APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1986, p. 366-370 Vol. 52, No. 20099-2240/86/080366-05$02.00/0Copyright X 1986, American Society for Microbiology

Transfer of Plasmids to an Antibiotic-Sensitive Mutant ofZymomonas mobilist

STEVEN E. BUCHHOLZ* AND DOUGLAS E. EVELEIGHDepartment ofBiochemistry and Microbiology, Cook College, Rutgers University, New Brunswick, New Jersey 08903

Received 5 May 1986/Accepted 14 May 1986

Wild-type strains of Zymomonas mobilis exhibit multiple antibiotic resistance and thus restrict the use ofmany broad-host-range plasmids in them as cloning vehicles. Antibiotic-sensitive mutants of Z. mobilis wereisolated and used as hosts for the conjugal transfer of broad-host-range plasmids from Escherichia coli. Suchantibiotic-sensitive strains can facilitate the application of broad-host-range plasmids to the study of Z. mobilis.

Zymomonas mobilis is an aeroduric gram-negativeanaerobe which produces ethanol as a major product offermentation (3, 27). Considerable interest has been gener-ated in the potential of Z. mobilis for commercial ethanolproduction (10, 13), since it can produce alcohol at doublethe rate that yeasts can (17, 19, 22). Though Z. mobilis is anatural fermentative agent in the production of palm winesand pulque (25), the possible large-scale production ofethanol by Z. mobilis has been restricted in part because itssubstrate range is limited to glucose, fructose, and sucrose(27). It should be possible to increase its range of utilizationof substrates by transferring appropriate hydrolase genes toit with recombinant DNA approaches. However, geneticmanipulation of this microorganism has been impeded in partbecause it is resistant to several commonly employed anti-biotics, thus limiting the use of standard broad-host-rangeplasmids as vectors, as the expression of resistance genes isused to monitor the plasmid transfer (26, 28).

It is possible to circumvent this problem in a limitedmanner by using plasmids that code only for antibioticresistance genes to which Z. mobilis is sensitive, such aschloramphenicol or tetracycline. However, this approachlimits both the number of available cloning vectors (23) andthe number of sites available for insertional inactivation. Analternative approach is to develop strains of Z. mobilis thatare sensitive to routinely used antibiotics, thus facilitatingthe use of R-plasmids as cloning vehicles. This latter ap-proach (the selection of antibiotic-sensitive strains of Z.mobilis) and their potential as recipients of broad-host-rangevectors are presented.

MATERIALS AND METHODS

Strains and plasmids. Z. mobilis CP4 and Escherichia coliHB101 have been described (5, 20). Other bacterial strainsand plasmids are listed in Table 1.Media and reagents. Z. mobilis was grown in GYx medium

consisting of D-glucose (2%), yeast extract (2%), and potas-sium monophosphate (0.1%), and E. coli strains were grownin LB medium consisting of tryptone (1%), sodium chloride(1%), and yeast extract (0.5%), or on MacConkey lactoseagar (BBL Microbiology Systems, Cockeysville, Md.), eachin static culture at 28°C. Antibiotics were purchased fromSigma Chemical Co., St. Louis, Mo.

* Corresponding author.t New Jersey Agricultural Experiment Station publication

D-01111-01-85.

Antibiotic-sensitive mutants. Antibiotic-sensitive mutantswere selected after UV irradiation of the naturally antibiotic-resistant Z. mobilis CP4 such that 95% of the cells wereinactivated (21), followed by overnight growth at 28°C inGYx broth containing 3 ,ug of mitomycin C per ml andsubsequent plating on GYx agar. Antibiotic-sensitive mu-tants were detected after transfer of colonies onto platescontaining antibiotics (100 ,ug/ml), with aerobic incubation at280C.MICs. MICs of antibiotics for Z. mobilis CP4 were deter-

mined by spreading 0.1 ml of exponential-phase cells (ap-proximately 5 x 107) on the surface of GYx agar platescontaining antibiotics, followed by aerobic incubation at28°C. The MIC was defined as the lowest concentration ofantibiotic that resulted in lack of growth after 48 h. MICs(micrograms per milliliter) for the various antibiotics were asfollows: ampicillin, 250; gentamicin, 80; kanamycin, 200;nalidixic acid, 1,000; neomycin, 1,250; chloramphenicol,125; rifampin, 10; spectinomycin, 50; tetracycline, 20; strep-tomycin, >7,000.

Plasmid DNA extraction and characterization. Extractionof plasmid DNA from Z. mobilis was performed essentiallyas described by Stokes et al. (26). Plasmid DNA from E. coliwas extracted in the same manner except that 0.3% TritonX-300 was used in place of 2% sodium dodecyl sulfate duringcell lysis. Plasmid DNA was characterized electrophoreti-cally in 0.5% agarose gels by using TAE buffer (40 mM Trishydrochloride, 40 mM acetate, 2 mM EDTA [pH 7.8]) (20).

Transfer of plasmids to Z. mobilis CP4.45. Conjugation ofE. coli and Z. mobilis was facilitated by the use of amobilizing strain, E. coli HB101 containing pRK2013 (Table

TABLE 1. E. coli strains and plasmids

Strain Plasmid Characteristics Source or reference

J53 RP4 Amr Kmr Tcr 16

HB101 pGC91.14 Amr Kmr Tcr lac+ 4pRK2013 Mobilizing plasmid 6pKT230 Kmr Smr 1

CSH52 pSUP304 Amr Kmr 23; AgrigeneticsapSUP204 Amr Cmr Tcr 23; Agrigeneticsa

M182 pMON5003 Kmr Smr lac+ Monsanto Co.b

S17-1 Mobilizing strain 23; Agrigeneticsaa Agrigenetics, Boulder, Colo.b Monsanto Co., St. Louis, Mo.

366

on May 6, 2021 by guest

http://aem.asm

.org/D

ownloaded from

ANTIBIOTIC-SENSITIVE Z. MOBILIS 367

TABLE 2. MICs in relation to culture conditions

Antibiotic MIC (p.g/ml) for strain:

Z. mobilis culture type Ampicillin Kanamycin

CP4 CP4.45 CP4 CP4.45

Aerobic; surface culture 250 80 200 60Anaerobic; surface culture 400 150 350 150Aerobic; agar overlay 300 100 300 120

1). Donor, recipient, and mobilizing strain cells were grownin liquid medium overnight, and then 0.1 ml of each wasmixed on plates consisting of nutrient broth (1%), glucose(2%), yeast extract (2%), potassium monophosphate (0.1%),and agar (1.5%). The cells were incubated anaerobically for4 h, resuspended with 0.5 ml of 0.9% sodium chloride, andspread onto selective media. Nalidixic acid (75 ,uglml) orstreptomycin (300 ,ug/ml) was used to select against the E.coli donor strains. Antibiotic-sensitive Z. mobilis cells thatreceived the transferred plasmid were selected by theirresistance to ampicillin (500 ,ug/ml) or kanamycin (500jxg/ml).Transformation of Z. mobilis with plasmid DNA was

performed by a modification of the calcium chloride protocol(20) essentially as suggested by S. K. Walia (personal com-munication). Magnesium chloride (150 mM) was used inplace of calcium chloride, and Z. mobilis cells were incu-bated in GYx for 6 h after heat shock. Selection of trans-formants was as described for selection of Z. mobilis trans-conjugants.

Retransfer of plasmids to E. coli. Broad-host-range plas-mids in Z. mobilis CP4.45 were retransferred via transfor-mation to E. coli S17-1 by the calcium chloride/rubidiumchloride procedure of Maniatis et al. (20). Conjugal transferwas also performed as described previously, except thatMacConkey lactose agar was used to select against Z.mobilis.

RESULTSIsolation and characterization of antibiotic-sensitive strains.

Roughly 5% of 1,100 colonies screened exhibited reducedresistance to two antibiotics, ampicillin and kanamycin. Inall cases, the sensitivity was dual in nature and only toampicillin and kanamycin. One mutant, designated Z.mobilis CP4.45, was selected for further study, as it exhib-ited the greatest sensitivity to both antibiotics. Z. mobilisCP4.45 exhibited MICs (micrograms per milliliter) of 80 forampicillin and 60 for kanamycin, compared to MICs for thewild type of 250 for ampicillin and 200 for kanamycin (seeabove). Over 99% of Z. mobilis CP4.45 cells retained theirsensitivity to both ampicillin and kanamycin after threeconsecutive subcultures on GYx plates lacking antibiotics.In instances of reversion, increased dual resistance wasnoted to both ampicillin and kanamycin; no revertantsexhibiting sensitivity to a single antibiotic were detected.

It should be noted that the level of antibiotic resistance forboth the wild-type and mutant Z. mobilis cultures wasdependent on culture conditions. Z. mobilis, when grownaerobically on plates, is more sensitive to the antibioticstested (ampicillin and kanamycin) than when grownanaerobically (Table 2) and, analogously, when grown withinagar is more resistant than when grown aerobically on thesurface of agar plates.To investigate the possibility that the antibiotic sensitivity

of Z. mobilis CP4.45 was the result of plasmid loss, plasmids

A B C D E F

20.:

14.9Kb:11.8Kb9.5Kb

FIG. 1. Comparison by gel electrophoresis of the plasmid con-tent of the native and antibiotic-sensitive strains of Z. mobilis.Plasmids were isolated by the method of Stokes et al. (26), andelectrophoretic resolution was achieved by using a voltage of 2.5V/cm ir. 0.5% agarose for 18 h. Lanes: A, control containing purifiedplasmids pSUP104 (9.5 kilobases [kb]), pKT210 (11.8 kb), pKT230-Cmr (14.9 kb), and pMON5003 (20.3 kb); B, Z. mobilis CP4containing its native plasmids; C, antibiotic-sensitive strain Z.mobilis CP4.45; D, Z. mobilis CP4.45 containing pSUP304; E,purified pSUP304 from E. coli HB101; F, Z. mobilis CP4.45 con-taining pKT230.

of the mutant and of the wild-type strain (CP4) were com-pared. No differences in the electrophoretic banding patternswere observed between the plasmids of the two strains (Fig.1).

Plasmids transferred to Z. mobilis CP4.45. To determinewhether the antibiotic-sensitive mutant Z. mobilis CP4.45would be useful in further plasmid transfer experiments,broad-host-range plasmids RP4, pGC91.14, pKT230, andpSUP304 (Table 1) were conjugally transferred from E. colito Z. mobilis at rates of 4.9 x 10-6, 4.0 x 10-6, 3.6 x 10-7,and 1.3 x 10-6 transconjugants per donor, respectively. Thetransfer of each plasmid to Z. mobilis CP4.45 was evidencedby expression of antibiotic resistance genes (Table 3). Trans-conjugants were further characterized by correlation withother plasmid-encoded antibiotic resistances (Tcr), plasmidmobility in agarose gel electrophoresis, and retransfer of theplasmids from Z. mobilis to E. coli.The transconjugants were first examined for phenotypic

expression of unselected marker genes (Table 3). Z. mobilisCP4.45 containing plasmids RP4 or pGC91.14 (Amr Km)also expressed enhanced resistance to tetracycline (65,ug/ml) compared with wild-type Z. mobilis (20 pug/ml), inaddition to the selected ampicillin or kanamycin resistances.Plasmids pSUP304 (Amr Km') and pKT230 (Kmr Smr) haveno phenotypic marker genes other than those used in theselection of the plasmids for expression in Z. mobilis.To further test for evidence of the conjugal transfer of

plasmids to Z. mobilis CP4.45, plasmids from both the donorand the Z. mobilis transconjugants were compared byagarose gel electrophoresis. Plasmid DNA of RP4,

VOL. 52, 1986

on May 6, 2021 by guest

http://aem.asm

.org/D

ownloaded from

368 BUCHHOLZ AND EVELEIGH

pGC91.14, pKT230, and pSUP304 from E. coli comigratedwith the plasmids in the Z. mobilis CP4.45 transconjugants(Fig. 1).To evaluate the state of the plasmids conjugally trans-

ferred to Z. mobilis, they were transferred back to E. coli(strains HB101 and S17-1). Plasmids RP4 and pGC91.14were each conjugally retransferred to E. coli HB101 withselection by ampicillin, kanamycin, or tetracycline on Mac-Conkey lactose plates. The transconjugants of E. coli HB101contained only plasmids which appeared phenotypicallyindistinguishable from the plasmids originally transferred toZ. mobilis CP4.45 when compared by expression of antibi-otic resistance and mobility on agarose gels.Although we were unable to accomplish conjugal retrans-

fer of plasmids pKT230 and pSUP304 from Z. mobilis to E.coli, retransformation of E. coli S17-1 by plasmid DNA fromZ. mobilis containing either pKT230 or pSUP304 yielded E.coli transformants. The transformed E. coli strain containedplasmids which appeared indistinguishable from the origi-nally transferred plasmids, as evidenced by plasmid-encodedantibiotic resistances and plasmid mobility on agarose gels.However, no transfer of native Z. mobilis plasmids to E. coliS17-1 by transformation was observed in these experiments.A similar result has been noted for several E. coli strains(28).

Transformation of Z. mobilis by pKT230, pSUP304, andRP4 was accomplished, albeit at a much lower frequency oftransfer (<100/jig of DNA) than was observed for transfor-mation of E. coli (>105/,ug of DNA) by pKT230 andpSUP304. As is routinely found in other bacteria, the effi-ciency of transformation in Z. mobilis is dependent on thesize of the transforming plasmid, with larger plasmids beingtransferred less often than smaller plasmids (data notshown), which appears to be similar to the phenomenonobserved in E. coli (15). Successful transformants of Z.mobilis were found only with those plasmids that were alsoconjugally transferred from E. coli strains (Table 3).

Plasmids such as pMON5003 (B. C. Hemming, personalcommunication) and pSUP204 (23) were not maintained in Z.mobilis when transferred by either method. Conjugation ofZ. mobilis CP4.45 with E. coli M182 containing plasmidpMON5003 resulted in Z. mobilis cells which expressedresistance to kanamycin, but which did not express I-galactosidase. Plasmid pMON5003 could not be demon-strated in Z. mobilis CP4.45 by agarose gel electrophoresisor retransfer to E. coli. Plasmid pSUP204 transferred to Z.mobilis resulted in cells which rapidly lost resistance tounselected antibiotics in serial culture.

DISCUSSION

Z. mobilis is naturally resistant to several antibiotics; thus,the introduction of genes into it through the aegis of broad-host-range plasmids has been restricted by the limited num-ber of antibiotic resistance genes that can be used asselective markers. To facilitate the use of such commonbroad-host-range plasmids, we decided to use antibiotic-sensitive mutants. Since the resistance to antibiotics variedaccording to culture conditions (Table 2), we defined theMIC as the lowest concentration of antibiotic which resultedin lack of growth of cells spread on the surface of agar plates,after 48 h of aerobic incubation at 28°C. The lowered level ofresistance was arbitrarily selected as sensitivity to 100 ,ug ofantibiotic per ml (GYx plates). This level is considerablyreduced from the natural levels of resistance to ampicillinand kanamycin (250 and 200 ,ug/ml, respectively). Such

TABLE 3. Antibiotic resistance conferred by broad-host-rangeplasmids

Antibiotic and MIC (,iLg/ml) for strain:plasmid used Z. mobilis CP4.45 E. coli

AmpicillinNo plasmid 80RP4 750 1,500pGC91.14 750 1,500pSUP304 >2,000 1,000

KanamycinNo plasmid 60RP4 1,200 250pGC91.14 1,200 250pSUP304 1,500 1,250pKT230 1,750 1,000

mutants with simply lowered resistance were quite adequatefor our purposes of gaining the use of broad-host-rangeplasmids.

Z. mobilis is well known to exhibit poor susceptibility tomutagenesis. A variety of mutants of Z. mobilis have beenreported, but these have not been used extensively in geneticanalysis of this organism (2, 12, 14, 22, 29, 30). We wereinitially unsuccessful in isolating antibiotic-sensitive mutantsof Z. mobilis CP4 by use of a single mutagen. However,when we used UV light and mitomycin C in succession, 54mutants were detected in our screening of only 1,100 colo-nies. All antibiotic-sensitive mutants isolated exhibited dualreduced resistance to both ampicillin and kanamycin, withthe most sensitive strain, Z. mobilis CP4.45, exhibiting MICsfor ampicillin and kanamycin of 80 and 60 ,ug/ml, respec-tively. The basis for the dual sensitivity is unknown. How-ever, the usual mechanisms of resistance to ampicillin andkanamycin are clearly distinct (11, 18), and it is unlikely thattwo distinct genetic resistance loci were inactivated in all theantibiotic-sensitive mutants. Eveleigh et al. (9) discussed thepossibility that resistance by Z. mobilis to several antibiotics(see above) may be the result of the inability of the organismto transport aminoglycosides across the cellular membrane.The combined sensitivity to both ampicillin and kanamycinof the antibiotic-sensitive Z. mobilis mutant may be due tosuch an effect on a general mechanism of resistance (forexample, cellular permeability). Mutants exhibiting modifiedpermeability can also exhibit other pleiotropic changes.These changes could be detrimental to future genetic engi-neering and in commercial application. However, cell per-meability mutants are used in large-scale industrial fermen-tations, e.g., the production of monosodium glutamate. Inour instance, no deleterious effects are apparent, and thegrowth rates of the wild-type and antibiotic-sensitive mu-tants are very similar.No change in plasmid content was associated with the

antibiotic-sensitive phenotype of strain CP4.45 (Fig. 1).However, Drainas et al. (7, 8) have isolated, via plasmidcuring, Z. mobilis strains which are sensitive to streptomy-cin and neomycin. Plasmid-associated antibiotic resistance(gentamicin, kanamycin, streptomycin) of Z. mobilis haspreviously been established (28). Thus, antibiotic resistancecould be determined by two distinct mechanisms; one is achromosomally encoded trait evidenced in strain CP4.45,and the other is plasmid encoded.The lowered resistance to ampicillin and kanamycin per-

mitted monitoring of both conjugation and transformation

APPL. ENVIRON. MICROBIOL.

on May 6, 2021 by guest

http://aem.asm

.org/D

ownloaded from

ANTIBIOTIC-SENSITIVE Z. MOBILIS 369

via antibiotic resistance (Table 3). Z. mobilis CP4.45 recip-ients were selected on antibiotic levels which were actuallygreater than the MIC for the wild-type cells to avoid selec-tion of resistant revertants. However, it was also noted thatantibiotic-sensitive strain recipients containing broad-host-range plasmids exhibit markedly higher antibiotic resistancelevels than do the wild-type cells. Thus, it should also bepossible to directly screen for plasmid transfer to the nativecells, provided that they will also exhibit markedly greaterantibiotic resistance. This was confirmed by selecting forplasmid RP4 after conjugal transfer from E. coli J53 to Z.mobilis CP4 by selection on kanamycin (500 p.g/ml) withcounterselection against the E. coli donor by streptomycin(200 ,ug/ml). The frequency of plasmid transfer by thismethod (5 x 10-8) was notably reduced from the frequencyobserved by conjugation with the antibiotic-sensitive mutantZ. mobilis CP4.45 (4 x 10-6) and from similar valuesreported by other workers (24). Application of this techniquefor direct selection of strains with greater antibiotic resist-ance than the wild-type strains without using antibiotic-sensitive strains as hosts is thus possible in Z. mobilis, butexceedingly large amounts of antibiotics (500 ,ug/ml) com-bined with the reduced frequency of plasmid transfer some-what mar this approach.The antibiotic resistances of Z. mobilis containing the

newly acquired plasmids were generally greater than thoseexhibited by the E. coli donor strains containing the sameplasmids (Table 3). The reason for the greater resistance inZ. mobilis compared with E. coli is unknown, but explana-tions could include physiologic distinctions in the promoterrecognition capabilities of the two organisms or perhapsphysical differences resulting in modified permeability of thecell envelopes.Although it is possible that some of the plasmids trans-

ferred to Z. mobilis underwent rearrangements which pre-vented reconjugation to E. coli, the transfer of plasmids to Z.mobilis has been demonstrated without major change in theirstatus (19, 22, 24). However, since plasmids RP4 andpGC91.14 could be reconjugated to E. coli and thenonreconjugal plasmids could be transferred to E. coli bytransformation without any change in phenotype, this maybe simply a problem in the mobilization of plasmids from Z.mobilis to E. coli.The transfer of plasmids such as pKT230 and pSUP304 to

antibiotic-sensitive strains of Z. mobilis, by using ampicillinor kanamycin for selection, extends the application of suchuseful plasmid cloning vehicles (1, 23, 24) in genetic studiesof this ethanologen by allowing the additional use of select-able antibiotic resistance genes from these two broad-host-range plasmid classes.

ACKNOWLEDGMENTS

We thank M. Dooley, Z. Shalita, and M. Yablonsky for helpfuldiscussions.

This study was supported by funds from the state of New Jersey,the U.S. Department of Agriculture (grant CRSR 4-26004), and theU.S. Department of Energy (subcontract XX-4-04150-1).

LITERATURE CITED

1. Bagdasarian, M., and K. N. Timmis. 1981. Host:vector systemsfor gene cloning in Pseudomonas. Curr. Top. Microbiol. Immu-nol. 96:47-67.

2. Bringer-Meyer, S., M. Schollar, and H. Sahm. 1985.Zymomonas mobilis mutants blocked in fructose utilization.Appl. Microbiol. Biotechnol. 23:134-139.

3. Carr, J. G. 1981. The genus Zymomonas, p. 1349-1354. In

M. P. Starr, H. Stolp, H. G. Truper, A. Balows, and H. G.Schlegel (ed.), The prokaryotes; a handbook on habitats, isola-tion and identification of bacteria, vol. 2. Springer-Verlag, NewYork.

4. Cornelis, G., D. Ghosal, and H. Saedler. 1978. Tn951: a newtransposon carrying a lactose operon. Mol. Gen. Genet.160:215-224.

5. Daily, E. L., H. W. Stokes, and D. E. Eveleigh. 1982. A geneticcomparison of strains of Zymomonas mobilis by analysis ofplasmid DNA. Biotechnol. Lett. 4:91-96.

6. Ditta, G., S. Stanfield, D. Corbin, and D. R. Helinski. 1980.Broad host range DNA cloning system for gram-negative bac-teria: construction of a gene bank of Rhizobium meliloti. Proc.Natl. Acad. Sci. USA 77:7347-7351.

7. Drainas, C. 1984. Studies on the natural plasmids ofZymomonas mobilis, no. 26. In Genetic engineering of micro-organisms important for agro-food industries. Commission ofthe European Communities, Marseille, France.

8. Drainas, C., M. A. Typas, and J. R. Kinghorn. 1984. A deriva-tive of Zymomonas mobilis ATCC 10988 with impaired ethanolproduction. Biotechnol. Lett. 6:37-42.

9. Eveleigh, D. E., H. W. Stokes, and E. L. Daily. 1983. Recombi-nant DNA approaches for enhancing the ethanol productivity ofZymomonas mobilis, p. 69-91. In D. L. Wise (ed.), Organicchemicals from biomass. Benjamin-Cummings, Publishing Co.,Menlo Park, Calif.

10. Fein, J. E., D. L. Barber, R. C. Charley, T. J. Beveridge, and H.Lawford. 1984. Effect of commercial feedstocks on growth andmorphology of Zymomonas mobilis. Biotechnol. Lett.6:123-128.

11. Gale, E. F., E. Cundliffe, P. E. Reynolds, M. H. Richmond, andM. J. Waring. 1981. The molecular basis of antibiotic action.John Wiley & Sons, Inc., New York.

12. Goodman, A. E., P. L. Rogers, and M. L. Skotnicki. 1982.Minimal medium for isolation of auxotrophic Zymomonas mu-tants. Appl. Environ. Microbiol. 44:496-498.

13. Grote, W., L. J. Lee, and P. L. Rogers. 1980. Continuousethanol production by immobilized cells ofZymomonas mobilis.Biotechnol. Lett. 2:481-486.

14. Haffie, T. L., P. W. Louie, and G. G. Khachatourians. 1985.Isolation of noninhibitory strains of Zymomonas mobilis. Appl.Environ. Microbiol. 49:1007-1009.

15. Hanahan, D., and The Cold Spring Harbor Laboratory. 1983.Studies on transformation of Escherichia coli with plasmids. J.Mol. Biol. 166:557-580.

16. Holloway, B. W. 1979. Plasmids that mobilize bacterial chromo-somes. Plasmid 2:1-19.

17. Krug, T. A., and A. J. Daugulis. 1983. Ethanol production usingZymomonas mobilis immobilized on an ion exchange resin.Biotechnol. Lett. 5:159-184.

18. Kurytowicz, W. (ed.). 1976. Antibiotics: a critical review. PolishMedical Publishers, Warsaw.

19. Lawford, G. R., B. H. Lavers, D. Good, R. Charley, J. Fein, andH. G. Lawford. 1983. Zymomonas ethanol fermentations: bio-chemistry and bioengineering, p. 482-507. In H. E. Duckworthand E. A. Thompson (ed.), Ethanol from biomass, internationalsymposium. Royal Society of Canada, Ottawa.

20. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

21. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

22. Rogers, P. L., M. L. Skotnicki, and K. J. Lee. 1984. Recentdevelopments in the Zymomonas process for ethanol produc-tion. Crit. Rev. Biotechnol. 1:273-288.

23. Simon, R., U. Priefer, and A. Puihler. 1983. Vector plasmids forin-vivo and in-vitro manipulation of gram-negative bacteria, p.98-106. In A. Piihler (ed.), Molecular genetics of bacteria-plantinteraction. Springer-Verlag, Berlin.

24. Skotnicki, M. L., D. E. Tribe, and P. L. Rogers. 1980. R-plasmidtransfer in Zymomonas mobilis. Appl. Environ. Microbiol.40:7-12.

25. Stokes, H. W., E. L. Daily, D. T. Spaniel, R. L. Williams, B. S.

VOL. 52, 1986

on May 6, 2021 by guest

http://aem.asm

.org/D

ownloaded from

370 BUCHHOLZ AND EVELEIGH

Montenecourt, and D. E. Eveleigh. 1981. Zymomonas: thephantom ethanol producer. Biosources Digest 3:124-132.

26. Stokes, H. W., E. L. Dally, M. D. Yablonsky, and D. E. Eveleigh.1983. Comparison of plasmids of strains ofZymomonas mobilis.Plasmid 9:138-146.

27. Swings, J., and J. De Ley. 1977. The biology of Zymomonas.Bacteriof. Rev. 41:1-46.

28. Walia, S. K., V. C. Carey, B. P. All III, and L. 0. Ingram. 1984.

APPL. ENVIRON. MICROBIOL.

Self-transmissible plasmid in Zymomonas mobilis carrying an-tibiotic resistance. Appl. Environ. Microbiol. 47:198-200.

29. Warr, R. G., A. E. Goodman, P. L. Rogers, and M. L. Skotnicki.1984. Isolation and characterization of highly productive strainsof Zymomonas mobilis. Microbios 41:71-78.

30. Yanase, H., M. Yasui, T. Mayazaki, and K. Tonomura. 1985.Fusion of spheroplasts and genetic recombination ofZymomonas mobilis. Agric. Biol. Chem. 49:133-140.

on May 6, 2021 by guest

http://aem.asm

.org/D

ownloaded from