To identify the molecular networks in Pseudomonas fluorescens that convey resistance to toxic concentrations of Zn, a common pollutant and hazard to biological systems.
To identify the molecular networks in Pseudomonas fluorescens that convey resistance to toxic concentrations of Zn, a common pollutant and hazard to biological systems.
Pseudomonas fluorescens strain ATCC 13525 was cultured in growth medium with millimolar concentrations of Zn. Enzymatic activities and metabolite levels were monitored with the aid of in-gel activity assays and high-performance liquid chromatography, respectively. As oxidative phosphorylation was rendered ineffective, the assimilation of citric acid mediated sequentially by citrate lyase (CL), phosphoenolpyruvate carboxylase (PEPC) and pyruvate phosphate dikinase (PPDK) appeared to play a key role in ATP synthesis via substrate-level phosphorylation (SLP). Enzymes generating the antioxidant, reduced nicotinamide adenine dinucleotide phosphate (NADPH) were enhanced, while metabolic modules mediating the formation of the pro-oxidant, reduced nicotinamide adenine dinucleotide (NADH) were downregulated.
Pseudomonas fluorescens reengineers its metabolic networks to generate ATP via SLP, a stratagem that allows the microbe to compensate for an ineffective electron transport chain provoked by excess Zn.
The molecular insights described here are critical in devising strategies to bioremediate Zn-polluted environments.
Zn is an essential micronutrient in most if not all organisms as it is involved in a variety of biochemical processes. It plays a pivotal catalytic, structural and regulatory role (Stefanidou et al. 2006). This metal is an important cofactor of numerous enzymes that are critical in metabolism, transcription and cell signalling (Haase and Rink 2009; Sera 2009; Bong et al. 2010). However, when present in elevated amounts, Zn interferes with numerous biological pathways. This metal can interact strongly with enzymatic imidazole and sulfhydryl groups and arrest their reactions (Duruibe et al. 2007). Furthermore, it disturbs the redox potential of the cell and helps to create an oxidative environment. Indeed, oxidative stress is an important feature of Zn toxicity (Lemire et al. 2008).
Hence, to survive a Zn-polluted environment, a situation common due to industrialization and anthropogenic activity, organisms have evolved a plethora of intricate strategies to combat this divalent metal (Bondarenko et al. 2008). Its intracellular immobilization in phytochelatins, metalothioneins and other cysteine-rich moieties has been observed (Blindauer et al. 2002; Di Baccio et al. 2005). Microbes living in Zn-polluted environments are known to precipitate the metal as sulfide- or phosphate-containing moieties and actively efflux Zn (Radhika et al. 2006). The synthesis of antioxidants, such as glutathione and NADPH, is also part of the arsenal at the disposal of bacteria to counter Zn (Lemire et al. 2010a; Poirier et al. 2013). Although some of the molecular pathways that mediate Zn detoxification have been reported, the involvement of metabolic networks in the adaptation to elevated levels of Zn has yet to be fully uncovered.
As part of our effort to decipher the molecular mechanisms Pseudomonas fluorescens utilizes to adapt to metal stress and to develop metal bioremediation technologies, we have evaluated the metabolic networks evoked in response to toxic levels of Zn. Although the ability of Ps. fluorescens to degrade Zn/citrate complexes has been reported (Joshi-Tope and Francis 1995), the biochemical pathways that enable this organism to survive Zn toxicity have yet to be fully understood. Here, we report on an alternative ATP-generating machinery in Ps. fluorescens subjected to Zn-induced oxidative stress. This metabolic module obviates the need to utilize the NADH- generating tricarboxylic acid (TCA) cycle. The enhanced production of NADPH via malic enzyme (ME), glucose-6-phosphate dehydrogenase (G6PDH) and isocitrate dehydrogenase (ICDH-NADP+) is also discussed. This study provides a global view of metabolic reconfiguration involved in adaptation to Zn toxicity and its implication in Zn waste management is also discussed.
Pseudomonas fluorescens strain ATCC 13525 was obtained from the American Type Culture Collection. It was maintained and grown in a mineral medium as described in the study of (Singh et al. 2008). Citrate (19 mmol l−1) was utilized as a carbon source, and Zn was added in varying concentrations (0·1, 1, 5 and 10 mmol l−1). This tricarboxylic acid is known to play a key role in the mobilization of Zn (Francis et al. 1992; Gadd 2004). The pH was adjusted to 6·8 with 2N NaOH. The medium was inoculated with 1 ml of stationary-phase cells grown in a Zn-free medium in an aerated gyratory water bath shaker, model 76 (New Brunswick Scientific, Enfield, CT) at 26°C and 140 rpm. The bacterial cells were harvested at similar growth phases (28 h for control and 38 h for Zn stressed) (Appanna and Whitmore 1995). Following centrifugation at 10 000 g for 15 min at 4°C, cells were washed with 0·85% NaCl and resuspended in a cell storage buffer (CSB) consisting of 50 mmol l−1 Tris-HCl, 5 mmol l−1 MgCl2, 1 mmol l−1 phenylmethylsulfonylfluoride (PMSF) and 1 mmol l−1 dithiothreitol (DTT), pH 7·3. The soluble cell-free extract (CFE) and membrane CFE were obtained by sonication followed by centrifugation at 180 000 g for 3 h at 4°C (Singh et al. 2008). Purity was assessed by the activity of glucose-6-phosphate dehydrogenase for soluble CFE and succinate dehydrogenase for membrane CFE. The protein content was quantified by the Bradford assay. These CFE fractions were kept at 4°C for up to 5 days, and various enzymatic activities were monitored (Chenier et al. 2008).
To evaluate oxidized lipid content in the membrane, the thiobarbituric acid reactive species assay (TBARS) was performed as described (Kayashima and Katayama 2002). The fraction containing the membrane portion was solubilized in a mixture containing 15% trichloroacetic acid, 0·37% trichlorobarbituric acid and 0·25 N HCl. Following heating at 100°C for 20 min and precipitation of the pellet, the supernatant was monitored at 532 nm. Reaction mixtures lacking trichlorobarbituric acid were used as negative controls. Protein carbonyl content was determined by performing a dinitrophenyl hydrazine (DNPH) assay as described in (Frank et al. 2000). One milligram of soluble protein was homogenized with 1 ml of 2% (w/v) DNPH and allowed to react for 1 h. The protein was subsequently precipitated, and the pellet was washed thrice with ethylacetate/ethanol (1 : 1). One millilitre of 6 N guanidine-HCl was added to the mixture and read spectrophotometrically at 360 nm. Reaction mixtures lacking DNPH were used as negative controls. The Fe/S cluster in the control and Zn-stressed soluble CFE (3 mg protein equivalent) was monitored spectrophotometrically (Middaugh et al. 2005). Total sulfhydryl was measured by adding 20 μl of sample to 75 μl of Tris-HCl (pH 8·2), 25 μl 5, 5′-dithio-bis 2-nitrobenzoic acid (DTNB) and 400 μl methanol. Samples were then spun down at 3000 g for 5 min at room temperature. The absorbance was measured at 412 nm (Hoenicka 1968; Hansen and Winther 2009). Reduced glutathione was utilized to obtain the standard curve.
The relative levels of various metabolites from the spent fluid and soluble CFE in control and Zn-stressed cultures were discerned by high-performance liquid chromatography (HPLC). The supernatant was filtered and injected into an Alliance HPLC equipped with C18 reverse-phase column at a flow rate of 0·7 ml min−1. The samples were diluted 10-fold using Milli-Q water and run in mobile phase containing 20 mmol l−1 KH2PO4, pH 2·9 prepared in Milli-Q water. The samples were loaded into the HPLC (Waters 2695 separation module) and Empower software (Milford, MA) was used for the automatic injection protocol. The organic metabolites and nucleotides were detected by using a Waters model 2487 UV-Vis dual wavelength detector operating at 210 and 254 nm, respectively. Peaks were identified by comparing to known standards and by spiking the samples with standard solutions (Lemire et al. 2010a,b,c). To assess how ATP was being generated, 2 mg ml−1 protein equivalent of soluble CFE was incubated for 1 h at 26°C in reaction buffer (25 mmol l−1 Tris, 5 mmol l−1 MgCl2, pH 7·0) containing oxaloacetate (2 mmol l−1), AMP (0·5 mmol l−1) and Pi (0·5 mmol l−1).
Blue native polyacrylamide gel electrophoresis (BN PAGE) was performed as described previously (Singh et al. 2005a; Wittig et al. 2006; Kikuchi et al. 2011) to evaluate various enzymes involved in the metabolism of citrate. To ensure optimal protein separation, 4–16% linear gradient gels were cast with the Bio-Rad MiniProteanTM 2 (Hercules, CA) system using 1 mm spacers. Sixty micrograms of soluble or membrane protein was loaded into the wells and the gels were electrophoresed under native conditions (50 mmol l−1 Bis-Tris, 500 mmol l−1 ε-aminocaproic acid, pH 7·0, 4°C) at a final concentration of 4 mg of protein per ml. Membrane proteins were prepared in a similar manner except 1% (v/v) β-dodecyl-d-maltoside was added to the preparation to facilitate the solubilization of the membrane-bound proteins. The blue cathode buffer [50 mmol l−1 Tricine, 15 mmol l−1 Bis-Tris, 0·02% (w/v) Coomassie G-250 (pH 7)] at 4°C was changed to a colourless cathode buffer [50 mmol l−1 Tricine, 15 mmol l−1 Bis-Tris, (pH 7)] at 4°C when the running front was half-way through the resolving gel. The in-gel visualization of enzyme activity was checked with the aid of formazan precipitation. The gels were incubated in the reaction mixture containing equilibration buffer, 5 mmol l−1 substrate, 0·5 mmol l−1 cofactor with 0·2 mg ml−1 of phenazine methosulfate (PMS) or DCPIP and 0·4 mg ml−1 of iodonitrotetrazolium (INT). Isocitrate dehydrogenase (ICDH), NAD+/NADP+, malate dehydrogenase (MDH), fumarase (FUM), isocitrate lyase (ICL), glucose-6-phosphate dehydrogenase (G6PDH) and malic enzyme (ME) activities were detected as described (Singh et al. 2005a,b, 2008; Chenier et al. 2008; Auger et al. 2011). ICDH-NAD activity was confirmed via spectrophotometric studies, by monitoring the appearance of NADH at 340 nm. While Complex I was measured with 5 mmol l−1 KCN, 5 mmol l−1 NADH and INT, cytochrome C oxidase (Complex IV) activity was deduced with the utilization of equilibration buffer supplemented with 10 mg ml−1 of diaminobenzidine, 10 mg ml−1 cytochrome C and 562·5 mg ml−1 of sucrose (Eubel et al. 2004; Auger et al. 2011). The activity of catalase was measured using p-anisidine, and the absorbance was monitored at 458 nm (Igamberdiev et al. 1995). One unit of catalase is defined as the amount that decomposes 1 μmol of H2O2 in 1 min in a solution containing one mg of protein at pH 7·3 at 26°C. The activities of citrate lyase (CL) and phosphoenolpyruvate carboxykinase (PEPCK) were visualized by utilizing enzyme-coupled assays as described (Auger et al. 2011). CL activity was measured spectrophotometrically by coupling the activity of this enzyme to exogenous MDH and monitoring the oxidation of NADH at 340 nm (Linn and Srere 1979). PEPC was assayed in a similar manner as PEPCK except no ADP was present in the mixture. Pyruvate kinase (PK) and PPDK were probed as described (Chastain et al. 2011; Auger et al. 2012). PPDK activity was further confirmed spectrophotometrically by coupling the activity of this enzyme to exogenous LDH and monitoring the oxidation of NADH at 340 nm (Sandoval et al. 2011). Reactions were stopped using destaining solution (40% methanol, 10% glacial acetic acid) once the activity bands reached their desired intensity. To ensure equal loading, control CFE and stressed CFE were electrophoresed and stained with Coomassie blue. The activity of pyruvate carboxylase (PC), an enzyme that does not appear to change significantly, was also probed as described (Singh et al. 2005a,b), as a control. To ensure proper loading, a gel was stained with Coomassie blue during each electrophoretic experiment.
In an effort to verify whether indeed Zn was causing the observed metabolic shift, Zn-stressed cells were incubated in control media, and control cells were subjected to Zn media to reverse the observed changes. Ammonium chloride (NH4Cl) was omitted from media for this regulation experiment to ensure that bacteria remain in their initial growth phase. Following an 8-h incubation, cells were harvested and the cellular fractions were isolated and assayed for enzymatic activities as described before.
All experiments were performed in triplicate and at least in biological duplicate. Data were expressed as mean ± standard deviation (SD). Statistical correlations of data were checked for significance using the student t test.
As 5 mmol l−1 Zn has been shown to result in a significant impact on the growth of Ps. fluorescens, this concentration was routinely utilized to probe the metabolic networks conferring the microbe resistance to the Zn challenge. The microbe could not tolerate concentrations of Zn higher than 5 mmol l−1 (Data not shown). Citrate is completely consumed and direct interaction between the microbe and the divalent metal has been observed (Appanna and Whitmore 1995). To confirm whether indeed Zn toxicity was leading to oxidative stress, oxidized lipids and proteins were analysed. A 2-fold increase in oxidized proteins in the Zn-challenged cells as compared to the control was observed (Table 1). There was no significant change in oxidized lipids, an observation that may be due to the ability of Zn to bind the membrane components of bacteria, rendering it resistant to lipid oxidation (Zago et al. 2000). The ability of menadione and such metals like Al and Ga to promote the formation of oxidized lipids and proteins has previously been shown (Bériault et al. 2007; Singh et al. 2007; Chenier et al. 2008). Total sulfhydryl was also found to be markedly diminished in the Zn-challenged cells (Table 1). The pyruvate carboxylase activity band was also used as a marker for equal protein loading as this enzyme did not change significantly in these cultures (Fig. 1d). A marked increase in the NADPH-generating enzymes was observed. There was a sharp rise in the activities of ICDH-NADP+, ME and G6PDH. The latter two enzymes were barely evident in the control cultures (Fig. 1a). The iron/sulfur (Fe/S) clusters in proteins are known to be a target of metal toxicity. Zinc may perturb these clusters and displace Fe, an event associated with oxidative stress (Lemire et al. 2008). The Fe/S cluster was indeed shown to be perturbed. To deal with this oxidative environment, the activity of catalase was approximately 2-fold higher in the stressed cells (Table 1).
|Assay||Oxidized lipidsa||Oxidized proteinsb||Total sulfhydrylc||Catalase activityd|
|Control||50 ± 1·3||11 ± 3·0||50 ± 14||8·93 ± 0·66|
|Zn stress||30 ± 1·5||20.5 ± 6·3e||4 ± 1·4e||18·04 ± 0·73e|
As Fe/S clusters were disrupted in the Zn-stressed cells, there was a significant inhibition of the TCA cycle and oxidative phosphorylation (Lemire et al. 2008). The activities of TCA cycle enzymes, such as isocitrate dehydrogenase (ICDH-NAD+), α-ketoglutarate dehydrogenase (α-KGDH) and malate dehydrogenase (MDH), were markedly diminished (Fig. 1b, Table 2). These enzymes generate NADH, a potential source of ROS and their reduction would indeed limit the further exacerbation of oxidative stress induced by Zn (Dineley et al. 2003). Complex I and Complex IV, two important components of the electron transport chain (ETC.) were also severely affected in the bacteria isolated from the Zn cultures (Fig. 1c). Hence, the ATP-generating machinery propelled by oxidative phosphorylation was impeded by Zn toxicity (Fig. 1e). To survive, Ps. fluorescens has to resort to alternative ATP-producing pathways. Indeed, the presence of pyruvate and phosphoenolpyruvate (PEP) will argue for such a possibility. These peaks were more prominent in the CFE of the stressed cells (Fig. 2).
|Control||0·011 ± 0·00431||0·074 ± 0·0093||0·059 ± 0·0038|
|Zn stress||0·0055 ± 0·0084b||0·017 ± 0·0061b||0·019 ± 0·0017b|
As key enzymes of the TCA cycle were ineffective, it was important to decipher how Ps. fluorescens was metabolizing citrate, the sole carbon source under the challenge of Zn. As pyruvate was an important constituent of the soluble CFE, a metabolite that may be liberated from the tricarboxylic acid, the presence of the enzyme CL was assessed. Indeed, while a sharp CL activity was evident in the Zn-stressed cells, no such band was evident in the control cells (Fig. 3b, Table 2). This enzyme apparently did not require ATP and was found to increase with the concentration of Zn in the culture medium. When the soluble CFE was incubated in the presence of oxaloacetate, AMP and Pi, the formation of ATP was readily detected in soluble CFE and was markedly higher in the Zn-stressed cells than in the control cells (Fig. 3a). Pyruvate and PEP were also present in increased concentrations under these conditions (Fig. 3a). Hence, it became obvious that substrate-level phosphorylation may be fulfilling the ATP need of these Zn-stressed bacteria. PEPC, an enzyme that mediates the synthesis of PEP from oxaloacetate with the participation of Pi was upregulated in activity under the influence of Zn toxicity (Fig. 3b). PEPCK, another enzyme that converts oxaloacetate into PEP in the presence of ATP was also upregulated in activity (Fig. 3b). These enzymes that interact with PEP with the concomitant formation of ATP were found to have enhanced activities in the CFE from the Zn-stressed cells. There appeared to be no significant change in PC, an enzyme that mediates the transformation of pyruvate to oxaloacetate. This enzyme was also routinely utilized as a loading control.
PK, which transforms PEP into pyruvate and ATP with the assistance of ADP, showed a marked increase in activity as seen in gel (Fig. 3b). PPDK, an enzyme responsible for the conversion of PEP into pyruvate and ATP utilizes AMP and PPi as co-substrates. This enzyme that was very prominent in the Zn cultures was barely evident in the control bacteria (Table 2). Also, the effects of various concentration of Zn on PEPC and other enzymes were investigated. Soluble CFEs obtained at the same growth phase from cells grown in different concentration of Zn were probed. There was an increase in the activity of PEPC with increasing concentrations of the divalent metal in the culture medium. However, at 10 mmol l−1 Zn, a decrease was observed (Fig. 3c). Furthermore, soluble CFE from Zn-stressed cultures obtained at different time intervals revealed a concomitant increase in PPDK activity with the time of growth (Fig. 3d). The enhanced activities of PPDK and other enzymes were dependent on Zn as the incubation of Zn-stressed cells in a control medium had a reverse effect (Fig. 3d).
The aforementioned data point to a marked reprogramming of the metabolic networks involved in the survival of Ps. fluorescens when exposed to Zn. To mitigate the oxidative environment promoted by Zn due to its interaction with sulfhydryl groups, catalase activity and NADPH-producing enzymes are enhanced. These enzymes are known to play a critical role in the homoeostasis of the cellular redox state (Cabiscol et al. 2000). The perturbation of Fe/S clusters has a major impact on the O2-dependent ATP-producing machinery of organisms. Aconitase is a key target of Zn as are enzymes like α-KGDH and FUM (Chenier et al. 2008). The interference with aconitase activity by aluminium has been known to elicit the activation of two downstream enzymes, namely ICDH-NADP and isocitate lyase (ICL). This molecular arrangement enables Ps. fluorescens to metabolize citrate, the sole carbon source (Hamel et al. 2004; Middaugh et al. 2005). Although the role of aconitase in the degradation of citrate in the Zn-stressed Ps. fluorescens cannot be ruled out, in this instance, it appears that the microbe invokes the participation of CL to degrade the tricarboxylic acid. In fact, the activity of this enzyme increased with the presence of Zn in the growth medium (Joshi-Tope and Francis 1995). This enzyme mediates the production of oxaloacetate and acetate without the requirement for ATP. Indeed numerous microbes that proliferate in anaerobic or O2-limiting environments have been shown to utilize this enzyme (Korithoski et al. 2005). This pathway may provide an evolutionarily benefit as it limits the utilization of ATP especially in situations where production of this triphosphate in an O2-dependent manner is severely compromised. Indeed, the inhibition of Complex I and IV will have a drastic impact on the ATP-budget in bacteria challenged by Zn (Dineley et al. 2003; Kuznetsova et al. 2005).
Hence, it is evident that the organism has invoked an alternative mechanism to generate ATP and limit the further imposition of oxidative stress. NADH is a pro-oxidant as its oxidation via the electron transport chain liberates ROS. The TCA cycle, the main provider of reducing factors for oxidative energy synthesis was markedly diminished as numerous key enzymes in this metabolic network were ineffective. It is not uncommon for an organism to adopt this strategy with a concomitant increase in NADPH formation (Fuhrer and Sauer 2009; Sandoval et al. 2011). Indeed, three NADPH-generating enzymes in the Zn-stressed microbe had increased activity. The modulation of NADH and NADPH is a common stratagem afforded to numerous organisms to maintain the redox balance when affected by an oxidative challenge (Lemire et al. 2010a,b,c).
If any organism is to proliferate in a Zn-containing environment, the balancing of the redox potential has to be coupled with an alternative machinery to fulfil its energy need (Lemire et al. 2008). In this instance, the enhanced production of PEP appears to be an important contributor to the ATP budget. This is achieved by the increased activity of PEPC and PEPCK in Zn-stressed cells. While the latter enzyme may help provide PEP with the involvement of a high energy triphosphate, the former is known to produce PEP with the involvement of Pi only. Hence, it is more energy efficient than PEPCK. Indeed, PEPC was barely discernable in the control cultures in this study. The fixation of PEP into ATP was attained by two mediators, namely PPDK and PK. The latter utilizes ADP, while the former invokes the participation of AMP to generate ATP (Auger et al. 2012). This ATP-producing machinery devoid of O2 consumption appears to be quite effective as PEP was generated and PPDK allowed for the synthesis of ATP from AMP. Such substrate-level phosphorylation aimed at ATP homoeostasis has also been observed in infectious organisms and in microbes surviving in extreme environments (Husain et al. 2012). Figure 4 depicts a scheme on how this microbe may be adapting to Zn toxicity.
This study provides critical molecular insights into a pivotal feature on the adaptation to Zn toxicity that has hitherto not been explored. An understanding of the metabolic networks involved in the homoeostasis of ATP in Zn-polluted environments will allow the tailoring of Ps. fluorescens aimed at metal waste management technology. Unraveling of these metabolic pathways lays the foundation for the bioengineering of microbes designed to decontaminate metal-polluted environments.
This study was funded by Laurentian University. Azhar Alhasawi is a recipient of a graduate scholarship from the Ministry of higher education of Saudi Arabia; Christopher Auger is a recipient of the NSERC post graduate scholarship at the doctoral level.
No conflict of interest declared.