Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. 2003 Jun 24; 100(13): 7883–7888.
Published online 2003 Jun 13. doi:  10.1073/pnas.1230487100
PMCID: PMC164682

Gene function analysis in environmental isolates: The nif regulon of the strict iron oxidizing bacterium Leptospirillum ferrooxidans


A random genomic library from an environmental isolate of the Gram-negative bacterium Leptospirillum ferrooxidans has been printed on a microarray. Gene expression analysis was carried out with total RNA extracted from L. ferrooxidans cultures in the presence or absence of ammonium as nitrogen source under aerobic conditions. Although practically nothing is known about the genome sequence of this bacterium, this approach allowed us the selection and sequencing of only those clones bearing genes that showed an altered expression pattern. By sequence comparison, we have identified most of the genes of nitrogen fixation regulon in L. ferrooxidans, like the nifHDKENX operon, encoding the structural components of Mo-Fe nitrogenase; nifSU-hesB-hscBA-fdx operon, for Fe-S cluster assembly; the amtB gene (ammonium transporter); modA (molybdenum ABC type transporter); some regulatory genes like ntrC, nifA (the specific activator of nif genes); or two glnB-like genes (encoding the PII regulatory protein). Our results show that shotgun DNA microarrays are very powerful tools to accomplish gene expression studies with environmental bacteria whose genome sequence is still unknown, avoiding the time and effort necessary for whole genome sequencing projects.

Keywords: shotgun DNA microarrays, gene expression, nitrogen fixation, iron oxidizer

Gene expression analysis by DNA microarrays (biochips) has enjoyed a tremendous increase during the last few years (1). As whole genome sequences are available at increasing rates, DNA microarrays are used for the functional analysis of the newly and already characterized genes. A previous knowledge of DNA sequence usually is a prerequisite for biochip construction from already known genes. It is possible to spot all ORFs of complete genomes from bacteria to humans. This strategy requires one to undertake expensive and time-consuming genome sequencing projects to study the function of just a particular set of genes. This is a genuinely tedious task if we want to apply this approach to those bacterial strains isolated from the environment. Here we use a shotgun DNA microarray strategy (2) for gene expression analysis in environmental isolates, bypassing the need for previous knowledge of the genome sequence.

The Gram-negative bacterium Leptospirillum ferrooxidans is one of the most abundant bacteria in mineral-processing bioreactors, as well as in certain natural environments like the extremely acidic Tinto River (in the southwestern region of Spain), which has pH values between 0.8 and 2.4 and high metal concentrations along the entire 90 km of its length (3, 4). L. ferrooxidans is a strict iron-oxidizing acidophile considered one of the main responsible agents for maintaining the pH balance and hence the physicochemical properties of the ecosystem. This process is achieved by accelerating the limiting step for pyrite (FeS2) leaching: the production of ferric iron (Fe3+) (5, 6). L. ferrooxidans is also very important because of its capacity to extract heavy metals from minerals or contaminated soils, and because it is directly involved in acid mine drainage (7). In addition, L. ferrooxidans is of great interest in astrobiology, because its metabolism could help us understand some relevant aspects of the origin and evolution of life on Earth (8). Its nutrient requirements are very simple: CO2 (carbon source), O2 (respiration), equation M1 (nitrogen source), minerals like pyrite to obtain energy, and some additional salts. Diazotrophic growth (nitrogen fixation) of L. ferrooxidans was inferred from an increase in iron oxidation in ammonium-free medium under limited oxygen (9). Despite all these features, very little is known about L. ferrooxidans metabolism. Here we study a very important aspect of L. ferrooxidans physiology, nitrogen source utilization, by a shotgun DNA microarray strategy.

Materials and Methods

Strains, Growth Conditions, and Plasmids. L. ferrooxidans L3.2, a natural isolate from the Tinto River (kindly provided by E. González-Toril and R. Amils, Centro de Biología Molecular “Severo Ochoa,” Madrid), was cultivated in Mackintosh medium (10) in the absence of an added nitrogen source at 30°C with 160 rpm of agitation (aerobic conditions) in an orbital incubator (Novotron HT, Bottmingen, Switzerland). Cells were recovered by filtration through 0.22-μm filters (Millipore). SmaI-digested pENTR1A (Invitrogen) was used as the cloning vector for L. ferrooxidans L3.2 gene library construction in Escherichia coli DH5α (Invitrogen). DNA manipulation was carried out as described (11).

DNA Library Construction. L. ferrooxidans L3.2 chromosomal DNA was digested with the blunt end generating restriction enzymes AluI, HaeIII, and RsaI at the same time, and agarose gel-purified fragments from 1 to 5 kb were ligated to SmaI or EcoRV-DraI-digested plasmid pENTR1A. The ligation mixture was used to transform E. coli DH5α by following the manufacturer's instructions. The number of clones needed to obtain a four to five redundancy gene library was estimated by the formula n = ln(1-P)/ln(1-I/GS) (11), where I is the insert site (kbp) and GS is the genome size (kbp). More than 6,000 colonies were plated; among them >90% carried a DNA insert as checked by minipreps.

DNA Microarray Construction. Colonies were cultivated, processed, and PCR amplified as described (12). Two oligonucleotide pairs were used: ccdF 5′(C6aminoGAGAGAGCCGTTATCGTCTGTTTGTGG)3′ and ccdR 5′(C6aminoATATGCACCACCGGGTAAAGTTCACGG)3′ for SmaI digested vector; and pen1 5′(C6aminoCTGTTAGTTAGTTACTTAAGCTCGGGC)3′ and pen2 5′(C6aminoTAACATCAGAGATTTTGAGACACGGGC)3′ for EcoRV-DraI digested plasmid. The L. ferrooxidans L3.2 genome size was estimated to be ≈2.5 Mbp (E. González-Toril and R. Amils, personal communication), so the total clone number needed for a three-to-four-times genome redundancy, with 2.3-kbp average DNA fragment size, was estimated to be ≈4,300. More than 90% of the clones had a DNA insert between 1 and 5 kbp, but PCR efficiency was ≈85%, so we PCR amplified >5,300 clones to obtain the required clone number. PCR products were checked by agarose gel electrophoresis in a mini-ready-to-run system (Amersham Biosciences) and purified by using a 100-μl PCR purification kit (TeleChem International, Sunnyvale, CA). Spotting was carried out by duplicate spots on silylated slides (TeleChem International) with a 417 Arrayer (Affymetrix, Santa Clara, CA) at 45–50% humidity, and slides were processed by following the supplier's recommendations. A total array containing 10,752 spots (including a replica) was constructed with controls located on different parts of the chip. Some of these control spots included PCR-amplified DNA fragments codifying the 16S and part of the 23S rRNA of L. ferrooxidans L3.2 as positive controls and herring sperm DNA as a negative control. Previous experiments allowed us to identify by similarity several L. ferrooxidans L3.2 genes involved in primary metabolism, like those encoding the isocitrate dehydrogenase and the elongation factors EF-G and EF-Tu. The whole PCR-amplified DNA fragments containing these genes were also spotted at different concentrations (from 250 to 50 ng/μl) in different places on the chip and used as internal markers for primary metabolism.

RNA Extraction, Labeling, and Chip Hybridization. Total RNA from L. ferrooxidans L3.2 cultures was extracted by acid phenol (13) or with a Qiagen (Valencia, CA) RNA extraction kit. RNA was treated with DNaseI, RNase-free, followed by phenol and chloroform extraction or treated directly on the same column during extraction by the Qiagen DNaseI kit. Total RNA quality was checked by spectrometry (BioPhotometer, Eppendorf) and by Agilent Technologies (Van Nuys, CA) Bioanalyzer 2100 and labeled by cDNA synthesis with reverse transcriptase (SuperScriptII, Invitrogen) and Cy3- or Cy5-dUTP (Amersham Bio-sciences) primed with random hexamers as described (http://cmgm.stanford.edu/pbrown/protocols/4_Ecoli_RNA.txt). Chip hybridization (55°C from 6 to 12 h) and wash were carried out as described (www.arrayit.com).

Scanning and Data Analysis. Slides were scanned for Cy3 and Cy5 in a 418 Array Scanner (Affymetrix). Images were overlapped and analyzed with SCANALYZE2 software (http://rana.lbl.gov/EisenSoftware.htm).

DNA Sequencing and Analysis. Sequence reactions were made from plasmid minipreps using dye terminator cycle sequencing reactions (14) and were run in an ABI Prism 377 Sequencer (Applied Biosystems). Sequences were analyzed by using DNASTAR (Madison, WI) package software (LASERGENE), which permits one to run the FASTA programs to search for similarities on the nonredundant protein and DNA sequence database at the National Center for Biotechnology Information (National Institutes of Health, Bethesda).


L. ferrooxidans Growth Under Ammonium Starvation. L. ferrooxidans L3.2 growth was monitored by colorimetric detection of iron oxidation as well as cell counting (Fig. 1). Under standard culture conditions (2% wt/vol of iron and 2.42 mM ammonium sulfate), the cell number increased exponentially from 0.01 to ≈1.5 OD at 410 nm, although more slowly in the absence of ammonium, where ferric iron (Fe3+) accumulated in the culture while the cell number was less than five times lower than that of ammonium rich cultures (Fig. 1B). This fact indicated a different growth rate due to the change from an easily assimilating nitrogen source, like ammonium, to a more energy-requiring one, like atmospheric nitrogen (N2). All these cultures were subjected to aerobic conditions, indicating that a limited oxygen concentration was not necessary for diazotrophic growth of L. ferrooxidans L3.2, as shown for other strains (9).

Fig. 1.
L. ferrooxidans growth curves under ammonium-rich (squares) and ammonium-starved (circles) conditions. Arrows indicate the points at which total RNA was extracted.

Ammonium Starvation-Induced nif Genes Under Aerobic Conditions. To identify those genes involved in nitrogen metabolism, 2 μg of total RNA from two exponentially growing L. ferrooxidans L3.2 cultures (Fig. 1B), with and without added ammonium sulfate, were labeled with different fluorochromes (Cy3 for starved and Cy5 for ammonium-containing cultures), mixed, and hybridized with an L. ferrooxidans L3.2 shotgun DNA microarray. Image analysis (Fig. 2) allowed us to select ≈80 spots having a ratio ≥1.8 when RNA came from ammonium-starved cultures. Images were normalized by using positive and primary metabolism indicators, and, as expected for two cultures in the same growth phase, the vast majority of the spots showed a yellow color (ratio ≈1); that is, there was no transcription induction in either nutritional condition. Similar results were obtained when the same RNA samples were cross-labeled with the other fluorochrome (Cy5 for starved and Cy3 for ammonium, Fig. 3). The experiment was repeated twice, and a total of 76 different spots with an average ratio between the two cross-labeled experiments of >1.8 were selected. Sequencing of corresponding clones was achieved in both senses by using the same oligonucleotides as for PCR amplification. DNA sequences were compared with databases by using blast programs, and those sequences with a significant score were annotated (Table 1, which is published as supporting information on the PNAS web site, www.pnas.org). We identified by similarity several genes related to nitrogen fixation (for a review, see refs. 1517) in other bacteria. Some of these genes are: the whole nifHDKENX operon, encoding the structural components of the molybdenum–iron (Mo-Fe) nitrogenase; another putative operon composed of nifS-nifU-hesB-hscB-hscA-fdx-orf1-orf2-mrp, very similar both in deduced amino acid sequence and organization to the E. coli iscS-iscU-iscA-hscB-hscA-fdx-ORF3 gene cluster (18, 19) involved in the assembly of Fe-S clusters and also codifying chaperons and factors involved in Mo-Fe nitrogenase synthesis; a glnB-like encoding factor PII (20) for nitrogen uptake regulation [>60% identity to PII protein (GlnB) of Rhodobacter sphaeroides] is located downstream of the ammonium transporter (amtB); a second glnB-like gene is located upstream but independently of another operon composed of a two-component response regulator, similar to nasT (21), and a putative nifV gene; the periplasmic molybdatebinding protein (ModA, >40% identity to the corresponding one of Helicobacter pylori) is located downstream of the nasT-lfe112g3-nifV putative operon; several putative two-component response regulatory systems; genes involved in chemotaxis, like an osmosensitive K+ channel His kinase sensor domain (KdpD2) and the ATPase domain (KdpA) of the two-component sensor KdpD (22, 23); a probable copper-transporting ATPase similar to FixI of Rhizobium leguminosarum, which is involved in symbiotic nitrogen fixation (24); several ABC-type transport systems, as well as cation-efflux systems; and, finally, several new genes encoding unknown hypothetical proteins that somehow must be related to nitrogen metabolism, because their transcription is neatly induced under ammonium starvation. One of the latter, and among the most induced, was highly similar to hypothetical proteins of other bacteria and to a putative nitrogen fixation-positive activator from Synechocystis sp. (GenBank accession no. S74707). We designated this L. ferrooxidans gene as nfaP, for nitrogen fixation activator protein.

Fig. 2.
L. ferrooxidans DNA microarray. (A) Differential gene expression after hybridization with a labeled cDNA sample mix prepared from RNA isolated from ammonium-rich (Cy5) or ammonium-starved (Cy3) culture. Spots correspond to DNA fragments preferentially ...
Fig. 3.
Identification of several operons involved in nitrogen fixation in a cross-labeled experiment. Green (Cy3-labeled) and red (Cy5-labeled) spots correspond to two different hybridized cDNAs from the same total RNA sample extracted from ammonium-starved ...

Experiments under ammonium starvation in aerobic conditions failed to detect the positive regulator of nif genes, nifA, or other regulators involved in nitrogen metabolism like ntrC, suggesting that there were no differences in transcription levels of these genes or they were below the threshold for selection (at least a 1.8-fold increase). However, nifA and ntrC genes were selected in two different DNA fragments that showed an increase in transcription level when L. ferrooxidans L3.2 was subjected to energetic stress by iron limitation (unpublished results). The corresponding spots were checked in the ammonium experiment chips, and indeed there were no differences in the transcription level between starved and enriched ammonium cultures (1.3 and 1.4 ratios for nifA- and ntrC-containing fragments, respectively).

Efficiency of Shotgun DNA Microarray. The main limitation of shotgun DNA microarrays is that the spotted DNA fragment may correspond to more than one unrelated transcript. In those cases, it would be necessary to recheck each gene with specific probes to identify those whose expression was truly altered. Our results indicated that shotgun DNA microarrays are very powerful tools for both gene expression and gene discovery, not only because we have identified a large number of known genes that are or might be involved in the processes under study by just sequencing a small number of clones, but also because most of these clones overlap in contiguous fragments (contigs) (Fig. 3). This fact is strong evidence in favor of this strategy, because (i) in a shotgun sequencing project, it would be highly improbable to get contigs with such a small number of sequences; (ii) some of the clones are small enough to contain only one ORF, or as many as two or three related ones; and (iii) some of the clones are long enough to contain complete operons, which can also aid in joining and overlapping smaller DNA fragments. Moreover, the use of only one or two pairs of primers for PCR library amplification to construct the microarray supposes a significant cost advantage.

Identification of nif Regulatory Sequences. Upstream sequences of the identified genes and operons can be analyzed to search for similarities and special features that could assist in assigning gene function. This is the case for the operons nifHDKENX and nasT-lfe112g3-nifV and nfaP gene (Fig. 4), where the upstream regions are specially A+T rich (≈60%) with respect to coding regions (≈47.5%). They also contain conserved -24/-12 consensus sequences for σ54-dependent promoters (25) and two or three putative NifA-binding sites (26, 27). Moreover, in the nifH upstream region, there are two more interesting features that stress the relevance of its regulatory function: an inverted repeat comprising the ATG translation start codon and a DNA region with high bending index between the two NifA-binding sites (data not shown).

Fig. 4.
Regulatory region analysis of some genes and operons whose transcription is induced under nitrogen-fixing conditions. These regions contain regulatory signals characteristic of nitrogen fixation (nif) genes: NifA upstream activating sequences (UAS); ...


For those bacteria capable of fixing the atmospheric nitrogen, PII proteins play a significant role in the regulation of this process (for a review, see refs. 20 and 2830). This is done by controlling the transcription of, among others, the nif genes encoding the enzymatic system (nitrogenase) for dinitrogen reduction. Such control is mediated by regulating either the expression or the activity of NifA in response to oxygen and fixed nitrogen. For free-living diazotrophs, the availability of fixed nitrogen, rather than oxygen availability, for controlling nifA expression is the critical factor. However, even inside this group, regulation of nifA transcription varies considerably, and both NtrC and GlnB are involved.

Gene expression analysis using our L. ferrooxidans L3.2 shotgun DNA microarray allowed us to identify most of the genes, up to the present time, that have been proven or assumed to be involved in nitrogen fixation in bacteria (15). Our results confirm L. ferrooxidans L3.2 as a free-living diazotroph. Fig. 5 shows a picture of the L. ferrooxidans nif regulon under oxygenic conditions.

Fig. 5.
L. ferrooxidans nitrogen fixation regulon. Green arrows and green figures represent genes and proteins, respectively, up-regulated under nitrogen fixation conditions. Yellow arrows indicate genes with the same level of expression in ammonium-rich and ...

Interestingly, nifA seems to be transcribed efficiently, as indicated by the relatively high spot intensities shown by the DNA fragment bearing it. The nifA transcription level would be independent of nitrogen sources, as is known to be the case in the photosynthetic bacterium Rhodospirillum rubrum (31), suggesting that nifA is regulated posttranslationally by a mechanism that controls NifA activity. NifA is inactive in the presence of ammonium, being activated when ammonium is depleted, probably through the action of overexpressed PII (GlnB). This mechanism has been reported in Azospirillum brasilense (32) and Herbaspirillum seropedicae (28), where PII is in turn activated by uridylylation (29). We have identified two glnB-like genes in L. ferrooxidans L3.2. Usually, the glnB gene associated with amtB is designated as glnK (20), but in the case of L. ferrooxidans L3.2, we kept the glnB name because its deduced amino acid sequence is more similar to GlnB than to GlnK. Additionally, it contains a lysine at position 3 and a glutamate at position 5, conserved in all GlnB proteins. Moreover, it does not seem to form an operon with amtB, as in most of the amtB-glnK pairs found in bacteria (33, 34), due to the presence of a putative Rho-independent transcription terminator (ΔG = -17.36 kcal·mol-1, 1 cal = 4.184 J) downstream of amtB. This observation suggests that both genes (amtB-glnB) are independently transcribed. No σ54-dependent promoter consensus sequences were found in the intergenic region, although a σ70 -10 consensus region (TATAAc) is located at -29 nt with respect to the ATG glnB translation start codon.

Nitrogen fixation is a high-energy-consuming process requiring ATP and the supply of Fe-S clusters not only for the structural nitrogenase components but also for many other proteins involved in electron transfer, redox and nonredox catalysis, or sensing for regulatory processes. Hence, it is quite logical that genes responsible for Fe-S cluster assembly (nifS-nifU-hesB-hscB-hscA-fdx-orf1; refs. 19, 35) were among the most induced (Fig. 3). Apart from transporters like ModABC, AmtB, FixI, or sensor systems like KdpD2, other genes involved in the synthesis of cell envelope components are induced, like glycerol-3-phosphate cytidyltransferase (tagD), involved in teichoic acid synthesis (36), trehalose-6-phosphate synthase/phosphatase (otsA, otsB) for trehalose synthesis (37), or 1-deoxy-D-xylulose-5-phosphate synthase (38) necessary for the biosynthesis of many essential aspects of the cell wall (Table 1). Interestingly, several hydrogenase genes were also induced (Table 1), which are likely to be involved in the oxidation of the hydrogen produced as a consequence of nitrogenase activity, as has been described in rhizobia (39).

In conclusion, adaptation to a more energy-demanding nitrogen source forces the cell to alter several aspects of its metabolism like energy production, cell envelope modification, chemotactic and pumping mechanisms, and sensing and signal transduction systems, as well as the specific nif genes.

Apart from its minimal nutrient requirements, our results showing that L. ferroxidans L3.2 can fix atmospheric N2 strongly suggest that some aspects of L. ferrooxidans metabolism are among the most primitive on Earth and merit further investigation. Nitrogen fixation could have played a key role in the early evolution of life not only on Earth but elsewhere (40). Shotgun DNA microarrays from plasmid libraries of a single unsequenced genome or metagenome from different environments are useful tools in molecular ecology and astrobiology.

Supplementary Material

Supporting Table:


We thank Marina Postigo and Miriam García for excellent technical assistance; Ricardo Amils and Elena González-Toril for providing the L. ferrooxidans L3.2 strain as well as valuable information; Victor de Lorenzo for helpful discussions; and David Hochberg for proofreading the manuscript. We especially thank Juan Pérez-Mercader for tremendous support. This work was supported by Ministerio de Ciencia y Tecnología Grant BXX2000-1385. V.P., M.M.-P., and Marina Postigo are fellows from Instituto Nacional de Técnica Aeroespacial, and Miriam García has an I3P contract from Consejo Superior de Investigaciones Científicas.


This paper was submitted directly (Track II) to the PNAS office.

Data deposition: The DNA sequences reported in this work have been deposited in the GenBank database (accession nos. AY204356–AY204460).


1. Duggan, D. J., Bittner, M., Chen, Y., Meltzer, P. & Trent, J. M. (1999) Nat. Genet. 21, 10-14. [PubMed]
2. Hayward, R. E., Derisi, J. L., Alfadhli, S., Kaslow, D. C., Brown, P. O. & Rathod, P. K. (2000) Mol. Microbiol. 35, 6-14. [PubMed]
3. Lopez-Archilla, A. I., Marin, I. & Amils, R. (2001) Microb. Ecol. 41, 20-35. [PubMed]
4. Lopez-Archilla, A. I. & Amils, R. (1999) Microb. Ecol. 38, 146-156. [PubMed]
5. Boon, M., Ras, C., Hansford, G. S. & Heijnen, J. J. (1995) in Biohydrometallurgical Processing, eds. Vargas, T., Jerez, D. A., Wiertz, J. V. & Toledo, H. (Univ. of Chile Press, Santiago, Chile), Vol. 1, pp. 153-181.
6. Sand, W., Gerke, T., Hallmann, R. & Schippers, A. (1995) Appl. Microbiol. Biotechnol. 43, 961-966.
7. Schrenk, M. O., Edwards, K. J., Goodman, R. M., Hamers, R. J. & Banfield, J. F. (1998) Science 279, 1519-1522. [PubMed]
8. Cairns-Smith, A. G., Hall, A. J. & Russell, M. J. (1992) in Origins of Life and Evolution of the Biosphere (Kluwer, Dordrecht, The Netherlands), pp. 161-180.
9. Norris, P. R., Murrell, J. C. & Hinson, D. (1995) Arch. Microbiol. 164, 294-300.
10. Mackintosh, M. E. (1978) J. Gen. Microbiol. 105, 215-218.
11. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY), 2nd Ed.
12. Hegde, P., Qi, R., Abernathy, K., Gay, C., Dharap, S., Gaspard, R., Hughes, J. E., Snesrud, E., Lee, N. & Quackenbush, J. (2000) BioTechniques 29, 548-562. [PubMed]
13. Parro, V., Schacht, S., Anne, J. & Mellado, R. P. (1999) Microbiology 145, 2255-2263. [PubMed]
14. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. [PMC free article] [PubMed]
15. Merrick, M. J. & Edwards, R. A. (1995) Microbiol. Rev. 59, 604-622. [PMC free article] [PubMed]
16. Masepohl, B., Drepper, T., Paschen, A., Gross, S., Pawlowski, A., Raabe, K., Riedel, K. U. & Klipp, W. (2002) J. Mol. Microbiol. Biotechnol. 4, 243-248. [PubMed]
17. Pedrosa, F. O., Benelli, E. M., Yates, M. G., Wassem, R., Monteiro, R. A., Klassen, G., Steffens, M. B., Souza, E. M., Chubatsu, L. S. & Rigo, L. U. (2001) J. Biotechnol. 91, 189-195. [PubMed]
18. Takahashi, Y. & Nakamura, M. (1999) J. Biochem. (Tokyo) 126, 917-926. [PubMed]
19. Tokumoto, U., Nomura, S., Minami, Y., Mihara, H., Kato, S., Kurihara, T., Esaki, N., Kanazawa, H., Matsubara, H. & Takahashi, Y. (2002) J. Biochem. (Tokyo) 131, 713-719. [PubMed]
20. Arcondéguy, T., Jack, R. & Merrick, M. (2001) Microbiol. Mol. Biol. Rev. 65, 80-105. [PMC free article] [PubMed]
21. Gutiérrez, J. C., Ramos, F., Ortner, L. & Tortolero, M. (1995) Mol. Microbiol. 18, 579-591. [PubMed]
22. Steyn, A. J., Joseph, J. & Bloom, B. R. (2003) Mol. Microbiol. 47, 1075-1089. [PubMed]
23. Heermann, R., Fohrmann, A., Altendorf, K. & Jung, K. (2003) Mol. Microbiol. 47, 839-848. [PubMed]
24. Kahn, D., David, M., Domergue, O., Daveran, M. L., Ghai, J., Hirsch, P. R. & Batut, J. (1989) J. Bacteriol. 171, 929-939. [PMC free article] [PubMed]
25. Barrios, H., Valderrama, B. & Morett, E. (1999) Nucleic Acids Res. 27, 4305-4313. [PMC free article] [PubMed]
26. Souza, E. M., Funayama, S., Rigo, L. U., Yates, M. G. & Pedrosa, F. O. (1991) J. Gen. Microbiol. 137, 1511-1522. [PubMed]
27. Souza, E. M., Pedrosa, F. O., Rigo, L. U., Machado, H. B. & Yates, M. G. (2000) Microbiology 146, 1407-1418. [PubMed]
28. Benelli, E. M., Souza, E. M., Funayama, S., Rigo, L. U. &. Pedrosa, F. O. (1997) J. Bacteriol. 179, 4623-4626. [PMC free article] [PubMed]
29. Benelli, E. M., Buck, M., de Souza, E. M., Yates, M. G. & Pedrosa, F. O. (2001) Can. J. Microbiol. 47, 309-314. [PubMed]
30. Jiang, P., Peliska, J. A. & Ninfa, A. J. (1998) Biochemistry 37, 12782-12794. [PubMed]
31. Zhang, Y., Pohlmann, E, L., Luddenm P. W. & Roberts, G. P. (2000) J. Bacteriol. 182, 983-992. [PMC free article] [PubMed]
32. Arsene, F., Kaminski, P. A. & Elmerich, C. (1996) J. Bacteriol. 178, 4830-4838. [PMC free article] [PubMed]
33. Meletzus, D., Rudnick, P., Doetsch, N., Green, A. & Kennedy, C. (1998) J. Bacteriol. 180, 3260-3264. [PMC free article] [PubMed]
34. Thomas, G., Coutts, G. & Merrick, M. (2000) Trends Genet. 16, 11-14. [PubMed]
35. Nakamura, M., Saeki, K. & Takahashi, Y. (1999) J. Biochem. (Tokyo) 126, 10-18. [PubMed]
36. Mauel, C., Young, M. & Karamata, D. (1991) J. Gen. Microbiol. 137, 929-941. [PubMed]
37. Giaever, H. M., Styrvold, O. B., Kaasen, I. & Strom, A. R. (1988) J. Bacteriol. 170, 2841-2849. [PMC free article] [PubMed]
38. Bailey, A. M., Mahapatra, S., Brennan, P. J. & Crick, D. C. (2002) Glycobiology 12, 813-820. [PubMed]
39. Baginsky, C., Brito, B., Imperial, J., Palacios, J. M. & Ruiz-Argueso, T. (2002) Appl. Environ. Microbiol. 68, 4915-4924. [PMC free article] [PubMed]
40. Klingler, J. M., Mancinelli, R. L. & White, M. R. (1989) Adv. Space Res. 9, 173-176. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...