Contextual Flexibility in Pseudomonas aeruginosa Central Carbon Metabolism during Growth in Single Carbon Sources

Abstract

Pseudomonas aeruginosa is an opportunistic human pathogen, particularly noted for causing infections in the lungs of people with cystic fibrosis (CF). Previous studies have shown that the gene expression profile of P. aeruginosa appears to converge toward a common metabolic program as the organism adapts to the CF airway environment. However, we still have only a limited understanding of how these transcriptional changes impact metabolic flux at the systems level. To address this, we analyzed the transcriptome, proteome, and fluxome of P. aeruginosa grown on glycerol or acetate. These carbon sources were chosen because they are the primary breakdown products of an airway surfactant, phosphatidylcholine, which is known to be a major carbon source for P. aeruginosa in CF airways. We show that the fluxes of carbon throughout central metabolism are radically different among carbon sources. For example, the newly recognized "EDEMP cycle" (which incorporates elements of the Entner-Doudoroff [ED] pathway, the Embden-Meyerhof-Parnas [EMP] pathway, and the pentose phosphate [PP] pathway) plays an important role in supplying NADPH during growth on glycerol. In contrast, the EDEMP cycle is attenuated during growth on acetate, and instead, NADPH is primarily supplied by the reaction catalyzed by isocitrate dehydrogenase(s). Perhaps more importantly, our proteomic and transcriptomic analyses revealed a global remodeling of gene expression during growth on the different carbon sources, with unanticipated impacts on aerobic denitrification, electron transport chain architecture, and the redox economy of the cell. Collectively, these data highlight the remarkable metabolic plasticity of P. aeruginosa; that plasticity allows the organism to seamlessly segue between different carbon sources, maximizing the energetic yield from each.IMPORTANCEPseudomonas aeruginosa is an opportunistic human pathogen that is well known for causing infections in the airways of people with cystic fibrosis. Although it is clear that P. aeruginosa is metabolically well adapted to life in the CF lung, little is currently known about how the organism metabolizes the nutrients available in the airways. In this work, we used a combination of gene expression and isotope tracer ("fluxomic") analyses to find out exactly where the input carbon goes during growth on two CF-relevant carbon sources, acetate and glycerol (derived from the breakdown of lung surfactant). We found that carbon is routed ("fluxed") through very different pathways during growth on these substrates and that this is accompanied by an unexpected remodeling of the cell's electron transfer pathways. Having access to this "blueprint" is important because the metabolism of P. aeruginosa is increasingly being recognized as a target for the development of much-needed antimicrobial agents.

Keywords: Pseudomonas aeruginosa; acetate metabolism; carbon flux; carbon metabolism; denitrification; glycerol metabolism; proteomics.

FIG 1

Biochemical pathways involved in central carbon catabolism in P. aeruginosa PAO1. The metabolic network was constructed around six main metabolic blocks, which are identified with different colors as follows: (i) the peripheral pathways that encompass the oxidative transformation of glucose, acetate, and glycerol (orange); (ii) the Embden-Meyerhoff-Parnas pathway (EMP; nonfunctional in P. aeruginosa due to the absence of 6-phosphofructo-1-kinase) (purple); (iii) the pentose phosphate pathway (PPP) (red); (iv) the Entner-Doudoroff pathway (EDP) (green); (v) the tricarboxylic acid cycle and glyoxylate shunt (blue); (vi) the anaplerotic and gluconeogenic bioreaction metabolic block (gray).

FIG 2

Comparison between protein and transcript fold changes (FC) for selected P. aeruginosa enzymes involved in central carbon metabolism. The figure shows the log₂ fold changes in protein and transcript levels in (i) the peripheral pathways that encompass the oxidative transformation of glucose, acetate, and glycerol and the corresponding phosphorylated derivatives of these metabolites (orange); (ii) EMP pathway (nonfunctional, due to the absence of 6-phosphofructo-1-kinase activity) (purple); (iii) the pentose phosphate (PP) pathway (red); (iv) the upper ED pathway (green); (v) the tricarboxylic acid cycle and glyoxylate shunt (blue); and (vi) anaplerotic and gluconeogenic bioreactions (gray). RNA-Seq and proteomic data are shown in Data Set S1. Correlation plots are shown in Data Set S3.

FIG 3

Growth on different carbon sources primarily affects metabolism. The statistically significant changes (P value of ≤ 0.01; fold change value of greater or equal to 1 or less than or equal to −1) during growth on glycerol and acetate are illustrated as Voronoi tree maps, which were created using the proteomaps Web service (15). Most of the proteomic changes were centered on “metabolism,” notably, “central carbon metabolism,” ‘biosynthesis,” “signaling and cellular process,” and “energy metabolism.” Pathway assignment was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) data set. Proteome alterations which could not be assigned to a specific pathway (uncharacterized/hypothetical proteins) are indicated as “Not Mapped.”

FIG 4

In vivo carbon flux distributions in central metabolism of P. aeruginosa PAO1 during growth on glycerol (A) or acetate (B) as the sole carbon source. Flux is expressed as a molar percentage of the average glycerol (9.2 mmol g⁻¹ h⁻¹) or acetate (30.4 mmol g⁻¹ h⁻¹) uptake rate, calculated from the individual rates in Data Set S2. Anabolic pathways from 11 precursors to biomass are indicated as filled blue triangles. The flux distributions with bidirectional resolution (i.e., net and exchange fluxes), including the drain from metabolic intermediates to biomass and confidence intervals of the flux estimates, are provided in Data Set S2. In agreement with previous studies of flux in P. aeruginosa and Pseudomonas fluorescens (20, 22, 23), we found no evidence for significant metabolite export during exponential growth in minimal media. The errors given for each flux reflect the corresponding 90% confidence intervals. The full flux data sets are presented in Data Set S2. Colors qualitatively indicate fluxomic correlation with changes on the protein/transcript level during growth in acetate (light green or red arrows indicate significant upregulation or downregulation, respectively; dark green or red arrows less-significant upregulation or downregulation, respectively.

FIG 5

Quantitative analysis of NADPH supply and demand (redox) for glycerol-grown (A) and acetate-grown (B) P. aeruginosa. Data represent ATP (energy metabolism) supply and demand for glycerol-grown (A) and acetate-grown (B) P. aeruginosa (see Fig. 6). Values representing reactions linked to NADPH (A and B) and ATP (C and D) metabolism were calculated from the obtained fluxes (see Fig. 4). Values are given as absolute fluxes (in millimoles per gram per hour) and are related to the specific carbon uptake rate (see Data Set S2). G6PDH, glucose 6-phosphate dehydrogenase; MAE, malic enzyme; ICDH, isocitrate dehydrogenase; Ox-P, oxidative phosphorylation; NGAM, non-growth-associated maintenance needs.

FIG 6

Maximal measured NADH/NAD⁺ and NADPH/NADP⁺ ratios in P. aeruginosa grown in the indicated sole carbon sources. Exponentially growing cells in MOPS-glycerol showed a significantly lower NADH/NAD⁺ ratio (P < 0.01) than cells grown in MOPS-acetate, MOPS-glucose, or MOPS-succinate and a significantly lower NADPH/NADP⁺ ratio than cells grown in MOPS-glucose (P = 0.0064). The total NAD(P)(H) concentrations seen at each time point and with each carbon source are shown in Fig. S2 (see also Fig. S4). The data were analyzed using GraphPad Prism (v 6.01) and t test statistical analysis (MOPS-glycerol versus MOPS-acetate, MOPS-glucose, or MOPS-succinate).

FIG 7

Coenzyme reoxidation is impaired in a dnr mutant. The NADH/NAD⁺ (A and B) and NADPH/NADP⁺ (C and D) ratios were measured in cultures of wild-type (WT) PAO1 (A and C) and in cultures of an isogenic Δdnr mutant (B and D). Cultures were grown in MOPS-acetate with or without 20 mM KNO₃, as indicated. The corresponding CFU counts are shown in Fig. S4. The data were analyzed using GraphPad Prism (v 6.01) using t test statistical analysis at 5 h of growth. Nitrate addition to the PAO1 wild-type strain resulted in significant reductions in the NADH/NAD⁺ ratio (**, P = 0.0017) and the NADPH/NADP⁺ ratio (*, P = 0.0202). Nitrate addition to a Δdnr mutant did not significantly alter either ratio (P = 0.8065 and P = 0.1862, respectively).