Cross-talk between type three secretion system and metabolism in Yersinia

Abstract

Pathogenic yersiniae utilize a type three secretion system (T3SS) to inject Yop proteins into host cells in order to undermine their immune response. YscM1 and YscM2 proteins have been reported to be functionally equivalent regulators of the T3SS in Yersinia enterocolitica. Here, we show by affinity purification, native gel electrophoresis and small angle x-ray scattering that both YscM1 and YscM2 bind to phosphoenolpyruvate carboxylase (PEPC) of Y. enterocolitica. Under in vitro conditions, YscM1, but not YscM2, was found to inhibit PEPC with an apparent IC(50) of 4 mum (K(i) = 1 mum). To analyze the functional roles of PEPC, YscM1, and YscM2 in Yop-producing bacteria, cultures of Y. enterocolitica wild type and mutants defective in the formation of PEPC, YscM1, or YscM2, respectively, were grown under low calcium conditions in the presence of [U-(13)C(6)]glucose. The isotope compositions of secreted Yop proteins and nine amino acids from cellular proteins were analyzed by mass spectrometry. The data indicate that a considerable fraction of oxaloacetate used as precursor for amino acids was derived from [(13)C(3)]phosphoenolpyruvate by the catalytic action of PEPC in the wild-type strain but not in the PEPC(-) mutant. The data imply that PEPC is critically involved in replenishing the oxaloacetate pool in the citrate cycle under virulence conditions. In the YscM1(-) and YscM2(-) mutants, increased rates of pyruvate formation via glycolysis or the Entner-Doudoroff pathway, of oxaloacetate formation via the citrate cycle, and of amino acid biosynthesis suggest that both regulators trigger the central metabolism of Y. enterocolitica. We propose a "load-and-shoot cycle" model to account for the cross-talk between T3SS and metabolism in pathogenic yersiniae.

FIGURE 1.

Native gel electrophoresis reveals binding of PEPC to YscM1 and YscM2, respectively. Recombinantly expressed and purified proteins were mixed, incubated at room temperature for 5 min, and loaded on 6% HEPES-buffered native polyacrylamide gels, as indicated. A and B, Coomassie-stained native gels; C, Western blot analysis after native gel electrophoresis. A, GST-YscM1 (dimeric) was in 2-fold excess over PEPC (tetrameric). The arrowhead indicates formation of GST-YscM1-PEPC complexes. The excess load of GST demonstrates that GST on its own exhibits no PEPC binding activity. B, YscM1-PEPC complex formation is indicated by an arrowhead. YscM1 (dimeric) was in 3-fold excess over PEPC (tetrameric). C, Western blot analysis of a native gel to demonstrate the YscM2/PEPC interaction. Following electroblotting, the same blot was developed with anti-PEPC and with anti-YscM2 sera successively. Data are representative of 3–5 independent experiments.

FIGURE 2.

Effect of YscM1 on PEPC activity. PEPC activity was monitored spectrophotometrically by coupling the PEPC reaction with the malate dehydrogenase (MDH) reaction recording NADH oxidation at 340 nm. Relative PEPC activities ΔA_i/ΔA₀ in the presence of varying concentrations of YscM1 were plotted. All proteins were recombinantly expressed and purified to homogeneity. The inhibitory effect of YscM1 on PEPC activity was invariably demonstrated by four experimenters independently.

FIGURE 3.

Small angle x-ray scattering experiments reveal PEPC conformational changes mediated by YscM1 and YscM2, respectively. Three views rotated 90°. A, low resolution model of the PEPC tetramer calculated from the scattering data assuming p22 symmetry. B, model calculated from the scattering data obtained from a YscM1/PEPC mixture. C, model calculated from the scattering data obtained from a YscM2/PEPC mixture. The scattering experiments were performed in duplicate with comparable results.

FIGURE 4.

Effect of YscM1/YscM2 overproduction on growth of Y. enterocolitica WA-314 under PEPC-requiring conditions at 27 °C. Yersiniae harboring plasmids pWS-YscM1, pWS-YscM2, and pWS were grown in rich medium overnight. The next day, bacteria were resuspended in M9 minimal salts supplemented with 1% glucose, 0.01% casamino acids, and 10 μg/ml thiamine. 1 mm isopropyl β-d-thiogalactoside was added to all cultures at time 0 to induce overproduction of YscM1/YscM2 or as control (pWS). The optical density was determined at 600 nm (A), and bacteriostasis was checked by plating and colony counting. All experiments in triplicate from independent cultures (error bars, ± one S.D.). B, to control overproduction of YscM proteins, whole cell lysates from all cultures were loaded on denaturing SDS gels. After electroblotting, the membranes were developed with a serum raised against YscM1 or with a serum raised against YscM2, respectively. The experiment as represented here was repeated more than five times with similar results.

FIGURE 5.

MALDI-TOF mass spectrometry of peptide fragments derived from secreted YopM after feeding of yersiniae cultures with uniformly ¹³C-labeled glucose. Y. enterocolitica WA-314 and mutant strains ppc, yscM1, and yscM2, respectively, were grown in BHI medium overnight. The next day, bacteria were resuspended in F-12 medium supplemented with 0.2 mm CaCl₂ and 2% [U-¹³C₆]glucose (¹³C) or 2% unlabeled glucose (¹²C). Cultures were incubated at 37 °C for 2 h, and then Yop secretion was induced, and cultivation was continued for 2 h at 37 °C. The supernatant of the cultures was collected, and proteins were precipitated by tricarboxylic acid and subjected to SDS-PAGE. The gel was Coomassie-stained, and the protein band corresponding to YopM was excised to perform MALDI-TOF mass spectrometry analysis. A, spectra of two YopM fragments from the parental strain WA-314 grown without labeling. B, spectra of the same two tryptic fragments of YopM when the strain was cultured in the presence of [U-¹³C₆]glucose. C–E, corresponding spectra after [U-¹³C₆]glucose labeling from the mutant strains as indicated. Corresponding analyses of YopE and YopH fragments are shown in Figs. S6 and S7. Vertical light gray lines were introduced to facilitate comparison of spectra. The data shown are representative of two independent experiments.

FIGURE 6.

Overall ¹³C excess (mol %) of labeled isotopologues in amino acids derived from bacterial protein after feeding of yersiniae with 0.2% [U-¹³C₆]glucose. The color map indicates ¹³C excess in quasilogarithmic form in order to visualize relatively small ¹³C excess values. For each individual labeling experiment, samples were measured three times.

FIGURE 7.

Isotopologue excess (mol %) of biosynthetic Ala (A), Asp (B), Ser (C), and Phe (D) derived from bacterial protein after feeding of yersiniae with 0.2% [U-¹³C₆]glucose. The colors indicate different strains. Each sample was measured three times. (M + 3) or (M + 2), molecule with three or two ¹³C atoms, respectively.

FIGURE 8.

Reaction network of the central carbon metabolism in Y. enterocolitica WA-314 The model is based on labeling experiments with [U-¹³C₆]glucose and isotopologue profiling of protein-derived amino acids. (M + 6), (M + 4), (M + 3), and (M + 2) subscripts indicate isotopologues containing six, four, three, or two ¹³C atoms in excess, respectively. Green arrows and green subscripts indicate isotopologues due to reactions in the citrate cycle and the malic enzyme. PGA, 3-phosphoglycerate.

FIGURE 9.

Reaction network of the central carbon metabolism in Y. enterocolitica PEPC^-. Red arrows indicate increased flux conducive of the isotopologues shown as red subscripts. For more details, see the legend to Fig. 8.

FIGURE 10.

Reaction network of the central carbon metabolism in Y. enterocolitica YscM1^- or YscM2^-. For more details, see the legend to Fig. 9.