sequencing and promoter analysis of the nifenxorf3orf5fdxanifq

1379

Braz J Med Biol Res 34(11) 2001

Azospirillum brasilense nifENXorf3orf5fdxAnifQ operonBrazilian Journal of Medical and Biological Research (2001) 34: 1379-1395ISSN 0100-879X

Sequencing and promoter analysis ofthe nifENXorf3orf5fdxAnifQ operonfrom Azospirillum brasilense Sp7

Departamentos de 1Genética and 2Biologia Molecular e Biotecnologia,Centro de Biotecnologia, Universidade Federal do Rio Grande do Sul,Porto Alegre, RS, Brasil

D.P. Potrich1,T.A. Bressel2, I.S. Schrank2

and L.M.P. Passaglia1

Abstract

A 40-kb DNA region containing the major cluster of nif genes hasbeen isolated from the Azospirillum brasilense Sp7 genome. In thisregion three nif operons have been identified: nifHDKorf1Y,nifENXorf3orf5fdxAnifQ and orf2nifUSVorf4. The operons contain-ing nifENX and nifUSV genes are separated from the structuralnifHDKorf1Y operon by about 5 kb and 10 kb, respectively. Thepresent study shows the sequence analysis of the 6045-bp DNA regioncontaining the nifENX genes. The deduced amino acid sequences fromthe open reading frames were compared to the nif gene products ofother diazotrophic bacteria and indicate the presence of seven ORFs,all reading in the same direction as that of the nifHDKorf1Y operon.Consensus s54 and NifA-binding sites are present only in the promoterregion upstream of the nifE gene. This promoter is activated by NifAprotein and is approximately two-times less active than the nifHpromoter, as indicated by the ß-galactosidase assays. This resultsuggests the differential expression of the nif genes and their respec-tive products in Azospirillum.

CorrespondenceL.M.P. Passaglia

Centro de Biotecnologia

Departamento de Genética, UFRGS

Av. Bento Gonçalves, 9500

Prédio 43421

Caixa Postal 15005

91501-970 Porto Alegre, RS

Brasil

E-mail: [email protected]

Research supported by CNPq and

FAPERGS.

Received April 12, 2001

Accepted July 25, 2001

Key words· Azospirillum brasilense· nif genes· nifENX operon· Promoter activity

Introduction

The biological process of nitrogen fixa-tion is catalyzed by the nitrogenase enzyme,an enzyme complex containing the nitroge-nase reductase (Fe-protein) and dinitroge-nase (MoFe-protein). The Fe-protein (NifH)is a dimer of identical subunits, which con-tains a single 4Fe-4S cluster. The MoFe-protein is an a2ß2 tetramer (NifD and NifK)including two iron-molybdenum cofactors(FeMoco) and about 16 additional iron andacid-labile sulfur irons (1). Twenty nif geneshave been described in the extensively stud-ied nitrogen-fixing organism Klebsiella pneu-

moniae, but they are not all essential for N2-fixing activity (2). These genes are locatedon a contiguous 24-kb chromosomal DNAfragment and organized into several operons(2). The observation that the nucleotide se-quence and protein structure of dinitroge-nase and dinitrogenase reductase are con-served among all nitrogen-fixing organismsallowed the identification and cloning ofhomologous sequences from several otherdiazotroph organisms (2-6). However, theorganization of the nif genes has been ob-served to diverge among the different diazo-trophic bacteria. In some cases, the nifHDKgenes are transcribed as a single unit, as

1380


D.P. Potrich et al.

observed in several fast-growing rhizobiaincluding Sinorhizobium meliloti (7) andRhizobium leguminosarum (8) and in thenewly described nif cluster from Glucon-acetobacter diazotrophicus (9). In othercases, the nifHDK operon contains other nifgenes, as reported for Azotobacter vinelan-dii, where they constitute an operon togetherwith nifTY, orf1 and orf2 (3), Herbaspiril-lum seropedicae, where they form a largeoperon composed by nifHDKENXorf1orf2(10), and Azospirillum brasilense, where theyare clustered with orf1 and nifY (4). In othercases, they are separated into two differentregions (nifH and nifDK), as shown in theslow-growing species Bradyrhizobium ja-ponicum (11) and Bradyrhizobium sp (Vigna)(cowpea Bradyrhizobia) (12,13).

In all diazotrophics several other geneshave been reported to be essential for pro-ducing an active nitrogenase. The gene prod-ucts of nifE and nifN are involved in synthe-sis of the FeMoco of nitrogenase and arestructurally related to the products of thenitrogenase genes nifDK (14). The nifENgenes from K. pneumoniae are clustered intoa single operon, together with nifX, down-stream from the nifHDKTY operon (2), andare expressed from the promoter located up-stream of the nifE gene (14,15). In Rhodo-bacter capsulatus, the nifENXorf4orf5nifQoperon contains the nifX and nifQ genes(16). Transposon insertions into the nifE andnifN genes yielded the Nif- phenotype in K.pneumoniae (17) and A. brasilense (18,19).However, transposon insertion in R. capsu-latus nifX showed that its product is notessential for nitrogen fixation (20). Recentresults indicate that the Azotobacter vinelan-dii NifX protein participates in FeMoco syn-thesis in vitro (21). The orf4 of R. capsulatusencodes a ferredoxin-like protein and is ho-mologous to orf3 found in A. vinelandii aspart of the operon nifENXorf3orf4 (3).

Bacteria of the genus Azospirillum arediazotrophs capable of fixing nitrogen infree-living state or associated with roots of

economically important grasses (22). In A.brasilense three nif operons, nifHDKorf1Y,nifENXorf3orf5fdxAnifQ and orf2nifUSVorf4, have been identified and their sequenceshave been determined (4,23, and the presentstudy). The operons encompassing thenifENX and nifUSV genes are separated fromthe structural nifHDKorf1Y operon by about5 and 10 kb, respectively (23). A putative-24/-12 promoter element has been found inthe promoter region of the nifH, nifE andnifU genes. The promoter site of nifH wasstudied in more detail and showed two over-lapping NifA-binding sites, where the ex-amination of activation of the mutant nifHpromoter by NifA revealed that the integrityof the NifA-binding site closer to nifH isrequired for the most efficient activation(24).

In the present study, we have determinedthe nucleotide sequence of the A. brasilenseSp7 nifENX genes, three open reading frames(orf3, orf5 and fdxA) and the nifQ gene,which constitute an operon transcribed froma single promoter upstream of the nifE gene.We have also shown that this promoter isactivated by NifA protein and is less activethan the nifH promoter, as measured by ß-galactosidase promoter fusion.

Material and Methods

Bacterial strains, plasmids and growthconditions

The bacterial strains and plasmids usedin the present study are listed in Table 1. LBmedium (25) was used for growing E. colistrains at 37ºC. A. brasilense Sp7 strain wasgrown at 30ºC in either LB medium or MMAbminimal medium (26) using 0.5% malate ascarbon source. For the ß-galactosidase as-says, A medium was used for growing E. coliMC1061 strain at 28ºC, as indicated bySambrook et al. (25). The medium wassupplemented with the antibiotics tetracy-cline and/or ampicillin when necessary, at

1381


Azospirillum brasilense nifENXorf3orf5fdxAnifQ operon

concentrations of 10 and 100 g/ml, respec-tively.

DNA manipulations and sequencing

Plasmid DNA preparation, restrictionenzyme analysis, transformation and elec-trophoresis on agarose or polyacrylamidegel were performed as described by Sam-brook et al. (25). Restriction endonucleasesand other enzymes were purchased fromPharmacia (Uppsala, Sweden) or Gibco/BRL(Gaithersburg, MD, USA) and used accord-ing to the manufacturer’s instructions.

The nucleotide sequence was determinedby the chain-termination method of Sangeret al. (as described in Ref. 27) using 33P-

ddNTPs and the ThermoSequenase radiola-beled terminator cycle sequencing kit (Amer-sham Pharmacia Biotech, Uppsala, Sweden).Defined restriction fragments from the EcoRI/PstI A. brasilense DNA region shown inFigure 1 were subcloned to generate severalsequencing plasmids (Table 1). The junc-tions of all subclones were checked by se-quencing directly from the larger parentalplasmid, pWY1 (this laboratory), using oli-gonucleotides generated from the sequencesalready obtained. All manual sequencing datawere confirmed by using the Molecular Ge-netics Instrumentation Facility, Universityof Georgia, Atlanta, GA, USA. Analysis ofDNA sequences and comparison with nucleo-tide and deduced protein sequences from

Table 1. Bacterial strains and plasmids used in the present study.

Bacteria Strain Relevant characteristics Reference

Azospirillum brasilense Sp7 ATCC 29145, AmpR 22

Escherichia coli XL1-Blue supE44 hsdR17 recA1 endA1 25gyrA46 thi relA1 lac-F’[proAB+

lacIq lacZD15 Tn10(tetr)]

MC1061 hsdR mcrB araD 139D 28(araABC-leu) 7679DlacX74 galU galK rpsL thi

Plasmid Relevant characteristics Reference

pBluescript AmpR Stratagene

pUC18 AmpR 25

pCK3 pRK290 derivative containing the 29Klebsiella pneumoniae nifA gene

pMC1403 translational fusion lacZ plasmid 28

pMCH pMC1403 derivative containing 24the A. brasilense nifH promoter

pMCE pMC1403 derivative containing the Present paperA. brasilense nifE promoter

pWY1 pUC18 derivative containing a 5.8-kb This laboratoryEcoRI/PstI A. brasilense DNA fragment

pKS2.2 pBSKS+ derivative containing a 2.2-kb Present paperEcoRI/HindIII DNA fragment

pKS0.8 pBSKS+ derivative containing a 0.8-kb Present paperHindIII/SalI DNA fragment

pKS1.5 pBSKS+ derivative containing a Present paper1.5-kb SalI DNA fragment

pKS1.3 pBSKS+ derivative containing a Present paper1.3-kb SalI/PstI DNA fragment

1382


D.P. Potrich et al.

other organisms were performed using theGCG (Wisconsin Package Version 9.0,Genetics Computer Group, Madison, WI,USA) computer programs (licensed toCENARGEN-EMBRAPA, Brasília, DF, Bra-zil).

PCR

The 590-bp DNA fragment containingthe promoter region of the nifE gene wasobtained by PCR with A. brasilense Sp7total DNA as a template. The oligonucleo-tides used for PCR were 5' CGCCGCCAACGACGAGGTCAAGAA 3', which corre-

sponds to the C-terminal region of the mcpAbgene (Schneider C, unpublished results), and5' GGTGGAGCAACCCGGCTCGTT 3',which corresponds to the beginning of thenifE-coding region. The amplified fragmentwas cloned into the HincII site of thepBluescript KS+ vector and completely se-quenced to check the absence of mutations.

Plasmid construction and ß-galactosidaseassays

The strategy used to obtain the pMCEplasmid is shown in Figure 1. DNA from thepBluescript KS+ plasmid carrying the 590-

E E E E E H P P P B E

draTG nifHDKorf1Y mcpAb nifENXorf3orf5fdxAnifQ

orf2nifUSVorf4

nifW fixABC

UAS (upstream activator sequence)nif promoter element

E E E S S P

nifE nifN nifX orf3 orf5 fdxA nifQ

PCR

E

590-bp DNA fragment

Cloning into pBSKS+ (Smal)Digestion with XholMung bean deletionDigestion with BamHI

Blunt end BamHI

Cloning into pMC1403 Sma l/BamHI

Plasmid pMCE (nifE::lacZ fusion)

1 kb

A

B

Figure 1. Physical map of the nifgene cluster in the genome ofAzospirillum brasilense (A) andcloning strategy of the nifE pro-moter region (B). A, Representsthe organization of nif and othergenes in the chromosome of A.brasilense. Arrows represent thepositions and direction of tran-scription of the operons. ThenifENXorf3orf5fdxAnifQ operonis shown in detail in the lowerpart of the figure. Each restric-tion fragment was cloned intothe pUC18 plasmid, as describedin Table 1. B, Cloning strategy ofthe nifE promoter region into thepMC1403 plasmid. E, EcoRI; H,HindIII; P, PstI; B, BamHI; S, SalI.

1383



bp PCR amplified DNA fragment (describedabove) was digested with XhoI (vector site)followed by mung bean deletion to generatea shortened blunt-end fragment. This linearpBluescript recombinant plasmid was fur-ther digested with BamHI and the blunt-end/BamHI fragment was cloned into thepMC1403 SmaI/BamHI-digested vector con-taining the lacZ translational fusion. Therecombinant pMCE was used to monitor theactivity of the nifE promoter by measuring ß-galactosidase activity in E. coli MC1061(28). The K. pneumoniae NifA protein syn-thesized constitutively was provided by thepCK3 plasmid (29). Plasmid pMCH, con-taining a nifH::lacZ fusion (24), was used asa positive control.

The ß-galactosidase activities of strainscarrying the nif-lacZ promoter fusions weredetermined according to the procedure ofMiller et al. (as described in 25) at 28ºC.

Nucleotide sequence accession number

The nucleotide sequence of the nifENXorf3orf5fdxAnifQ operon of A. brasilenseSp7 has been deposited in Genbank underaccession No. AF361867 along with the pre-dicted amino acid sequence.

Results

The nifENXorf3orf5fdxAnifQ genes of A.brasilense Sp7 are in the same transcrip-tional unit and are expressed from a s54

promoter upstream of nifE.Tn5 mutagenesis of the A. brasilense

DNA regions downstream from the nifHDKoperon revealed a region of approximately6000 bp containing the nifENX genes (19).In this region seven complete open readingframes (ORFs) were identified, all reading inthe same direction as that of the nifHDKoperon, as indicated in Figure 1. The assign-ment of genes was based on deduced aminoacid sequence identities to the nif gene prod-ucts of K. pneumoniae and several other

diazotrophic bacteria. The likely initiationcodon for all seven genes or ORFs is an ATGpreceded in each case by a characteristicAG-rich ribosome-binding site.

Only one region within the 6045-bp se-quence displays close similarity to the s54

recognition consensus sequence (15) andoccurs at 45 bp upstream from the transla-tional initiation codon of nifE. We identifiedtwo nif-specific upstream activator se-quences, TGT-N10-ACA, characteristic ofNifA-dependent promoters (30), located 72and 45 bp upstream of the putative s54-dependent promoter, respectively.

The A. brasilense nifE-, nifN- and nifX-coding regions are 1413, 1398 and 471nucleotides long, respectively, and predictpolypeptides of 417 residues correspondingto NifE, 466 residues corresponding to NifN,and 157 residues corresponding to NifX pro-teins. An overlapping coding region wasobserved between nifN and nifX. The pre-dicted amino acid sequences from A. brasi-lense nifE, nifN and nifX genes were com-pared to their counterpart sequences fromother diazotrophic organisms, as shown inAppendix 1 (I, II and III) (see pages 1388-1393). The identity of the deduced A. brasi-lense NifE, NifN and NifX amino acid se-quences is distributed over the entire lengthof corresponding proteins. Conserved cys-teine residues (marked by black dots in Ap-pendix 1) are present in all NifE, NifN andNifX proteins, except for some cysteine resi-dues that were not found in the H. seropedi-cae NifE protein (11; Appendix 1(I)). Thehighest similarity level was found betweenA. brasilense and G. diazotrophicus NifE (9)proteins (60.8%), A. brasilense and B. japo-nicum NifN (31) proteins (50.3%), and A.brasilense and H. seropedicae NifX (10)proteins (55%).

In addition to the nifENX genes, fourother ORFs were identified (Figure 1). Thefirst one, orf3, revealed high similarity withG. diazotrophicus orf1 (59.6%) (9). The con-tiguous ORF, orf5, showed 61.8% similarity

1384


D.P. Potrich et al.

to Azorhizobium caulinodans orf1 (32) and61.2% similarity to Acetobacter diazotro-phicus orf2 (9). The third ORF was homolo-gous to a ferredoxin-like protein from R.capsulatus (54.9% similarity) (16) and aferredoxin III protein from the cyanobacter-ium Plectonema boryanum (57.3% simi-larity) (33) and was assigned to fdxA. Thecomparison between these ORFs and theirhomologous counterparts is shown in Ap-pendix 2 (see pages 1394 and 1395).

The last ORF identified within the se-quenced region of A. brasilense encoded aprotein of 196 amino acids and the align-ment of the deduced amino acid sequence ofthis A. brasilense ORF with that of NifQproteins from K. pneumoniae (34), R. capsu-latus (16), Acetobacter diazotrophicus (9),Enterobacter agglomerans (35), Azotobactervinelandii (4) and Rhizobium sp (36) re-vealed an overall ranging of homology from30.5 to 37.5% of similarity (Appendix 1(IV)).The identity between the NifQ proteins wasmainly restricted to the C-terminal part, in-cluding a typical cysteine motif (marked inAppendix 1) found in all NifQ proteins iden-tified to date.

Inverted repeat structures were detectedonly in two regions at 110 and 173 nucleo-

tides from the nifQ stop codon. MessengerRNA transcribed from these regions couldpotentially form a characteristic stem-and-loop secondary structure and may be in-volved in the termination of transcription.No other ORF was found in the region be-tween the end of the nifQ gene and thebeginning of the orf2nifUSVorf4 operon.

The nifE promoter activity is dependent onNifA protein

To verify the functionality of the putativenif promoter identified in the region up-stream of the nifE initiation codon we ampli-fied by PCR a 590-bp DNA fragment con-taining the regulatory region of the nifENXorf3orf5fdxAnifQ operon. This fragment wascloned into the pMC1403 translational fu-sion plasmid and the recombinant pMCEwas used to monitor the activity of the nifEpromoter by measuring ß-galactosidase ac-tivity in E. coli MC1061 (as described inMaterial and Methods). As shown in Figure2, the ß-galactosidase activity driven by theA. brasilense nifE promoter was only de-tected if the NifA protein was provided intrans via the pCK3 plasmid that promotesconstitutive expression of the K. pneumo-niae NifA. In fact, E. coli MC1061 harbor-ing pMCE, but not pCK3, showed only back-ground levels of ß-galactosidase activity (Fig-ure 2). These results demonstrate that theactivity of the nifE promoter is dependent onNifA.

We also used the pMCH plasmid (24) asa positive control and we were able to com-pare the activities of both A. brasilense nifpromoters. The nifE promoter is approxi-mately two times less active than the nifHpromoter, as indicated by the ß-galactosi-dase assays (Figure 2, compare E + NifAwith H + NifA).

Discussion

Nitrogen fixation genes, nif genes, are

Figure 2. In vivo expression ofnif’-’lacZ from the nifE (E) andnifH (H) promoters in the E. coliMC1061 strain in the presence(+ NifA) or absence of the NifAactivator protein provided bypCK3 plasmid. Expression ofnif’-’lacZ fusion was measuredas a function of the OD600 forcultures grown in A medium(see Material and Methods). Ineach case, three data points (be-tween OD600 of 0.2 and 0.6)were used to determine theslope of the line, which reflectsthe differential rate of nif’-’lacZexpression. Curve labeled MC(which overlaps the curve fromthe nifH promoter in the ab-sence of NifA) gives data for theE. coli MC1061 host strain(negative control).

ß-G

alac

tosi

dase

act

ivity

(U

/ml)

250

200

150

100

50

00.2 0.4 0.6

OD600

H + NifA

E + NifA

EMC/H

1385



frequently clustered into different transcrip-tional units. The overall organization of nifoperons also shows some conservationamong the genomes of diazotrophic bacte-ria. In A. brasilense the distribution of thenifENX genes within the operon resemblesthe organization described for other diazo-trophic organisms. NifE and NifN proteinswere found to be similar to those present inother a-group Proteobacteria. NifX was moresimilar to the gene product of H. seropedi-cae, a member of the ß-group of Proteobac-teria, than the product of the g-group mem-bers. However, NifQ was similar to the NifQfrom Azotobacter vinelandii which belongsto the g-group of Proteobacteria.

In some members of the Proteobacteriathe nif genes, initially described in K. pneu-moniae, are organized in operons togetherwith different ORFs. In A. brasilense theorganization of nifENXorf3orf5fdxAnifQ issimilar to that found in G. diazotrophicus. Infact, the organization of the nif gene in G.diazotrophicus and A. brasilense seems tobe highly similar: the nifENXorf3orf5fdxAnifQ operon containing orthologous pro-teins similarly organized, the presence of themcpA gene in the surroundings of the nifcluster and the organization of the orf2nifUSVoperon may indicate that both microorgan-isms share common characteristics, espe-cially concerning their ability to enhanceplant growth through the transfer of bacteri-ally fixed nitrogen and the production ofplant growth-stimulating factor(s) (9,22). Inaddition, A. brasilense ORF3 was found tobe similar to ORF1 from H. seropedicaepresent within the related nifENXorf1orf2gene cluster (10) and to ORF3 from A. vine-landii also located downstream to the nifXgene (3).

Several studies have demonstrated thatnifEN are essential for the nitrogen fixationprocess. However, determination of the roleof NifX during nitrogen fixation shows some

differences concerning the diazotrophic bac-teria. Araújo et al. (19), using insertionalmutagenesis, obtained five Tn5 insertions inthe region adjacent to the nifHDK genes ofthe A. brasilense genome. Four of them werelocated in the region corresponding to thenifEN genes and gave a Nif- phenotype. Thefifth insertion which gave a Nif+ phenotypewas within the corresponding region of theA. brasilense nifX gene, as further confirmedby sequencing analysis (present paper), andlike the nifX gene from the ß-group memberH. seropedicae it proved not to be essentialfor nitrogen fixation (10). In contrast, NifXwas shown to be essential for nitrogenaseactivity in an in vitro system in A. vinelandii(21), although under laboratory conditions,A. vinelandii nifX showed a Nif+ phenotype(3). To date, the nifX mutant in all diazotrophsdescribed has wild-type nitrogenase activity(3,10,20).

Transcriptional and translational organi-zation of the nif gene cluster of A. brasilenserevealed conserved features. As found forthe other A. brasilense nif/fix operons, se-quence analysis of the nifENXorf3orf5fdxAnifQ revealed s54 and NifA-binding sitesupstream of nifE, which are required, respec-tively, for nif promoter recognition and for nifgene transcriptional activation. Similarly toother nif genes, the A. brasilense nifE pro-moter is positively controlled by the activatorNifA protein.

The nucleotide sequence and promoteranalysis of the nifENXorf3orf5fdxAnifQ op-eron in A. brasilense revealed the presenceof typical features in the deduced proteincommon among the related proteins in otherorganisms, as well as sequences upstream ofnifE indicating a NifA-dependent transcrip-tional activation. Moreover, ORFs similar toorf3, orf5, and the putative fdxA were alsodescribed in other groups of the Proteo-bacteria.

1386


D.P. Potrich et al.

References

1. Orme-Johnson WH (1985). Molecular ba-sis of nitrogen fixation. Annual Review ofBiophysics and Biophysical Chemistry, 14:419-459.

2. Arnold W, Rump A, Klipp W, Priefer O &Pühler A (1988). Nucleotide sequence ofa 24,206 base-pair DNA fragment carryingthe entire nitrogen fixation gene clusterof Klebsiella pneumoniae. Journal of Mo-lecular Biology, 203: 715-738.

3. Jacobson MR, Brigle KE, Bennett LT,Setterquist RA, Wilson MS, Cash VL,Beynon JL, Newton WE & Dean DR(1989). Physical and genetic map of themajor nif gene cluster from Azotobactervinelandii. Journal of Bacteriology, 171:1017-1027.

4. Passaglia LMP, Nunes CP, Zaha A &Schrank IS (1991). The nifHDK operon inthe free-living nitrogen-fixing bacteria Azo-spirillum brasilense sequentially com-prises genes H,D,K, an 353 bp orf andgene Y. Brazilian Journal of Medical andBiological Research, 24: 649-675.

5. Machado IMP, Yates MG, Machado HB,Souza EM & Pedrosa FO (1996). Cloningand sequencing of the nitrogenase struc-tural genes nifHDK of Herbaspirillumseropedicae. Brazilian Journal of Medicaland Biological Research, 29: 1599-1602.

6. Kessler PS & Blank CLJA (1998). The nifgene operon of the methalogenic ar-chaeon Methanococcus maripaludis.Journal of Bacteriology, 180: 1504-1511.

7. Corbin D, Barran L & Ditta GS (1983).Organization and expression of Rhizobiummeliloti nitrogen fixation genes. Proceed-ings of the National Academy of Sciences,USA, 80: 3005-3009.

8. Krol AJM, Hontelez JGJ, Roozendaal B &Kammen AV (1982). On the operon struc-ture of the nitrogenase genes of Rhizo-bium leguminosarum and Azotobactervinelandii. Nucleic Acids Research, 10:4147-4157.

9. Lee S, Reth A, Meletzus D, Sevilla M &Kennedy C (2000). Characterization of amajor cluster of nif, fix, and associatedgenes in a sugarcane endophyte, Aceto-bacter diazotrophicus. Journal of Bacteri-ology, 182: 7088-7091.

10. Klassen G, Pedrosa FO, Souza EM, YatesMG & Rigo LU (1999). Sequencing andfunctional analysis of the nifENXorf1orf2gene cluster of Herbaspirillum seropedi-cae. FEMS Microbiology Letters, 181:165-170.

11. Fischer HM & Hennecke H (1984). Link-

age map of the Rhizobium japonicum nifHand nifDK operons encoding the polypep-tides of the nitrogenase enzyme complex.Molecular and General Genetics, 196:537-540.

12. Jagadish M, Yun AC, Noti JD, Folkerts O& Szalay AA (1985). Structural and func-tional organization of the nif region in theslow-growing, broad host-range cowpeaRhizobium strain IRc78. In: Szalay A &Legocki RP (Editors), Advances in Molec-ular Genetics of Bacterial-Plant Interac-tions. Cornell University Publishers,Ithaca, NY, 27-31.

13. Yun AC & Szalay AA (1984). Structuralgenes of dinitrogenase and dinitrogenasereductase are transcribed from two sepa-rate promoters in the broad host rangecowpea Rhizobium strain IRc78. Proceed-ings of the National Academy of Sciences,USA, 81: 7358-7362.

14. Brigle KE, Weiss MC, Newton WE & DeanDR (1987). Products of the iron-molybde-num cofactor-specific biosynthetic genes,nifE and nifN, are structurally homologousto the products of the nitrogenase molyb-denum-iron protein genes, nifD and nifK.Journal of Bacteriology, 169: 1547-1553.

15. Beynon JL, Cannon MC, Buchanan-Wollaston AV & Cannon FC (1983). Thenif promoters of Klebsiella pneumoniaehave a characteristic primary structure.Cell, 34: 665-671.

16. Moreno-Vivian C, Hennecke S, Pühler A &Klipp W (1989). Open reading frame 5(ORF5), encoding a ferredoxin-like pro-tein, and nifQ are cotranscribed with nifE,nifN, nifX, and ORF4 in Rhodobactercapsulatus. Journal of Bacteriology, 171:2591-2598.

17. Roberts GP, MacNeil T, MacNeil D & BrillWJ (1978). Regulation and characteriza-tion of protein products coded by the nif(nitrogen fixation) genes of Klebsiellapneumoniae. Journal of Bacteriology, 136:267-279.

18. Galimand M, Perroud B, Delorme F,Paquelin A, Vieille C, Bozouklian H &Elmerich C (1989). Identification of DNAregions homologous to nitrogen fixationgenes nifE, nifUS and fixABC in Azospiril-lum brasilense Sp7. Journal of GeneralMicrobiology, 135: 1-13.

19. Araújo EF, Zaha A, Schrank IS & SantosDS (1988). Characterization of DNA seg-ments adjacent to the nifHDK genes ofAzospirillum brasilense Sp7 by Tn5 site-directed mutagenesis. In: Klingmüller W

(Editor), Azospirillum IV: Genetics, Physi-ology and Ecology. Proceedings of theFourth Bayreuth Azospirillum Workshop,Bayreuth, Germany, 16-25.

20. Moreno-Vivian C, Schmehl M, MasepohlB, Arnold W & Klipp W (1989). DNA se-quence and genetic analysis of the Rho-dobacter capsulatus nifENX gene region:homology between NifX and NifB sug-gests involvement of NifX in processingof the iron-molybdenum cofactor. Molec-ular and General Genetics, 216: 353-363.

21. Shah VK, Rangaraj P, Ranjini C & AllenRM (1999). Requirement of NifX and othernif proteins for in vitro biosynthesis of theiron-molybdenum cofactor of nitrogenase.Journal of Bacteriology, 181: 2797-2801.

22. Tarrand JJ, Krieg NR & Döbereiner J(1978). A taxonomic study of the Spirillumlipoferum group, with descriptions of anew genus, Azospirillum gen. nov. andtwo species, Azospirillum lipoferum(Beijerinck) comb. nov. and Azospirillumbrasilense sp nov. Canadian Journal ofMicrobiology, 24: 967-980.

23. Frazzon J & Schrank IS (1998). Sequenc-ing and complementation analysis of thenifUSV genes from Azospirillum brasi-lense. FEMS Microbiology Letters, 159:151-158.

24. Passaglia LMP, Schrank A & Schrank IS(1995). The two overlapping Azospirillumbrasilense upstream activator sequenceshave differential effects on nifH promoteractivity. Canadian Journal of Microbiology,41: 849-854.

25. Sambrook J, Fritsch EF & Maniatis T(1989). Molecular Cloning: A LaboratoryManual, 2. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY, USA.

26. Vanstockem M, Michiels KW, Vanderley-den J & Gool APV (1987). Transposonmutagenesis of Azospirillum brasilenseand Azospirillum lipoferum: physical anal-ysis of Tn5 and Tn5-Mob insertion mu-tants. Applied and Environmental Micro-biology, 53: 410-415.

27. Sanger F, Nicklen S & Coulson AR (1977).DNA sequencing with chain-terminatinginhibitors. Proceedings of the NationalAcademy of Sciences, USA, 74: 5463-5467.

28. Casadaban MJ, Martinez-Arias A, ShapiraS & Chou JH (1983). ß-galactosidase genefusions for analysing gene expression inEscherichia coli and yeast. Methods inEnzymology, 100: 293-308.

29. Kennedy C & Drummond MH (1985). Use

1387



of cloned nif regulatory elements fromKlebsiella pneumoniae to examine nifregulation in Azotobacter vinelandii. Jour-nal of General Microbiology, 131: 1787-1795.

30. Buck M, Miller S, Drummond MH & DixonRA (1986). Upstream activator sequencesare present in the promoters of nitrogenfixation genes. Nature, 320: 374-378.

31. Aguilar OM, Taormino JP, Thöny B,Ramseier T, Hennecke H & Szalay AA(1990). The nifEN genes participating inFeMo cofactor biosynthesis and genesencoding dinitrogenase are part of thesame operon in Bradyrhizobium species.Molecular and General Genetics, 224:413-420.

32. Arigoni F, Kaminski PA, Hennecke H &Elmerich C (1991). Nucleotide sequenceof the fixABC region of Azorhizobiumcaulinodans ORS571: similarity of the fixBproduct with eucaryotic flavoproteins,characterization of fixX, and identificationof nifW. Molecular and General Genetics,225: 514-520.

33. Schrautemeier B, Neveling U & Schmitz S(1994). Characterization of the genomeregion encoding an fdxH-type ferredoxinand a new 2[4Fe-4S] ferredoxin from thenonheterocystous, nitrogen-fixing cyano-bacterium Plectonema boryanum PCC73110. Journal of Bacteriology, 176: 1037-1046.

34. Buikema WJ, Klingensmith JA, GibbonsSL & Ausubel FM (1987). Conservation ofstructure and location of Rhizobium meli-loti and Klebsiella pneumoniae nifBgenes. Journal of Bacteriology, 169: 1120-1126.

35. Siddavattan D, Singh M & Klingmüller W(1993). Structure of the nifQ gene fromEnterobacter agglomerans 333 and itsoverexpression in Escherichia coli. Mo-lecular and General Genetics, 239: 435-440.

36. Freiberg C, Fellay R, Bairochi A, Brough-ton WJ, Rosenthal A & Perret R (1997).Molecular basis or symbiosis betweenRhizobium and legumes. Nature, 387:394-401.

37. Setterquist RA, Brigle KE, Beynon JL,Cannon MC, Ally A, Cannon FC & DeanDR (1988). Nucleotide sequence of thenifE and nifN genes from Klebsiella pneu-moniae. Nucleic Acids Research, 16:5215.

38. Beynon JL, Cannon MC, Buchanan-Wol-laston AV, Ally A, Setterquist RA, DeanDR & Cannon FC (1988). The nucleotidesequence of the nifT, nifY, nifX and nifWgenes of K. pneumoniae. Nucleic AcidsResearch, 16: 9860.

39. Nicholas KB & Nicholas Jr HB (1997).GeneDoc: Analysis and visualization ofgenetic variation. Available at http://www.cris.com/~Ketchup/genedoc.shtml.Accessed March 2001.

40. Schrautemeier B, Neveling U & Schmitz S(1995). Distinct and differently regulatedMo-dependent nitrogen-fixing systemsevolved for heterocysts and vegetativecells of Anabaena variabilis ATCC 29413:characterization of the fdxH1/2 gene re-gions as part of the nif1/2 gene clusters.Molecular Microbiology, 18: 357-369.

1388


D.P. Potrich et al.

Appendix 1(I) continued on next page

APPENDIX 1(I)APPENDIX 1(I)APPENDIX 1(I)APPENDIX 1(I)APPENDIX 1(I)

Comparison of the predicted amino acid sequences of the Azospirillum brasilense (ab) NifE (I), NifN (II), NifX (III) andNifQ (IV) proteins with analogous gene products from Azotobacter vinelandii (av) (4), Klebsiella pneumoniae (kp)(37,38), Gluconacetobacter diazotrophicus (ad) (9), Rhodobacter capsulatus (rc) (16), Bradyrhizobium japonicum (bj)(31), Herbaspirillum seropedicae (hs) (10), Rhizobium sp (rsp) (36), and Enterobacter agglomerans (ea) (35). A blackbackground indicates conserved residues in all aligned sequences, dark grey indicates conserved residues in at least 80%of the aligned sequences, and light grey indicates conserved residues in at least 60% of the aligned sequences. Conservedcysteine residues are indicated by black dots. Multiple alignment was done using the PILEUP program, University ofWisconsin Genetics Computer Group. The alignment editing was done using the GeneDoc program and the Dayhoff PAM250 score table (39).

Appendix 1

1389



Appendix 1(I) continued

Appendix 1 continued on next page

1390


D.P. Potrich et al.

Appendix 1(II) continued on next page

APPENDIX 1(II)APPENDIX 1(II)APPENDIX 1(II)APPENDIX 1(II)APPENDIX 1(II)

1391



Appendix 1(II) continued


1392


D.P. Potrich et al.

APPENDIX 1(IV)APPENDIX 1(IV)APPENDIX 1(IV)APPENDIX 1(IV)APPENDIX 1(IV)

APPENDIX 1(III)APPENDIX 1(III)APPENDIX 1(III)APPENDIX 1(III)APPENDIX 1(III)

Appendix 1(IV) continued on next page

1393



Appendix 1(IV) continued

1394


D.P. Potrich et al.

APPENDIX 2(I)APPENDIX 2(I)APPENDIX 2(I)APPENDIX 2(I)APPENDIX 2(I)

Comparison of the predicted amino acid sequences of the Azospirillum brasilense (ab) Orf3 (I), Orf5 (II) and FdxA (III)proteins with analogous gene products from Azotobacter vinelandii (av) (3), Anabaena variabilis (anb) (40), Plectonemaboryanum (pb) (33), Gluconacetobacter diazotrophicus (ad) (9), Rhodobacter capsulatus (rc) (16), Azorhizobiumcaulinodans (ac) (32) Rhizobium sp NGR234 (rsp) (36), and Herbaspirillum seropedicae (hs) (10). A black backgroundindicates conserved residues in all aligned sequences, dark grey indicates conserved residues in at least 80% of the alignedsequences, and light grey indicates conserved residues in at least 60% of the aligned sequences. Multiple alignment wasdone using the PILEUP program, University of Wisconsin Genetics Computer Group. The alignment editing was doneusing the GeneDoc program and the Dayhoff PAM 250 score table (39).

Appendix 2


1395



APPENDIX 2(III)APPENDIX 2(III)APPENDIX 2(III)APPENDIX 2(III)APPENDIX 2(III)

APPENDIX 2(II)APPENDIX 2(II)APPENDIX 2(II)APPENDIX 2(II)APPENDIX 2(II)

sequencing and promoter analysis of the nifenxorf3orf5fdxanifq

Documents