A Novel Growth-Based Selection Strategy Identifies New Constitutively Active Variants of the Major Virulence Regulator PrfA in Listeria monocytogenes

Abstract

Listeria monocytogenes is a Gram-positive pathogen able to cause severe human infections. Its major virulence regulator is the transcriptional activator PrfA, a member of the Crp/Fnr family of transcriptional regulators. To establish a successful L. monocytogenes infection, the PrfA protein needs to be in an active conformation, either by binding the cognate inducer glutathione (GSH) or by possessing amino acid substitutions rendering the protein constitutively active (PrfA*). By a yet unknown mechanism, phosphotransferase system (PTS) sugars repress the activity of PrfA. We therefore took a transposon-based approach to identify the mechanism by which PTS sugars repress PrfA activity. For this, we screened a transposon mutant bank to identify clones able to grow in the presence of glucose-6-phosphate as the sole carbon source. Surprisingly, most of the isolated transposon mutants also carried amino acid substitutions in PrfA. In transposon-free strains, the PrfA amino acid substitution mutants displayed growth, virulence factor expression, infectivity, and DNA binding, agreeing with previously identified PrfA* mutants. Hence, the initial growth phenotype observed in the isolated clone was due to the amino acid substitution in PrfA and unrelated to the loci inactivated by the transposon mutant. Finally, we provide structural evidence for the existence of an intermediately activated PrfA state, which gives new insights into PrfA protein activation.IMPORTANCE The Gram-positive bacterium Listeria monocytogenes is a human pathogen affecting mainly the elderly, immunocompromised people, and pregnant women. It can lead to meningoencephalitis, septicemia, and abortion. The major virulence regulator in L. monocytogenes is the PrfA protein, a transcriptional activator. Using a growth-based selection strategy, we identified mutations in the PrfA protein leading to constitutively active virulence factor expression. We provide structural evidence for the existence of an intermediately activated PrfA state, which gives new insights into PrfA protein activation.

Keywords: ActA; LLO; Listeria monocytogenes; PrfA; PrfA*; crystal structure.

FIG 1

Growth of Listeria monocytogenes in defined medium (DM) supplemented with glucose-6-phosphate (G-6-P). (A) Four strains were tested for growth in DM with 0.2% G-6-P as the sole carbon source. These were (i) the wild-type EGDe strain (WT), (ii) the strain carrying an in-frame deletion of the hpt gene (Δhpt), (iii) the strain carrying the Gly to Ser substitution of codon 145 in the PrfA protein (PrfA_G145S), and (iv) the PrfA_G145S strain carrying an in-frame deletion of the hpt gene (PrfA_G145S Δhpt). Growth was monitored by viable count for 3 days and shown as mean values with standard deviations (n = 3). Statistical analysis was used to compare growth of the mutant strains with that of the wild-type strain at each time point (Student’s t test [two tailed; *, P < 0.05; **, P < 0.01]). (B) Fosfomycin resistance of the indicated strains. Fosfomycin discs were used, and the clearing zone was measured. The radius of the clearing zone is indicated relative to the wild type as an average of three independent experiments. Statistical analysis was used to compare fosfomycin sensitivity of the PrfA* strain with the WT and the PrfA* and Δhpt strains (Student’s t test [two tailed; ***, P < 0.001]). (C) (Top) Expression of hpt in the indicated strains grown in BHI until the OD₆₀₀ is 1. RNA was isolated and hpt expression was examined with Northern blotting using radiolabeled probes against hpt and tmRNA (control). A representative of three independent experiments is shown. (Bottom) Measurement of hpt expression from the top panel. Expression is relative to WT (set to 1). Student’s t test (two tailed; ***, P < 0.001).

FIG 2

Effect of the amino acid substitutions on PrfA activity. (A) Transposon-free strains expressing PrfA proteins with the wild-type sequence or indicated amino acid substitutions were tested for growth in DM supplemented with 0.2% G-6-P for 72 h. A representative of five independent experiments is shown. (B) Fosfomycin resistance of the indicated strains. Fosfomycin discs were used, and the clearing zone was measured. The radius of the clearing zone is indicated relative to the wild type as an average of three independent experiments. Statistical analysis was used to compare fosfomycin sensitivity of the PrfA* strain with the that of the wild type and the PrfA* and Δhpt strains (Student’s t test [two tailed; *, P < 0.05; **, P < 0.01; ***, P < 0.001]). (C) (Top) Expression of hpt in the indicated strains grown in BHI until the OD₆₀₀ reached 1. RNA was isolated and hpt expression was examined with Northern blotting using radiolabeled probes against hpt and tmRNA (control). A representative of three independent experiments is shown. (Bottom) Quantification of hpt expression from the top panel. Expression is relative to the WT (set to 1). Student’s t test (two tailed; *, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 3

The identified point mutations in PrfA give rise to PrfA* phenotypes. (A) ActA, PrfA, and LLO virulence factor expression of the indicated strains examined using Western blots. The strains were grown in BHI until the OD₆₀₀ reached 1. RNA polymerase beta (RpoB) was used as a loading control for whole-cell fraction samples (ActA and PrfA), and P60 was used as a loading control for the secreted fraction (LLO). A representative of four independent experiments is shown. (B) Expression of ActA and PrfA virulence factors in the presence of 1% Amberlite. The strains were grown in BHI with or without 1% Amberlite XAD4 until the OD₆₀₀ reached 1 before sample preparation and Western blotting. RNA polymerase beta (RpoB) was used as a loading control. A representative of four independent experiments is shown. See also Fig. S3 for quantification of ActA, PrfA, and LLO expression levels from panel A and Fig. S4 for quantification of ActA expression levels from panel B.

FIG 4

The PrfA* mutant strains show a larger uptake and cell-to-cell spread compared to a wild-type strain. (A) Intracellular growth of the indicated strains was tested by viable count in the colon epithelial cell line Caco-2. An average with standard deviation of three independent experiments is shown. Statistical analysis (inset) compared the number of CFU for the wild-type strain with that of the mutant strains at 2, 4, and 6 h postinfection (Student’s t test, [two tailed; *, P < 0.05; ***, P < 0.001; ns, no significant difference]). (B) The indicated strains were tested for infection (adhesion/invasion and/or cell-to-cell spread) of the Caco-2 derivative TC7 cell line using a multiplicity of infection of 1:500. The number of plaques formed was determined and represented as the number of plaques relative to the WT (set to 1). An average with standard deviation from three independent experiments is shown. Statistical analysis compared infection of the wild-type strain with that of the mutant strains (Student’s t test [two tailed; **, P < 0.01; ***, P < 0.001]).

FIG 5

Superimposed structures of PrfA. All superimpositions are based on residues 2 to 237 in monomer B. For root-mean-square (rms) deviations, see Table S2. (A) The PrfA_WT homodimer (PDB code 2BEO [17]). The HTH motifs are shown in dark blue; only the recognition helix (αF) is folded. (B) Superimposed structures of the four intermediately activated structures of PrfA studies: PrfA_A94V, PrfA_L140H, PrfA_L140F, and PrfA_A218G. Only the HTH motifs in monomer B are folded. (C) Superimposed structures of PrfA_WT, PrfA_G145S, and one representative of the intermediately activated fold (PrfA_A218G). Noteworthy is the shift with up to 6 Å in the position of the recognition helix in monomer A in PrfA_A218G compared to that of monomer A in PrfA_G145S. The C-alpha trace of PrfA_G145S is shown in green, with the folded HTH motifs shown in orange. (D) Seven structures of PrfA in complex with DNA: PrfA_WT (PDB code 5LEJ [10]), PrfA_WT-GSH (PDB code 5LRS [10]), PrfA_G145S (PDB code 5LEK [10]), PrfA_A94V, PrfA_L140H, PrfA_L140F, and PrfA_A218G. Note the close to identical structures of all PrfAs when in complex with DNA (rms deviations < 0.5 Å).