Transposon mutagenesis and strain construction inZymomonas mobilis
K.-M. Pappas, I. Galani and M.A. TypasDepartment of Biochemistry, Cell and Molecular Biology, and Genetics, University of Athens, Panepistemiopolis,Athens, Greece
5784/06/96: received 7 June 1996, revised 16 August 1996 and accepted 27 August 1996
K.-M. PAPPAS, I . GALANI AND M.A. TYPAS. 1997. Conjugative or mobilizable plasmidscarrying the transposable elements Tn5, Tn501 or mini Mu were readily transferred fromEscherichia coli donors into Zymomonas mobilis recipients with frequencies dependingboth on donor and recipient strain used. With the exception of pULB113 (RP4 : : mini Mu),all foreign plasmids exhibited high instability in Z. mobilis transconjugants under bothselective and non-selective conditions. Transposition events and consequentmutagenesis occurred readily in Z. mobilis transconjugant strains, with Tn5 and Tn501being far less successful than mini Mu. Transposon mutagenesis with the help ofmini Mu resulted in the isolation of a large number of independent auxotrophs withpolyauxotrophs, cysteine, methionine and isoleucine requiring-isolates being themost frequent. When chromosomal DNA from all these mutants was digested withvarious restriction enzymes and the resulting restriction patterns were hybridizedwith a mini Mu probe, the majority of these mutants appeared to have insertions atdifferent sites of the chromosome. Thus, transposon mutagenesis by mini Mu isproven to be a simple and efficient tool for mutant production and the genetic analysisof Z. mobilis.
INTRODUCTION the stability of many types of mutants produced (Typas andGalani 1992). Moreover, the genetic analysis of Z. mobilis hasZymomonas mobilis is a Gram-negative, facultatively anaerobicbeen limited by inefficient methods of gene transfer (Skot-
bacterium which has considerable potential for industrialnicki et al. 1982 ; Montenecourt 1985 ; Buchholz and Eveleighethanol production (Karsch et al. 1993). Although several1986, 1990 ; Sprenger et al. 1993). Although some reports onresearch groups are working with this organism, the methodstransformation of Z. mobilis have appeared (Browne et al.employed for mutant production and the range of mutants1984 ; Yanase et al. 1986) this system is still at a rudimentaryavailable are still very limited. Conventional mutagens suchstage. Thus, the only means by which foreign DNA can beas u.v.-irradiation and N?-methyl-N?-nitrosoguanidineintroduced into this organism is direct or helped conjugation(MNNG) have been used almost exclusively for the pro-with Inc-P group plasmids (Skotnicki et al. 1982 ; Stokes etduction of mutants (Skotnicki et al. 1982 ; Stokes et al. 1983;al. 1983 ; Typas 1986 ; Afendra and Drainas 1987).Eveleigh et al. 1983 ; Goodman et al. 1984 ; Montenecourt
Transposon mutagenesis generally leads to stable, non-1985 ; Typas and Galani 1992), but an increasing number ofleaky polar mutations through insertional inactivation ofreports raise the problem of marker stability of such mutantschromosomal genes and additionally it provides readily(Walia et al. 1984 ; Ingram et al. 1984 ; Buchholz and Eveleighidentifiable genetic marker(s) carried by the transposon (Berg1990 ; Typas and Galani 1992). Additionally, the organism isand Berg 1983). It is considered, therefore, a powerful geneticsuspected of possessing several repair systems which affectsystem which can be used alternatively in Z. mobilis to over-come the problems of marker stability and/or chemicalCorrespondence to : Dr Milton A. Typas, Department of Biochemistry, Cell(MNNG) mutagenesis, which apart from point mutationsand Molecular Biology, and Genetics, University of Athens,
Panepistemiopolis, Kouponia, 157 01 Athens, Greece. induces multiple lesions in the genome (Cerda-Olmedo et al.
1968). Although several transposons, e.g. Tn5 and Tn10 potassium acetate (pH 4·8) and an equal volume of phenol :chloroform were added and the mixture was thoroughly(Skotnicki et al. 1982), Tn951 (Carey et al. 1987 ; Goodman
et al. 1984) and Tn1725 (Walker and Pemberton 1988), car- mixed by inversion and kept in ice for 10 min. The emulsionwas centrifuged for 10 min and the supernatant fluid wasried by broad host range plasmids have been introduced into
Z. mobilis via conjugation, they have exclusively been used as phenol : chloroform extracted as before. The chromosomalDNA was precipitated by standard methods (Sambrook et al.foreign gene markers and no transposition phenomena have
been reported so far in this organism. The purpose of this 1989). A modified Birboim and Doly method in which thelysis solution was 1% SDS, 0·2 mol l−1 NaOH, 5 mmol l−1work, therefore, was to produce transposon-induced mutants
(auxotrophs and/or antibiotic sensitive) using transposons EDTA and a phenol/chloroform (1 : 1) extraction precededthe isopropanol precipitation step (Sambrook et al. 1989) wasTn5, Tn501 and mini Mu carried on promiscuous IncP1
plasmids. For comparative purposes, the small mobilizable used throughout for both small and large scale Z. mobilisplasmid DNA isolations. Large and mini plasmid prep-plasmids pBR322 and pACYC184 carrying Tn5tac1 and
Tn501 respectively, were used as potential ‘suicide’ vectors arations from E. coli were made by the standard alkaline lysismethod (Sambrook et al. 1989). Large plasmid isolation fromwith the help of Escherichia coli strain S17.1 containing the
tra functions of RP4 on its chromosome (Simon et al. 1983). Ps. aeruginosa was made according to Sinclair and Holloway(1982).
MATERIALS AND METHODSConjugation and strain construction
Bacteria and growth conditionsEscherichia coli strains carrying plasmids of the IncP1 incom-patibility group are known to conjugate efficiently only onBacterial strains, plasmids and transposons used in this study
are described in Table 1. Zymomonas mobilis strains were solid surfaces (Guiney and Lanka 1989). All matings wereperformed on filters, at a ratio of donor/recipient of 1 : 3–grown at 30°C, on a chemically-defined minimal medium
(MM) which allows for the isolation of auxotrophs of all 1 : 10, at 30°C. The filters were placed on CM plates for6 h, washed with sterile Ringer’s solution and the bacterialkinds (Galani et al. 1985). This medium with the addition
of 0·5% yeast extract served as a complete medium (CM). suspensions were serially diluted and finally transferred ontothe appropriate selective media. Plasmids pMO75 andEscherichia coli strains were routinely cultured at 37°C, on
Luria agar (LA) or Luria broth (LB) and M9 supplemented pMO22 originally harboured in Ps. aeruginosa PAO5 andPAO11 respectively, were kindly provided by Professor B.W.with the appropriate nutrients when auxotrophy was exam-
ined (Sambrook et al. 1989). For Pseudomonas aeruginosa Holloway (Monash University, Australia) and are derivativesof plasmid R91.5 carrying Tn5 (Kmr) and Tn501 (Merr)strains, minimal medium, nutrient agar and nutrient yeast
broth were as described by Haas and Holloway (1976). The respectively (Whitta et al. 1985). Plasmid pULB113, an RP4-miniMu3A fusion product (Van Gijsengen and Toussaintantimicrobial agents used and their final concentrations (mg
ml−1) were : for Z. mobilis markers, carbenicillin (Cb) 500, 1982), was carried by E. coli MXR strain (kindly provided byDrs F. Van Gijsegen and A. Toussaint, Universite Libre dekanamycin (Km) 100, tetracycline (Tc) 20, rifampicin (rif) 20,
chloramphenicol (Cm) 100, novobiocin (nov) 40 and mercuric Bruxelles, Belgium). Plasmids pBR322 : :Tn5tac1 (Chow andBerg 1988) and pACYC184 : :Tn501 (Osborn et al. 1993)chloride (Mer) 40 ; for E. coli and Ps. aeruginosa ampicillin
100, carbenicillin 500, kanamycin 100, tetracycline 20 and were kindly provided by Dr D. Berg, Washington UniversityMedical School, St Louis, USA, and Dr A.M. Osborn,mercuric chloride 40.School of Life Sciences, University of Liverpool, UK, respec-tively. Plasmids pRK2013 and RP4 were harboured in E.
DNA isolationcoli ED8767, and similarly pMO75 and pMO22 were alsotransferred by transformation into E. coli strain ED8767Chromosomal DNA was extracted from parental Z. mobilis
stains or transconjugants showing stability for the transposon (Cohen et al. 1972), whereas plasmids pBR322 : : Tn5tac1 andpACYC184 : : Tn501 were transferred to the mobilizing E.markers by the method of Byun et al. (1986) for large scale
preparations. For mini screens of chromosomal DNA the coli strain S17.1 (Simon et al. 1983). The resulting strainswere used as donors in all subsequent conjugation experi-following method was employed : 3 ml of culture cells were
pelleted in microfuge tubes and the pellet was resuspended ments with Z. mobilis strains as recipients. All Z. mobilisstrains used were spontaneous double mutants, resistant toin 200 ml of TAE buffer 1×, pH 7·9 (Tris–acetate 40 mmol
l−1, EDTA 1 mmol l−1). Two volumes (400 ml) of lysis novobiocin and rifampicin, produced stepwise. Strain CU10was further subjected to MNNG treatment (Typas and Galanisolution (50 mmol l−1 Tris–Cl, 3% SDS, pH 12·6) were
added, the suspension was briefly agitated and incubated for 1992) and an isolate which had lost simultaneously plasmidspZMO1–5, has been used throughout this work. Matings15 min at 65°C. To the clear lysate 50ml of ice-cold 5 mol l−1
Table 1 Bacterial strains, plasmids and transposons used in this study (the transposon genetic marker is underlined)—––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––
StrainBacterial species designation Chromosomal markers* Plasmids and related markers* Source of reference—––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––Zymomonas mobilis ATCC 10988 rif r novr pZMO1, pZMO2, pZMO3, this work
pZMO4, pZMO5, pZMO6CU1 rif r novr pZMO1, pZMO2, pZMO3, —, Drainas et al. 1984
pZMO5, pZMO6CU10 rif r novr —, —, —, —, —, pZMO6 this workCP4 rif r novr pZMO6a† this work
Pseudomonas aeruginosa PAO 5 trpC/D 54, rif-5, ton-1 pMO75 (R91.5 : :Tn5), CbrKmr Whitta et al. 1985tra¦¦
PAO 11 trpC/D 54, nal-19 pMO22 (R91.5 : :Tn501) Cbr Carrigan et al. 1978Merr tra¦¦
Escherichia coli ED 8767 F−, recA56, supE44, supF58, Sambrook et al. 1989hsd3(r−
B , m−B ), galK2, metB1
— — RP4 Apr Kmr Tcr tra¦¦ Thomas 1989— — pRK2013 Kmr tra¦¦ Figurski and Helinski 1979— — pMO75 (R91.5 : :Tn5), Cbr Kmr this work— — tra++
pMO22 (R91.5 : :Tn501) Cbr this workMerr tra¦¦
S17.1 RP4-2-Tc : :Mu-Km : :Tn7 Simon et al. 1983— — pBR322 : :Tn5tac1, Kmr Chow and Berg 1988— — pACYC184 : :Tn501 Apr Merr Osborn et al. 1993
Cmr
MXR F−, galE, recA456, Dprolac Van Gijsengen andToussaint 1982
* Gene marker symbols and inhibitor concentrations are shown in Materials and Methods. The transposon marker is underlined in allcases.† pZMO6a, Very similar to pZMO6 of ATCC 10988 according to restrictional analysis (Pappas and Typas, unpublished).
with the Z. mobilis recipient strains and conjugation fre- natural resistance (Sprenger et al. 1993), to screen for anti-biotic sensitivity. Loss of resistance to these antibiotics couldquencies obtained are listed in Table 2.be due to insertion of the transposable element at the respec-tive locus(i). Enrichment of auxotrophs produced by mini
Transposon insertion mutagenesis and foreignMu insertions was achieved as follows : cells of pULB113-
plasmid stabilitytransconjugant cultures were pelleted, washed and used toinoculate liquid Z. mobilis MM supplemented with a mixtureTransconjugants selected from plates containing the appro-
priate antibiotics were further examined for transposition of amoxycillin (400 mg ml−1) and clavulanic acid (100 mgml−1), with an inoculum size not exceeding 0·5% of the freshevents. At least 500 transconjugant colonies from each mating
were inoculated into multiwell plates containing 200 ml of medium. The cell suspension was kept for 20 h at 27°C,during which no growth occurred. The cells were then har-CM with the addition of the appropriate antibiotic for each
transposable element. The cultures were incubated at 30°C vested, washed and plated on CM selective media containingkanamycin and tetracycline, from where independent col-for 8–12 h and were then replicated with a 48-pin replica
plater into new multiwell plates until they had reached a onies could be examined for auxotrophic traits.Foreign plasmid marker stability in Z. mobilis was esti-minimum of 50 division cycles in the presence of the anti-
biotics. At this stage they were replica-transferred onto CM mated by measuring the inheritance of the antibiotic resist-ance plasmid marker for up to 200 generations, after growthand MM, to screen for auxotrophy, or CM with the addition
of one of the several antibiotics to which Z. mobilis exhibits under selective or non-selective conditions. The percentage
Table 2 Conjugation frequencies of matings with Escherichia coli DNA labelling and hybridizationdonor strains which contained plasmids carrying transposons and
DNA was Southern transferred onto Hybond-N (Amersham)Zymomonas mobilis recipientsnylon using a Vacugene blotter (Pharmacia LKB) according—–––––––––––––––––––––––––––––––––––––––––––––––––––––to the manufacturer’s instructions. To use as gene probes forDonor Recipient Conjugation
(E. coli) (Z. mobilis)* frequencies† Tn1, Tn5, Tn501 and mini Mu, the following fragments—––––––––––––––––––––––––––––––––––––––––––––––––––––– were isolated and labelled : the 2·7 kb PstI fragment of Tn1ED 8767 pMO75 ATCC 10988 3×10−7 from RP4, the 2·7 kb Bgl/II of Tn5 from pMO75, the 2·9 kb
CU1 5×10−7PstI fragment of mini Mu from pULB113, and the 1·9 kb
CU10 7×10−7
SalI of Tn501 from pACYC184 : :Tn501. The latter wasCP4 1×10−6
found to bear the Tn501 insertion approximately 50 bp down-stream of the SalI site of pACYC184. As a result the probeED 8767 pMO22 ATCC 10988 1×10−7
almost exclusively contains Tn501 sequences. The fragmentsCU1 6×10−7
were digoxigenin-UTP-labelled using the Boehringer non-CU10 2·2×10−7
radioactive labelling kit according to the manufacturer’sCP4 1×10−6
instructions (Anon. 1989). Pre-hybridization and hybrid-ED 8767 RP4 ATCC 10988 2×10−5 ization were performed by standard methods (Boehringer
CU1 3·3×10−5 Mannheim, protocol) using high stringency conditions atCU10 1·5×10−4 68°C. 32P standard labelling methods were employed whereCP4 3·2×10−3
necessary (Sambrook et al. 1989). Colony hybridization wasperformed as previously described (Scordaki and Drainas
MXR pULB113 ATCC 10988 1·2×10−5
1990).CU1 2×10−4
CU10 3·4×10−4
CP4 1×10−3
RESULTS
S17.1 pBR322 : :Tn5tac1 ATCC 10988 ³10−8
Inter-species conjugationCU1 2·3×10−7
CU10 6·1×10−7
The choice of appropriate complementary markers for Z.CP4 1·2×10−6
mobilis, Ps. aeruginosa and E. coli was a difficult task as theformer exhibits a high level of natural resistance to severalS17.1 pACYC184 : :Tn501 ATCC 10988 ³10−8
antimicrobial agents (Typas and Galani 1992) and the plas-CU1 ³10−8
mid-carrying E. coli or Ps. aeruginosa strains listed in TableCU10 1·3×10−7
CP4 3·1×10−7 1 are naturally resistant to a number of complementary markers.—––––––––––––––––––––––––––––––––––––––––––––––––––––– From over 30 different antibiotics for which Z. mobilis is* All Z. mobilis recipient strains used in matings with Tn501- naturally resistant, only tobramycin (at 20 mg ml−1) could becontaining donors were sensitive to HgCl2 (Mers mutants, see also used as a counter selection agent against Ps. aeruginosa, intext). matings between these bacterial species. In general, con-† Conjugation frequencies represent the average of at least five jugations with Ps. aeruginosa donors and Z. mobilis recipientsindependent experiments in each case.
failed to give transconjugants. To examine the nature of thisfailure, drops from the supernatant fluid of each parentalculture were added on a plate overlayed with the other strain.It was observed that the Z. mobilis supernatant fluid caused aof transconjugant Z. mobilis colonies, which hybridized to the
appropriate transposon probe by in situ colony hybridization strong inhibition of Ps. aeruginosa growth. Such a phenom-enon has never been observed before when E. coli single oror dot blot was also determined.double donors were used. As all the Z. mobilis strains used inthis work produce a minimum of 5% (v/v) ethanol at the
Restriction enzyme digestions and agarose gelspecific growth stage used for matings, a 5% ethanol water
electrophoresissolution was used against Ps. aeruginosa to examine its effect.The suspension used did not inhibit Ps. aeruginosa to theAll enzymes were obtained from Boehringer Mannheim,
Biolabs, BRL or Minotech and digestions were made accord- same extent as the Z. mobilis supernatant fluid, and thus, thesuggested excretion of bacteriocins by Z. mobilis (Haffie et al.ing to manufacturers’ instructions. Digests were separated in
0·8% BRL ultrapure agarose gels using standard methods 1985) may be an explanation for this inhibitory effect. Toavoid such ‘incompatibility’ problems between Z. mobilis and(Sambrook et al. 1989).
Ps. aeruginosa, plasmids pMO75 and pMO22 were transferred contrast, transconjugants resistant to the pULB113 selectionmarkers (TcrKmr) were very stable from the beginningby transformation into E. coli strain ED8767. As shown in
Table 2, donor strains with pMO75, pMO22, (99%), under both selective or non-selective conditions, andcould be maintained as such for at least 200 generations thatpBR322 : : Tn5tac1 and pACYC184 : : Tn501 gave trans-
conjugants at similar frequencies, varying from 10−6 to 10−8 were examined. This high maintenance level of pULB113was in sharp contrast with the instability exhibited by RP4depending on the recipient strain, with the mobilizable plas-
mids showing slightly lower frequencies when compared with which was lost after only 40 generations under non-selectiveconditions (Fig. 1).the tra proficient pMO75 and pMO22. Attempts to increase
the conjugal transfer of pMO75 and pMO22 with the help ofpRK2013 gave similar results in triparental matings and thus,
Tn5 and Tn 501 transpositionsonly results of biparental matings with Z. mobilis are describedin Table 2. Contrary to the above plasmids, both RP4 and Zymomonas mobilis transconjugants grown under selective
pressure for pBR322 : : Tn5tac1 and pACYC184 : : Tn501RP4 : :mini Mu (pULB113) Z. mobilis transconjugantsranged at much higher frequencies (3·2×10−3–1·2×10−5). were not only rare but also extremely unstable, hence the
absence of any of these in further examinations. Out of 1800From the conjugation frequencies obtained it is clear that thebest Z. mobilis recipient was strain CP4 and the worst strain KmrCbr pMO75 and 1500 MerrCbr pMO22 transconjugant
colonies selected for the transposon marker for over 200ATCC 10988, with the two cured strains CU1 and CU10giving intermediate conjugation frequencies. generations, only 45 (2·5%) for pMO75 and 27 (1·8%) for
pMO22 remained stable. Colony hybridization of all thesetransconjugants with the Tn5 and Tn501 probes indicated
Plasmid stabilitythe presence of the corresponding transposable elements. TheTn1 probe, indicative of the carbenicillin resistance locus ofForeign plasmid stability was examined for up to 200 gen-
erations, by analysing over 100 colonies of randomly isolated plasmid R91.5, failed to hybidize with any of the colonies,thus suggesting that plasmid R91.5 was not maintained in Z.transconjugants from each mating amongst E. coli donors and
all the Z. mobilis strains listed in Table 1 used as recipients. mobilis cells.To detect possible transpositional insertions of Tn5 andThe majority of Z. mobilis transconjugants carrying pMO75,
pMO22, pACYC184 : : Tn501 or pBR322 : : Tn5tac1 were Tn501 into the DNA of Z. mobilis native plasmids, plasmidDNA extracts from all stable transconjugants were examinedextremely unstable under both selective or non-selective con-
ditions. Both the carrier plasmid markers and the transposon either electrophoretically or through dot blot hybridizationsfor the presence of Tn5 or Tn501. It appeared that no inser-markers were lost after only 5–15 generations (see Fig. 1). Intion had occurred in the Z. mobilis plasmids and therefore itcan be concluded that the transposons had been transferredonto the chromosome. It should be noted here that the nativeplasmid pZMO6 of ATCC 10988 and the corresponding insize of CP4, were found to hybridize strongly with the Tn501probe (see Fig. 2), thus indicating that the above plasmidscarry homologous regions with Tn501. However, there isgood evidence (data not shown), that the mercuric chlorideresistance locus in Z. mobilis is cryptic and not expressedunder high mercuric salt concentrations. Nevertheless, toraise no ambiguity, all Z. mobilis recipients in mating pairswith donors carrying Tn501 were selected to be Mers (Table2).
EcoRI chromosomal DNA digestions of all Tn5- and20
0
100
0
Number of generations
% P
lasm
id m
ark
er
sta
bil
ity
50
90
80
70
60
40
30
20
10
0
10
20
30
40
50
60
70
80
90
10
0
11
0
12
0
13
0
14
0
15
0
16
0
17
0
18
0
19
0
Tn501-carrying Z. mobilis strains were Southern-transferredFig. 1 Foreign plasmid marker stability in transconjugants of and hybridized with the corresponding probes. In all cases aZymomonas mobilis ATCC 10988 grown under non-selective high molecular weight band appeared to hybridize strongly.conditions for 200 generations. One hundred randomly isolated
In the case of Tn501, EcoRI digestion should normally gen-transconjugants from each mating were examined. Resultserate an internal Tn501 2·3 kb fragment hybridizing with therepresent the average of three independent experiments.probe, which was not detected (data not shown). Further-pACYC184 and pBR322 curves were identical, thus formore, HindIII digestions resulted in all cases in a 4·4 kbreasons of clarity only the latter is presented. ×,fragment, which also could not be accounted for based on thepBR322 : :Tn5tac1 ; E, pMO22 ; r, pMO75 ; Ž, RP4 ; �,
pULB113 (RP4 : :mini Mu) transposon map (Whitta et al. 1985), if the transposon had
vast majority of Tn5- and Tn501-carrying stable trans-conjugants gave identical hybridization patterns with theenzyme used. From all Tn5 and Tn501 transconjugantsexamined for auxotrophy or antibiotic resistance only fourauxotrophs of the former (½0·22%) and one auxotroph andone cephalosporin-sensitive mutant (½0·1%) of the latterremained stable.
mini Mu transpositions and mutagenesis
Once the stability of pULB113 (mini Mu) transconjugantswas established (Fig. 1), large numbers of independent trans-conjugants were exposed to auxotrophic enrichment (seeMaterials and Methods). The enrichment procedureincreased the numbers of putative auxotrophs by 15-fold (i.e.from ½2·6% without enrichment to 30% with enrichment).Many of these mutants were leaky auxotrophic, and afterseveral screenings, a total of 156 independent stable auxo-
Fig. 2 HindIII chromosomal DNA digestions of Zymomonas trophs were isolated and their auxotrophy was determined.mobilis strain CU1 hybridized with the 1·9 kb SalI-labelled
They were found to belong to different types with polyauxo-Tn501-specific probe. Lanes : 1, l HindIII ; 2–6, Tn501-carryingtrophs (52%), cysteine (18·6%), methionine (9%), isoleucinestrains ; 7, parental strain CU1 ; 8, HindIII restriction pattern of(7%) requiring auxotrophs being the most predominant. ToCU1 plasmid pZMO6. Arrowhead indicates the 4·4 kb band-verify the type of insertion, chromosomal DNA from ran-signal unique to CU1 isolates carrying the Tn501 insertiondomly selected pULB113-containing prototrophic isolatesand from all the 156 auxotrophs was PstI digested and sub-jected to hybridization with both the mini Mu and Tn1
integrated as a whole (Fig. 2). Similarly, BglII chromosomal probes. Signals appeared as expected. On the contrary, dotDNA digestions of all Tn5-carrying Z. mobilis strains failed blot analysis of plasmid preparations from each of the exam-to generate the internal 2·7 kb BglII of Tn5, except in one ined isolates failed to give positive hybridization signals withcase (Fig. 3, lane 6), which presented an even lower molecular the same probes. This, along with the retention of all RP4weight signal which could not be accounted for. In total, the antibiotic resistance markers, strongly indicates that the entire
pULB113 had been incorporated onto the Z. mobilis chro-mosome.
The independently isolated auxotrophs provided an excel-lent material to examine if mini Mu had a preferential site ofinsertion, especially within the group of auxotrophs for thesame requirement. Of all restriction endonucleases used itwas SmaI, KpnI, PvuII, ClaI, SalI and SphI which alloweddifferent polymorphic patterns to appear. The enzymes usedhere were selected in such a way that they cleaved the RP4backbone, as close to the mini Mu insertion site as possible,in order to obtain a range of polymorphic fragments. Chro-mosomal DNA from all the cysteine (28) and methionine(14) requiring CP4 auxotrophs were digested with variousenzymes and were hybridized with the mini Mu probe. ThePvuII hybridization patterns of the above mentioned auxo-trophs are presented in Figs 4 and 5, respectively. Thehybridization patterns in the same auxotrophic group clearlyindicate that the vast majority of these auxotrophs were pro-Fig. 3 BglII chromosomal DNA digestion of Zymomonas mobilisduced by insertions at different sites of the chromosome, asstrain ATCC 10988 hybridized with the 2·7 kb BglII probededuced from a combination of restriction enzymes used.from Tn5. Lanes : 1, l HindIII ; 2–10, nine independentlyOnly two pairs of isolates from the cysteine (Fig. 4, lanes 5isolated Tn5-carrying Z. mobilis transconjugants ; 11, parental strain
ATCC 10988 and 12, 6 and 23) and the methionine (Fig. 5, lanes 7 and
Fig. 4 PvuII chromosomal digestionsfrom independently isolated cysteine-requiring Zymomonas mobilis CP4transconjugants hybridized with thelabelled PstI 2·9 kb fragment of mini Mu.Lanes : 1, l HindIII ; 2–29, the independentCP4 transconjugants with pULB113insertions ; 30, parental strain CP4 restrictedwith the same enzyme
8, 11 and 14) requiring auxotrophs had similar patterns, (Skotnicki et al. 1982 ; Ingram et al. 1984 ; Buchholz andEveleigh 1986 ; Brestic-Goachet et al. 1987). This was furthersuggesting similar insertion sites.substantiated from our experiments with pBR322 : : Tn5tac1and pYCAC184 : : Tn501. It was also shown that the R91.5
DISCUSSIONconstructs (IncP10) containing Tn5 and Tn501, which wereused with great success in many other bacteria (Carrigan etConjugal transfer of the broad-host-range plasmids R68.45
and RP1 (both IncP1 type) from E. coli donors to Z. mobilis al. 1978 ; Whitta et al. 1985), were very unstable in Z. mobilistransconjugants. In our hands, transconjugants with trans-recipients has been reported previously (Skotnicki et al. 1982 ;
Stokes et al. 1983 ; Walia et al. 1984 ; Buchholz and Eveleigh poson insertions were isolated from all conjugation pairs withIncP1 transposon-carrying plasmids used here. As shown in1986). In all these experiments the transfer frequencies were
extremely variable depending on the donor strain that was Table 2, the conjugation frequencies of RP4 and RP4 : :miniMu (pULB113) were very similar. Thus, the structure of theused. Our results here clearly show that it is not only the
donor strain (for example no interspecies conjugation RP4 derivative used did not effect the efficiency of its transferas it does in other bacteria (Saulnier et al. 1987). The sur-between Ps. aeruginosa and Z. mobilis can take place), but also
the recipient strain used which influences the conjugation prisingly high stability of pULB113, therefore, as comparedwith the great instability of its RP4 parental plasmid (Fig. 1)frequencies. The plasmid-lacking recipients CU1 and CU10
exhibited a small but reproducible increase in conjugation can only be attributed to the presence of the mini Musequences. As no autonomous R91.5 or pULB113 plasmidsfrequencies compared with their parental strain ATCC
10988. The reasons for this increase are not easily understood were detected in plasmid preparations of Z. mobilis stabletransconjugants and the hybridization patterns of restrictedas the former lacks only one plasmid species (pZMO4 ; Drainas
et al. 1984) and the latter five plasmid species out of six chromosomal DNA invariably gave positive signals with thetransposon probes used, it is evident that transposition(pZMO1–5 ; see Table 1). However, as the strain CP4, which
contains only one large size plasmid (very similar to pZMO6, phenomena occur readily in Z. mobilis. In the majority of thestable Tn5- or Tn501-carrying transconjugants examined, noaccording to restriction analysis experiments ; Pappas and
Typas, unpublished results), clearly showed the best recipient multiple transpositions were observed. This is in agreementwith the mode of action of both Tn5 and Tn501 as it is knownability in conjugation, it remains to be seen if a plasmidless
Z. mobilis CP4 strain will be the best recipient for this bacterial they both integrate only once into the genome of bacteria(Berg and Berg 1983 ; Whitta et al. 1985). However, the factspecies. Foreign plasmids of any other incompatibility group
apart from IncP and IncQ invariably failed to propagate in Z. that all the Tn5- and Tn501-carrying stable isolates gaveunexpected hybridization patterns (according to the restric-mobilis recipients after biparental or triparental conjugations
and Eveleigh 1986 ; Typas and Galani 1992). The efficiencyof auxotroph mutant production with the use of mini Muand the stability of these mutants unequivocally proves thattransposon mutagenesis is an extremely powerful tool formutant production in Z. mobilis. The fact that auxotrophs ofthe same requirement gave different patterns of hybridizationwith mini Mu (Figs 4 and 5) clearly indicates that thistransposable element has no insertional specificity in thebacterium. Thus, the ability to integrate gene fusion plasmidDNA into the chromosome as described in this study pro-vides a simple method for analysing cis-acting regulatoryelements in this organism and significantly increases the rep-ertoire of tools available for its genetic analysis. As thesemethods for chromosomal integration of foreign DNA do notrequire homologous recombination and all the auxotrophmutants have acquired a scorable marker on their chro-mosome, the possible transfer of chromosomal markers orretrotransfer mediated by pULB113 previously observed inE. coli (Mergeay et al. 1987) can also be examined and areunder investigation currently in this laboratory.
ACKNOWLEDGEMENTS
This work was partially supported by grant 4/1367 (ResearchCommittee, University of Athens). The authors wish to thankProfessor B.W. Holloway and Professor N. Panopoulos forcritically reading this manuscript. K-MP wishes to thank the
Fig. 5 PvuII chromosomal DNA digestions from independently National Hellenic Grant Institute (I.K.Y.) for a scholarshipisolated methionine-requiring Zymomonas mobilis CP4 during the course of this worktransconjugants hybridized with the labelled 2·9 kb PstI fragmentof mini Mu. Lanes : 1, parental strain CP4 ; 2–15, the independentCP4 transconjugants with pULB113 insertions
REFERENCES
Afendra, A.S. and Drainas, C. (1987) Expression and stability of ation maps of the corresponding plasmids and transposons)recombinant plasmid in Zymomonas mobilis and Escherichia coli.clearly indicates that genetic rearrangements take place in Z.Journal of General Microbiology 133, 127–134.
mobilis. Similarly, the identical hybridization patterns usually Anon. (1992) Biochemica Boehringer Mannheim. Digoxigeninobserved with all restriction enzymes used within each class DNA labeling and detection nonradioactive : Application Manual.of these transconjugants suggests a preferential insertion site Berg, D.E. and Berg, C.M. (1983) The prokaryotic transposablefor Tn5 and Tn501. In the case of Tn501, Ogale and Deob- element Tn5. Biotechnology 1, 417–435.agkar (1988) have previously reported the presence of a region Brestic-Goachet, N., Gunasekaran, P., Cami, B. and Baratti, J.
(1987) Transfer and expression of broad-host-range plasmids inconferring resistance to HgCl2 on a plasmid of approximatelyZymomonas mobilis. Biotechnology Letters 9, 13–18.20 kb in ATCC 10988. However, this estimation is not in
Browne, G.M., Skotnicki, M.L., Goodman, A.E. and Rogers, P.L.agreement with the size of any of the native plasmids of the(1984) Transformation of Zymomonas mobilis by a hybrid plasmid.above Z. mobilis strain. The strong hybridization of the largePlasmid 12, 211–214.plasmid (pZMO6) of this strain with the Tn501 probe there-
Buchholz, S.E. and Eveleigh, D.E. (1986) Transfer of plasmids tofore provides a readable marker for this plasmid which wasan antibiotic-sensitive mutant of Zymomonas mobilis. Applied and
thought previously to be cryptic. Environmental Microbiology 52, 266–270.Auxotroph mutants of Z. mobilis have been previously Buchholz, S.E. and Eveleigh, D.E. (1990) Genetic modification of
produced by different research groups and these mutants Zymomonas mobilis. Biotechnology Advances 8, 547–581.were invariably derived by MNNG mutagenesis. Never- Byun, M.O.K., Kaper, T.B. and Ingram, L.O. (1986) Constructiontheless, problems of mutant stability were apparent in all of a new vector for the expression of foreign genes in Zymomonas
mobilis. Journal of Indian Microbiology 1, 9–15.these cases (Goodman et al. 1982 ; Walia et al. 1984 ; Buchholz
Carey, T., Sewell, G.M., Osman, Y.A. and Ingram, L.O. (1987) Montenecourt, B.S. (1985) Zymomonas, a unique genus of bacteria.Expression of the lactose transposon (Tn951) in Zymomonas In Biology of Industrial Microorganisms ed. Demain, A.L. andmobilis. Applied and Environmental Microbiology 45, 1163–1168. Solomon, N.A. pp. 261–289. Menlo Park, CA : Benjamin-Cum-
Carrigan, J.M., Helman, Z.M. and Krishnapillai, V. (1978) Trans- mings.fer-deficient mutants of the narrow-host-range plasmids R91.5 of Ogale, S.S. and Deobagkar, D.N. (1988) A high molecular weightPseudomonas aeruginosa. Journal of Bacteriology 135, 911–919. plasmid of Zymomonas mobilis harbors genes for HgCl2 resistance.
Cerda-Olmedo, E., Hanawalt, P.C. and Guerola, N. (1968) Muta- Biotechnology Letters 10, 43–48.genesis of the replication point by nitrosoguanidine : map and Osborn, A.M., Bruce, K.D., Strike, P. and Ritchie, D.A. (1993)pattern of replication of the Escherichia coli chromosome. Journal Polymerase chain-restriction fragment length polymorphismof Molecular Biology 33, 705–710. analysis shows divergence among mer determinants from gram–
Chow, W.-Y. and Berg, D.E. (1988) Tn5tac1, a derivative of trans- negative soil bacteria indistinguishable by DNA-DNA hybrid-poson Tn5 that generates conditional mutations. Proceedings of ization. Applied and Environmental Microbiology 59, 4024–4030.the National Academy of Sciences USA 85, 6468–6472. Sambrook J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Clon-
Cohen, S.N., Chang, A.C.Y. and Hsu, L. (1972) Nonchromosomal ing : A Laboratory Manual, 2nd edn. Cold Spring Harbor, NY :antibiotic resistance in bacteria : transformation of E. coli by R- Cold Spring Harbor Laboratory Press.factor DNA. Proceedings of the National Academy of Sciences USA Saulnier, P., Hanquier, J., Jaoua, S., Reichenbach, H. and Guespin-69, 2110–2114. Michel, J.F (1988) Utilization of Inc-P1 plasmids as vectors for
Drainas, C., Typas, M.A. and Kinghorn, J.R. (1984) A derivative transposon mutagenesis in Mycobacteria. Journal of Generalof Zymomonas mobilis ATCC 10988 with impaired ethanol pro- Microbiology 134, 2889–2895duction. Biotechnology Letters 5, 37–42. Schoonejans, E. and Toussaint, A. (1983) Utilization of plasmid
Eveleigh, D.E., Stokes, H.W. and Dally, E.L. (1983) Recombinant pULB113 (RP4 : :mini Mu) to construct a linkage map of ErwiniaDNA approaches for enhancing the ethanol productivity of Zymo- carotovora subsp. chrysanthemi. Journal of Bacteriology 154, 1489–monas mobilis. In Organic Chemicals from Biomass ed. Wise, D.L. 1492.pp. 69–91. Menlo Park, CA : Benjamin-Cummings. Scordaki, A. and Drainas, C. (1990) Analysis and stability of Zymo-
Figurski, D.H. and Helinski, D.R. (1979) Replication of an origin- monas mobilis ATCC 10988 plasmid pZMO3. Plasmid 23, 59–66.containing derivative of plasmid RK2 dependent on a plasmid Simon, R., Priefer, U. and Puhler, A (1983) A broad host rangefunction provided in trans. Proceedings of the National Academy mobilization system for in vivo genetic engineering : transposonof Sciences USA 76, 1648–1652. mutagenesis in gram-negative bacteria. Bio/Technology 1, 784–
Galani, I., Drainas, C. and Typas, M.A. (1985) Growth require- 790.ments and the establishment of a chemically defined minimal Sinclair, M.I. and Holloway, B.W. (1982) A chromosomally locatedmedium in Zymomonas mobilis. Biotechnology Letters 7, 673–678.
transposon in Pseudomonas aeruginosa. Journal of Bacteriology 151,Goodman, A.E., Rogers, P.L. and Skotnicki, M.L. (1982) Minimal
569–579.medium for the isolation of auxotrophic Zymomonas mobilis.
Skotnicki, M.L., Lee, K.J., Tribe, D.E. and Rogers, P.L. (1982)Applied and Environmental Microbiology 44, 496–498.
Genetic alteration of Zymomonas mobilis for ethanol production.Goodman, A.E., Strzelecki, A.T. and Rogers, P.L. (1984) For-
In Genetic Engineering of Microorganisms for Chemicals ed. Hollen-mation of ethanol from lactose by Zymomonas mobilis. Journal ofder, A., DeMoss, R.D., Kaplan, S., Koninsky, J., Savage, D. andBiotechnology 1, 219–228.Wolfe, R.S. pp. 271–290. New York : Plenum Press.Guiney, D.G. and Lanka, E. (1989) Conjugative transfer of IncP
Sprenger, G.A., Typas, M.A. and Drainas, C. (1993) Geneticsplasmids. In Promiscuous Plasmids of Gram-negative Bacteria ed.and genetic engineering of Zymomonas mobilis. World Journal ofThomas, C.M. pp. 27–56. Academic Press.Microbiology and Biotechnology 9, 17–24.Haas, D and Holloway, B.W. (1976) R factor variants which
Stokes, H.W., Dally, E.L., Yablonsky, M.D. and Eveleigh, D.E.enhanced sex factor activity in Pseudomonas aeruginosa. Molecular(1983) Comparison of plasmids in strains of Zymomonas mobilis.and General Genetics 144, 243–251.Plasmid 9, 138–146.Haffie, T.L., Louie, P.W. and Khachatourians, G.G. (1985) Iso-
Thomas C.M. (ed.) (1989) Promiscuous Plasmids of Gram-negativelation of non inhibitory strains of Zymomonas mobilis. Applied andBacteria. Academic Press.Environmental Microbiology 49, 1007–1009.
Typas, M.A. (1986) Physiological and genetical studies of the bac-Ingram, L.O., Carey, V.C., Dombeck, K.M., Holt, A.S., Holt,terium Zymomonas mobilis for increased ethanol productionW.A., Osman, Y.A. and Walia, S.K. (1984) Biochemical andand/or SCP production following the fermentation of low valuegenetical improvement of Zymomonas mobilis. Biomass 6, 131–agricultural products. In Biotechnology Engineering Programme,143.Final Report ed. Magnien, E. pp. 673–678. European CommunityKarsch, T., Stahl, U. and Esser, K. (1983) Ethanol production byPublications.Zymomonas and Saccharomyces, advantages and disadvantages.
Typas, M.A. and Galani, I. (1992) Chemical and UV mutagenesisEuropean Journal of Applied Microbiology and Biotechnology 18,in Zymomonas mobilis. Genetica 87, 37–45.387–391.
Van Gijsengem, F. and Toussaint, A. (1982) Chromosome transferMergeay, M., Lejeune, P., Sadouk, A., Gerits, J. and Fabry, L.and R-prime formation by an RP4 : :mini Mu derivative in Esch-(1987) Shuttle transfer (or retrotransfer) of chromosomal markerserichia coli, Salmonella typhimurium, Klebsiella pneumoniae andmediated by plasmid pULB113. Molecular and General Genetics
Walia, S.K., Carey, V.C., All, B.P. and Ingram, L.O. (1984) Self- Whitta, S., Sinclair, M.I. and Holloway, B.W. (1985) Transposontransmissible plasmid in Zymomonas mobilis carrying antibiotic mutagenesis in Methylobacterium AM1 (Pseudomonas AM1). Jour-resistance. Applied and Environmental Microbiology 47, 198–200. nal of General Microbiology 131, 1547–1549.
Walker, M.J. and Pemberton, J.M. (1988) Construction of trans- Yanase, H., Kotani, T. and Tonomura, K. (1986) Transformation ofposons encoding genes for b-glucosidase, amylase and poly- Zymomonas mobilis with plasmid DNA. Agricultural and Biologicalgalacturonate trans-eliminase from Klebsiella oxytoca and their Chemistry 59, 3139–3144.expression in a range of gram-negative bacteria. Current Micro-biology 17, 69–75.
