Lactoylglutathione lyase, a critical enzyme in methylglyoxal detoxification, contributes to survival of Salmonella in the nutrient rich environment

Abstract

Glyoxalase I which is synonymously known as lactoylglutathione lyase is a critical enzyme in methylglyoxal (MG) detoxification. We assessed the STM3117 encoded lactoylglutathione lyase (Lgl) of Salmonella Typhimurium, which is known to function as a virulence factor, due in part to its ability to detoxify methylglyoxal. We found that STM3117 encoded Lgl isomerises the hemithioacetal adduct of MG and glutathione (GSH) into S-lactoylglutathione. Lgl was observed to be an outer membrane bound protein with maximum expression at the exponential growth phase. The deletion mutant of S. Typhimurium (Δlgl) exhibited a notable growth inhibition coupled with oxidative DNA damage and membrane disruptions, in accordance with the growth arrest phenomenon associated with typical glyoxalase I deletion. However, growth in glucose minimal medium did not result in any inhibition. Endogenous expression of recombinant Lgl in serovar Typhi led to an increased resistance and growth in presence of external MG. Being a metalloprotein, Lgl was found to get activated maximally by Co(2+) ion followed by Ni(2+), while Zn(2+) did not activate the enzyme and this could be attributed to the geometry of the particular protein-metal complex attained in the catalytically active state. Our results offer an insight on the pivotal role of the virulence associated and horizontally acquired STM3117 gene in non-typhoidal serovars with direct correlation of its activity in lending survival advantage to Salmonella spp.

Figure 1 (See previous page).

Compromised survival and membrane architecture in Δlgl cells when grown in rich medium. (A) Genetic organization of the STM3117-STM3120 locus in S. Typhimurium. STM3117–3120 genes are co-transcribed as a single mRNA. Amplicons representing the 3 intergenic loci (indicated by flanking gene number) spanning the 4 genes, generated with specific primers against the cDNA. Quantitative RT-PCR analysis of transcripts of STM3117 and junctions between STM3117–3118, STM3118–3119, STM3119–3120 and STM3120–3121. The last junction region was kept as negative control. The relative abundance of the transcripts was normalized to the level of STM3117. The values are the mean ± SD of 2 independent cDNA samples. The RT-PCR samples were run in gel to validate product formation. (B) Survival rate of bacteria in presence of MG (0.2 mM). Survival upon exposure to MG was calculated by dividing the growth rate of the treated culture upon the untreated one. The CFU values at each time point are the mean±SD of quadruplate samples, representative of 3 independent experiments. (C) Relative survival of WT, Δlgl and the complement in presence of indicated concentration of external MG in LB medium compared to the untreated control, analyzed by O.D₆₀₀ measurement. Representative graphs are shown here. Δlgl shows a significant decrease in survival in the initial 2 h of exposure to 10 μM MG. (D) Growth of WT, Δlgl and/or complement strain in LB, TB (terrific broth), M9 (minimal medium with 0.3% carbon) and F (phagosome mimicking) media as monitored by O.D₆₀₀ measurement. Results of representative experiments performed in triplicates are shown. Error bars indicate standard deviations. (E) Enumeration of CFU of WT and Δlgl at the indicated time points during growth in LB medium. WT and Δlgl grown on LB agar plate for 12 hr depicting the difference in colony size. (F) Estimation of non viable population after 3 hr of growth using propidium iodide (1 μg/ml). Upper right quadrant shows the mean percentage of non viable cells in the indicated strains from triplicates. Unstained WT cells as control. Mutant culture exhibited a slight increase in non-viable bacterial population. (G) Representative scanning electron micrographs of WT and Δlgl grown in LB till exponential (-L) or stationary phase (-S) and in M9 medium till exponential phase. Adjacent images in the first and middle rows represent 2 different magnifications. Arrows indicate membrane disruption and leakage of cytosolic contents in Δlgl. Bottom panel shows enlarged images highlighting the damage in Δlgl. Graph shows the percentage of mutant bacteria affected. Lipid peroxidation levels in WT and Δlgl strains at exponential phase of growth in LB. The samples were evaluated for malondialdehyde (MDA) production using the lipid peroxidation assay kit from Abcam. Results are the means ± SEM of 3 independent determinations. (H) Expression level of STM3117 in WT S. Typhimurium grown till log, late log and stationary phase in LB. *P < 0 .05; **P < 0 .01; ***P < 0 .001 (Students t-test).

Figure 1 (See previous page).

Figure 2.

Endogenous methylglyoxal toxicity in Δlgl strain causes oxidative DNA damage. (A) Representative histograms of TUNEL stained WT and Δlgl cells after 4 h of growth in LB with respect to unstained WT cells. The dotted line at 10³ fluorescence unit shows the cutoff of TUNEL positive cells. 5 μg/ml Ciprofloxacin (c) treated (for 2 h) bacteria were taken as positive controls. The mean ± SEM fluorescence intensities of FITC are shown graphically for 3 independent experiments. Representative images of WT and Δlgl bacteria which were processed the same way as in FACS for detection of TUNEL signal (green) by confocal microscopy. Scale bar 10 μm. (B) TUNEL stained WT and Δlgl (and complement) cells after 3 h of growth in M9 minimal or LB with or without exposure (1 hr or 30 min) to 0.2 μM methylglyoxal. Numbers on each histogram indicate the weighted mean fluorescence of TUNEL, done in quadruplets. (C) The release of ³(H)- Thymidine labeled nucleotides in culture supernatants of WT and Δlgl upon 0.2 mM MG exposure (2 h). The percentage increase in the level of ³(H)- Thymidine in cell free filtrate of indicated strains, determined by dividing the radioactivity (cpm) present in cell free supernatant of treated sample to that present in untreated one. Bars represent the mean ± SEM from 3 independent experiments performed in triplicates. **P < 0 .01; ***P < 0 .001 (Students t-test). (D) HPLC chromatograms of bacterial culture (perchloric acid extract) after derivatisation of cellular MG into 2-methylquinoxaline (2-MQ) which eluted at 6 min. Five-methylquinoxaline (27.7 nmoles) (5-MQ) was used as the internal standard eluting at 8.75 min. Two-MQ content corresponded to 0.03 and 0.016 mV height in WT and Δlgl cells respectively. The peak heights were converted to nmoles/g bio mass as per the calibration curve. Bottom panel shows chromatograms where the bacterial extracts were spiked with 0.1 nmol of MG. Δlgl chromatogram shows an increase in the analyte (2-MQ) peak height compared to that in the WT chromatogram.

Figure 2.

Figure 2.

Figure 3.

Biochemical characterization of recombinant Lgl. (A) Multiple sequence alignment of Lgl with glyoxalase I of E. coli, S. Typhimurium, S. Typhi, H. sapiens and hypothetical protein (HP, another putative glyoxalase I) of S. Typhimurium. Phylogenetic tree using the amino acid sequences of glyoxalase I, Lgl and the second putative GlxI (HP) from S. Typhimurium and other organisms. (B) Affinity purification, using Ni-NTA resin, of the supernatant (post lysis) and urea soluble (pellet) fractions of the lysate containing recombinant Lgl, with different concentrations of imidazole (Imd). A polypeptide of expected molecular weight (∼17 kDa) was detected by using anti-His antibody in the induced lysate of pPROEX:lgl transformed E. coli BL21 cells (Un-uninduced, In-induced with IPTG). Mass spectrometric profile of the pure recombinant Lgl validated the molecular weight to be ∼16.7 kDa.

Figure 3.

.

Figure 4.

Localization of recombinant Lgl in S. Typhimurium (A) Sub-cellular fractions of Δlgl expressing pPROEX:lgl subjected to immunoblot analysis with anti-histidine antibody. Presence of His tagged Lgl in Cytosolic (cyto) and crude membrane (CM) fraction but not in periplasmic fraction (PP). Lgl localizes to the outer membrane fraction (OM) upon further separation of crude membranes into outer and inner membranes (IM). Purified recombinant Lgl was run in parallel to validate the presence of recombinant Lgl in the fractionated samples. DnaK and OmpX proteins were probed as positive controls representative of the corresponding sub-cellular fraction. (B) Prediction of sub-cellular localization by CELLO (subcellular localization predictor) of glyoxalase I and Lgl from S. Typhimurium. Numbers depict the probability of the prediction based on amino acid sequence screened under multiple parameters.

Figure 5.

Expression of Lgl in S. Typhi confers partial protection from external methylglyoxal toxicity. (A) Growth of WT S. Typhi and S. Typhi expressing pQE:lgl upon exposure to different concentrations of MG in LB medium. (B) Growth rate of WT S. Typhi and S. Typhi (pQE:lgl) in presence or absence of 0.2 mM MG in LB medium. From the time of MG exposure, at every hour interval (for 2 h), fixed volume of cultures were plated to enumerate the CFUs. Growth rate was calculated by dividing the CFUs of 2 consecutive hours. The CFU values at each time point were the mean ± SD of quadruplet samples.