A molecular comparison of [Fe-S] cluster-based homeostasis in Escherichia coli and Pseudomonas aeruginosa

Abstract

Iron-sulfur [Fe-S] clusters are essential protein cofactors allowing bacteria to perceive environmental redox modification and to adapt to iron limitation. Escherichia coli, which served as a bacterial model, contains two [Fe-S] cluster biogenesis systems, ISC and SUF, which ensure [Fe-S] cluster synthesis under balanced and stress conditions, respectively. However, our recent phylogenomic analyses revealed that most bacteria possess only one [Fe-S] cluster biogenesis system, most often SUF. The opportunist human pathogen Pseudomonas aeruginosa is atypical as it harbors only ISC. Here, we confirmed the essentiality of ISC in P. aeruginosa under both normal and stress conditions. Moreover, P. aeruginosa ISC restored viability, under balanced growth conditions, to an E. coli strain lacking both ISC and SUF. Reciprocally, the E. coli SUF system sustained growth and [Fe-S] cluster-dependent enzyme activities of ISC-deficient P. aeruginosa. Surprisingly, an ISC-deficient P. aeruginosa strain expressing E. coli SUF showed defects in resistance to H₂O₂ stress and paraquat, a superoxide generator. Similarly, the P. aeruginosa ISC system did not confer stress resistance to a SUF-deficient E. coli mutant. A survey of 120 Pseudomonadales genomes confirmed that all but five species have selected ISC over SUF. While highlighting the great versatility of bacterial [Fe-S] cluster biogenesis systems, this study emphasizes that their contribution to cellular homeostasis must be assessed in the context of each species and its own repertoire of stress adaptation functions. As a matter of fact, despite having only one ISC system, P. aeruginosa shows higher fitness in the face of ROS and iron limitation than E. coli.

Importance: ISC and SUF molecular systems build and transfer Fe-S cluster to cellular apo protein clients. The model Escherichia coli has both ISC and SUF and study of the interplay between the two systems established that the ISC system is the house-keeping one and SUF the stress-responding one. Unexpectedly, our recent phylogenomic analysis revealed that in contrast to E. coli (and related enterobacteria such as Salmonella), most bacteria have only one system, and, in most cases, it is SUF. Pseudomonas aeruginosa fits the general rule of having only one system but stands against the rule by having ISC. This study aims at engineering P. aeruginosa harboring E. coli systems and vice versa. Comparison of the recombinants allowed to assess the functional versatility of each system while appreciating their contribution to cellular homeostasis in different species context.

Keywords: Escherichia coli; Pseudomonas aeruginosa; iron-sulfur biogenesis; stress adaptation.

Fig 1

Essentiality of the ISC system in P. aeruginosa. (A) Schematic representation of the P. aeruginosa isc operon in the wild-type strain PAO1 and in the arabinose (ARA)-dependent ΔiscU P_araiscU conditional mutant. Genes are not in scale. (B) Growth curves of PAO1 and ΔiscU P_araiscU at 37°C in MH supplemented or not with increasing concentrations of ARA. (C) Growth of PAO1 and ΔiscU P_araiscU at 37°C in MH upon subsequent refreshes with or without ARA (see Materials and Methods for details). (D) Western blot analysis for IscU, IscS, and the loading control LptC in IscU-replete and -depleted cells obtained as shown in panel C. (E) Colony growth of PAO1 and ΔiscU P_araiscU at 37°C on MH agar plates under aerobic and anaerobic conditions. (F and G) Biofilm formation by PAO1 and ΔiscU P_araiscU in (F) microtiter plates or (G) flow cells. The asterisk indicates a statistically significant difference (P < 0.05) with respect to the wild type. Data are the mean (±SD) or are representative of at least three independent assays.

Fig 2

Effect of IscU depletion on P. aeruginosa antibiotic susceptibility. (A) Western blot analysis for IscU and the loading control LptC in PAO1 and ΔiscU P_araiscU conditional mutant cells cultured with different arabinose (ARA) concentrations. (B) Inhibition halos in the Kirby-Bauer disc diffusion assay of gentamicin (Gm), kanamycin (Km), streptomycin (Sm), tetracycline (Tc), ciprofloxacin (Cip), ceftazidime (Caz), and imipenem (Ipm) for PAO1 and ΔiscU P_araiscU cells cultured in the presence of the indicated ARA concentrations. (C) Survival curves of IscU-replete and -depleted cells, obtained as described in Fig. 1C, exposed to 4× MIC of gentamicin, rifampicin, meropenem or colistin, corresponding to 2 µg/mL for all antibiotics against PAO1. Curves obtained for antibiotic treatments at 1× and 2× MIC are shown in Fig. S1. (D) Survival curves of metabolically inactive PAO1 cells exposed to the indicated antibiotics at 4 × MIC. Cells were resuspended in saline and incubated at 4°C for 14 h (29) and then subjected to antibiotic treatment in saline. Data are the mean (±SD) or are representative of at least three independent assays.

Fig 3

E. coli ISC and SUF systems functionally replace P. aeruginosa ISC. (A) Colony growth of the P. aeruginosa ΔiscU P_araiscU conditional mutant carrying the plasmid pME6032 with the entire isc operon (isc_EC), the entire suf operon (suf_EC), the iscU gene or sufBCD genes from E. coli under an IPTG-inducible promoter on MH agar plates. PAO1 and ΔiscU P_araiscU with the empty plasmid pME6032 were used as positive and negative controls, respectively. (B) Planktonic growth of PAO1 and ΔiscU P_araiscU carrying pMEsuf_EC or the empty plasmid at 37°C in MH. When indicated, arabinose (ARA) and IPTG were added at 0.5% and 0.5 mM, respectively. (C) Enzymatic activity, expressed as percentage relative to PAO1 pME6032, of succinate dehydrogenase (SDH), fumarase A (FumA), and aconitase (Acn) in ΔiscU P_araiscU pMEsuf_EC cultured in MH with 0.5 mM IPTG. IscU-depleted cells carrying the empty plasmid were obtained as described in Fig. 1C and used as the negative control. Data are the mean (±SD) or are representative of at least three independent assays. The asterisks indicate statistically significant differences (P < 0.05) with respect to the wild type carrying the empty plasmid.

Fig 4

Effect of E. coli ISC and SUF systems on P. aeruginosa oxidative stress resistance and anaerobiosis. (A) Colony growth of the P. aeruginosa ΔiscU P_araiscU conditional mutant or the ΔiscU deletion mutant carrying the plasmid pME6032 with the entire isc operon (isc_EC) or the entire suf operon (suf_EC) of E. coli under an IPTG-inducible promoter on MH agar plates supplemented with H₂O₂ or paraquat (PQ) at 0.25 × MIC for the wild type (corresponding to 0.25 mM and 0.125 mM, respectively). (B) Colony growth of the strains described in panel A grown under anaerobiosis conditions. PAO1 and ΔiscU P_araiscU with the empty plasmid pME6032 were used as positive and negative controls, respectively. When indicated, arabinose (ARA) and IPTG were added at 0.5% and 0.5 mM, respectively. Images are representative of three independent assays.

Fig 5

P. aeruginosa ISC system can replace the endogenous ISC system in E. coli. (A) Colony growth of the E. coli wild type (WT) and Δisc mutant, carrying the empty plasmid pME6032 or the one with the entire isc operon from P. aeruginosa (pMEisc_PA), on M9 glucose agar plates. (B) Planktonic growth of E. coli WT and Δisc strains carrying pMEisc_PA or the empty plasmid at 37°C in M9 glucose. (C) β-Galactosidase activity of the E. coli WT (black bars) and ΔiscU (gray bars) strains carrying the chromosomal fusions PhmpA::lacZ (left panel) or PiscR::lacZ (right panel) and the empty plasmid pME6032 or pMEisc_PA cultured in LB. The plasmid containing the E. coli iscU (pMEiscU_EC) was used as control. β-Galactosidase activity is expressed as Miller units. The asterisks indicate statistically significant differences (P < 0.05) with respect to the ΔiscU mutants carrying the empty plasmid. (D) Survival curves of WT and ΔiscUA mutant carrying the empty plasmid pME6032 or pMEisc_PA, exposed to 5 µg/mL gentamycin (Gm) in LB medium. The dotted lines represent the survival of the same strains without Gm addiction. For all the experiments, IPTG was added at the concentration of 100 µM. Data are the mean (±SD) or are representative of at least three independent assays.

Fig 6

P. aeruginosa ISC functionally replaces E. coli ISC and SUF systems. (A) Colony growth of the E. coli wild type (WT) and ΔiscUA Δsuf strains carrying the plasmid pME6032 empty or containing the entire isc operon from P. aeruginosa (pMEisc_PA), on LB agar plates. The single mutants ΔiscUA Δsuf carrying pME6032 were used as controls. (B) Planktonic growth of E. coli WT and ΔiscUA Δsuf strains carrying pMEisc_PA or the empty plasmid at 37°C in LB. For all the experiments, IPTG was added at the concentration of 100 µM. Data are the mean (±SD) or are representative of at least three independent assays.

Fig 7

P. aeruginosa ISC is not able to replace the E. coli SUF under oxidative stress conditions. (A) The E. coli wild type (WT) and Δsuf strains carrying the chromosomal fusion PsoxS::lacZ were transformed with the empty plasmid pME6032 or pMEisc_PA. At time zero, phenazine methosulfate (PMS, 30 µM) was added, and β-galactosidase activity was monitored at the indicated time points and expressed as Miller units. The dotted lines represent β-galactosidase activity of the same strains without PMS addiction. (B) Colony growth on LB agar, supplemented or not with PMS (50 µM) or dipyridyl (DIP, 300 µM), of the E. coli Δsuf mutant strain carrying the empty plasmid pME6032 or pMEisc_PA. WT or Δsuf transformed with the plasmid carrying the suf operon of E. coli (pMEsuf_EC) were used as controls. For all the experiments, IPTG was added at the concentration of 100 µM. Data are the mean (±SD) or are representative of at least three independent assays.

Fig 8

Comparison of E. coli and P. aeruginosa abilities to sustain stress. Planktonic growth of the E. coli MG1655 (left graphs) and P. aeruginosa PAO1 (right graphs) strains in LB supplemented or not with the concentrations indicated in the graphs (expressed in μM) of (A) PQ, (B) PMS, and (C) DIP. Data are the mean (±SD) of at least three independent assays.

Fig 9

Taxonomic distribution of ISC, SUF, and NIF components in Pseudomonadales. The reference tree has been inferred using a concatenation of If2, RpoB, and RpoC (IQ-TREE, LG + R10, 3,573 amino-acid positions, 201 sequences). The group of Oceanospirillales has been collapsed and used as an outgroup (represented as a gray triangle). The dots on branches indicate an ultrafast bootstrap value ≥0.95. The scale bar indicates the average number of substitutions per site. The numbers after species name correspond to the NCBI taxonomic ID. The presence of proteins in proteomes is indicated by dark-colored squares.