Molecular mechanisms underlying the positive stringent response of the Bacillus subtilis ilv-leu operon, involved in the biosynthesis of branched-chain amino acids

Abstract

Branched-chain amino acids are the most abundant amino acids in proteins. The Bacillus subtilis ilv-leu operon is involved in the biosynthesis of branched-chain amino acids. This operon exhibits a RelA-dependent positive stringent response to amino acid starvation. We investigated this positive stringent response upon lysine starvation as well as decoyinine treatment. Deletion analysis involving various lacZ fusions revealed two molecular mechanisms underlying the positive stringent response of ilv-leu, i.e., CodY-dependent and -independent mechanisms. The former is most likely triggered by the decrease in the in vivo concentration of GTP upon lysine starvation, GTP being a corepressor of the CodY protein. So, the GTP decrease derepressed ilv-leu expression through detachment of the CodY protein from its cis elements upstream of the ilv-leu promoter. By means of base substitution and in vitro transcription analyses, the latter (CodY-independent) mechanism was found to comprise the modulation of the transcription initiation frequency, which likely depends on fluctuation of the in vivo RNA polymerase substrate concentrations after stringent treatment, and to involve at least the base species of adenine at the 5' end of the ilv-leu transcript. As discussed, this mechanism is presumably distinct from that for B. subtilis rrn operons, which involves changes in the in vivo concentration of the initiating GTP.

FIG. 1.

B. subtilis ilv-leu operon and deletion analysis for the positive stringent response. The ilv-leu operon, consisting of seven genes (ilvBHC and leuABCD) (12, 20), was transcribed from the ilv-leu promoter (P_ilv-leu) to a terminator downstream of leuD (41). The locations of the TnrA box (40), a catabolite-responsive element (cre) for the binding of the complex of the CcpA and P-Ser-HPr proteins (36, 41), and the CodY binding sites (CodY-I, -II, -III, and -IV) (35) are indicated. To perform deletion analysis of the ilv-leu promoter region for the positive stringent response, the respective promoter regions comprising bases −248 to +26, −187 to +26, −100 to +26, and −55 to +26 were fused with lacZ and then integrated into the amyE locus, as described in the text.

FIG. 2.

Involvement of CodY and RelA in the positive stringent response of ilv-leu transcription upon lysine starvation. To monitor the ilv-leu promoter activity, all of the Lys⁻ strains used in this analysis carried the lacZ fusion with the promoter region comprising nucleotides −248 to +26 in the amyE locus. Cells of strains FU737 (relA⁺ tnrA⁺) and FU775 (relA⁺ tnrA::Tn917) (A), FU739 (relA1 tnrA⁺) and FU776 (relA1 tnrA::Tn917) (B), FU737 and FU773 (relA⁺ ccpA::neo) (C), FU739 and FU774 (relA1 ccpA::neo) (D), FU737 and FU745 (relA⁺ ΔcodY) (E), and FU739 and FU747 (relA1 ΔcodY) (F) were grown in MM medium containing a mixture of 16 amino acids (15 amino acids plus lysine) to the logarithmic growth phase (OD₆₀₀ of 0.5) and then spun down, as described in the text. A portion of the cells was suspended and further incubated in MM medium containing 15 amino acids plus lysine, and the other part of the cells was suspended and incubated in MM medium containing 15 amino acids to subject the cells to lysine starvation, as described in the text; arrows indicate the start times of incubation with and without lysine. Cell growth (OD₆₀₀, open symbols) and lacZ expression (β-Gal activity, filled symbols) were monitored during growth and lysine starvation; results for wild-type strains (tnrA⁺ ccpA⁺ codY⁺) with starvation (circles) and without starvation (squares) and mutant strains (tnrA::Tn917, ccpA::neo, or ΔcodY) without starvation (diamonds) and with starvation (triangles) are shown.

FIG. 3.

Deletion analysis of the ilv-leu promoter region as to its positive stringent response. (A) Lys⁻ strains (FU737 [relA⁺] and FU739 [relA1], FU771 [relA⁺] and FU772 [relA1], FU769 [relA⁺] and FU770 [relA1], and FU844 [relA⁺] and FU845 [relA1]), carrying the lacZ fusions of the ilv-leu promoter regions comprising nucleotides −248 to +26 (a), −187 to +26 (b), −100 to +26 (c), and −55 to +26 (d), were used for the analysis. (B) Lys⁻ relA⁺ strains (FU844 [codY⁺] and FU810 [ΔcodY]) (a) and Lys⁻ relA1 strains (FU845 [codY⁺] and FU809 [ΔcodY]) (b), carrying the ilv-leu promoter region (nucleotides −55 to +26), were used. Cells of each strain were grown and subjected to lysine starvation, as described in the legend to Fig. 2. Each arrow indicates the start time of further incubation with and without lysine after cell suspension. Cell growth (OD₆₀₀, open symbols) and lacZ expression (β-Gal activity, filled symbols) were monitored during growth and lysine starvation. (A) relA⁺ strains with starvation (circles) and without starvation (squares) and relA1 strains with starvation (triangles) and without starvation (diamonds). (B) codY⁺ strains with starvation (circles) and without starvation (squares) and ΔcodY strains with starvation (triangles) and without starvation (diamonds).

FIG. 4.

Effect of decoyinine addition on ilv-leu expression. (A) relA⁺ strains (FU737 [codY⁺] and FU745 [ΔcodY]) (a) and relA1 strains (FU739 [codY⁺] and FU747 [ΔcodY]) (b), carrying the ilv-leu promoter region (nucleotides −248 to +26), and (B) relA⁺ strains (FU844 [codY⁺] and FU810 [ΔcodY]) (a) and relA1 strains (FU845 [codY⁺] and FU809 [ΔcodY]) (b), carrying the ilv-leu promoter region (nucleotides −55 to +26), were used for the analysis. Each of the codY⁺ ΔcodY set strains was grown as two cultures, and decoyinine was added to only one culture. Cell growth (OD₆₀₀, open symbols) and lacZ expression (β-Gal activity, filled symbols) were monitored during incubation before and after decoyinine addition (codY⁺ strains with decoyinine [circles] and without decoyinine [squares] and ΔcodY strains with decoyinine [triangles] and without decoyinine [diamonds]), as described in the text.

FIG. 5.

Positive stringent response of the ilv-leu operon to amino acid (lysine) starvation involves two molecular mechanisms. As shown at the top, when B. subtilis cells grow in MM medium containing glucose and glutamine and supplemented with 16 amino acids, the CodY protein interacting with GTP and branched-chain amino acids, corepressors of CodY, represses the transcription from the ilv-leu promoter through interference of RNAP entry to it by the complex. As shown at the bottom, when cells are subjected to lysine starvation, the RelA protein synthesizes (p)ppGpp from GTP. The promoter (p)ppGpp inhibits IMP dehydrogenase, forming XMP in the de novo synthesis pathway starting from 5-phosphoribosyl-1-pyrophosphate (PRPP), resulting in lowering of the in vivo GMP and subsequent GTP concentrations. On the other hand, addition of decoyinine to the same medium causes the inhibition of GMP synthase without the production of (p)ppGpp, resulting in lowering of the in vivo GMP and subsequent GTP concentrations. The lowering of the GTP concentration most likely causes relief from CodY repression of the ilv-leu promoter (mechanism 1), whereas lowering of the GTP concentration as well as raising of the ATP and CTP concentrations likely causes the activation of the transcription from the ilv-leu promoter catalyzed by RNAP (mechanism 2).

FIG. 6.

Nucleotide sequence dependency of the CodY-independent positive stringent response in the vicinity of the transcription initiation site. (A) The wild-type ilv-leu TTCA sequence (CA) (nucleotides −2 to +2) in the vicinity of the transcription initiation site was randomly substituted, as described in the text. As a result, the promoter possessing the TTGG sequence did not exhibit the CodY-dependent positive stringent response at all. Based on this fact, we constructed a series of strains carrying the promoter region (nucleotides −55 to +26) possessing the TTGG (GG), TTGA (GA), or TTCG (CG) sequence, which was fused with lacZ in addition to the wild-type strain possessing the CA sequence. (B) Cells of strains FU844 (CA) (circles), FU859 (GA) (triangles), FU905 (CG) (diamonds), and FU904 (GG) (squares) were grown in MM medium supplemented with 16 amino acids as two cultures. When the OD₆₀₀ of the cultures reached 0.5, decoyinine was added to one culture, and then the cultures with and without decoyinine were incubated further. On the other hand, the four strains were grown in the same medium until the OD₆₀₀ was 0.5 as two cultures, and then one culture was subjected to lysine starvation for more than 1 h, as described in the text. Cell growth (OD₆₀₀, dotted lines) and lacZ expression (β-Gal activity, solid lines) were monitored without stringent treatment (open symbols) and with treatment (filled symbols). Arrows indicate the start times of stringent treatment. The numbers in parentheses are ratios, which were obtained by dividing the β-Gal activities of the cells subjected to stringent treatment for 1 h by those of the cells with no treatment after 1 h.

FIG. 7.

Mapping of the 5′ end of the ilv-leu transcripts derived from the base-substituted promoters by means of primer extension analysis. Total RNAs from strains FU844 (CA), FU895 (GA), FU905 (CG), and FU904 (GG) grown in MM medium supplemented with 16 amino acids were annealed with the PEpR primer (Table 2), and then primer extension was performed as described in the text. Lanes A, T, G, and C contained the products of the respective dideoxy sequencing reactions, with the PCR product as the template, as described in the text. The part of the wild-type nucleotide sequence of the coding strand corresponding to the ladder is shown with the transcription initiation base (+1) (enlarged CA sequence [nucleotides +1 to +2]), and the corresponding −10 and −35 regions for the ilv-leu promoter are underlined.

FIG. 8.

In vitro transcription from base-substituted ilv-leu promoters. (A) In vitro transcription from the wild-type ilv-leu promoter. His-tagged B. subtilis RNAP was prepared, and the DNA template used for in vitro transcription was a PCR product covering the ilv-leu promoter region (nucleotides −55 to +26) and a 5′ portion of lacZ, as described in the text. Multiple rounds of transcription producing a 250-base runoff transcript were performed with the substrate concentrations of 200 μM ATP, CTP, and GTP and 12 μM of UTP, as described in the text. The runoff transcripts were electrophoresed on a urea-polyacrylamide gel and then quantified with an image analyzer, as described in the text. (B) (Top) In vitro transcription using a CA template (wild type) possessing C and A at nucleotides +1 and +2 was carried out with various concentrations—0 μM (lane 1), 0.02 μM (lane 2), 0.2 μM (lane 3), 2 μM (lane 4), 20 μM (lane 5), 200 μM (lane 6), and 2,000 μM (lane 7)—of GTP, ATP, and CTP, and fixed concentrations of the other three NTPs. (Bottom) Concentration dependencies for GTP (filled circles), ATP (open circles), and CTP (open squares) of in vitro transcription on the CA, GA, CG, and GG templates, where a transcription ratio of 1 indicates the maximal synthesis of the runoff transcript.