SEARCH

SEARCH BY CITATION

Keywords:

  • [Fe–S] cluster assembly;
  • Firmicutes;
  • Proteobacteria;
  • genetic complementation;
  • SUF operon

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Iron–sulfur [Fe–S] clusters are inorganic prosthetic groups that play essential roles in all living organisms. Iron and sulfur mobilization, formation of [Fe–S] clusters, and delivery to its final protein targets involves a complex set of specific protein machinery. Proteobacteria has three systems of [Fe–S] biogenesis, designated NIF, ISC, and SUF. In contrast, the Firmicutes system is not well characterized and has only one system, formed mostly by SUF homologs. The Firmicutes phylum corresponds to a group of pathological bacteria, of which Enterococcus faecalis is a clinically relevant representative. Recently, the E. faecalis sufCDSUB [Fe–S] cluster biosynthetic machinery has been identified, although there is no further information available about the similarities and/or variations of Proteobacteria and Firmicutes systems. The aim of the present work was to compare the ability of the different Proteobacteria and Firmicutes systems to complement the Azotobacter vinelandii and Escherichia coli ISC and SUF systems. Indeed, E. faecalis sufCDSUB is able to complement the E. coli SUF system, allowing viable mutants of both sufABCDSE and iscRSU-hscBA-fdx systems. The presence of all E. faecalis SUF factors enables proper functional interactions, which would not otherwise occur in proteins from different systems.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Iron–sulfur [Fe–S] clusters are inorganic prosthetic groups, widely distributed in nature, that play essential roles in diverse biological processes such as electron transfer, redox and nonredox catalysis, and gene regulation, and as sensors within all living organisms (Frazzon & Dean, 2003; Johnson et al., 2005). The biosynthetic process of iron and sulfur mobilization and formation of [Fe–S] clusters, and delivery of these clusters to their final destination involves the recruitment of iron (ferrous or ferric forms) from their storage sources, cysteine desulfurase-catalyzed release of sulfide ions, their association, and transport and transfer of the [Fe–S] clusters to the final molecular destinations, mainly within polypeptide chains. [Fe–S] clusters have the characteristic of being chemically assembled by the reductive coupling of [2Fe–2S] units, despite their structural diversity, reactivity, electronic properties, and molecular environments (Kiley & Beinert, 2003). Although assembly of [Fe–S] clusters is an efficient process in vitro (Malkin & Rabinowitz, 1966), several kinds of specific protein machinery are required for this process in vivo due to the cellular toxicity of Fe2+/3+ and S2−. Bacterial microorganisms, and most specifically the Proteobacteria phylum, are the most studied organisms inside the [Fe–S] cluster biosynthesis machinery field. There are three kinds of [Fe–S] biogenesis machinery described in bacteria, designated NIF, ISC, and SUF. The NIF system, first described in Azotobacter vinelandii, is formed by structural and regulatory genes involved in the specific task of performing specialized functions in nitrogen fixation and subsequent maturation of the nitrogenase (Jacobson et al., 1989a, b; Rubio & Ludden, 2008). The ISC system, encoded by the iscRSUA-hscBA-fdx gene cluster, is the housekeeping system for the [Fe–S] protein maturation (Zheng et al., 1998) and is highly conserved in Proteobacteria. ISC is probably the most substantial machinery in living organisms, as it can be found in a wide variety of cells, including numerous bacteria, archaea, and plants (Takahashi & Tokumoto, 2002). The SUF system, first described in Escherichia coli, comprises proteins encoded by the sufABCDSE operon, and is expressed under stress growth conditions such as oxidative stress, NO stress, and iron starvation (Fontecave et al., 2005).

Firmicutes are predicted to contain only one kind of biosynthetic machinery for [Fe–S] cluster assembly. This is formed mostly by E. coli SUF homologs (sufC, sufD, sufS, sufB) and is completed by the presence of sufU, an iscU E. coli homolog (Fig. 1), although Enterococcus faecalis lacks the A-type of scaffold (ATC) sufA and the desulfurase activator sufE (Riboldi et al., 2009). Recently, SufU emerged as a candidate for desulfurase activator in Bacillus subtilis (Selbach et al., 2010; Albrecht et al., 2011). The Firmicutes phyla are a group of bacteria that participate extensively in virulence episodes and pathological processes in the host organism. Enterococcus spp. comprises commensal microorganisms that colonize the gastrointestinal and vaginal tract and, occasionally, the oral cavity in humans. Enterococcus faecalis is a clinically relevant bacterium, responsible for 80–90% of clinical isolates in nosocomial infections (Tendolkar et al., 2003). Pathological processes of these microorganisms include infections of the urinary tract, wounds, bloodstream, and endocardium (Kauffman, 2003). The pathogenic phenotype is mainly due to virulence factors such as cytolysin, aggregation substance, proteases, hyaluronidase, and bacteriocins, which enable the microorganism to adhere to host tissues, facilitating tissue invasion and causing immunomodulation and toxin-mediated damage. A second clinically important characteristic of the Enterococcus spp. is resistance to a wide range of antimicrobial agents (Shepard & Gilmore, 2002).

image

Figure 1.  Genetic organizations of genes involved in [Fe–S] cluster assembly. Azotobacter vinelandii has the NIF system involved in nitrogenase maturation and biological nitrogen fixation, and the housekeeping genes of the ISC system, involved in maturation of general [Fe–S] proteins, for example aconitase. The same ISC system was also verified for housekeeping maturation in Escherichia coli, whose genome also has the SUF system, which is involved in oxidative stress and iron-depletion conditions. The Enterococcus faecalis machinery for [Fe–S] cluster formation seems to combine several E. coli SUF homologs with the ISC homolog sufU. Regions with homologous sequences or similar functions between the systems are color coded, and different colors within nifU indicate different domains within this protein.

Download figure to PowerPoint

Considering the high conservation of the SUF system among the Firmicutes, and as E. faecalis only has the SUF system, it is intriguing to determine the similarities and divergences of this system with ISC and SUF systems previously described in Proteobacteria. Thus, the present work aims to elucidate in vivo the capacity of the E. faecalis SUF operon to complement the ISC and SUF systems from the Proteobacteria representatives A. vinelandii and E. coli.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Strains, media, and plasmid construction

The Azotobacter vinelandii and Escherichia coli strains used in this study are listed in Table 1, and plasmids used for in vivo experiments in Table 2. Escherichia coli were grown in the following media: Luria broth (10.0 g L−1 tryptone, 5.0 g L−1 yeast extract, 5.0 g L−1 NaCl), and M9-glycerol minimal medium, supplemented as needed with 5.0 mM adenine, 0.3 mM leucine, 0.3 mM isoleucine, 0.1 mM nicotinic acid, 0.3 mM thiamine, and 0.3 mM valine. Azotobacter vinelandii was grown in Burk's minimal medium (BN) containing 2.0% sucrose as the carbon source and 13.0 mM ammonium acetate as nitrogen source (Strandberg & Wilson, 1968). The following antibiotics were used in this study: ampicillin (100 μg mL−1), rifampicin (100 μg mL−1), kanamycin (50 μg mL−1), gentamicin (50 μg mL−1), tetracycline (50 μg mL−1), and vancomycin (30 μg mL−1). Arabinose was used at 0.3% w/v for expression in E. coli and A. vinelandii under arabinose promoter (pBAD). X-gal at a final concentration of 0.6 mg mL−1 was used for cloning insertion determination.

Table 1.   Strains used in this study
StrainGenotype or descriptionAntibiotic resistanceDerivation or references
  • *

    pDB303: congression plasmid with 1.7-kb EcoRI fragment from λ24-9 subcloned into pUC8 (RifR determinant rpoB113).

  • VAN, vancomycin; RIF, rifampicin; TET, tetracycline; KAN, kanamycin.

Enterococcus faecalis
 FA22Laboratory strainJacob et al. (1975)
 JH22Laboratory strainYagi & Clewell (1980)
 X1Clinical outbreak strainVANSwenson et al. (1995)
Azotobacter vinelandii
 DJ1418ΔscrXlacZ:KanRKANJohnson et al. (2006)
 AES1ΔscrX∷(araC sufC)RIFDJ1418 × pEFSE3 × pDB303*
 AES2ΔscrX∷(araC sufD)RIFDJ1418 × pEFSE13 × pDB303*
 AES3ΔscrX∷(araC sufS)RIFDJ1418 × pEFSE24 × pDB303*
 AES4ΔscrX∷(araC sufU)RIFDJ1418 × pEFSE33 × pDB303*
 AES5ΔscrX∷(araC sufB)RIFDJ1418 × pEFSE53 × pDB303*
 AES6ΔscrX∷(araC sufSU)RIFDJ1418 × pEFSE73 × pDB303*
 AES7ΔscrX∷(araC sufCDSUB)RIFDJ1418 × pEFSE121 × pDB303*
Escherichia coli
 TB1FaraΔ(lac-proAB) [80dlac D(lacZ)M15] hsdR rpsLNew England Biolabs
 CAG18470purCTn10TETNichols et al. (1998)
 TL254MG1655 Δ(lacZYA-argF) U169Donahue et al. (2000)
 PJ23TL524 iscSΔpurC80Tn10TETP1(PJ19)[RIGHTWARDS ARROW]TL523
 MC1061FaraD139Δ(ara, leu)7696Δ(lacY74)Lauhon & Kambampati (2000)
 CL100MC1061 ΔiscSLauhon & Kambampati (2000)
 PJ19CL100 purC80Tn10TETP1(CAG18470)[RIGHTWARDS ARROW]CL100
 JW1670-1Δ(araD-araB)567ΔlacZ4787(∷rrnB-3), λ-, ΔsufS755kanKANBaba et al. (2006)
 EESC41JW1670-1 ΔsufSFRTFLP removal of KanR
 GSO97MG1655 ΔsufSEFRTOutten et al. (2004)
 GSO92MG1655 ΔsufABCDSEFRTOutten et al. (2004)
 CAG18481λ-, zfh-208Tn10, rph-1TETNichols et al. (1998)
 JW2514-4Δ(araD-araB)567ΔlacZ4787(∷rrnB-3),λ-, ΔiscS776kanKANBaba et al. (2006)
 EESC42JW2514-4 ΔiscS776kan zfh-208Tn10KAN, TETP1(CAG18481)[RIGHTWARDS ARROW]JW2514-4
 EESC54GSO92 ΔiscS776kan zfh-208Tn10 (pEFSE121)KAN, TETGSO92 × EESC42+pEFSE121
Table 2.   Plasmids used in this study
PlasmidGenotypeMarkerReferences
  1. AMP, ampicillin; GEN, gentamicin.

pCP20flpTemperature sensitiveCherepanov & Wackernagel (1995)
pDB943PT7iscSAMPZheng et al. (1993)
pDB551PT7nifSAMPZheng et al. (1998)
pDB1568pBADAMPDos Santos et al. (2007)
pDB303rpoB113RIFBrigle et al. (1987)
pEFSE3pBADsufCAMPThis work
pEFSE13pBADsufDAMPThis work
pEFSE24pBADsufSAMPThis work
pEFSE33pBADsufUAMPThis work
pEFSE53pBADsufBAMPThis work
pEFSE73pBADsufSUAMPThis work
pEFSE121pBADsufCDSUBAMPThis work
pDB1005hscAkanRKANZheng et al. (1998)
pDB933KiscSkanRKANThis work
pDB954KiscSUkanRKANThis work
pDB1018iscUkanRKANJohnson et al. (2006)
pDB1370iscSUA-hscBA-fdxgenRGENJohnson et al. (2006)
pCRBluntCloning vectorKANInvitrogen
pTOPO4Cloning vectorKANInvitrogen
pEFSC1sufCKANThis work
pEFSC11sufDKANThis work
pEFSC22sufSKANThis work
pEFSC31sufUKANThis work
pEFSC51sufBKANThis work
pEFSC72sufSUKANThis work

Genetic manipulations

Recipient strains used in this work have been described previously (Table 1) and confirmed in terms of promoter region arrangements; modifications (either insertions or mutations) carried out in this work did not alter any characteristic of expression which could result in polar effects.

Azotobacter vinelandii strains were constructed by transformation experiments in which homologous reciprocal recombination occurred between cloned, isolated A. vinelandii DNA in a recombinant plasmid and a corresponding genome region. As an example, the vector pEFSC31 was constructed first using PCR (Epicentre's Failsafe PCR kit) to isolate the sufU gene from the chromosomal DNA of E. faecalis. The PCR primers were designed to add an NdeI restriction enzyme site at the N-terminus of sufU and a BglII restriction enzyme site at the C-terminus. The 0.7-kb PCR product was ligated into the pCR4-TOPO vector (Invitrogen TOPO TA sequencing kit) or pCR-Blunt vector (Invitrogen). This plasmid was digested with NdeI and BglII and the resulting DNA fragment was ligated into the NdeI–BglII sites of pDB1568, putting expression of the SufU protein under control of the pBAD in a region of DNA containing the scrX gene. Other plasmids used in this study (Table 2) were constructed in a similar fashion.

Incorporation of the SUF genes into the A. vinelandii genome was achieved as described by Jacobson et al. (1989a, b). DJ1418, used as the parent strain, contains the complete endogenous ISC operon and a lacZ:kanamycin resistance cartridge inserted into scrX. This insertion causes the strain to turn blue when plated on BN media containing X-gal, and thus allows for a blue-white screening of the new strains. AES4, which contains ΔscrX∷(araC sufU), was constructed by transforming DJ1418 by congression (coincidental transfer of genetic markers) with pEFSC31 and pDB303 (containing the rifampicin resistance marker). The double reciprocal recombination event was selected by screening for white colonies on BN plates containing both rifampicin and X-gal. In this way, the ISC operon was intact in both DJ1418 and the only recombinant changes were downstream in the sucrose scrX region. When this strain is grown on BN plus arabinose media, the SufU protein should be expressed. Other strains used in this study (AES1–7) were constructed in a similar fashion.

To explore the ability of the E. faecalis SUF genes to complement the activity of the ISC genes in A. vinelandii, a second round of transformations was performed to remove the ISC gene of interest from the A. vinelandii chromosome. For example, AES14, which should contain iscU∷kanamycin resistance cartridge and ΔscrX∷(araC sufU), was attempted by transforming A. vinelandii strain AES7 with pDB1018, and screening for colonies on BN plus kanamycin and arabinose. Other strains constructed in this study were submitted to the same type of experiment.

The ability of the E. faecalis machinery to complement the activity of both SUF and ISC genes in E. coli was tested by complementation with pEFSE24, pEFSE73, and pEFSE121. Previously constructed single mutant E. coliΔiscS strains (CL100 and PJ23) were submitted to complementation to achieve ISC complementation. Controls were performed using parental strains (MC1061 and TL254, respectively). Competent E. coli strains were transformed to acquire pEFSE24, pEFSE73, and pEFSE121 vectors, coding for sufS, sufSU, and sufCDSUB, respectively. The plasmids pDB551 (coding for A. vinelandii NifS) and pDB943 (coding for A. vinelandii IscS) were used as positive controls and the expression vector pDB1568 as a negative control for the complementation experiment. After transformation and selection on Luria broth-Amp plates, colonies were picked and plated on either M9-glycerol minimal modified media (by the addition of adenine, isoleucine, leucine, valine, and arabinose) or M9-glycerol minimal modified media supplemented with thiamine and nicotinic acid. Addition of adenine was necessary due to purC modification. Isoleucine, leucine, and valine were used to counteract the lag time verified for E. coliΔiscS growing on minimal media, as without them it grows at half the rate of the parental strain. The auxotrophy for thiamine and nicotinic acid caused by the lack of IscS was used for screening of complementation by comparative growth on either supplemented or nonsupplemented M9-glycerol modified minimal media. Although positive controls were cloned into vectors under lactose promoter control (pT7), the expression of IscS and NifS was high enough to allow complementation.

Double mutant E. coli strains lacking both ΔiscS and sufS, sufSE or sufABCDSE were constructed and submitted to complementation for SUF complementation. The E. coli strains CAG18481, JW1670-1, and JW2514-4 were obtained from the E. coli Genetic Stock Center, Yale University. JW2514-4 derivatives were constructed via bacteriophage P1 transduction, using pCP20 for phage lambda Red (FLP)-mediated removal of cassettes when necessary, as described by Datsenko & Wanner (2000). This methodology provided an E. coliΔsufS (EESC41) strain without kanamycin resistance. The same protocol was performed previously for E. coliΔsufSE (GSO97) and E. coliΔsufABCDSE (GSO92) strains (Outten et al., 2004). P1 phage infection of CAG18481 was performed, and phages containing the zfh-208Tn10 region were used in a transduction experiment using JW2514-4 as the recipient. Strains were selected using Luria broth plates containing both kanamycin and tetracycline, which produced strain EESC42, also submitted to P1 phage infection. EESC41 was transformed with pEFSE24, pEFSE73, pEFSE121, pDB943, and pDB15668 and selected for ampicillin resistance. Each was infected with EESC42-P1 phage lysate, and transductants were selected for kanamycin resistance (on Luria broth plates containing kanamycin, citrate, and arabinose or lactose) and scored for tetracycline resistance (on the same Luria broth plate above, plus tetracycline) for determination of cotransduction frequency. The same protocol was followed for GSO97 and GSO92, for a total of 15 transductions (Table 3). Viable cells were screened for auxotrophic phenotype by plating them on both M9-glycerol minimal modified medium and M9-glycerol minimal modified supplemented with thiamine and nicotinic acid.

Table 3.   Tetracycline selection and kanamycin scoring results for Escherichia coliΔiscSΔsufS double mutants containing potential complementing plasmids
Escherichia coli strain (genotype)PlasmidCotransduction frequency (%) (iscSKanR zfh-208Tn10)
  • For each transduction, 50–75 tetracycline-resistant colonies were screened for resistance to kanamycin.

  • *

    Positive control, iscS from Azotobacter vinelandii under pT7.

  • Negative control, empty vector.

EESC41 (ΔsufS)pDB943*75
pDB15680
pEFSE240
pEFSE730
pEFSE1210
GSO97 (ΔsufSE)pDB94396
pDB15680
pEFSE240
pEFSE730
pEFSE1210
GSO92 (ΔsufABCDSE)pDB94371
pDB15680
pEFSE240
pEFSE730
pEFSE12193

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Testing the complementation of the A. vinelandii ISC system using E. faecalis SUF genes

Azotobacter vinelandii uses the NIF and ISC systems, with the NIF system involved in maturation of nitrogenase. Genome sequences of the Firmicutes predict only the presence of the SUF machinery, with the SUF genes likely having the same functions as the ISC representatives in Proteobacteria. Therefore, A. vinelandii-containing SUF homologs were constructed to test possible complementation between the Proteobacteria ISC and Firmicutes SUF systems.

The A. vinelandii strains expressing the E. faecalis SUF homologs were constructed by homologous recombination, starting from A. vinelandii strain DJ1418 (Table 1, Fig. 2a). Several strains were created that contain all of the ISC genes and various combinations of the E. faecalis SUF genes under pBAD control. The genes inserted into the scrX region of A. vinelandii included sufS, sufSU, sufC, sufD, sufU, sufB, or the entire sufCDSUB. Phenotypic (Lac, KanS, RifR) and genotypic (PCR verification) characterizations confirmed the insertion of each of the SUF regions into the A. vinelandii DJ1418 host chromosome, yielding strains AES1 to AES7 (Fig. 2b). Strains were constructed by recombinant events using the sequenced vectors which were recombinant expressed, and derivative peptides were confirmed by mass spectra analyses. In the same way, single-site recombination events were avoided and the correct double recombinant event guaranteed by means of phenotypic and genotypic analyses. Furthermore, sequences flanking the recombinant sites of the lineages constructed and the confirmation that the regions of interest were indeed correctly in frame was possible by sequencing experiments. This demonstrates that the promoter region of the recombinant lineage was correct and that the gene could be expressed without any problems. Proper expression of these proteins was confirmed (Riboldi, Oliveri & Frazzon, unpublished data).

image

Figure 2. Azotobacter vinelandii ISC complementation. (a) Genetic methodology for obtaining recombinant strains containing Enterococcus faecalis SUF elements. (b) Azotobacter vinelandii recombinant strain-confirming phenotype containing sufC (AES1), sufD (AES2), sufS (AES3), sufU (AES4), sufSU (AES5) sufB (AES6), or sufCDSUB (AES7). DJ1418 recipient strain was also verified. Specific media with respective features are depicted. The phenotype described corresponds to the first recombination event, related to the insertion of SUF representatives under arabinose control in A. vinelandii chromosome. Truncated A. vinelandii ISC elements from the second homologous recombinant event do not present any viable cells when grown in the presence of arabinose.

Download figure to PowerPoint

To determine whether the SUF genes could complement ISC elements in the [Fe–S] cluster assembly in A. vinelandii, attempts were made to inactivate various ISC genes in the above strains. Plasmids containing kanamycin resistance cartridges truncating the housekeeping ISC gene were used to transform the above strains, with selection for kanamycin resistance on media containing arabinose. The combinations tested included: sufU or sufB as scaffolds, instead of iscU (AES4 or AES5 × pDB1018); sufS as desulfurase, instead of iscS (AES3 × pDB933K); sufSU as desulfurase, instead of iscS (AES6 × pDB933K); sufC as the ATPase partner of the system, instead of hscA (AES1 × pDB1005); sufD against all biological possibilities (AES2 × pDB1018, pDB933K, or pDB1005); and finally, the entire operon sufCDSUB, instead of iscSUA-hscBA-fdx (AES7 × pDB1370). No viable kanamycin-resistant strains were obtained, indicating that the inactivation of the ISC protein was lethal despite expression of the SUF-correspondent factor, and suggesting that the SUF operon of E. faecalis is not able to complement the ISC elements of A. vinelandii.

Complementation of the E. coliΔiscS by E. faecalis SUF genes

Escherichia coli corresponds to a Proteobacteria representative that possesses both ISC and SUF systems for [Fe–S] cluster formation. As in A. vinelandii, the ISC system serves as the housekeeping machinery, but instead of having the NIF system, E. coli possesses the SUF system as an alternative system induced in cases of oxidative stress and iron limitation.

To determine whether the E. faecalis SUF operon is able to complement the ISC system of E. coli, in vivo experiments were performed using mutants lacking iscS (see Table 1). The iscS mutants require thiamine, nicotinic acid, and branched chain amino acids for growth. This auxotrophic phenotype eliminates the need for E. coli SUF mutation to verify E. coli ISC complementation. Thus, the strains will only be viable if there is some component complementing iscS functions related to the amino acid homeostasis (classic function of type I of cysteine desulfurase related to [Fe–S] cluster formation), as much as for [Fe–S] cluster formation. Two strains of differing genetic background were utilized – PJ23 and CL100. The respective parental strains (TL254 and MC1061, respectively) were also assayed. After transformation with plasmids expressing sufS, sufSU, or sufCDSUB, strains were plated on M9-glycerol minimal medium containing arabinose, either with or without required nutritional supplements. Empty vector (pDB1568) was used as negative control and plasmids containing iscS or nifS from A. vinelandii as positive controls. No growth was observed on nonsupplemented medium after 72 h at 37 °C, although control strains grew as expected (Fig. 3a). These results indicate the E. faecalis SUF machinery is not able to complement the ISC system of Proteobacteria, even in E. coli, which is slightly evolutionarily different from A. vinelandii in terms of the presence of SUF machinery in the latter.

image

Figure 3. Escherichia coliΔiscS complementation. (a) Escherichia coli strains correspond to parental strains TL254 and MC1061 and the iscS mutants ΔPJ23 and ΔCL100, respectively. Complementation vectors correspond to empty vector pDB1568, positive controls pDB551 (Azotobacter vinelandii nifS) and pDB943 (A. vinelandii iscS), pEFSE24 (Enterococcus faecalis sufS), pEFSE73 (E. faecalis sufSU), pEFSE121 (E. faecalis sufCDSUB). M9-glycerol modified medium (M9) and M9-glycerol modified supplemented medium (Suppl.) were performed as described in Materials and methods. (b) Escherichia coli SUF complementation: E. coli K 12, E. coli EESC42 (ΔiscS), E. coli GSO92 (ΔsufABCDSE), and E. coli EESC54 (ΔiscsufABCDSE) double mutant complemented with pEFSE121 plated on M9-glycerol modified medium and supplemented M9-glycerol modified medium.

Download figure to PowerPoint

Complementation of the E. coli SUF machinery by E. faecalis SUF genes

Several Proteobacteria representatives possess the SUF. genes together with the housekeeping ISC machinery. However, E. faecalis possess the only SUF system with high homology with the corresponding E. coli SUF genes, with the addition of sufU, similar to E. coli iscU.

Genetic experiments were performed to assess the possibility that the cloned E. faecalis SUF genes can complement E. coli mutants lacking one or more of the components of the SUF system. SUF mutants of E. coli have no apparent growth phenotype. However, combination of an SUF mutation (or mutations) with an iscS mutation is lethal unless a plasmid is present in trans that provides either iscS or the missing SUF function(s) (Trotter et al., 2009). To guarantee the complementation of the iscS mutant, the complementing element needs to fill the gaps caused by the absence of iscS.

This is what seems to occur in vivo when the E. coli sufABCDSE system produces viable strains of E. coli ISC mutants (Takahashi & Tokumoto, 2002). This system plays roles related not only to [Fe–S] cluster formation, but also to nicotinic acid and thiamine biosynthesis. Escherichia coli strains JW1670-1 (ΔsufS), GSO97 (ΔsufSE), and GSO92 (ΔsufABCDSE) were used as recipient strains for phage P1 transduction experiments in which the donor strain (EESC42) contained ΔiscSkan and a tightly linked Tn10, which confers tetracycline resistance. In each transduction, tetracycline resistance was selected and kanamycin resistance scored as described by Outten et al. (2004). The appearance of viable kanR transductants would indicate complementation of either iscS or SUF function(s) by the resident plasmid. As negative and positive control plasmids, the empty vectors pDB1568 and pDB943 (which encodes iscS from A. vinelandii) were used.

Azotobacter vinelandii IscS was able to complement all double mutants, whereas the only complementation observed using the test strains was with strain GSO92 (ΔsufABCDSE), containing pEFSE121 (which encodes sufCDSUB). Tetracycline-resistant transductants were obtained that displayed resistance to kanamycin and ampicillin, and grew on glucose minimal medium (containing arabinose) after 48 h of incubation (Fig. 3b). A control reaction was carried out for EESC42 (iscSkanTn10) complemented with pEFSE121, which grew on Luria broth media (plus ampicillin and arabinose) and supplemented M9-glycerol modified minimum media but did not grow on unsupplemented M9-glycerol modified minimum media. These data demonstrate that total deletion of the SUF machinery from Proteobacteria can be complemented using the entire SUF operon from Firmicutes.

It is quite remarkable that the complemented strain was able to grow on unsupplemented glycerol minimal medium. This indicates that complementation of iscSkan by the sufCDSUB operon of E. faecalis is also occurring. As complementation did not occur using the sufS or sufSE recipients, it is clear that there are differences between the sufSE and sufSU complexes, perhaps with respect to their mechanisms of action and/or interaction between each other and with other SUF proteins.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The present paper discusses the possibility of genetic complementation among Proteobacteria [Fe–S] cluster biosynthetic machinery and the E. faecalis sufCDSUB operon. Complementation was not observed when individual proteins from the E. faecalis SUF system were expressed in E. coli strains lacking putative homolog proteins. In contrast, complementation was verified when the E. faecalis SUF system was inserted into the E. coli strain lacking both ISC and SUF systems. It appears that the presence of all complements of a given system enables proper functional interactions, which do not otherwise occur among proteins from different systems, even though these proteins are predicted to have similar functions.

The first aspect addressed by the authors was to check the capacity of E. faecalis sufCDSUB operon to replace functions of the ISC system from Proteobacteria. For this purpose, A. vinelandii, the model organism from which the ISC system was first identified, was used for recombinant events. Azotobacter vinelandii are nitrogen-fixing bacteria, containing the NIF system for nitrogenase maturation (Jacobson et al., 1989a, b); however, the NIF system is active only under nitrogen-fixing conditions. In contrast, the ISC system of A. vinelandii contains the housekeeping iscRSUA-hscBA-fdx genes for [Fe–S] cluster formation (Zheng et al., 1998). Whole sufCDSUB was not able to complement ISC operon. Several matches for specific homologous gene complementation were tried but all of them were synthetically lethal. This is in accordance with the vast diversity found between the systems analyzed. In the E. faecalis SUF operon, sufU is the only ortholog of the ISC system and, although sharing conserved cysteine residues, sufU and iscU show several structural dissimilarities, mainly in key protein–protein interaction sites (Riboldi et al., 2009). Likewise, E. faecalis do not have any ATC that could mimic iscA and/or sufA functions. In addition, the primary structure of SufB from E. faecalis is not similar to E. coli SufB, as it lacks several conserved cysteine residues responsible for the [Fe-S] cluster assembly in Proteobacteria. These differences could explain the lack of complementation observed for the Proteobacteria ISC system. Even though it shows several other genes with putatively similar functions to ISC representatives, such as IscS desulfurase (represented by SufS) and HscA with its ATPase activity (represented by SufC), studies have described the importance of specific protein–protein interactions for ISC representatives in terms of [Fe–S] cluster assembly and delivery to its final targets (Tokumoto et al., 2002; Lill, 2009; Py & Barras, 2010), which may explain the lethal pattern observed. To confirm this ISC specificity, E. coli iscS mutant strains were tested for ISC complementation, in which sufCDSUB, sufS, or sufS plus the putative desulfurase activator sufU plasmids was unable to complement ISC as well. This result agrees with data described above: indeed, neither sufCDSUB or any other gene alone is able to complement Proteobacteria ISC elements, demonstrating the conservancy of the ISC system. Escherichia coli iscS mutants were chosen for this type of experiment because the auxotrophic phenotype can be distinguished by supplemented media and parental strains, and because it also permits the verification of complementation on further deletions, as verified for the SUF system.

Because the E. faecalis operon shares major ortholog elements with the SUF system, we verified the possibility of E. coli sufABCDSE complementation. Escherichia coliΔiscSTn10∷ΔABCDSE complemented with sufCDSUB was able to grow on Luria broth plates containing arabinose. It was also able to grow on M9-glycerol modified media in the absence of iscS, albeit with a weaker phenotype and requiring 48 h to grow. In this way, the entire sufCDSUB could complement the whole sufABCDSE system, not just replacing this system but also contributing to maturation of proteins linked to the ISC system, perhaps due to the presence of SufU and its [Fe–S] cluster assembly characteristics similar to IscU. As the entire sufCDSUB system is able to provide viable E. coli strains, it is able to perform the necessary functions for nicotinic acid and thiamine homeostasis and the relevant processes in [Fe–S] cluster homeostasis.

However, sufCDSUB is not able to complement E. coliΔiscS strains (Fig. 3a). This may be related to the presence of E. coli SUF components, in which protein complexes are essential for proper SUF function in E. coli. The presence of these elements and/or complexes could be either inhibiting or obstructing the actuation of the in trans operon. This hypothesis is based on data found in this work, where (1) neither E. coliΔiscS∷ΔsufS or E. coliΔiscS∷ΔsufSE could be complemented by sufS, sufSU, or sufCDSUB, and (2) E. faecalis sufCDSUB was not able to complement E. coliΔiscS strains but could complement E. coliΔiscSΔsufABCDSE.

In fact, several specific protein–protein interactions involving E. coli SUF system partners have been described: SufE and SufBCD acting synergistically to modulate SufS activity (Outten et al., 2003); SufBC2D complex acting as scaffold and mobilizing sulfur from SufSE, using FADH2 for reductive mobilization of iron (Layer et al., 2007; Wollers et al., 2010); SufD and the ATPase activity of SufC required for in vivo iron acquisition (Saini et al., 2010). Finally, Takahashi & Tokumoto (2002) described the requirement of a plasmid containing the whole sufABCDSE to construct an iscRSUA mutant in a sufABCDSE background. The same pattern was observed here, where plasmids coding for sufS and sufSU did not complement any strain tested. This indicates the requirement for the entire SUF operon and the importance of protein–protein interactions, still to be determined, for the actuation of the complex during [Fe–S] cluster formation processes and maturation of target proteins. These data are very exciting, as they are the first record of complementation between Proteobacteria and Firmicutes [Fe–S] cluster elements.

In summary, the present work found that neither the sufCDSUB whole operon or specific genes contained in this system are able to complement ISC systems from A. vinelandii and E. coli, but that the entire E. faecalis SUF operon is able to complement the E. coli SUF system, producing viable mutants of both sufABCDSE and iscRSU-hscBA-fdx systems.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We would like to thank Prof. Dennis R. Dean for supporting G.P.R. in his lab and for all discussions during that period. Also thanks to Prof. Wayne F. Outten, who kindly provided E. coli SUF. mutant strains, and Valerie L. Cash for expert technical assistance. This work was supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq –#306397/2006-4, #473769/2007-7) and Coordenação de Aperfeiçoamento de Pessoal de Nivel Superior (PDEE/CAPES) of the Brazil government.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References