ribB and ribBA genes from Acidithiobacillus ferrooxidans: expression levels under different growth conditions and phylogenetic analysis

Research in Microbiology 159 (2008) 423e431www.elsevier.com/locate/resmic

ribB and ribBA genes from Acidithiobacillus ferrooxidans: expressionlevels under different growth conditions and phylogenetic analysis

Fabio H.P. Knegt a, Luciane V. Mello b, Fernanda C. Reis a, Marcos T. Santos a,Renato Vicentini c, Lucio F.C. Ferraz a, Laura M.M. Ottoboni a,*

a Centro de Biologia Molecular e Engenharia Genetica (CBMEG), Universidade Estadual de Campinas (UNICAMP),

C.P. 6010, 13083-875 Campinas, S.P., Brazilb NIBHI e Northwest Institute for Bio-Health Informatics, University of Liverpool, Liverpool L69 7ZB, UK

c Departamento de Genetica e Evoluc~ao, Instituto de Biologia (IB), Universidade Estadual de Campinas (UNICAMP),13083-970 Campinas, S.P., Brazil

Received 20 December 2007; accepted 8 April 2008

Available online 23 April 2008

Abstract

Acidithiobacillus ferrooxidans is a Gram-negative, chemolithoautotrophic bacterium involved in metal bioleaching. Using the RNA arbi-trarily primed polymerase chain reaction (RAP-PCR), we have identified several cDNAs that were differentially expressed when A. ferrooxidansLR was submitted to potassium- and phosphate-limiting conditions. One of these cDNAs showed similarity with ribB. An analysis of theA. ferrooxidans ATCC 23270 genome, made available by The Institute for Genomic Research, showed that the ribB gene was not located inthe rib operon, but a ribBA gene was present in this operon instead. The ribBA gene was isolated from A. ferrooxidans LR and expressionof both ribB and ribBA was investigated. Transcript levels of both genes were enhanced in cells grown in the absence of K2HPO4, in the presenceof zinc and copper sulfate and in different pHs. Transcript levels decreased upon exposure to a temperature higher than the ideal 30 �C and at pH1.2. A comparative genomic analysis using the A. ferrooxidans ATCC 23270 genome revealed similar putative regulatory elements for bothgenes. Moreover, an RFN element was identified upstream from the ribB gene. Phylogenetic analysis of the distribution of RibB and RibBAin bacteria showed six different combinations. We suggest that the presence of duplicated riboflavin synthesis genes in bacteria must providetheir host with some benefit in certain stressful situations.� 2008 Elsevier Masson SAS. All rights reserved.

Keywords: Acidithiobacillus ferrooxidans; ribB; ribBA; Gene expression; Phylogenetic analysis

1. Introduction

Riboflavin is the precursor molecule for synthesis of twocoenzymes, flavin mononucleotide (FMN) and flavin adeninedinucleotide (FAD), both essential for cellular metabolism.Many enzymes are involved in synthesis of riboflavin fromguanosine triphosphate (GTP) and ribulose-5-phosphate, amongthem GTP cyclohydrolase II (GCHII, RibA) and 3,4-dihydroxy-2-butanone 4-phosphate synthase (DHBP synthase, RibB) [3,30].

The organization of riboflavin synthesis genes in bacterialgenomes does not follow a common pattern. In the genome

* Corresponding author.

E-mail address: [email protected] (L.M.M. Ottoboni).

0923-2508/$ - see front matter � 2008 Elsevier Masson SAS. All rights reserved.

doi:10.1016/j.resmic.2008.04.002

of Bacillus subtilis, for example, these genes are organizedin an operon, while in Escherichia coli they are distributedthroughout the genome [11]. Also, in B. subtilis and otherGram-positive bacteria, RibB and RibA are encoded by a sin-gle gene (ribBA) [11]. Besides the ribBA gene, Helicobacterpylori and Photobacterium phosphoreum also carry an addi-tional copy of ribA [5,18]. In Enterobacteriaceae, the genesribB and ribA remain separated [3].

Very little is known about the rib genes from Acidithioba-cillus ferrooxidans. This bacterium is Gram-negative, non-sporulating, rod-shaped and derives energy from the oxidationof reduced sulfur compounds and ferrous iron [20]. Thisg-proteobacterium is commonly found in a wide range of en-vironments including mining areas, sewage treatment plants

424 F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

and marine habitats. A. ferrooxidans has gained attention dueto its involvement in the bioleaching of metals, a process inwhich metal sulfides are converted to water-soluble metalsulfates [6]. This work addresses the cloning and transcriptlevel determination of riboflavin synthesis genes ribB andribBA from this microorganism in different stressful situations.An in silico analysis of the upstream genomic region of ribBand ribBA was performed to identify regulatory motifs. Thedistribution of RibB and RibBA in bacteria was also analyzed.

2. Materials and methods

2.1. Bacterial strain and growth conditions

The A. ferrooxidans Brazilian strain LR [14] was selectedfor differential gene expression analysis in response to potas-sium- and phosphate-limiting conditions. These bacteria weregrown at 30 �C 250 rpm in a salt solution of modified T&Kliquid medium [28] containing (in g/l): (0.4) K2HPO4 $ 3H2O,(0.4) MgSO4 $ 7H2O, (0.4) (NH4)2SO4, and (33.4) FeS-O4 $ 7H2O, pH 1.8 adjusted with sulfuric acid. The K2HPO4

salt was omitted from the T&K medium for bacterial growthunder potassium- and phosphate-limiting conditions. Theomission of this salt from the medium does not prevent theA. ferrooxidans growth, because these bacteria accumulatepolyphosphate granules in high amounts [2].

2.2. RAP-PCR

RAP-PCR experiments were performed as described byPaulino and co-workers [22]. RNA was isolated according to

Fig. 1. RNA slot-blot hybridization using RNA isolated from A. ferrooxidans LR c

24 h. ribB (A) and ribBA (B) were used as probes. The amount of RNA in the memb

gene. Histograms show the relative expression value in each situation.

Winderickx and Castro [31] from cells grown in the presence(control) and in the absence of 1.75 mM K2HPO4 and thenused in first-strand cDNA synthesis using Ready-to-go RT-PCR beads (Amersham Biosciences). The arbitrary primersused were OPF01, OPJ06 and OPJ14 (Operon Technologies).For second-strand synthesis, 4 ml of the first-strand reactionwere mixed, in a 20 ml reaction, with 1 � PCR buffer,1.25 mM MgCl2, 2 mM of the same primer used in first-strandcDNA synthesis, 2 mM of each deoxynucleoside triphosphate(Amersham Biosciences), 1 mCi [a-33P]dCTP (AmershamBiosciences) and 1.5 U of Taq DNA polymerase (Invitrogen).The amplifications were performed in duplicate in a PerkinElmer 2400 thermal cycler using the following amplificationconditions: initial denaturation at 94 �C (5 min), 40 cycles at94 �C (30 s), 40 �C (2 min), 72 �C (30 s) and final extensionat 72 �C (20 min). The amplification products were submittedto electrophoresis on a 5% acrylamide e 50% urea e1 � Tris-borate-EDTA gel.

Differentially expressed cDNAs were excised from thegel and DNA was eluted and amplified according to Paulinoand co-workers [22]. The differential expression of thecDNAs was confirmed by slot-blot hybridization. ThecDNAs whose differential expression was confirmed werecloned into the pGEM-T Easy vector (Promega) followingthe manufacturer’s specifications. At least three clonesfrom each sample were sequenced (ABI Prism 377, AppliedBiosystems) and the obtained sequences were compared withGenBank sequences using the BLAST algorithm, version2.2.8 [1]. Among isolated cDNAs was one that exhibitedsimilarity with ribB that encodes the enzyme DHBPsynthase.

ultured in the absence and in the presence of 1.75 mM K2HPO4 for 12, 18 and

ranes was normalized by hybridization with the A. ferrooxidans LR 16S rRNA

425F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

2.3. Amplification of ribBA from A. ferrooxidans LR

The following primers were used for amplification ofribBA: ribBA_f (50-ATGAGCAGTGCCAGTATAAGTTC-30)and ribBA_r (50-TCATGACTCATCCTCGGGAA-30). For thePCR reaction, 50e100 ng of genomic DNA from A. ferrooxi-dans LR were mixed in a 20 ml reaction with 2.5 U Taq poly-merase (Invitrogen), 1.5 mM MgCl2, 20 mM dNTPs and

Fig. 2. RNA slot-blot hybridization using RNA isolated from A. ferrooxidans LR c

and B), 50 mM CuSO4, 37 �C (C and D) and at different pHs (1.2, 1.5, 1.8, 2.5 and 3

the hybridizations. The amount of RNA in the membranes was normalized by hyb

relative expression value in each situation.

4 pmoles of each primer. The amplification conditions con-sisted of an initial denaturation step of 5 min at 94 �C followedby 30 cycles of 30 s at 94 �C, 30 s at 55 �C, 45 s at 72 �C anda final extension of 30 min at 72 �C. The amplification productwas cloned and sequenced. ribB and ribBA sequences from A.ferrooxidans strain LR were identical to those reported byTIGR (http://www.tigr.org) for A. ferrooxidans strain ATCC23270.

ultured in the absence and in the presence of 100, 300 and 600 mM ZnSO4 (A

.0) (E and F). ribB (A, C and E) and ribBA (B, D and F) were used as probes in

ridization with the A. ferrooxidans LR 16S rRNA gene. Histograms show the

426 F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

2.4. Total RNA slot blots

The transcript levels of ribB and ribBA were analyzed byRNA slot-blot hybridization in the following conditions:absence and presence of different zinc sulfate concentrations(100, 300 and 600 mM), 50 mM copper sulfate, 37 �C, differ-ent pHs (1.2, 1.5, 1.8, 2.5 and 3.0) and a time course (12, 18and 24 h) in the presence and absence of K2HPO4. RNA wasisolated as described by Winderickx and Castro [31] from A.ferrooxidans LR cells grown in the conditions mentionedabove until late log-phase, except for the time-course experi-ment. Three mg of RNA were mixed with 10 mM TriseHCl/1 mM EDTA (TE) pH 8.0 to a final volume of 12.5 ml andthen mixed with 37.5 ml of a solution containing 500 ml ofdeionized formamide, 162 ml of formaldehyde 37% and100 ml of MOPS 10� [23]. The samples were incubated at65 �C for 5 min and 50 ml of 20� SSC were added. Duplicatesamples were transferred to nylon membranes (Hybond-N,Amersham Biosciences) using a slot-blot apparatus (GibcoBRL). After incubation at 80 �C for 2 h, membranes wereprehybridized for at least 2 h and hybridized overnight with107e108 cpm/ml of the denatured probe. The prehybridiza-tion, hybridization and wash solutions were done as describedby Paulino and co-workers [22]. Membranes were exposed toX-ray film and hybridization signals were quantified using Ko-dak Digital Science-1D image analysis software v. 2.0.3 (Ko-dak). The amount of RNA in the membranes was normalizedby hybridization with the A. ferrooxidans LR 16S rRNA gene.

2.5. Comparative genomics and phylogenetic analysis

Comparative analysis of the A. ferrooxidans non-codingupstream genomic region of ribB and ribBA was performedto predict new putative regulatory sites. In this sense, Phylo-Gibbs algorithm [25] was used to determine similar motifsamong the sequences. The motif width was set to 20 bp. All

Fig. 3. Comparative genomic analysis of ribB and ribBA (rib operon) and identific

genes. (A) Schematic representation of the genomic region of ribB and ribBA from A

and nrdR) around both the ribB gene and the rib operon. (B) Sequence logo of a n

from A. ferrooxidans. The height of each stack of letters represents the degree of s

motifs that showed a reliable track score were used in consen-sus determination. The sequence logo was created usingWebLogo software [9]. The Rfam program [16] was used tosearch for RFN elements in ribB and ribBA upstream regions.The RNA secondary structure of the RFN element was pre-dicted using the Mfold program [36].

An alignment of available RibB (DHBP synthase) domainsequences was retrieved from the corresponding entry(PF00926) in the Pfam database release 20.0 [4]. It contained387 sequences from bacteria. The Pfam alignment was refinedusing MUSCLE [10]. Closely related sequences were removedusing the redundancy removal option of JALVIEW [7] whichwas also used for general alignment manipulation. However,sequences from species with more than one RibB homologuewere kept in the alignment irrespective of their sequence iden-tity. Based on the alignment, a first phylogenetic tree of 169sequences was constructed using the maximum parsimonymethod in PAUP version 4.0b 10 with heuristic search [26].This first tree was used to choose representative species whichwould reveal the different combinations of RibB and RibBAdomains present in these species of bacteria. New phyloge-netic trees using sequences from representative species werecalculated using PHYLIP (both Protdist and Protpars methods;[12]) and CLUSTAL W [27]. Branch support was evaluated by1000 bootstrap replicates.

3. Results and discussion

In this work, a differentially expressed cDNA was isolated byRAP-PCR when A. ferrooxidans LR was submitted to potas-sium- and phosphate-limiting conditions. The deduced aminoacid sequence from this cDNA presented 76% similarity (e-value 7ee70) with the enzyme 3,4-dihydroxy-2-butanone 4-phosphate synthase (DHBP synthase). This enzyme is involvedin riboflavin biosynthesis. In a previous study realized in our lab-oratory, the deduced amino acid sequence of a cDNA presented

ation of the new putative conserved binding site located upstream from these

. ferrooxidans. Note the presence of transcriptional regulator genes (luxR, lysR,

ew putative conserved binding site upstream from the ribB and the rib operon

equence conservation. The letters are sorted with the most frequent at the top.

427F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

similarity with the bifunctional enzyme DHBP synthase/GTPcyclohydrolase II. These results suggested that the A. ferrooxi-dans genome contained both genes ribB and ribBA. An analysisof the A. ferrooxidans ATCC 23270 genome, sequenced byTIGR, confirmed our suspicion. These genes are 1500 kb apartin the genome and the ribBA gene is located in a rib operon.The nusB and thiL genes are located downstream from the riboperon, which is a common feature among several g-proteobac-teria like Vibrionaceae and Pseudomonadaceae, and glyA is lo-cated upstream from the operon as in Photobacteriumphosphoreum and Vibrio fischeri [18].

The presence of additional copies of genes from the riboperon in bacteria has been previously shown [18]. The extracopy of ribB in the genome of A. ferrooxidans might be relatedto the necessity for enhancing production of riboflavin in thisorganism in a shorter time.

To investigate expression of ribB and ribBA when A.ferrooxidans LR was submitted to different growth conditions,the complete sequence of the ribBA gene was isolated by PCR

Fig. 4. RNA secondary structure of the putative RFN element located upstream from

structure are shown by numbers (1e5) and by the variable stem-loop. The bars in

from A. ferrooxidans LR. The DHBP synthase domain ofRibBA was located in the N-terminal and the domain corre-sponding to GTP cyclohydrolase II was located in the C-terminal.The nucleotide sequences of ribB and the DHBP synthasedomain of ribBA presented 64% identity. The deduced aminoacid sequences presented 57% identity. Fassbinder and co-workers [11] found similar identity after comparing theRibB portion of RibBA from Helicobacter pylori with RibBfrom Escherichia coli.

It was observed that expression of ribB and ribBA couldchange in response to the presence or absence of K2HPO4

and also in response to bacterial growth phase (Fig. 1A,B).During growth in the presence of 1.75 mM of K2HPO4 (con-trol) the expression of ribB and ribBA showed almost nochanges in the beginning and mid-log phases. Between thelog and stationary phase, a decrease in expression of ribBand ribBA was observed. This could be due to a decrease inthe bacterial metabolic rate due to the imminent end of theoxidizable substrate.

the ribB gene in A. ferrooxidans. Complementary stems of the RNA secondary

dicate conserved nucleotides in the stems.

428 F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

In the absence of K2HPO4 the expression of ribB and ribBAwas even more dependent on the bacterial growth phase. Prob-ably due to an initial low metabolic rate in the absence ofK2HPO4, at 12 h of growth, the expression of ribB and ribBAwas lower than that observed in the presence of K2HPO4. Thehighest expression was observed in the mid-log phase. Thisexpression was followed by a decrease at 24 h of growth. At18 and 24 h of growth, expression of the genes was higherin the absence than in the presence of K2HPO4 (Fig. 1A,B).

As shown in Fig. 2A,B, the expression of both ribB andribBA was enhanced in the presence of zinc sulfate (ZnSO4),and as the concentration of this metal sulfate increased from100 to 600 mM the expression of the genes increased. The

Fig. 5. (A) Phylogenetic tree. All RibB and RibBA containing sequences from bacte

sequences were removed from the alignment maintaining representatives of each

uncorrected pair-wise distances between the aligned sequences using the tree option

to their domain architecture (see B and Table 1). (B) Schematic representation of do

gle, isolated DHBP synthase domain, RibB (black rectangle), (B) DHBP synthase an

(D) two copies of RibBA, (E) one copy of each, or (F) one copy of RibB and two

pattern of expression of the two genes was very similar,though ribBA showed relative expression values smaller thanribB.

It was reported that some proteins, such as cytochrome coxidase and rusticyanin that play a role in A. ferrooxidansrespiration [17], are involved in resistance to divalent cationslike mercury. Since riboflavin is a precursor of coenzymesof respiratory enzymes, we speculate that the observedincrease in the expression of these genes involved in thesynthesis of riboflavin was due to the necessity of providingcoenzymes to increase the levels of respiratory enzymes.

It was shown that riboflavin plays an important role in thereduction of Fe3þ in H. pylori [33]. Also, the enzyme mercury

ria from the PFAM databases were aligned using Clustal W [27]. Highly similar

domain architecture. A neighbor-joining tree was calculated from a matrix of

within Clustal W. Sequences used in the final alignment are labeled according

main architectures of selected bacterial species. The species contain (A) a sin-

d GTP cyclohydrolase II, RibBA (oval gray) domains, (C) two copies of RibB,

copies of RibBA.

Table 1

RibB and RibBA homologues analyzed in Fig. 5A,B. Species names are fol-

lowed by Swissprot/Trembl identifiers, except for E0 which are the proteins re-

ported here

Group Species names Swissprot/Trembl

identifiers

A1 Escherichia coli RIBB_ECOL6

A2 Ehrlichia ruminantium Q5HC97_EHRRW

A3 Anaplasma marginali Q5P9M6_ANAMM

B1 Neisseria meningitidis RIBAB_NEIMB

B2 Mycobacterium leprae Q9CCP4_MYCLE

B3 Campylobacter jejuni Q5HVJ8_CAMJR

B4 Caulobacter crescentus Q9A9S5_CAUCR

C Actinobacillus succinogenes Q3EHN0_ACTSC

and Q3EG58_ACTSC

D1 Mycobacterium bovis RIBAB_MYCBO

and Q7TZ92_MYCBO

D2 Corynebacterium jeikeium Q4JVI5_CORJK

and Q4JSS8_CORJK

E1 Shewanella putrefaciens Q2ZT24_SHEPU

and Q2ZQD6_SHEPU

E2 Shewanella amazonensis Q3QM96_9GAMM

and Q3QH34_9GAMM

E3 Vibrio cholerae RIBB_VIBCH

and Q9KPU3_VIBCH

E4 Desulfovibrio vulgaris Q72B63_DESVH and

Q72CT4_DESVH

E0 Acidithiobacillus ferrooxidans RibB and RibBA

F1 Pseudomonas putida RIBB_PSEPK,

Q88GB1_PSEPK

and Q88QH7_PSEPK

F2 Ralstonia eutropha Q46TZ9_RALEJ,

Q474N3_RALEJ and

Q471N1_RALEJ

429F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

reductase present in the mer operon from A. ferrooxidans [24]and other bacteria is a flavoprotein that uses FAD as a cofactor[13]. These facts suggest a probable link between riboflavinand resistance to metal ions in A. ferrooxidans.

The expression of ribB and ribBA was tested when bacteriawere grown in the presence of copper sulfate (CuSO4) and ata temperature higher than the optimum (Fig. 2C,D). Both ribBand ribBA were induced in the presence of copper sulfate andrepressed at 37 �C. It is interesting to note that, though bothgenes had their expression enhanced in the presence of coppersulfate, the expression of ribBA (approximately 3�) wasweaker than that observed for ribB (approximately 11�).

The A. ferrooxidans ideal pH is around 2.0 [19]. Since pHchange is one of the parameters that can affect bioleaching,expression of ribB and ribBA was analyzed during A. ferroox-idans growth at different pHs: 1.2, 1.5, 1.8 (control), 2.5 and3.0. As observed in the other situations tested, the expressionof the two genes at the different pHs was very similar(Fig. 2E,F). Expression was enhanced at pH higher than 1.8and at pH 1.5. Expression was repressed at pH 1.2.

Like other acidophilic organisms, A. ferrooxidans hasa cytosolic pH close to neutrality [8]. However, the periplasmand outer membrane are exposed to pH variations [29]. Sev-eral proteins involved in oxidation of Fe2þ in A. ferrooxidans,like rusticyanin and several cytochromes, are located in theouter membrane or periplasmic space [34,35]. In this way,pH changes can affect oxireduction reactions that occur duringcellular respiration and can alter, for example, the electro-chemical potential of the medium, the stability of the enzymesthat are part of that process and their cofactors, like FAD andFMN that derive from riboflavin. Therefore, we speculate thatthe increase in the expression of ribB and ribBA at pHs 1.5, 2.5and 3.0 is related to iron oxidation alterations at these pHs.

It is interesting to note that in all the situations tested thepattern of expression of ribB and ribBA was very similar,though these genes are in different locations in the genome.Thus, there is a possibility that ribB is under the same regula-tory control as ribBA (rib operon). Transcriptional regulatorgenes (luxR, lysR and nrdR) were found around the ribB andribBA genes in the genome of A. ferrooxidans (Fig. 3A). More-over, we identified a new putative conserved binding site thatis present three times upstream from the ribB gene and onetime upstream from the rib operon (Fig. 3B). We also foundputative known binding sites located upstream from the ribBgene and the rib operon (data not shown). The possibility ofcommon regulation for ribB and ribBA corroborates thehypothesis that RibB might work as a complement to RibBA,increasing the efficiency of riboflavin biosynthesis in A.ferrooxidans.

Previous reports have identified regulatory elements knownas RFN upstream from prokaryotic riboflavin biosynthesisgenes [15,30,32]. This conserved RNA secondary structureregulates expression of target genes by an FMN-mediatedmechanism [15,32]. Comparative sequence analysis wascarried out to search for RFN elements upstream from therib operon and the ribB gene from A. ferrooxidans. A putativeRFN element was identified (108.65 bits) only upstream from

the ribB gene (from nucleotides �248 to �100) (Fig. 4). Thisputative RFN element shows a highly conserved secondarystructure consisting of five conserved and one variable stem-loop [15,30]. To our knowledge, this is the first report of anRFN element upstream from a riboflavin biosynthesis genein A. ferrooxidans. According to Vitreschak and co-workers[30], the RFN element usually regulates single riboflavin bio-synthesis genes in proteobacteria, which is in agreement withour finding.

Phylogenetic analysis was performed to systematicallyanalyze the distributions of RibB and RibBA in bacteria(Fig. 5A). The results showed 6 different combinations, asillustrated in Fig. 5B. Letter codes in both figures are related(see Table 1). In Fig. 5B, (A) indicates proteins containingthe DHBP synthase domain only, while in (B) the proteinshave both DHBP synthase and GTP cyclohydrolase IIdomains. In both groups there is a single RibB domain. Actino-bacillus succinogenes (C) has two copies of RibB and speciesin D have two copies of RibBA. Group E, in which Acidithio-bacillus ferrooxidans is placed, consists of species containingone copy of RibB and one of RibBA. Species in group Fpresent a copy of RibB, besides the two copies of RibBA.As mentioned previously, extra copies of the genes that codefor these proteins might be related to different needs in ribofla-vin production.

Liao and co-workers [21] determined the crystal structureof the DHBP synthase, revealing conserved amino acid

430 F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

residues, the functional roles of which were proposed. Thefinal alignment used to calculate Fig. 5 was checked and all22 key residues are present in all proteins, indicating that iden-tical catalytic activity is expected for all of them. The fact thatduplicate copies frequently arising independently maintaintheir activity and are not lost from the genome suggests thatthey must provide their host with some benefit. If their cata-lytic activities are identical, it may be that the duplicate copieshave evolved to interact differently with other cellular compo-nents and therefore offer additional regulatory possibilities tothe cell.

Acknowledgements

This work was supported by the Fundac~ao de Amparo aPesquisa do Estado de S~ao Paulo (FAPESP; grant 02/07642-3). FHPK (01/12517-0) and FCR (05/00139-02) had fellow-ships from FAPESP. LMMO had a research fellowship fromthe Conselho Nacional de Desenvolvimento Cientıfico e Tec-nologico (CNPq). LFCF and MTS had fellowships from theFundac~ao de Apoio a Ciencia, Tecnologia e Educac~ao(FACTE).

References

[1] Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z.,

Miller, W., et al. (1997) Gapped BLAST and PSI-BLAST: a new gener-

ation of protein database search programs. Nucleic Acids Research 25,

3389e3402.

[2] Alvarez, S., Jerez, C.A. (2004) Copper ions stimulate polyphosphate

degradation and phosphate efflux in Acidithiobacillus ferrooxidans.

Applied and Environmental Microbiology 70, 5177e5182.

[3] Bacher, A., Eberhardt, S., Richter, G. (1996) Biosynthesis of riboflavin.

In: F.C. Neidhart, R. Curtiss III, J.L. Ingraham, E.C.C. Lin, K.B. Low,

B. Magasanik, W.S. Reznikoff, M. Riley, M. Schaechter, &

H.E. Umbarger (Eds.), Escherichia coli and Salmonella typhimurium,

Cellular and molecular biology (pp. 657e664). Washington, DC: Amer-

ican Society for Microbiology.

[4] Bateman, A., Coin, L., Durbin, R., Finn, R.D., Hollich, V., Griffiths-

Jones, S., et al. (2004) The Pfam protein families database. Nucleic Acids

Research 32, D138eD141.

[5] Bereswill, S., Fassbinder, F., Volzing, C., Covacci, A., Haas, R., Kist, M.

(1998) Hemolytic properties and riboflavin synthesis of Helicobacter

pylori: cloning and functional characterization of the ribA gene encoding

GTP-cyclohydrolase II that confers hemolytic activity to Escherichia

coli. Medical Microbiology and Immunology 186, 177e187.

[6] Bosecker, K. (1997) Bioleaching: metal solubilization by microorgan-

isms. FEMS Microbiology Reviews 20, 591e604.

[7] Clamp, M., Cuff, J., Searle, S.M., Barton, G.J. (2004) The Jalview Java

alignment editor. Bioinformatics 20, 426e427.

[8] Cobley, J.G., Cox, J.C. (1983) Energy conservation in acidophilic bacte-

ria. Microbiological Reviews 47, 579e595.

[9] Crooks, G.E., Hon, G., Chandonia, J.M., Brenner, S.E. (2004) WebLogo:

a sequence logo generator. Genome Research 14, 1188e1190.

[10] Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high

accuracy and high throughput. Nucleic Acids Research 32, 1792e1797.

[11] Fassbinder, F., Kist, M., Bereswill, S. (2000) Structural and functional

analysis of the riboflavin synthesis genes encoding GTP cyclohydrolase

II (ribA), DHBP synthase (ribBA), riboflavin synthase (ribC ), and

riboflavin deaminase/reductase (ribD) from Helicobacter pylori strain

P1. FEMS Microbiology Letters 191, 191e197.

[12] Felsenstein, J. (1989) PHYLIP e phylogeny inference package (version

3.2). Cladistics 5, 164e166.

[13] Fox, B., Walsh, C.T. (1982) Mercuric reductase. Purification and

characterization of a transposon-encoded flavoprotein containing an

oxidation-reduction-active disulfide. Journal of Biological Chemistry

257, 2498e2503.

[14] Garcia Jr., O., Silva, L.L. (1991) Differences in growth and iron oxida-

tion among Thiobacillus ferrooxidans cultures in the presence of some

toxic metals. Biotechnology Letters 13, 567e570.

[15] Gelfand, M.S., Mironov, A.A., Jomantas, J., Kozlov, Y.I.,

Perumov, D.A. (1999) A conserved RNA structure element involved

in the regulation of bacterial riboflavin synthesis genes. Trends in

Genetics 15, 439e442.

[16] Griffiths-Jones, S., Moxon, S., Marshall, M., Khanna, A., Eddy, S.R.,

Bateman, A. (2005) Rfam: annotating non-coding RNAs in complete

genomes. Nucleic Acids Research 33, D121eD124.

[17] Iwahori, K., Takeuchi, F., Kamimura, K., Sugio, T. (2000) Ferrous

iron-dependent volatilization of mercury by the plasma membrane of

Thiobacillus ferrooxidans. Applied and Enviromental Microbiology 66,

3823e3827.

[18] Kasai, S., Sumimoto, T. (2002) Stimulated biosynthesis of flavins in

Photobacterium phosphoreum IFO 13896 and the presence of complete

rib operons in two species of luminous bacteria. European Journal of

Biochemistry 269, 5851e5860.

[19] Kelly, D.P., Wood, A.P. (2000) Reclassification of some species of

Thiobacillus to the newly designated genera Acidithiobacillus gen.

nov., Halothiobacillus gen. nov. and Thermithiobacillus gen. nov. Interna-

tional Journal of Systematic and Evolutionary Microbiology 50,

511e516.

[20] Kuenen, J.G., Tuovinen, O.H. (1981) The genera Thiobacillus and Thio-

microspira. In: M.P. Starr, H. Stolp, H.G. Truper, A. Balows, &

H.G. Schegel (Eds.), The prokaryotes (pp. 1023e1036). Berlin:

Springer-Verlag.

[21] Liao, D.I., Calabrese, J.C., Wawrzak, Z., Viitanen, P.V., Jordan, D.B.

(2001) Crystal structure of 3,4-dihydroxy-2-butanone 4-phosphate

synthase of riboflavin biosynthesis. Structure 9(1), 11e18.

[22] Paulino, L.C., Mello, M.P., Ottoboni, L.M.M. (2002) Differential gene

expression in response to copper in Acidithiobacillus ferrooxidans

analyzed by RNA arbitrarily primed polymerase chain reaction. Electro-

phoresis 23, 520e527.

[23] Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular cloning: a lab-

oratory manual. Cold Spring Harbor, NY: Cold Spring Harbor

Laboratory.

[24] Shiratori, T., Inoue, C., Sugawara, K., Kusano, T., Kitagawa, Y. (1989)

Cloning and expression of Thiobacillus ferrooxidans mercury ion resis-

tance genes in Escherichia coli. Journal of Bacteriology 171,

3458e3464.

[25] Siddharthan, R., Siggia, E.D., van Nimwegen, E. (2005) PhyloGibbs:

a gibbs sampling motif finder that incorporates phylogeny. PLoS Compu-

tational Biology 1. e67.

[26] Swofford, D.L. (2002) PAUP*: phylogenetic analysis using parsimony

(*and other methods) version 4.0b 10. Sunderland, MA: Sinauer

Associates.

[27] Thompson, J.D., Higgins, D.G., Gibson, T.J. (1994) CLUSTAL W:

improving the sensitivity of progressive multiple sequence alignments

through sequence weighting, position specific gap penalties and weight

matrix choice. Nucleic Acids Research 22, 4673e4680.

[28] Tuovinen, O.H., Kelly, D.P. (1972) Biology of Thiobacillus ferrooxidansin relation to the microbiological leaching of sulphide ore. Zeitschrift fur

Allgemeine Mikrobiologie 12, 311e346.

[29] Valdes, J., Veloso, F., Jedlicki, E., Holmes, D. (2003) Metabolic recon-

struction of sulfur assimilation in the extremophile Acidithiobacillusferrooxidans based on genome analysis. BMC Genomics 4, 51e67.

[30] Vitreschak, A.G., Rodionov, D.A., Mironov, A.A., Gelfand, M.S. (2002)

Regulation of riboflavin biosynthesis and transport genes in bacteria by

transcriptional and translational attenuation. Nucleic Acids Research

30, 3141e3151.

[31] Winderickx, J., Castro, J.M. (1994) Practical course in molecular biology

of microorganisms, January 23eFebruary 11, Universidade Federal de

Ouro Preto e MG, pp. 59.

431F.H.P. Knegt et al. / Research in Microbiology 159 (2008) 423e431

[32] Winkler, W.C., Cohen-Chalamish, S., Breaker, R.R. (2002) An mRNA

structure that controls gene expression by binding FMN. Proceedings

of the National Academy of Sciences USA 99, 15908e15913.

[33] Worst, D.J., Gerrits, M.M., Vandenbroucke-Grauls, C.M., Kusters, J.G.

(1998) Helicobacter pylori ribBA-mediated riboflavin production is

involved in iron acquisition. Journal of Bacteriology 180, 1473e1479.

[34] Yarzabal, A., Brasseur, G., Bonnefoy, V. (2002a) Cytochromes c of

Acidithiobacillus ferrooxidans. FEMS Microbiology Letters 209,

189e195.

[35] Yarzabal, A., Brasseur, G., Ratouchniak, J., Lund, K., Lemesle-

Meunier, D., DeMoss, J.A., et al. (2002b) The high molecular weight

cytochrome c Cyc2 of Acidithiobacillus ferrooxidans is an outer

membrane protein. Journal of Bacteriology 184, 313e317.

[36] Zuker, M., Mathews, D.H., Turner, D.H. (1999) Algorithms and thermo-

dynamics for RNA secondary structure prediction: a practical guide. In:

J. Barciszewski, & B.F.C. Clark (Eds.), RNA biochemistry and biotech-

nology, number 70 in Nato Science Series (pp. 11e43). Dordrecht, The

Netherlands: Kluwer Academic Publishers.