A physical map of the megaplasmid pHG1, one of three genomicreplicons in Ralstonia eutropha H16

Edward Schwartz *, Ba«rbel FriedrichInstitut fu«r Biologie, Mikrobiologie, Humboldt-Universita«t zu Berlin, Chausseestr. 117, 10115 Berlin, Germany

Received 30 April 2001; accepted 25 May 2001

First published online 25 June 2001

Abstract

We have used pulsed field gel electrophoresis and megabase DNA techniques to investigate the basic genomic organization of Ralstoniaeutropha H16, and to construct a physical map of its indigenous megaplasmid pHG1. This Gram-negative, soil-dwelling bacterium is afacultative chemolithoautotroph and a denitrifier. In the absence of organic substrates it can grow on H2 as its sole energy source and CO2

as its sole source of carbon. Under anaerobic conditions it can utilize nitrate as a terminal electron acceptor, whereby dinitrogen is released.Essential genetic determinants of the enzyme systems responsible for these metabolic processes are linked to the 0.44-Mb conjugativemegaplasmid pHG1. Aside from pHG1, the genome of R. eutropha H16 is comprised of two circular chromosomes measuring 4.1 and 2.9Mb, adding up to a total genome size of 7.1 Mb. An estimated five copies of rDNA are distributed on the two chromosomes. Amacrorestriction map of pHG1 was derived for the endonucleases DraI and XbaI. Hybridization studies showed that genes for anaerobicmetabolism are located on all three genomic replicons. ß 2001 Federation of European Microbiological Societies. Published by ElsevierScience B.V. All rights reserved.

Keywords: Physical mapping; Megaplasmid; Hydrogenase ; Chemolithotrophy; Denitri¢cation; Ralstonia eutropha

1. Introduction

Ralstonia eutropha (formerly Alcaligenes eutrophus) H16is a Gram-negative, facultatively lithotrophic, strictly res-piratory bacterium, which is widely distributed in soil andfresh-water habitats. In the absence of organic com-pounds, R. eutropha can utilize H2 as its sole source ofenergy. In recent years interest in biological H2 oxidationhas been growing, in part due to steadily increasing e¡ortsaimed at the development of H2-based energy technolo-gies. Extensive studies on the genetics and biochemistry ofH2 oxidation in R. eutropha have made this bacterium themodel organism for the investigation of H2-based litho-trophy (reviewed in [1]). The genetic determinants of theH2-oxidizing system are located on a 0.4-Mb-large, au-tochthonous, conjugative megaplasmid designated pHG1.

In addition to H2-based lithotrophy, the genetic informa-tion carried on pHG1 contributes in other important waysto the metabolic diversity of its host bacterium. R. eutro-pha is capable of ¢xing CO2 via a ribulose bisphosphatecarboxylase and the other enzymes of the Calvin cycle.One of the two duplicate sets of genes for the Calvin-cycleenzymes is located on pHG1 [2]. The megaplasmid pHG1also carries determinants for anaerobic growth on nitrate.These include a periplasmic nitrate reductase, a £avohe-moprotein, a membrane-bound nitrate reductase, a nitritereductase, a nitric oxide reductase, a N2O reductase, ananaerobic ribonucleotide reductase and an O2-independentcoproporphyrinogen III reductase [3,4].

This report summarizes the results of an initial geno-mic study of R. eutropha H16 and its megaplasmidpHG1. We have determined the number, size and topol-ogy of the replicons constituting the R. eutropha genome.We also surveyed the genome for rRNA operons andinvestigated the distribution of some key genes of anaer-obic metabolism. To facilitate the ongoing molecular ge-netic investigations and as an aid for a genome sequencingproject, we have constructed a macrorestriction map ofpHG1.

0378-1097 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.PII: S 0 3 7 8 - 1 0 9 7 ( 0 1 ) 0 0 2 6 5 - 8

* Corresponding author. Tel. : +49 (30) 20938117;Fax: +49 (30) 20938102.

E-mail address: edward.schwartz@rz.hu-berlin.de (E. Schwartz).

FEMSLE 10019 16-7-01

FEMS Microbiology Letters 201 (2001) 213^219

www.fems-microbiology.org

2. Materials and methods

2.1. Bacterial strains and media

Strains of R. eutropha were grown at 30³C either inLuria^Bertani (LB) medium or in mineral salts mediumcontaining 0.4% (w/v) fructose [5]. Escherichia coli strainswere grown in LB medium at 37³C. When required, anti-biotics were used at the following concentrations (Wgml31) : ampicillin, 100 (E. coli) ; gentamicin, 50 (E. coli) ;kanamycin, 50 (E. coli) or 350 (R. eutropha) ; zeocin, 50 (E.coli).

2.2. Cloning and plasmid construction

An internal segment of a R. eutropha rrs gene was am-pli¢ed from total genomic DNA of R. eutropha H16 usingthe primer pair fD1 (5P-CCGAATTCGTCGACAACA-GAGTTTGATCCTGGCTCAG-3P) and rD1 (5P-CCCG-GGATCCAAGCTTAAGGAGGTGATCCAGCC-3P) [6],and the resulting 1.5-kb amplicon was subsequently ligatedto linear vector DNA of pGEM-T (Promega). Hybridplasmids carrying 1.5-kb inserts were selected and thepresence of rrs sequences was veri¢ed by sequencing.

One of the resulting plasmids, designated pCH854, wasused to screen the R. eutropha genome for rrn operons.

2.3. Construction of a megaplasmid variant with aunique restriction site

For mapping studies we engineered a megaplasmid de-rivative carrying a singular homing sequence for theintron-encoded endonuclease I-SceI. Brie£y, a duplexDNA was generated by annealing the oligonucleotidesSCE1A (5P-GATCTAGGGATAACAGGGTAATC-3P)and SCE1B (5P-GATCGATTACCCTGTTATCCCTA-3P)and was inserted into the BglII site of plasmid pCH855.This plasmid contains a segment spanning the 5P region ofthe R. eutropha operon encoding a membrane-bound hy-drogenase. In the resulting plasmids containing the syn-thetic DNA segment in both orientations, a single BglIIsite was recreated. An 865-bp BamHI fragment ofpUCGM [7] containing the aacC1 gene for gentamicinresistance was introduced into this site. Finally, a 1.7-kbPstI^XhoI fragment containing the modi¢ed sequence wastransplanted into vector pLO2. The resulting derivative(pCH856) was used to transfer the I-SceI site to pHG1by the previously described allelic-exchange technique [5].

Fig. 1. PFGE analysis of the R. eutropha H16 genome. A: Identi¢cation of large, circular replicons. Embedded DNA was separated via PFGE (con-stant pulses, 20 min; ¢eld strength, 2 V cm31 ; gel system, 0.8% (w/v) agarose in TAE; included angle, 106³; run time, 60 h). Native and irradiated (5^30 Gy) samples were electrophoresed on the same gel. Sizes of DNA marker fragments (Hansenula wingei and Schizosaccharomyces pombe chromo-somes, Bio-Rad) are given at the right. Arrowheads indicate the three DNA species (1 and 2, large and small chromosomes, respectively; pHG1, mega-plasmid). Sizes of marker fragments are given at the right. B: Restriction fragment patterns of the total genomic DNA of R. eutropha H16 (wild-type,WT) and of the megaplasmid-free derivative HF210 (-). Embedded DNA was digested with the rarely cutting enzymes AseI, DraI, SpeI and XbaI, andthe resulting fragments were separated via PFGE (linear ramping, 2^20 s; ¢eld strength, 180 V; included angle, 106³; gel system, 1.2% (w/v) agarose,0.5U TBE; temperature, 5³C). M: low-range marker (Biolabs). Arrowheads indicate megaplasmid-derived restriction fragments. Fragment sizes are giv-en at the left.

FEMSLE 10019 16-7-01

E. Schwartz, B. Friedrich / FEMS Microbiology Letters 201 (2001) 213^219214

2.4. Preparation and handling of intact genomic DNA

R. eutropha strains were grown to an OD436 of four to¢ve in mineral salts medium containing fructose and em-bedded in low-melting-point agarose following a standardprotocol [8]. High-molecular-mass DNA was separated viacontour-clamped homogeneous electric ¢eld (CHEF) elec-trophoresis using either a CHEF Mapper XA apparatus(Bio-Rad) or a Gene Navigator system (Amersham Phar-macia Biotech). In order to detect full-sized replicons,samples of embedded DNA were irradiated using a Gam-matron S apparatus (Siemens) containing a Co60 Q raysource.

3. Results and discussion

3.1. The genome of R. eutropha H16 consists of threecircular replicons

Prior to mapping the megaplasmid pHG1, we sought todetermine the size and topology of the R. eutropha H16genome. A previous study on Burkholderia cepacia notedthe existence of more than one Mbp-sized replicon instrains of R. eutropha [9]. To pursue this point we pre-pared agarose-embedded total genomic DNA of R. eutro-pha H16 and separated the native DNA by pulsed ¢eld gelelectrophoresis (PFGE). The resulting gel revealed threeethidium bromide-stained bands. A very faint band corre-sponding to a DNA molecule of approximately 0.44 Mbwas observed in the lower part of the gel (Fig. 1A). Thisband is probably attributable to pHG1, since its size isclose to the previously published estimate [10]. Two addi-tional bands corresponding to DNA molecules of approx-imately 3 and 4 Mb were visible. This suggested that R.eutropha contains two chromosome-sized replicons in ad-dition to the megaplasmid. Since circular molecules arenot resolved by CHEF electrophoresis, and the majorpart of the £uorescent material was retained in the agaroseplug, we assumed that the three £uorescent bands repre-sented a fraction of the circular DNA molecules whichhad su¡ered random breakage during sample preparation.To verify this, we exposed samples of embedded R. eutro-pha DNA to Q irradiation prior to separation by PFGE.Since Q irradiation causes random double-strand breakageof DNA, this treatment results in the linearization of cir-cular molecules. Doses of irradiation up to 10 Gy led to amarked increase in the £uorescence of all three bands inthe PFGE gels (Fig. 1A). Higher doses (20^30 Gy) re-sulted in a decline in the intensity of the £uorescence ofthe high-molecular-mass species and the concomitant ap-pearance of di¡usely distributed low-molecular-mass ma-terial. These observations are compatible with the notionthat the three R. eutropha H16 replicons are circular intheir native state. The two larger replicons, henceforthreferred to as chromosome 1 and chromosome 2, were

estimated to be 4.1 and 2.9 Mb, respectively. The size ofthe megaplasmid was determined to be 0.44 Mb. This¢gure is slightly lower than the estimate (0.45 Mb) ob-tained by conventional agarose-gel electrophoresis [10].Based on the above ¢gures, the genome of R. eutrophaH16 totals about 7.1 Mb, placing this organism near theupper end of the size range of bacterial genomes, forwhich physical data are available (reviewed in [11]). Multi-ple Mbp-sized replicons have been observed in severalother bacteria including Rhodobacter sphaeroides, Sinorhi-zobium meliloti and Brucella melitensis (reviewed in [12]).Complex genome organization appears to be typical of theBurkholderia/Ralstonia group. Each of two strains of B.cepacia examined contain four large replicons [9,13]. Ral-stonia solanacearum, a close relative of R. eutropha, con-tains two large replicans [14].

3.2. Physical mapping of megaplasmid pHG1

Five basic approaches were used to establish a macro-restriction map of pHG1:

1. Digests of embedded samples of total genomic DNA ofthe wild-type R. eutropha strain H16 and the megaplas-mid-free strain HF210 were analyzed by PFGE. Asexpected for a GC-rich (s 65%) genome, enzymeswith AT-biased recognition sites gave fewer productsthan enzymes recognizing GC-rich sequences. However,even the `rare cutters' yielded very complex fragmentpatterns. The octanucleotide cutters included in oursurvey (PacI, PmeI and SwaI) did not cut pHG1. Wechose four endonucleases for further analyses: AseI,DraI, SpeI and XbaI. Comparison of the restrictionpatterns of the wild-type and megaplasmid-free strainsled to the identi¢cation of some megaplasmid-derivedfragments (Fig. 1B). However, the restriction patternwas in many places so dense that individual bandswere not discernible.

2. Isolated megaplasmid DNA was labelled with 32P andused to probe blots of the PFGE gels described above.This procedure aided the identi¢cation of megaplasmid-derived fragments in areas where the banding patternwas particularly dense (data not shown).

3. Digests of the isolated megaplasmid DNA were sepa-rated via conventional agarose-gel electrophoresis. Thispermitted direct identi¢cation of restriction fragmentsin the size range 6 50 kb. The combined use of theseapproaches led to the catalogue of restriction fragmentsof the four mapping enzymes (Table 1). The restrictionpatterns for AseI and SpeI contained multiple smallfragments and/or superimposed fragments which posean obstacle to mapping. Therefore, we limited the map-ping project to the enzymes DraI and XbaI.

4. Following an approach pioneered by Thierry and Du-jon [15], we constructed a pHG1 derivative containing aunique homing site for the endonuclease I-SceI, and

FEMSLE 10019 16-7-01

E. Schwartz, B. Friedrich / FEMS Microbiology Letters 201 (2001) 213^219 215

used the resulting R. eutropha mutant, designatedHF530, for mapping. Embedded samples of HF530DNA were digested with I-SceI to completion. Subse-quently, the samples were treated with either DraI orXbaI under conditions which favored partial digestion.Following separation via PFGE, the gel was used toprepare a Southern blot. A digoxigenin (DIG)-labelled

internal fragment of the aacC1 gene, the marker imme-diately adjacent to the I-SceI site, was used to visualizemegaplasmid-derived restriction products (Fig. 2). Theintervals between the stained bands correspond to thesizes of the restriction fragments produced by the en-zymes DraI and XbaI. The order of restriction frag-ments in the circular map can be deduced from theseries of size intervals. In the case of the DraI partialdigest, the intervals were estimated to be (from bottomto top) 52, 46, 28 and 98 kb in size. This indicated thatthe DraI fragments map in the following order: C-D-E-B-A. The XbaI partial digest gave bands separated bythe following intervals (from bottom to top): 63, 23, 35,42, 8, 14, 102 and 16 kb. Matching the XbaI fragmentsto these intervals gave the following order: C-F-E-D-I-H-B-G-A.

5. To verify the restriction map obtained by the abovestrategies, we carried out Southern hybridizations usingselected pHG1-speci¢c probes. XbaI fragment I hybrid-ized to DraI fragment B. XbaI fragment G hybridizedto DraI fragment A. A 1.3-kb fragment containing gene

Fig. 2. Analysis of I-SceI/DraI and I-SceI/XbaI partial cleavage productsof pHG1. Embedded samples of total genomic DNA of the I-SceI inter-poson mutant HF530 were digested with I-SceI to completion, and thenbrie£y with either DraI (left panel) or XbaI (right panel), yielding partialreaction products. After separation by PFGE, the restriction productswere transferred to nylon membranes. Hybridization using a digoxi-genin-labelled probe speci¢c for the aacC1 gene revealed a series ofnested restriction fragments of pHG1 extending in one direction fromthe I-SceI site. The sizes of the intervals between the bands are given inkb.

Fig. 3. DraI and XbaI cleavage maps of the R. eutropha H16 megaplas-mid pHG1. The I-SceI homing sequence is indicated by a vertical ar-row. The various XbaI and DraI fragments are marked with the capitalletters as assigned in Table 1. The positions of the hoxKGZMLOQRTVand cbbLSXYEFPTZGKA operons are indicated and the direction oftranscription is indicated by arrows. The segments containing the napABand narG2 genes are indicated by solid and dashed arcs, respectively.

Table 1Restriction fragments generated from R. eutropha H16 megaplasmidpHG1 by rarely cutting enzymes

Fragment Fragment size (kb)

AseI DraI SpeI XbaI

A 240 210 180 145B 76 95 105 86C 41 54 40 62D 30 40 37 42E 20 23 14 (E1) 36

14 (E2)F 7 12 31G 6 8 19H 2 13I 9Total 422 422 410 443

FEMSLE 10019 16-7-01

E. Schwartz, B. Friedrich / FEMS Microbiology Letters 201 (2001) 213^219216

hoxK hybridized to DraI and XbaI fragments A.pGE49, a recombinant plasmid taken from a pHG1gene bank, was found to contain the 13-kb XbaI frag-ment (i.e. fragment H) within the 15-kb insert of pHG1DNA. The labelled 15-kb insert hybridized as expectedto XbaI fragment H and also to fragments B and I. A4.3-kb fragment of the megaplasmid narG2 locus en-coding a respiratory nitrate reductase hybridized toDraI fragment A and XbaI fragment B. This indicatesthat narG2 lies within the ca. 70-kb-long segment whereDraI fragment A and XbaI fragment B overlap. Thus,the results of the Southern experiments support themap deduced from the partial digests (Fig. 3).

3.3. Identi¢cation of rDNA in R. eutropha H16

The number and arrangement of rRNA (rrn) operonsare characteristic features of bacterial genomes and areconserved among related strains [11]. Aside from encodingvital components of the translational machinery, the rrnoperons demarcate units of a genome, which can undergoampli¢cation via duplication and reshu¡eling via transpo-sition [16]. Pursuant to an initial characterization of the R.eutropha genome, we investigated the number and distri-bution of rrn operons. A DIG-dUTP-labelled 900-bp SphIfragment of a cloned R. eutropha H16 rrs gene was used toprobe Southern blots of PFGE gels containing the linear-ized chromosomes and megaplasmid of R. eutropha H16(Fig. 4A,B). The blot gave hybridization signals corre-sponding to chromosomes 1 and 2. The region of theblot corresponding to the megaplasmid was masked byrandomly distributed hybridizing material, making an in-terpretation di¤cult. Nevertheless, the pattern of staining

suggests that rrn sequences are present on both chromo-somes, but not on the megaplasmid. To con¢rm this resultand to estimate the number of copies of the rrn operon inthe R. eutropha H16 genome, we carried out a restrictionanalysis using the intron-encoded endonuclease I-CeuI.Since the recognition sequence of this enzyme coincideswith a highly conserved region of bacterial rrl genes whichencode the 23S rRNA [17], I-CeuI analysis can be used toestimate rrn copy number. Embedded DNA from the wild-type R. eutropha strain H16 and from the plasmid-freederivative HF210 was digested with I-CeuI and separatedvia PFGE. The ethidium bromide-stained gel revealed sim-ilar patterns of £uorescent bands for the two strains (Fig.4C). The I-CeuI digest of H16 DNA contained an addi-tional band corresponding to a molecule of about 0.44Mb. This band could conceivably represent pHG1 linear-ized by I-CeuI. However, the 0.44-Mb band was muchfainter than the neighboring bands, indicating that it didnot represent a product of digestion by I-CeuI, but ratherresulted from random double-strand breakage of the meg-aplasmid DNA. The sizes of the I-CeuI fragments wereestimated as follows: 2.2, 1.85, 1.32, 0.55 and 0.21 Mb.The sum of these estimates is in good agreement with thesum of the sizes of the two chromosomes. Thus, the resultsof the I-CeuI analysis corroborate the Southern blot, in-dicating that copies of rrn genes are located on each of thechromosomes, but not on the megaplasmid, and further-more suggest, that there are in all ¢ve copies in the R.eutropha genome.

3.4. Distribution of genes for anaerobic metabolism in thegenome of R. eutropha

Genetic analysis of the R. eutropha H16 denitri¢cation

Fig. 4. Detection of rrn homologs in R. eutropha. A: Irradiated R. eutropha H16 DNA separated by PFGE (phase 1: 25-min pulses for 50 h; phase 2:linear ramping, 13^36-s pulses for 6 h; for other details see Fig. 1A). The embedded genomic DNA was irradiated with 10 Gy of Q radiation prior toelectrophoresis. 1 and 2: large and small chromosomes, respectively ; pHG1, megaplasmid. B: Southern blot of the gel shown in A after screening witha DIG-labelled probe consisting of a ca. 900-bp SphI fragment of the cloned R. eutropha rrs sequence. C: I-CeuI digests of R. eutropha H16 andHF210 DNA separated by PFGE (linear ramping, 200^600-s pulses for 48 h; ¢eld strength, 3 V cm31 ; gel system, 1.1% (w/v) agarose in TAE; includedangle, 106³). The arrowhead indicates a band representing linear pHG1 DNA arising from random double-strand breakage. Sizes of selected standardfragments are given. Hw, H. wingei chromosomes; Sp, Schizosaccharomyces pombe chromosomes; Sc, Saccharomyces cerevisiae chromosomes.

FEMSLE 10019 16-7-01

E. Schwartz, B. Friedrich / FEMS Microbiology Letters 201 (2001) 213^219 217

pathway revealed that some of the determinants of the keyenzymes are megaplasmid-borne, while others are chromo-somal [3,4]. We were curious to know whether the chro-mosomal loci mapped to one or to both chromosomes.Therefore, we screened a Southern blot of a PFGE gelcontaining linearized DNA of the three genomic repliconsof R. eutropha using probes speci¢c for anaerobiosis genes(Fig. 5). Hybridization with a probe derived from themegaplasmid narG2 locus, which encodes one of the twoisofunctional respiratory nitrate reductases, gave two sig-nals, one corresponding to the megaplasmid DNA and theother to chromosome 2. The chromosomal signal was lessintense than the megaplasmid signal, re£ecting di¡erencesin the nucleotide sequence of the two alleles. As expected,the nirS probe, which is speci¢c for a gene encoding a cd1-type nitrite reductase, gave only a single hybridizationsignal; the hybridizing material coincided with the bandrepresenting chromosome 2. Two hybridization signalswere obtained using the norB1 probe, con¢rming the pre-vious ¢nding [3] that this gene, which encodes a nitricoxide reductase, is duplicated. The chromosomal copymapped to chromosome 2. Finally, screening with a probe

for the chromosomal hemN gene, an essential determinantof anaerobic heme biosynthesis, gave only one prominentband. Here the hybridization signal correlated with chro-mosome 1. Thus, essential determinants of anaerobic me-tabolism are located on all three replicons of the R. eutro-pha H16 genome.

Acknowledgements

We thank H. Schneeweiss and A. Strack for skilledtechnical assistance and T. Eitinger, R. Cramm and A.Pohlmann for stimulating discussions. E.S. is indebted toM. Reh and J. Kalkus for introducing him to PFGE tech-niques and to U. Ro«mling and W.R. Hess for helpfuladvice. Special thanks to A. Berg for irradiating DNAsamples. This work was supported by the Fonds derChemischen Industrie.

References

[1] Friedrich, B., Bernhard, M., Dernedde, J., Eitinger, T., Lenz, O.,Massanz, C. and Schwartz, E. (1996) Hydrogen oxidation by Alcali-genes. In: Microbial growth on C1 compounds (Lidstrom, M.E. andTabita, F.R., Eds.), pp. 110^117. Kluwer Academic Publishers, Dor-drecht.

[2] Kusian, B. and Bowien, B. (1997) Organization and regulation of cbbCO2 assimilation genes in autotrophic bacteria. FEMS Microbiol.Rev. 21, 135^155.

[3] Cramm, R., Siddiqui, R.A. and Friedrich, B. (1997) Two isofunc-tional nitric oxide reductases in Alcaligenes eutrophus H16. J. Bacter-iol. 179, 6769^6777.

[4] Pohlmann, A., Cramm, R., Schmelz, K. and Friedrich, B. (2000) Anovel NO-responding regulator controls the reduction of nitric oxidein Ralstonia eutropha. Mol. Microbiol. 38, 626^638.

[5] Lenz, O., Schwartz, E., Dernedde, J., Eitinger, M. and Friedrich, B.(1994) The Alcaligenes eutrophus H16 hoxX gene participates in hy-drogenase regulation. J. Bacteriol. 176, 4385^4393.

[6] Weisburg, W.G., Barns, S.M., Pelletier, D.A. and Lane, D.J. (1991)16S ribosomal DNA ampli¢cation for phylogenetic study. J. Bacter-iol. 173, 697^703.

[7] Schweizer, H.P. (1993) Small broad-host-range gentamycin resistancegene cassettes for site-speci¢c insertion and deletion mutagenesis.BioTechniques 15, 831^834.

[8] Birren, B. and Lai, E. (1993) Pulsed ¢eld gel electrophoresis : A prac-tical guide. Academic Press, San Diego, CA.

[9] Rodley, P.D., Ro«mling, U. and Tu«mmler, B. (1995) A physical ge-nome map of the Burkholderia cepacia type strain. Mol. Microbiol.17, 57^67.

[10] Hogrefe, C. and Friedrich, B. (1984) Isolation and characterization ofmegaplasmid DNA from lithoauthotrophic bacteria. Plasmid 12,161^169.

[11] Cole, S. and Saint Girons, I. (1994) Bacterial Genomics. FEMS Mi-crobiol. Rev. 14, 139^160.

[12] Fonstein, M. and Haselkorn, R. (1995) Physical mapping of bacterialgenomes. J. Bacteriol. 177, 3361^3369.

[13] Cheng, H.P. and Lessie, T.G. (1994) Multiple replicons constitutingthe genome of Pseudomonas cepacia 17616. J. Bacteriol. 176, 4034^4042.

[14] Boucher, C., personal communication.[15] Thierry, A. and Dujon, B. (1992) Nested chromosomal fragmentation

Fig. 5. Assignment of genes of anaerobic metabolism to the three repli-cons of R. eutropha H16. Embedded DNA was irradiated with 10 Gy Qradiation and separated via PFGE (phase 1: 25-min pulses for 50 h;phase 2: linear ramping, 13^36-s pulses for 6 h; ¢eld strength, 2 Vcm31 ; included angle, 106³; gel system, 0.8% agarose in TAE). Follow-ing electrophoresis, DNA was blotted via alkaline transfer and subjectedto Southern analysis using 32P-labelled restriction fragments of clonedR. eutropha DNA as probes. A: Ethidium bromide-stained agarose gel.Arrowheads indicate the three DNA species (1 and 2, chromosomes 1and 2; pHG1, megaplasmid). B: Hybridization patterns of the gelshown in A. Hybridizations were done with the following probes: narG,a 4.3-kb EcoRI^BamHI fragment of the narG2 region in plasmidpCH810; nirS, a 0.7-kb SalI fragment of pCH448; norB, a 1.6-kbEcoRI^BamHI fragment of pCH444; hemN, a 4.4-kb EcoRI^BamHIfragment of pCH504.

FEMSLE 10019 16-7-01

E. Schwartz, B. Friedrich / FEMS Microbiology Letters 201 (2001) 213^219218

in yeast using the meganuclease I-SceI: a new method for physicalmapping of eukaryotic genomes. Nucleic Acids Res. 20, 5625^5631.

[16] Hill, C.W. and Harnish, B.W. (1982) Transposition of a chromoso-mal segment bounded by redundant rRNA genes into other rRNAgenes in Escherichia coli. J. Bacteriol. 149, 449^457.

[17] Liu, S.-L., Hessel, A. and Sanderson, K. (1993) Genomic mappingwith I-CeuI, an intron-encoded endonuclease speci¢c for genes forribosomal RNA, in Salmonella spp., Escherichia coli, and other bac-teria. Proc. Natl. Acad. Sci. USA 90, 6874^6878.

FEMSLE 10019 16-7-01

E. Schwartz, B. Friedrich / FEMS Microbiology Letters 201 (2001) 213^219 219