Identification of a regulator that controls stationary-phase expression of catalase-peroxidase in Caulobacter crescentus

Abstract

Expression of the catalase-peroxidase of Caulobacter crescentus, a gram-negative member of the alpha subdivision of the Proteobacteria, is 50-fold higher in stationary-phase cultures than in exponential cultures. To identify regulators of the starvation response, Tn5 insertion mutants were isolated with reduced expression of a katG::lacZ fusion on glucose starvation. One insertion interrupted an open reading frame encoding a protein with significant amino acid sequence identity to TipA, a helix-turn-helix transcriptional activator in the response of Streptomyces lividans to the peptide antibiotic thiostrepton, and lesser sequence similarity to other helix-turn-helix regulators in the MerR family. The C. crescentus orthologue of tipA was named skgA (stationary-phase regulation of katG). Stationary-phase expression of katG was reduced by 70% in the skgA::Tn5 mutant, and stationary-phase resistance to hydrogen peroxide decreased by a factor of 10. Like the wild type, the skgA mutant exhibited starvation-induced cross-resistance to heat and acid shock, entered into the helical morphology that occurs after 9 to 12 days in stationary phase, and during exponential growth induced katG in response to hydrogen peroxide challenge. Expression of skgA increased 5- to 10-fold in late exponential phase. skgA is the first regulator of a starvation-induced stress response identified in C. crescentus. SkgA is not a global regulator of the stationary-phase stress response; its action encompasses the oxidative stress-hydrogen peroxide response but not acid or heat responses. Moreover, SkgA is not an alternative sigma factor, like RpoS, which controls multiple aspects of starvation-induced cross-resistance to stress in enteric bacteria. These observations raise the possibility that regulation of stationary-phase gene expression in this member of the alpha subdivision of the Proteobacteria is different from that in Escherichia coli and other members of the gamma subdivision.

FIG. 1

Growth stage dependence of katG expression in skg mutants. LacZ activity from a chromosomal katG::lacZ translational fusion was measured with ONPG as substrate at the indicated times during growth in PYE. Bars: open, Tn5 control strain SGC125; hatched, skgA::Tn5 strain SGC127; filled, skgB::Tn5 strain SGC126. Values of OD₆₀₀ for exponential cultures of SGC125, SGC127, and SGC126 were 0.44, 0.55, and 0.29, respectively. We previously showed for wild-type C. crescentus a low and constant expression of katG::lacZ during the entire period of exponential growth in katG fusion strain SGC109. The parental strain for the Tn5 mutants was SGC109. Expon, exponential.

FIG. 2

Stationary-phase survival. Cultures in PYE were sampled at the indicated times, and titers were determined on PYE plates without antibiotics. The OD₆₀₀ values at 0 h were 0.06 to 0.11 and for exponential-phase samples (4.5 to 6 h) were 0.29 to 0.65. Saturation OD₆₀₀ values for skgA and skgB strains were slightly lower than those for the katG-null and Tn5 control strains, 1.0 to 1.2 versus 1.2 to 1.5. Strains used were SGC111 katG (■), SGC125 Tn5 control (●), SGC127 skgA::Tn5 (○), and SGC126 skgB::Tn5 (▴).

FIG. 3

Amino acid sequence alignment of SkgA with known transcriptional regulators. Asterisks designate putative helix-turn-helix DNA binding motifs in TipA_L from S. lividans and NolA from Bradyrhizobium sp. Identities with SkgA are indicated by vertical lines. TipA_S, lacking the amino-terminal DNA binding domain and containing the carboxyl-terminal ligand binding domain, corresponds to residues 110 to 253 of TipA_L. The Tn5 insertion in skgA occurs within the codon for the Ala residue designated by the arrowhead. The amino terminus of SkgA was located by homology and may be Val-1 or Val-5.

FIG. 4

Map of the C. crescentus skgA locus. The XhoI-HindIII and HindIII-HindIII fragments were subcloned from the original cosmid. EagI and PstI sites mark the ends of the sequenced region (dashed line) submitted to GenBank (accession no. AF170912) and are not unique sites. The Tn5 insertion site is designated by an arrowhead. The 5′ end of mutL was placed by assuming equal lengths for the C. crescentus and E. coli homologues.

FIG. 5

Growth stage dependence of skgA expression. LacZ activity in SGC130 skgA⁺ skgA::lacZ was measured by hydrolysis of CPRG during growth in PYE. Thick bars indicate LacZ activity as nanomoles of CPRG hydrolyzed per minute divided by OD₆₀₀ of the culture sampled; solid circles indicate OD₆₀₀.

FIG. 6

Resistance to heat and acid in exponential and stationary phase. (A) 50°C heat shock. (B) pH 4 shock. (C) 5 mM H₂O₂. See Materials and Methods for details. Circles, Tn5 control (strain SGC125). Squares, skgA::Tn5 (strain SGC127). Open symbols, exponential phase (OD₆₀₀ of 0.36 to 0.76). Filled symbols, stationary phase (20 h later). In panel C (5 mM H₂O₂), no CFU were seen for either exponential culture at time points beyond 1 h.

FIG. 6

Resistance to heat and acid in exponential and stationary phase. (A) 50°C heat shock. (B) pH 4 shock. (C) 5 mM H₂O₂. See Materials and Methods for details. Circles, Tn5 control (strain SGC125). Squares, skgA::Tn5 (strain SGC127). Open symbols, exponential phase (OD₆₀₀ of 0.36 to 0.76). Filled symbols, stationary phase (20 h later). In panel C (5 mM H₂O₂), no CFU were seen for either exponential culture at time points beyond 1 h.