regulon increases the cellular capacity for iron uptake and mobilizes an iron-sparing response mediated in large part by a small noncoding RNA named FsrA. FsrA functions, in collaboration with three small basic proteins (FbpABC), to repress many "low-priority" iron-containing enzymes. We have used transcriptome analyses to gain insights into the scope of the iron-sparing response and to define subsets of genes dependent for their repression on FsrA, FbpAB, and/or FbpC. Enzymes of the tricarboxylic acid (TCA) cycle, including aconitase and succinate dehydrogenase (SDH), are major targets of FsrA-mediated repression, and as a consequence, flux through this pathway is significantly decreased in a fur mutant. FsrA also represses the DctP dicarboxylate permease and the iron-sulfur-containing enzyme glutamate synthase (GltAB), which serves as a central link between carbon and nitrogen metabolism. Allele-specific suppression analysis was used to document a direct RNA-RNA interaction between the FsrA small RNA (sRNA) and the gltAB leader region. We further demonstrated that distinct regions of FsrA are required for the translational repression of the GltAB and SDH enzyme complexes.

Fig 1

Hierarchical cluster analysis of transcriptional changes in iron-sparing response mutant strains. A hierarchical cluster analysis (generated using Treeview) represents genes that are at least 1.25-fold repressed in the fur-versus-WT microarray experiment and induced at least 2.0-fold under at least one of the following conditions: fur versus fur fsrA, fur versus fur fbpAB, fur versus fur fbpC, or fur versus fur fbpABC. Clusters with similar regulatory patterns are indicated by brackets and represent genes mostly affected by FsrA (R), FbpC (C), FsrA and FbpAB (RAB), and FsrA and FbpABC (RABC). Red intensity indicates increasing expression, while green intensity indicates decreased expression. Black indicates no change or no data for the gene/array indicated. Genes are listed to the right of the cluster, and the corresponding regulators (where known) are indicated to the right of each gene.

Fig 2

Northern analysis of selected FsrA regulon members. Equal amounts of total RNA isolated from wild-type and fur, fur fsrA, and fur fbpABC mutant strains were hybridized with antisense RNA probes specific for sdhA, citB, and lutC, as indicated. Cells were grown in Belitsky minimal medium and harvested at OD₅₀₀s of 0.4, 1.0, and 2.0, corresponding to the log phase (lane 1), transition phase (lane 2), and stationary phase (lane 3). The arrows point to the expected sizes of the specific transcripts.

Fig 3

Predicted RNA pairings for FsrA and selected FsrA regulon members. (A) Secondary structure of FsrA, including the terminator hairpin (T). A region including the second of two tetracytidine repeats is underlined. (B) Predicted RNA-RNA hybrids between FsrA (3′ to 5′; with the CCCCUCU sequence in bold and underlined for indexing) and the TIR of the indicated target mRNA (5′ to 3′).

Fig 4

FsrA interacts directly with the gltAB leader region. Growth of the indicated strains was monitored in MM with ammonia as a sole nitrogen source. Cells were inoculated from overnight cultures into MM without glutamate, and the OD₆₀₀s were recorded using the Bioscreen C MBR system for 24 h. The wild type and the fur and fur fsrA mutants are shown in all panels. The fur fsrA mutant was complemented with a WT copy at the thrC locus as shown in panel A. Mutants in the gltAB leader region (designated 1, 2, and 3, as in Fig. 3) and the corresponding compensatory changes in FsrA (to restore the predicted pairing) are shown in panels A, B, and C, respectively.

Fig 5

Absolute metabolic fluxes in B. subtilis mutants during exponential growth in glucose batch culture. Selected fluxes are given for the enzymes that catalyze the indicated reactions: MDH, flux measured from malate to oxaloacetate; AKG:MAL, flux measured from α-ketoglutarate to malate; and TCA, flux through the TCA cycle. Strains tested for flux analysis are described below the graph. One of two replicate experiments is shown. Generally, the 95% confidence intervals were between 10 and 15% of the values shown for the major fluxes. Larger confidence intervals were estimated for reactions with low fluxes. The complete solution for the flux analyses in two replicate experiments is shown in the supplemental material.

Fig 6

Western analysis of DctP-FLAG. Western blot analysis was used to monitor the expression of DctP-FLAG in various mutant backgrounds. Ten micrograms total protein from membrane fractions was loaded per lane. Each lane is labeled with the strain background, and a wild-type strain carrying no FLAG construct was included as a negative control. The first lane is the molecular mass marker (in kDa). Development was carried out with anti-FLAG primary antibody and alkaline phosphatase-linked secondary anti-rabbit antibody.

Fig 7

Growth phenotypes of selected mutant strains. (A) Growth curves of mutant strains affected in the iron-sparing response in succinate MM. The strains used in this experiment are as follows: WT (open diamond), fur mutant (cross), fur fsrA mutant (open triangle), and fur fsrA +fsrA strain (HB12573, filled triangle). (B) Growth curves of mutant strains affected in the iron-sparing response in fumarate MM. Strain designations are as in panel A. (C and D) Comparison of growth in succinate MM (C) or fumarate MM (D). Column labels indicate amounts of IPTG added. Columns represent final OD₆₀₀s after 24 h of growth at 37°C. The strains used in this experiment are as follows: WT +P_spac-dctP-FLAG (HB12551, white) and fur +P_spac-dctP-FLAG (HB12558, gray). All growth data presented here are the averages and standard deviations for three biological replicates.

Fig 8

A simplified metabolic model highlighting key points of Fur regulation. Under iron-deficient conditions, the Fur-regulated fsrA, fbpAB, and fbpC genes are derepressed. FsrA is a primary effector in mediating the translational repression of two TCA cycle enzymes, aconitase (CitB) and succinate dehydrogenase (SdhCAB). FsrA additionally represses expression of glutamate synthase (GltAB), which synthesizes glutamate from glutamine and the TCA cycle intermediate α-ketoglutarate, and DctP, a component of the dicarboxylate permease which imports succinate and fumarate. FsrA together with FbpB also represses synthesis of the LutABC lactate oxidase enzymes, which allow the use of lactate as an energy and carbon source.