A kinetic platform to determine the fate of nitric oxide in Escherichia coli

Abstract

Nitric oxide (NO•) is generated by the innate immune response to neutralize pathogens. NO• and its autoxidation products have an extensive biochemical reaction network that includes reactions with iron-sulfur clusters, DNA, and thiols. The fate of NO• inside a pathogen depends on a kinetic competition among its many targets, and is of critical importance to infection outcomes. Due to the complexity of the NO• biochemical network, where many intermediates are short-lived and at extremely low concentrations, several species can be measured, but stable products are non-unique, and damaged biomolecules are continually repaired or regenerated, kinetic models are required to understand and predict the outcome of NO• treatment. Here, we have constructed a comprehensive kinetic model that encompasses the broad reactivity of NO• in Escherichia coli. The incorporation of spontaneous and enzymatic reactions, as well as damage and repair of biomolecules, allowed for a detailed analysis of how NO• distributes in E. coli cultures. The model was informed with experimental measurements of NO• dynamics, and used to identify control parameters of the NO• distribution. Simulations predicted that NO• dioxygenase (Hmp) functions as a dominant NO• consumption pathway at O2 concentrations as low as 35 µM (microaerobic), and interestingly, loses utility as the NO• delivery rate increases. We confirmed these predictions experimentally by measuring NO• dynamics in wild-type and mutant cultures at different NO• delivery rates and O2 concentrations. These data suggest that the kinetics of NO• metabolism must be considered when assessing the importance of cellular components to NO• tolerance, and that models such as the one described here are necessary to rigorously investigate NO• stress in microbes. This model provides a platform to identify novel strategies to potentiate the effects of NO•, and will serve as a template from which analogous models can be generated for other organisms.

Figure 1. Simplified diagram of the NO• biochemical reaction network in an E. coli culture.

The intracellular and extracellular (media) compartments are represented by the green and tan shaded regions, respectively. The lower-left corner represents the gas phase in direct contact with the liquid media. Species in bold text represent NO• and its reactive oxidized forms (NO₂•, N₂O₃, and ONOO⁻). Italic text indicates the enzyme or group of enzymes responsible for the associated reaction/pathway. Red reaction arrows represent NO• autoxidation; purple, enzymatic detoxification; blue, ONOO⁻ formation and degradation; tan, cytochrome inhibition; teal, [Fe-S] nitrosylation and repair; green, thiol nitrosation and denitrosation; gray, tyrosine nitration; orange, DNA deamination and repair.

Figure 2. Dynamics of NO• in aerobic wild-type E. coli cultures.

(A) NO• concentration following delivery of 0.5 mM DPTA to a culture of aerobic, exponential-phase, wild-type E. coli at an OD₆₀₀ of 0.05 was measured experimentally (solid red line) and predicted by the model (dashed black line). Error bars (light red) represent the standard error of the mean for 3 independent experiments. (B) Fold increase in SSR between the experimentally measured and predicted NO• concentration as a function of parameter value for uncertain parameters that significantly affected the SSR upon variation. The remaining 35 parameters exhibited a negligible increase in SSR when varied. (C) Simulated NO• concentration profile (black line) and corresponding cumulative distribution of total NO• consumption following the addition of 0.5 mM DPTA to wild-type E. coli. The stacked, shaded regions represent the predicted cumulative fraction of NO• consumed by each pathway, where the bar to the right of the plot represents the final distribution of NO• consumption at the limit t→∞. “Cellular” refers to NO• consumed by any intracellular pathway, “Gas” is loss of NO• to the gas phase, and “Autoxidation” is the reaction of NO• with O₂ in the growth media. (D) Predicted cumulative distribution of intracellular NO• consumption in wild-type culture following addition of 0.5 mM DPTA. Additional bar at far right shows the contribution of other pathways that are not visible on the full 0–100% scale. “Hmp” is detoxification of NO• by Hmp, “Oxidation” is NO• consumed through reaction with O₂ or O₂•⁻, “NorV/NrfA” is the reduction of NO• by NorV or NrfA, and “[Fe-S]” is NO• consumed by the nitrosylation of iron-sulfur clusters.

Figure 3. Effect of Δhmp on NO• dynamics in aerobic E. coli cultures.

(A) Simulated NO• profiles for wild-type (solid black line) and Δhmp (dashed red line) cultures following addition of 0.5 mM DPTA. (B) NO• concentration following delivery of 0.5 mM DPTA NONOate to a culture of exponential-phase, Δhmp E. coli at an OD₆₀₀ of 0.05 was measured experimentally (solid blue line) and predicted by the model (dashed black line). Error bars (light blue) represent the standard error of the mean for 3 independent experiments. The inset shows the measured (blue bars) and predicted (dashed white bars) NO₂ ⁻ and NO₃ ⁻ concentrations at 10 h after DPTA delivery to the Δhmp culture. Error bars represent the standard error of the mean for 3 independent experiments. The dotted red line represents the limit of detection for the assay, with the asterisk (*) indicating that the measured [NO₃ ⁻] was negligible, as it did not differ significantly from the detection limit (one-sample t-test, 95% confidence). (C) Simulated NO• concentration profile (black line) and corresponding cumulative distribution of total NO• consumption following addition of 0.5 mM DPTA to Δhmp E. coli. (D) Predicted cumulative distribution of intracellular NO• consumption in a Δhmp culture following addition of 0.5 mM DPTA.

Figure 4. Effect of experimentally-accessible parameters on NO• dynamics.

(A) Predicted total (left column) and intracellular (right column) NO• distributions corresponding to variation in 27 experimentally-accessible parameters. Each parameter name is listed adjacent to the 5 total and intracellular NO• distributions obtained as a result of varying that parameter logarithmically within its physiological range. The distributions are ordered such that the parameter is increasing in value from top to bottom. (B) Percentage of total NO• predicted to be consumed by Hmp as a function of NO• donor release rate. Points (green circles) correspond to the release rates measured for DPTA and PAPA. (C) Simulated NO• concentration profile for wild-type (solid black line) and Δhmp (red dashed line) cultures treated with 0.5 mM PAPA.

Figure 5. Increasing the NO• delivery rate alters the dynamics and distribution of NO• consumption.

(A) Simulated NO• concentration curve (black line) and corresponding cumulative distribution of total NO• consumption in a wild-type culture following addition of 0.5 mM PAPA. (B) Predicted cumulative distribution of intracellular NO• consumption in a wild-type culture following addition of 0.5 mM PAPA. (C) Simulated NO• concentration curve (black line) and corresponding cumulative distribution of total NO• consumption in a Δhmp culture following addition of 0.5 mM PAPA. (D) Predicted cumulative distribution of intracellular NO• consumption in a Δhmp culture following addition of 0.5 mM PAPA.

Figure 6. The utility of Hmp for NO• consumption and tolerance decreases with increased NO• delivery rate.

(A) Experimentally measured NO• concentration profiles following addition of 0.5 mM PAPA to a culture of wild-type (solid red line) or Δhmp (solid blue line) E. coli at an OD₆₀₀ of 0.05. Lines depict the mean of three independent experiments, and error bars (light red and light blue for wild-type and Δhmp, respectively) represent the standard error of the mean. Also shown are the corresponding model-predicted NO• concentration profiles for wild-type (dark red dashed line) or Δhmp (dark blue dotted line) cultures. (B). Predicted cumulative concentration (per unit cellular volume) of [2Fe-2S] and [4Fe-4S] clusters damaged following DPTA (purple) or PAPA (teal) treatment of wild-type (solid lines) and Δhmp cultures (dashed lines) is plotted as a function of the model parameter governing the rate of [Fe-S] nitrosylation by NO• (k _NO•-[Fe-S]). (C) Predicted durations of cytochrome bd (Cyd) inhibition by NO•, defined as the time required for the percentage of NO•-bound Cyd to drop below 50% of the total Cyd concentration. (D) Experimentally measured growth curves (quantified by OD₆₀₀) for wild-type and Δhmp cultures following treatment with 0.5 mM of DPTA or PAPA demonstrate more comparable duration of bacteriostasis between wild-type and Δhmp for treatment with PAPA than DPTA. Due to the faster NO• delivery kinetics associated with PAPA, OD₆₀₀ readings were taken more frequently (20 min intervals) than with DPTA (30 min intervals). After addition of DPTA to Δhmp cells, growth resumption was not observed within the 10 h timeframe of the experiment.

Figure 7. NO• dynamics and distribution in E. coli cultures under microaerobic conditions.

Shown along the left are the simulated and experimentally-measured NO• concentration profiles for (A) wild-type, (B) ΔnorV, and (C) Δhmp cultures following addition of 0.5 mM DPTA, where the O₂ concentration was maintained at 35 µM by bubbling with N₂. Error bars (light red, light purple, and light blue for wild-type, ΔnorV, and Δhmp, respectively) represent the standard error of the mean for at least 3 independent experiments. Shown directly to the right of the NO• concentration profiles are the corresponding predicted cumulative distributions of total (left) and intracellular (right) NO• consumption.