Lowering GTP level increases survival of amino acid starvation but slows growth rate for Bacillus subtilis cells lacking (p)ppGpp

Abstract

Bacterial cells sense external nutrient availability to regulate macromolecular synthesis and consequently their growth. In the Gram-positive bacterium Bacillus subtilis, the starvation-inducible nucleotide (p)ppGpp negatively regulates GTP levels, both to resist nutritional stress and to maintain GTP homeostasis during growth. Here, we quantitatively investigated the relationship between GTP level, survival of amino acid starvation, and growth rate when GTP synthesis is uncoupled from its major homeostatic regulator, (p)ppGpp. We analyzed growth and nucleotide levels in cells that lack (p)ppGpp and found that their survival of treatment with a nonfunctional amino acid analog negatively correlates with both growth rate and GTP level. Manipulation of GTP levels modulates the exponential growth rate of these cells in a positive dose-dependent manner, such that increasing the GTP level increases growth rate. However, accumulation of GTP levels above a threshold inhibits growth, suggesting a toxic effect. Strikingly, adenine counteracts GTP stress by preventing GTP accumulation in cells lacking (p)ppGpp. Our results emphasize the importance of maintaining appropriate levels of GTP to maximize growth: cells can survive amino acid starvation by decreasing GTP level, which comes at a cost to growth, while (p)ppGpp enables rapid adjustment to nutritional stress by adjusting GTP level, thus maximizing fitness.

FIG 1

Diagram of the de novo and salvage GTP biosynthesis pathways in B. subtilis. GTP is synthesized through the de novo pathway (blue) or the salvage pathway from guanosine (GUO) or guanine (GUA) (purple). Steps common to both pathways are in green. Enzymes that are inhibited by (p)ppGpp are underlined. The de novo enzyme GuaB is moderately inhibited by (p)ppGpp, while the salvage enzyme HprT is strongly inhibited. Gmk participates in both pathways and is strongly inhibited by (p)ppGpp. GTP level affects survival of amino acid starvation, including amino acid downshift and survival of arginine hydroxamate (RHX) treatment, and cell growth.

FIG 2

Growth rate of (p)ppGpp⁰ suppressor mutant cells negatively correlates with survival of amino acid downshift and survival of RHX treatment and positively correlates with GTP level. Pearson correlations of growth rate (μ) with survival of amino acid downshift (log₁₀ of plating efficiency on minimal medium) (A) and survival of RHX treatment (% survival) (B) for (p)ppGpp⁰ suppressor mutants are shown (see Table 1 for suppressor mutant alleles). Wild-type and (p)ppGpp⁰ values are plotted for comparison. Values for survival of amino acid downshift (A) and RHX survival (B) are from the work of Kriel et al. (9). For amino acid downshift, cells were grown in CAA medium, harvested during log phase, washed, and plated on minimal medium. For RHX treatment, cells were grown in CAA medium, treated with RHX for 20 min, and plated for enumeration of CFU. (C to E) Pearson correlation of growth rate with GTP/ATP ratio (C), GTP level (D), and ATP level (E) for (p)ppGpp⁰ guaB suppressor mutants. Average GTP/ATP ratios (n = 4) and nucleotide levels (n = 2) are plotted. Arbitrary units (A.U.) are determined from the phosphorimager counts normalized to cell density (OD₆₀₀). All growth rates were calculated from growth curves in low-phosphate CAA medium. Error for growth rates and GTP/ATP ratios for all suppressor mutants is shown in Fig. S1. Strains are YB886 (wild type), JDW755 [(p)ppGpp⁰], AK6 [(p)ppGpp⁰ codY], WZ11 [(p)ppGpp⁰ gmk], AK1 [(p)ppGpp⁰ guaA], AK22, AK14, AK2, WZ7, AK9, AK31, WZ8, WZ9, AK11, and WZ1 [(p)ppGpp⁰ guaB₁ through (p)ppGpp⁰ guaB₁₀].

FIG 3

Guanosine modulates growth of the (p)ppGpp⁰ guaB₁ strain in LB medium. (A) Average and standard deviation of three growth curves for AK22 [(p)ppGpp⁰ guaB₁] in LB medium supplemented with the indicated concentrations of GUO. (B and C) Average growth rates in LB for the (p)ppGpp⁰ guaB₁ strain (B) and the (p)ppGpp⁰ gmk suppressor mutant (WZ11) (C). Error bars indicate standard deviations (n ≥ 6).

FIG 4

Guanosine inhibits growth of (p)ppGpp⁰ guaB₁ cells in CAA medium. (A) Representative growth curve from three independent experiments. (p)ppGpp⁰ guaB₁ (AK22) cells were grown in CAA medium supplemented with the indicated concentrations of GUO. (B) Average growth rates for (p)ppGpp⁰ guaB₁ cells in CAA medium supplemented with GUO. Error bars indicate standard deviations (n ≥ 9). (C and D) (p)ppGpp⁺ cells containing a guaB depletion construct (JDW1397) grow at the same rate regardless of GUO concentration in CAA medium. (C) Representative growth curve of (p)ppGpp⁺ guaB^down (P_spac-guaB) cells in CAA medium supplemented with GUO (n ≥ 5). (D) Average growth rates of (p)ppGpp⁺ guaB^down cells in CAA medium supplemented with GUO. As a control for maximal growth rate when guaB is induced, cells were also grown in the same medium (without GUO) with 0.5 mM IPTG. Error bars indicate standard deviations (n ≥ 5). A comparison of growth rate and GTP level for (p)ppGpp⁺ guaB^down cells is shown in Fig. S1 in the supplemental material.

FIG 5

Adenine prevents GTP accumulation in (p)ppGpp⁰ cells and restores growth in the presence of guanosine. (A) GTP/ATP ratio, (B) change in GTP level (normalized to time zero [T₀]), and (C) change in ATP level (normalized to T₀) in wild-type, (p)ppGpp⁰, and (p)ppGpp⁰ guaB₁ cells treated with GUO and adenine (ADE). Cells were grown in low-phosphate CAA medium, sampled (T₀), and then treated for 20 min with 1 mM GUO/ADE or an equivalent volume of dimethyl sulfoxide (DMSO; control). DMSO is the solvent for ADE. Nucleotide levels were quantified by TLC. Error bars indicate ranges (n = 2). Growth of wild-type (D), (p)ppGpp⁰ (E), and (p)ppGpp⁰ guaB₁ (F) cells upon treatment with 1 mM GUO/ADE or DMSO (control) is shown. Cells were grown to early log phase and then treated. Addition of GUO and/or ADE is indicated with an arrow. One representative growth curve from two replicates is shown. Strains are YB886 (the wild type), JDW755 [(p)ppGpp⁰] and AK22 [(p)ppGpp⁰ guaB₁].

FIG 6

Manipulation of GTP/ATP ratio or GTP level in (p)ppGpp⁰ guaB₁ (AK22) cells alters growth rate. Growth rate is positively correlated with GTP/ATP ratio (A) and GTP level (B) when nucleotide level is modulated by ADE and GUO concentration in CAA medium. (C) ATP level is not correlated with growth rate. GTP and ATP levels are relative to the control in medium lacking GUO or ADE. Values are Pearson correlation (r) and P value. Average values are plotted for growth rate (n ≥ 6) and nucleotide levels (n = 2 for most conditions). Details on ADE and GUO concentrations, with averages and error for all measurements, are in Fig. S4 in the supplemental material.

FIG 7

Guanosine modulates growth of (p)ppGpp⁰ guaB₁ ΔcodY cells in LB medium. (A) Averages and standard deviations of three growth curves for the (p)ppGpp⁰ guaB₁ ΔcodY mutant (JDW2208) in LB medium supplemented with the indicated concentrations of GUO. (B) Average growth rates in LB for the (p)ppGpp⁰ guaB₁ ΔcodY mutant. Error bars indicate standard deviations (n ≥ 6).

FIG 8

Model of the quantitative relationship between GTP level and stress survival (dashed line) or growth (solid line) in the absence of (p)ppGpp.