cloning of the alcaligenes eutrophus genes for synthesis of phb

Vol. 170, No. 12JOURNAL OF BACTERIOLOGY, Dec. 1988, p. 5837-58470021-9193/88/125837-11$02.00/0Copyright ©D 1988, American Society for Microbiology

Cloning of the Alcaligenes eutrophus Genes for Synthesis ofPoly-3-Hydroxybutyric Acid (PHB) and Synthesis of

PHB in Escherichia coliPETER SCHUBERT, ALEXANDER STEINBUCHEL,* AND HANS G. SCHLEGEL

Institut fur Mikrobiologie der Universitat Gottingen, Grisebachstrasse 8, D-3400 Gottingen, Federal Republic of Germany

Received 3 August 1988/Accepted 13 September 1988

Eight mutants of Alcaligenes eutrophus defective in the intracellular accumulation of poly-fi-hydroxybutyricacid (PHB) were isolated after transposon TnS mutagenesis with the suicide vector pSUP5011. EcoRI fragmentswhich harbor TnS-mob were isolated from pHC79 cosmid gene banks. One of them, PPT1, was used as a probeto detect the intact 12.5-kilobase-pair EcoRI fragment PP1 in a XL47 gene bank ofA. eutrophus genomic DNA.In six of these mutants (PSI, API, GPI, GPIV, GPV, and GPVI) the insertion of TnS-mob was physicallymapped within a region of approximately 1.2 kilobase pairs in PP1; in mutant API, cointegration of vectorDNA has occurred. In two other mutants (GPII and GPIII), most probably only the insertion element hadinserted into PP1. All PHB-negative mutants were completely impaired in the formation of active PHBsynthase, which was measured by a radiometric assay. In addition, activities of f-ketothiolase and ofNADPH-dependent acetoacetyl coenzyme A (acetoacetyl-CoA) reductase were diminished, whereas the activityof NADH-dependent acetoacetyl-CoA reductase was unaffected. In all PHB-negative mutants the ability toaccumulate PHB was restored upon complementation in trans with PP1. The PHB-synthetic pathway of A.eutrophus was heterologously expressed in Escherichia coli. Recombinant strains of E. coli JM83 and K-12,which harbor pUC9-1::PP1, pSUP202::PP1, or pVK101::PP1, accumulated PHB up to 30% of the cellular dryweight. Crude extracts of these cells had significant activities of the enzymes PHB synthase, f-ketothiolase, andNADPH-dependent acetoacetyl-CoA reductase. Therefore, PP1 most probably encodes all three genes of thePHB-synthetic pathway in A. eutrophus. In addition to PHB-negative mutants, we isolated mutants whichaccumulate PHB at a much lower rate than the wild type does. These PHB-leaky mutants exhibited activitiesof all three PHB-synthetic enzymes; TnS-mob had not inserted into PP1, and the phenotype of the wild typecould not be restored with fragment PP1. The rationale for this mutant type remains unknown.

Poly-3-hydroxybutyric acid (PHB), which had been de-tected in 1926 by Lemoigne (31), is a polymer of D-(-)-1-hydroxybutyric acid and is a widespread intracellular stor-age compound typical of procaryotic organisms (10). Formany bacteria, PHB is the principal storage compound.Depending on the organism and on the physiological condi-tions applied to the cells, PHB may function as a carbon and/or energy storage compound or as a sink for reducingequivalents (9, 48). Biochemical studies revealed two dif-ferent pathways for the synthesis of PHB. (i) In mostorganisms, e.g., Azotobacter beijerinckii and Zoogloea ra-migera, a three-step metabolic pathway is realized. The firststep is catalyzed by the enzyme 1-ketothiolase (EC2.3.1.16), which condenses acetyl coenzyme A (acetyl-CoA)to acetoacetyl-CoA (38). This intermediate is reduced toD-(-)-P3-hydroxybutyryl-CoA by an NADPH-dependentacetoacetyl-CoA reductase (EC 1.1.1.36). In the last step theenzyme PHB synthase catalyzes the head-to-tail polymer-ization of the monomer to PHB. (ii) In Rhodospirillumrubrum PHB is synthesized via a five-step synthetic path-way. An NADH-dependent acetoacetyl-CoA reductase (EC1.1.1.35) catalyzes the formation of L-(+)-3-hydroxybutyryl-CoA, which is subsequently converted to D-(-)-P-hydroxy-butyryl-CoA by two stereospecific enoyl-CoA hydratasesprior to polymerization (37).

Despite intensive research on the physiology of PHBformation, relatively little was known about the enzymeswhich catalyze its synthesis, and almost no genetic studies

* Corresponding author.

have been performed on this important bacterial pathway inthe past. This has changed drastically, since PHB is ofincreasing interest to the chemical industry as an alternativematerial to classic polymers. PHB, which is produced fromAlcaligenes eutrophus and distributed on the market underthe trade name Biopol, has properties comparable to those ofpolypropylene. It can be manufactured from renewablefeedstock and is biodegradable (7). Copolymers of 3-hy-droxybutyric acid and other P3-hydroxyalkanoic acids suchas 3-hydroxyvaleric acid are of especial interest becausethey confer distinct properties on the polyester (11, 24).Recently the genes for ,-ketothiolase and acetoacetyl-CoAreductase from Z. ramigera have been cloned and se-quenced; the reductase gene maps immediately downstreamof the thiolase gene (44). Detailed biochemical studies on thecorresponding enzymes have been performed (8, 14). Onlythe thiolase gene from A. eutrophus has been cloned (S.Slater and D. Dennis, Abstr. Annu. Meet. Am. Soc. Micro-biol. 1987, H123, p. 139).

In contrast to P-ketothiolase and acetoacetyl-CoA reduc-tase, we have scant knowledge about the PHB synthase, themost important enzyme of the synthetic pathway. For Ba-cillus megaterium (36), R. rubrum (36), and Zoogloea rami-gera (14, 57), it was shown that PHB synthase is associatedwith phospholipids on the surface of the PHB granules undercertain conditions of growth. Almost nothing is known aboutthe mechanism of the synthase reaction and other propertiesof this enzyme. Griebel and Merrick proposed a protein, A-I,which, in B. megaterium, mediates between the monomer

5837

5838 SCHUBERT ET AL.

and the growing chain of the polymer and which functions asan acyl carrier (17).Our studies aimed at the cloning of the genes in A.

eutrophus, which are involved in the synthesis of PHB. Inthis report we describe the isolation of mutants of A.eutrophus which carry the transposon TnS-mob inserted intothe PHB synthase gene. One of these mutants provided aprobe for the detection of the intact gene in a K gene bank ofA. eutrophus H16 genomic DNA. The heterologous expres-sion of the A. eutrophus PHB synthase gene in Escherichiacoli and the formation of PHB granules in recombinantstrains of E. coli provide some evidence that all three genesof the PHB-synthetic pathway are clustered in A. eutrophus.

(Part of this work was communicated to the Akademie derWissenschaften zu Gottingen on 10 June 1988 and waspresented at the Second International Symposium on Over-production of Microbial Products, Ceske Budejovice,Czechoslovakia, on 7 July 1988.)

MATERIALS AND METHODSBacterial strains and plasmids. The strains of A. eutrophus

and E. coli, as well as the plasmids and bacteriophage usedin this study and mutant strains of A. eutrophus isolatedfollowing transposon mutagenesis, are listed in Table 1.Growth of bacteria. E. coli was grown at 37°C in complex

Luria-Bertani (LB) medium (34), which in some experimentswas supplemented with 0.5 to 1.0% (wt/vol) glucose. A.eutrophus was grown either in a complex medium of nutrientbroth (NB; 0.8%, wt/vol) or in a mineral salts medium (MM)(49). To allow extensive accumulation of intracellular PHB,the concentration of NH4Cl in the MM was reduced to 0.05or 0.005% (wt/vol). For heterotrophic growth, the mineralsalts medium was supplemented with filter-sterilized solu-tions of the carbon sources as indicated below. Media weresolidified with 1.5% (wt/vol) agar, and if necessary, antibi-otics were added as indicated.TnS mutagenesis and mutant selection. For the isolation of

transposon-induced mutants, A. eutrophus HF39 was matedwith E. coli S17-1(pSUP5011) as described below. Twodifferent methods were used to isolate mutants, which wereimpaired in the formation of PHB.

(i) Kanamycin-resistant transconjugants of A. eutrophuswere selected on selective plates (NB containing 500 ,ug ofstreptomycin per ml and 160 pug of kanamycin per ml). Fromthese plates they were transferred with a toothpick toMM-fructose agar plates containing either a high (0.1%, wt/vol) or low (0.005%, wt/vol) concentration of NH4Cl. Thosetransconjugants, which formed diaphanous instead ofopaque colonies, were isolated.

(ii) The cell suspension washed off the mating agar wastransferred to MM-fructose medium containing 0.005% (wt/vol) NH4Cl and was aerobically incubated at 30°C for about48 h. Cells which had not accumulated PHB were separatedfrom those which had by centrifugation in a Percoll densitygradient (43). Percoll density gradients were made from 75%(vol/vol) Percoll in 150 mM NaCl, and 0.2 ml of cellsuspension (approximately 108 cells) was loaded onto the topof the gradient. These were spun for 30 min at 4°C and 21,000rpm in an OTD50B ultracentrifuge (Ivan Sorvall, Inc., Nor-walk, Conn.), and the gradient was fractionated by platingportions of 0.1 ml onto MM-fructose plates which contained0.005% (vol/vol) NH4Cl. Mutants which appeared diapha-nous were subjected to further studies.

Isolation of DNA. Total genomic DNA of A. eutrophus wasisolated from cells grown in fructose (0.2% wt/vol) MM at30°C by the procedure described by Marmur (35).

A DNA phage particles were prepared by the method ofManiatis et al. (34) from plates exhibiting confluent lysis ofE. coli WL87. The phages were harvested by high-speedcentrifugation (48,000 x g for 3 h at 4°C), suspended in SMbuffer (34), and purified from debris by repeated centrifuga-tion steps (5 to 10 s) in an Eppendorf centrifuge. Theextraction of phage DNA followed the procedure describedby Maniatis et al. (34), including the proteinase K treatment.

Plasmid DNA was prepared from crude lysates by thealkaline extraction procedure (5, 34). DNA restriction frag-ments were isolated from agarose gels by electroelution intoa sodium acetate solution in an apparatus obtained fromBiometra, Gottingen, Federal Republic of Germany.

Analysis of plasmid DNA. Crude lysates were separated byelectrophoresis in horizontal slab gels containing 0.8% (wt/vol) agarose in TBE buffer (50 mM Tris hydrochloride, 50mM boric acid, 1.25 mM disodium EDTA [pH 8.5]). Elec-trophoresis was done at 150 V and 40 mA for 6 h.

Isolated plasmid DNA was digested with various restric-tion endonucleases under the conditions described by Ma-niatis et al. (34) or by the manufacturer. DNA restrictionfragments were separated in TBE buffer in horizontal slabgels containing 0.8 to 2.0% (wt/vol) agarose or in verticalpolyacrylamide gels containing 8.0% (wt/vol) acrylamide.The conditions used for separation followed the recommen-dations of Maniatis et al. (34). The molecular weights ofthese fragments were estimated by comparing their migra-tion with that of standard fragments obtained from K DNA.DNA bands were stained with ethidium bromide and visu-alized on a UV transilluminator.

Hybridization of Southern filters. DNA restriction frag-ments were separated by horizontal electrophoresis in 0.8%(wt/vol) agarose gels as described above for the analysis ofplasmid DNA. Conditions for the transfer of denatured DNAto nitrocellulose filters (BA85: pore size, 0.45 ,um; Schlei-cher and Schull, Dassel, Federal Republic of Germany), fortheir hybridization with biotinylated probes (labeled withbio-11-dUTP), and for the detection of biotinylated DNAhave been described previously (30).

Hybridization of plaques. KL47 libraries were screened onfilters as described by the manufacturer, Bethesda ResearchLaboratories, Inc., Gaithersburg, Md. (BRL Focus 7:11,1985). Conditions for hybridization with biotinylated probesand for the detection of hybrids were as described forSouthern filters. About 500 to 1,000 plaques per plate werescreened.DNA ligation. Restricted DNA was ligated as described by

Maniatis et al. (34). Restriction endonucleases were inacti-vated by heat (10 min at 65°C) before ligation. Vectors weredephosphorylated by treatment with bacterial alkaline phos-phatase prior to ligation.

Transformation. For transformation, E. coli was grownaerobically in LB medium containing 20 mM MgCl2 (18) at37°C. Competent cells were prepared and transformed byusing the calcium chloride procedure described by Maniatiset al. (34).

Conjugation. Matings of A. eutrophus (recipient) with E.coli S17-1 harboring hybrid donor plasmids were performedon solidified NB medium as described by Friedrich et al.(13). After 20 h of incubation at 30°C, the cells were washedfrom the agar and plated onto media suitable for isolation ofthe transconjugants.

Construction of cosmid and lambda L47 libraries. GenomicDNA (about 600 jig/ml) of TnS-induced mutants was par-tially digested by EcoRI to generate a high portion ofrelatively large fragments. These fragments were extracted

J. BACTERIOL.

CLONING OF A. EUTROPHUS PHB GENES 5839

TABLE 1. Bacterial strains, plasmids, and bacteriophage used in this study

Strain, plasmid or Relevant characteristics Source or referencebacteriophage

Wild type, prototrophicPHB-negative mutant of H16Smr mutant of H16

DSM 428, ATCC 17699DSM 541 (50)55

PHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr KmrPHB-negative mutant of HF39, Smr Kmr

PHB-leaky mutant of HF39, Smr KmrPHB-leaky mutant of HF39, Smr KmrPHB-leaky mutant of HF39, Smr KmrPHB-leaky mutant of HF39, Smr KmrPHB-leaky mutant of HF39, Smr KmrPHB-leaky mutant of HF39, Smr KmrPHB-leaky mutant of HF39, Smr Kmr

Spontaneous PHB-negative mutant of H16Spontaneous PHB-negative mutant of H16

This studyThis studyThis studyThis studyThis studyThis studyThis studyThis study

This studyThis studyThis studyThis studyThis studyThis studyThis study

This studyThis study

Wild type, prototrophicPHB-negative mutant of N9APHB-negative mutant of N9APHB-negative mutant of N9APHB-negative mutant of N9APHB-negative mutant of N9APHB-negative mutant of N9A

Wild type, prototrophicPHB-negative mutant of B19PHB-negative mutant of B19PHB-negative mutant of B19PHB-negative mutant of B19PHB-negative mutant of B19PHB-negative mutant of B19

Wild type, prototrophicPHB-negative mutant of G27PHB-negative mutant of G27PHB-negative mutant of G27PHB-negative mutant of G27PHB-negative mutant of G27PHB-negative mutant of G27PHB-negative mutant of G27PHB-negative mutant of G27

Wild type, prototrophicPHB-negative mutant of G29PHB-negative mutant of G29PHB-negative mutant of G29PHB-negative mutant of G29

DSM 518DSM 1348H. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublished

DSM 51559H. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublished

DSM 51659H. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublished

DSM 51759H. G. Schlegel, unpublishedH. G. Schlegel, unpublishedH. G. Schlegel, unpublished

recA and tra genes of plasmid RP4 are integrated into the chromosome;auxotrophic for proline and thiamine

recAl, auxotrophic for prolinerecBCara A(lac-proAB) thi +80 dlacZAMlSIacAM15

53

18Amersham Buchler58

Continued on following page

A. eutrophusH16H16-PHB-4HF39

PSIAPIGPIGPIIGPIIIGPIVGPVGPVI

RT01RT05RT06RT16RT18RT31RT49

SPiSP2

N9AN9A-PHB-02N9A-PHB-03N9A-PHB-05N9A-PHB-06N9A-PHB-07N9A-PHB-08

B19B19-PHB-01B19-PHB-02B19-PHB-03B19-PHB-04B19-PHB-05B19-PHB-06

G27G27-PHB-01G27-PHB-02G27-PHB-03G27-PHB-04G27-PHB-05G27-PHB-06G27-PHB-07G27-PHB-08

G29G29PHB-01G29PHB-02G29PHB-03G29PHB-04

E. coliS17-1

DH1WL87JM83

VOL. 170, 1988


TABLE 1-Continued

Strain, plasmid orbacteriophage Relevant characteristics Source or reference

PlasmidspSUP5011 Cmr Apr Kmr, harbors TnS::mob 52, 54pHC79 Cosmid, Tcr Apr 23pVK101 Tcr Kmr 29pUC9-1 Apr lacPOZ 19pBR325 Apr Cmr TCr 6

BacteriophageXL47 32

with phenol, phenol-chloroform (1:1), chloroform, and etherand precipitated with ethanol. They were ligated to theEcoRI-digested, dephosphorylated cosmid pHC79 (about250 ,uglml) at a ratio of 1:4 (insert to vector) with a total DNAconcentration of 250 ,ug/ml. The DNA was packaged with Xcoat proteins by using an in vitro packaging kit and trans-fected into E. coli DH1 by the methods of Hohn and Collins(22, 23). Cells were plated on LB medium containing antibi-otics as indicated.

Preparation of crude extracts. Approximately 0.2 to 1.0 g(wet weight) of cells was suspended in 2 ml of 100 mMpotassium phosphate buffer (pH 7.0) and disrupted by soni-cation (2 min) by using an MSE (150 W) ultrasonic disinte-grator with a probe 9.5 mm in diameter. To obtain the crudecellular extract, we removed unbroken cells by centrifuga-tion for 5 min at 1,000 x g in a bench centrifuge. The solubleprotein fraction the supernatant fraction was recovered aftercentrifugation at 100,000 x g for 60 min in an OTD centrifuge(Sorvall) with a TFT 65.13 rotor.

Determination of enzyme activities. The assay mixture forthe determination of PHB synthase activity contained 50mM Tris hydrochloride (pH 7.5), 10 mM dithiothreitol, 10mM MgCl2, and 5 to 30 ,g of protein from crude cellularextract in a total volume of 100 RI. It was preincubated at30°C for 2 min before the reaction was started by the additionof D,L-[3H]f-hydroxybutyryl-CoA to a final concentration of0.6 mM. After a 5-min incubation at 30°C the reaction wasstopped by the addition of 200 ,u of trichloroacetic acid (5%,wt/vol). After the addition of 0.9 ml of H20 the tritiatedpolymer was separated from tritiated monomer by extractionwith 0.6 ml of chloroform which contained 0.02% (wt/vol)PHB (14). The chloroform layer was washed three timeswith 0.9 ml of H20. A portion of the chloroform phase (0.5ml) was evaporated at room temperature overnight. Evapo-rated material was dissolved in chloroform and transferredto a counting vial. It was again evaporated overnight, and theremaining material was dissolved in 0.1 ml of chloroform-10ml of a Quickszint 2000 scintillant (Zinsser AnalyticalGmbH, Frankfurt, Federal Republic of Germany). Radioac-tivity was counted in an LS7800 liquid scintillations spec-trometer (Beckman Instruments, Inc., Fullerton, Calif.).The activity of ,-thiolase (EC 2.3.1.9) in the soluble

protein fraction was determined in the presence of 2 U ofP-hydroxyacyl-CoA dehydrogenase from porcine heart bymeasuring the oxidation of NADH concomitant with thereduction of the acetoacetyl-CoA, which is formed duringthe thiolase reaction, as described by Oeding and Schlegel(41).

Activities of NADH- or NADPH-dependent acetoacetyl-CoA reductase (EC 1.1.1.35 and EC 1.1.1.36, respectively)in the soluble protein fraction were determined by a modifi-cation of the method described by Lynen and Wieland (33).

The reaction proceeded in 100 mM Tris hydrochloride buffer(pH 7.8) in the presence of 30 ,uM acetoacetyl-CoA and of125 ,uM reduced pyridine nucleotide.

Protein determination. The protein contents in crude ex-tracts and in the soluble fractions were determined by themethod of Beisenherz et al. (3). To sediment interferingmaterial, we centrifuged the samples for 30 min at 3,000 x gin a bench centrifuge prior to the determination of theabsorbance.

Determination of PHB. Quantitative determination ofPHBwas performed by the infrared spectroscopy method de-scribed by Juttner et al. (27) after extraction of lyophilizedcells with chloroform. In colonies grown on mineral mediumcontaining a permissive concentration of NH4Cl, PHB wasvisualized by staining with Sudan black as described bySchlegel et al. (50).

Synthesis of metabolites. Acetyl-CoA was synthesizedfrom CoA and acetic anhydride as described by Ochoa (40).Acetoacetyl-CoA was synthesized from CoA and diketene inaccordance with a method described by Simon and Shemin(51).For the synthesis of D,L-[3H]-p-hydroxybutyryl-CoA, 167

,umol of acetoacetyl-CoA in 16 ml of H20-3 ml of 100 mMsodium phosphate buffer was gently stirred at room temper-ature (pH 7.5). With the pH kept between 7.4 and 8.0 with0.1 N HCI, 0.28 ,umol of [3H]NaBH4 dissolved in 280 RI of0.1 N NaOH was added dropwise. The reaction was chasedwith 260 ,umol ofNaBH4 which was dissolved in 630 pul of 0.1N NaOH. For further formation of ,B-hydroxybutyryl-CoAwas detected after 100 min. The pH was adjusted to 5.1 with1 N HCI. About 60% of the radioactivity was released as H2;it was oxidized to H20 by passage along a column of CuOand was recovered from the spent air in a trap of liquidnitrogen. To dispose of tritiated water which has beenformed by exchange reaction, we lyophilized the solution.The lyophilized sample was dissolved in 30 ml of H20 andstored at aliquots at -20°C.

Chemicals. Restriction endonucleases, biotin-11-dUTP,the nick translation kit, the DNA detection kit, T4 DNAligase, lambda DNA, and substrates used in the enzymeassays were obtained from C. F. Boehringer & Soehne,Mannheim, Federal Republic of Germany, or from GIBCO/BRL-Bethesda Research Laboratories GmbH, Eggenstein,Federal Republic of Germany. Agarose type NA was pur-chased from Pharmacia, Uppsala, Sweden. Percoll, antibi-otics, and ethidium bromide were from Sigma Chemical Co.,St. Louis, Mo. Sodium [3H]borohydride was purchased fromAmersham Buchler, Brunswick, Federal Republic of Ger-many. All complex media were from Difco Laboratories,Detroit, Mich. Most other chemicals were obtained from E.Merck AG, Darmstadt, Federal Republic of Germany.

J. BACTERIOL.


RESULTS

Isolation of spontaneous PHB-negative mutants. A largenumber of mutants ofA. eutrophus defective in the synthesisof PHB have been isolated in the past following mutagenesiswith nitrous acid or alkylating agents (50). To examinewhether such mutants would also occur spontaneously, wecultivated A. eutrophus for 48 h in nitrogen-poor MM(0.005% [wt/vol] NH4Cl) with fructose as the carbon sourceto allow the wild-type cells to accumulate large amounts ofPHB. Approximately 108 cells were then loaded onto aPercoll density gradient and centrifuged to separate wild-type cells from cells of spontaneous PHB-negative mutants.We were able to isolate PHB-negative mutants from theregion of the gradient which was above the band represent-ing the wild-type cells. Spontaneous mutants such as SP1 orSP2 occurred at a frequency of 7.5 x 10-8.

Isolation of TnS-induced PHB-negative mutants. To labelgenes which are essential for the synthesis of PHB in A.eutrophus, we applied transposon mutagenesis to generatePHB-negative mutants. The transposable element TnS-mobwas introduced into streptomycin-resistant A. eutrophusmutant HF39 by conjugation with E. coli S17-1 whichharbors the suicide vehicle pSUP5011. Mutant strains PSIand API were isolated after transfer of approximately 35,000kanamycin-resistant transconjugants to nitrogen-poor andnitrogen-rich MM-fructose medium with toothpicks duringscreening for transparent colonies. A number of additionalPHB-negative mutants (GPI, GPII, GPIII, GPIV, GPV, andGPVI) were identified on nitrogen-poor MM-fructose agarplates after enrichment in a Percoll density gradient.During screening of kanamycin-resistant transconjugants

for PHB-negative mutants, we isolated numerous mutantswhich grew normally on fructose but accumulated PHB at amuch lower rate, as was obvious from the delayed appear-ance of opaqueness of the colonies. These so-called PHB-leaky mutants (RT01, RT05, RT06, etc.) appeared at a

threefold-higher frequency than PHB-negative mutants.Biochemical characterization of spontaneous and Tn5-in-

duced PHB-negative mutants. To localize the defect in thethree-step PHB-synthetic pathway in PHB-negative mu-

tants, a radiometric assay for the determination of PHBsynthase activity was elaborated (see Materials and Meth-ods); activities of ,-ketothiolase and of acetoacetyl-CoAreductase were determined photometrically. All TnS-in-duced PHB-negative mutants isolated in this study, mutantH16-PHB-4, as well as all other PHB-negative mutantslisted in Table 1 which appeared diaphanous on nitrogen-poor fructose mineral agar plates and which could not bestained with Sudan black, lack PHB synthase activity. Noneof the transposon-induced PHB-negative mutants exhibitedPHB synthase activity and lacked thiolase or NADH- or

NADPH-dependent acetoacetyl-CoA reductase activity.However, activities of 1-ketothiolase and ofNADPH-depen-dent acetoacetyl-CoA reductase were greatly diminished(Table 2). The activity of NADH-dependent acetoacetyl-CoA reductase was unaffected in all mutants examined. Withthe infrared spectroscopy method, no PHB could be de-tected in cells which were isolated under nitrogen starvation.Genomic probing for fragments labeled with TnS-mob. To

determine whether the insertion sites of TnS-mob wererestricted to one specific region of the mutant genomes, we

hybridized Southern blots with a probe specific for TnS-mob.Genomic DNAs of the wild type and the PHB-negativemutant strains PSI, API, GPI, GPII, GPIII, GPIV, GPV, andGPVI were completely digested with EcoRI, and the frag-

TABLE 2. Specific activities of PHB-synthetic enzymes invarious species and mutant strains of A. eutrophus"

Sp act (U/g of protein) of":

Species or Acetoacetyl-Co Astrain PHB reductase

synthase f3-Ketothiolase NADH NADPHdependent dependent

H16 180 428 932 768HF39 NDb 475 949 527N9A 191 ND ND NDG29 177 ND ND NDB19 101 ND ND ND

H16-PHB-4 <0.1 22 1,080 42

PSI <0.1 19 1,078 38API <0.1 39 1,091 45GPI <0.1 58 1,153 40GPII <0.1 50 ND 90GPIII <0.1 130 ND 70GPIV <0.1 31 849 36GPV <0.1 46 731 27GPVI <0.1 65 651 38

RT01 112 412 627 538RT05 86 315 923 353RT06 104 345 261 340RT16 70 308 713 371RT18 ND 747 553 553RT31 ND 481 772 639RT49 ND 521 621 822

" Cells were grown at 30°C in nitrogen-poor (0.05% [wt/vol] NH4CI)MM-fructose medium. Cells were harvested approximately 6 h after depletionof the nitrogen source. Activities of PHB synthase, ,B-ketothiolase, andacetoacetyl-CoA reductases in the crude cellular extract or in the solubleprotein fraction were determined as described in Materials and Methods.

b One unit of enzyme activity is defined as the transformation of 1 ,umol ofsubstrate per min.

C ND, Not determined.

ments were separated electrophoretically, blotted onto ni-trocellulose, and hybridized with a TnS probe (center 4,300-base-pair [bp] HindlIl fragment of Tn5-mob which harborsapproximately 300 bp each of IS5R and IS50L). Withmutants PSI, GPI, GPIV, GPV, and GPVI, hybridization toa 20-kbp EcoRI fragment was observed; no Hybridizationsignal was observed with DNA of the wild type. GenomicDNA of all TnS-induced PHB-negative mutants was incom-pletely digested with EcoRI and was ligated to pHC79 DNA.Kanamycin-resistant clones which harbor the EcoRI frag-ment with Tn5-mob were isolated from cosmid libraries in E.coli DH1. EcoRI fragments harboring Tn5-mob were differ-entiated from other EcoRI fragments which had been ligatedto pHC79 by hybridization with the Tn5 probe. The 20-kbpEcoRI fragments cloned were referred to as PPT1 (frommutant PSI), PPT3 (GPI), PPT4 (GPIV), PPT5 (GPV), andPPT6 (GPVI). Taking into consideration that the size ofTnS-mob is 7.5 kbp (52) and that it has no restriction site forEcoRI (26), these results indicate that the size of the corre-sponding wild-type EcoRI-fragment is approximately 12.5kbp.With mutant API, hybridization to a 15-kbp EcoRI frag-

ment was observed with the TnS probe. In addition, weakhybridization occurred to an 11-kbp EcoRI fragment. Thisresult indicated the presence of different amounts of TnSDNA in two different EcoRI fragments of mutant strain API.From mutant strain API a 15-kbp EcoRI fragment conferring

VOL. 170, 1988


kanamycin resistance was cloned; it was referred to asPPT2a. In addition to PPT2a, one of the hybrid cosmidsharbored the 11-kbp EcoRI fragment mentioned above; itwas referred to as PPT2b.With genomic DNA from mutant strains GPII and GPIII,

hybridization to a 21- and a 17-kbp EcoRI fragment, respec-tively, was observed with the TnS probe. The correspondingEcoRI fragments conferring kanamycin resistance werecloned and referred to as PPT7 (from mutant GPII) and PPT8(from mutant GPIII).The EcoRI fragment PPT1 was ligated to the suicide

vector pSUP202. When pSUP202::PPT1 was mobilized fromE. coli S17-1 into A. eutrophus, HF39 homogenotic (pheno-type, Kmr Tcs PHB-) and heterogenotic (phenotype, KmrTcr PHB+) transconjugants appeared at a ratio of approxi-mately 1:10. In addition, transconjugants exhibiting thephenotype Kmr Tcs PHB+ appeared at a threefold-higherfrequency. It is unknown why these transconjugants, whichprobably rely on transposition, appeared at a threefold-higher frequency than homogenotic and heterogenotic trans-conjugants, which rely on homologous recombination.

Cloning of the intact gene for PHB synthase. To clone theEcoRI fragments encoding the intact genes for the PHB-synthetic pathway, we used biotinylated PPT1 and PPT2aDNA to detect the corresponding native fragments in a XL47gene bank of A. eutrophus 1116. After infection with recom-binant XL47 phages, E. coli WL87 was plated on LBmedium. Plates were blotted to nitrocellulose membranesand hybridized with biotinylated PPT1 and PPT2a DNA asdescribed in Materials and Methods. In two experiments, 2and 3 of approximately 2,000 plaques were positive. Allpositive rec nbinant phages contained a 12.5-kbp EcoRIfragment; they were referred to as PP1 (detected with PPT1)or PP2 (detected with PPT2a). By performing hybridizationstudies and by comparing the pattern of restriction frag-ments, we demonstrated the identity of PP1 and PP2. PP1was purified by electroelution and then ligated to pVK101,pSUP202, and pUC9-1.

Analysis of cloned fragments and physical mapping oftransposon insertions. Fragment PP1 and fragments harbor-ing TnS-mob were isolated from the respective hybrid plas-mids by electroelution after electrophoretic separation ofEcoRI-restricted plasmid DNA in agarose gels. They werethen treated with HindlIl, XhoI, BamHI, EcoRV, BglII,KpnI, SmaI, and Sall, whose cleavage sites in PP1 weredetermined (Fig. 1). On the basis of the known restrictionmap for TnS (26), which harbors the RP4 mob site (53), theinsertions of TnS-mob into PP1 were physically mapped.TnS-mob had inserted into a 2,200-bp SailI fragment in themutant strains PSI, GPI, GPV, and GPVI, whereas it hadinserted into an adjacent 140-bp Sall-fragment in strainGPIV. Insertions cover a region of approximately 1,200 bp inall five mutants isolated in this study (Fig. 1).The EcoRI fragments PPT2a and PPT2b, which had been

isolated from mutant strain API, both hybridized with bio-tinylated PP1 DNA. Although PPT2a contained the 140-,1,000-, and 2,100-bp Sall fragments of PP1 in addition to a4,360-bp Sall fragment, PPT2b contained the 300-, 2,600-,3,000-, and 1,150-bp Sall fragments of PP1 in addition to a5,600- and a 6,200-bp Sall fragment. In contrast to the 5,600-and the 6,200-bp SailI fragments, which gave a strong hybrid-ization signal, the 4,360-bp SailI fragment gave a weakhybridization signal with the TnS probe. With a pBR325probe (linearized pBR325 DNA) only the 4,360- and the6,200-bp SailI fragments gave a signal, whereas all three Sallfragments (4,360, 5,600, and 6,200 bp) gave a signal with an

IS50 probe (1,000-bp HpaI-HindIII fragment of TnS). Con-sidering the physical data for Tn5, the results may beexplained by the cointegration of pSUP5011 which followedthe insertion of one single insertion element. The primaryinsertion occurred in the 2,200-bp SalI fragment of PP1 andmapped between the insertions of TnS-mob in mutant strainsGPV and PSI (Fig. 1). The cointegration of vector DNAresulted in the introduction of a cleavage site for EcoRI inthe derivative PP1 EcoRI fragment.Not only did TnS-harboring EcoRI fragments PPT7 and

PPT8, which had been cloned from mutant strains GPII andGPIII, differ in size from PPT1, but neither fragment exhib-ited any homology to PP1, as revealed by hybridization withbiotinylated PP1 DNA and by the occurrence of differentSall restriction fragments. Hybridization of biotinylated PP1to genomic DNA of both mutants produced only one signalwith a 14-kbp EcoRI fragment, whereas hybridization withthe IS50 probe, as well as with biotinylated PPT1 DNA,yielded a signal with a 14- and a 17-kbp EcoRI fragment ofmutant GPIII genomic DNA or with a 14- and a 20-kbpEcoRI fragment of mutant GPII genomic DNA. TnS-mobhad probably inserted into PP1; thereafter, the transposonmay have been excised, leaving one insertion element inPP1, and then may have been inserted into a random part ofthe genome.

Cointegrate formation usually occurs only with transpo-sons, which transpose by a replicative mechanism (4), butwas also reported for TnS, which transposes by a conserv-ative mechanism (12). The precise mechanism of cointegra-tion of vector DNA in mutant API and of parts of TnS-mobin mutants GPII and GPIII is unknown and remains to beelucidated. However, the high proportion of such irregularmutants among transposon-induced PHB-negative mutantsis remarkable.Complementation of PHB-negative mutants ofA. eutrophus.

Plasmid pVK1O1::PP1 was transferred by conjugation fromE. coli S17-1 to different mutants of A. eutrophus, whichwere impaired in the synthesis of PHB. Transconjugantswere collected on mineral agar plates containing 0.5% (wt/vol) sodium succinate and 12.5 jig of tetracycline per ml. Theability to accumulate PHB and to synthesize active PHBsynthase (Table 3) could be restored without any exceptionin all 35 PHB-negative mutants tested. The complementationexperiments included all eight transposon-induced PHB-negative mutants.

Biochemical and genetic characterization of mutants leakyfor synthesis of PHB. Mutants which were leaky for thesynthesis of PHB appeared at a threefold-higher frequencythan PHB-negative mutants did. They exhibited normalgrowth on fructose; however, PHB accumulated much moreslowly than in the wild-type. In PHB-leaky mutants PHBsynthase activity was present, although at a slightly lowerlevel; activities of thiolase and of acetoacetyl-CoA reduc-tases were not affected (Table 2).

Hybridization experiments clearly demonstrated that inPHB-leaky mutants Tn5 had not inserted into the EcoRIfragment PP1. With biotinylated PP1-DNA, genomic DNAof PHB-leaky mutants RT18 and RT49 hybridized with a12-kbp EcoRI fragment, whereas two fragments reacted withbiotinylated PPT1 DNA: one with a 12-kbp EcoRI fragment,which is probably PP1, and the other with a different EcoRIfragment, which probably harbors the locus responsible forthe PHB-leaky phenotype. Normal synthesis of PHB couldnot be restored when plasmid pVK101::PP1 was transferredto PHB-leaky mutant RT6, RT16, RT18, RT31, or RT49.

Heterologous expression of the PHB-synthetic genes in E.

J. BACTERIOL.


Hind Ill

Xho I

BamH I 870

EcoR V

Bgl 11 440a

Sma I ON

Sal I O 2600

-S S S S

260 302600 3000 1150

mob

IS5QL S\IS50 EEZ

-PPT2b

=SN'''''ISSOR SS S

1 000 21 00

__PPT2a

FIG. 1. Physical map of the 12.5-kbp EcoRI fragment PP1. The insertions of TnS-mob in six mutants are shown. At the bottom,cointegration of pSUP5011 DNA into PP1 in mutant API and the generation of EcoRI fragments PPT2a and PPT2b are shown. Numbersindicate the sizes of individual DNA fragments in base pairs. Abbreviations: E, EcoRI; S, Sall. Symbols: , PP1 DNA; _, pBR325DNA; E, Tn5-mob DNA; E, IS50 DNA; V, mob site.

coli. The hybrid plasmids pUC9-1.:PP1, pSUP202::PP1, andpVK101::PP1 were isolated and transferred to E. coli JM83and K-12 by transformation. After prolonged incubation on

LB agar plates, colonies of recombinant clones appearedmuch more opaque than colonies of the nonrecombinantparent strains. The opacity was enhanced by the addition of0.5% (wt/vol) glucose to the LB medium. When cells fromthese colonies were investigated under the light microscope,light-scattering inclusion bodies were visible. If cells harbor-

TABLE 3. Activity of PHB synthase in PHB-negative mutantsafter complementation by the plasmid pVK101::PP1

Strain Sp act of PHB synthaseStrain ~~~~~~~~(U/gof protein)"

H16 ................................... 106H16-PHB-4.................................. <0.1H16-PHB-4(pVK101::PP1) ............... ......... 109PSI ......... ......................... <0.1PSI(pVK101::PP1) .................................. 95API................................... <0.1API(pVK101::PP1) .................................. 127

" One unit of enzyme activity is defined as the transformation of 1 ,umol ofsubstrate per min.

ing plasmids with PP1 inserts were incubated in liquid LBmedium supplemented with 0.5%9 (wt/vol) glucose, theyaccumulated PHB up to 30% of the cell dry weight.The accumulation of PHB in the recombinant strains

indicated heterologous expression of the A. eutrophus PHB-synthetic genes in E. coli and prompted us to measure theactivities of the corresponding enzymes in crude extracts.Crude extracts of these cells exhibited not only the activityof PHB synthase but also that of ,B-ketothiolase andNADPH-dependent acetoacetyl-CoA reductase (Table 4).The activity of NADH-dependent acetoacetyl-CoA reduc-tase was only marginal. The orientation of PP1 in pUC9-1had no influence on the synthesis of PHB or on the activityof the three enzymes examined. If PP1 was not present, no

PHB synthase activity was detectable in E. coli, and theactivities of the other enzymes were only marginal (Table 4).

DISCUSSION

We have isolated eight transposon-induced mutants of A.eutrophus H16 which are impaired in the intracellular accu-

mulation of PHB. Insertions of Tn5-mnob were physicallymapped within the 12.5-kbp EcoRI fragment PP1. The clon-ing of TnS-labeled fragments provided a specific DNA probe

E E-I

VOL. 170, 1988


TABLE 4. Heterologous expression of the A. eutrophus PHB-synthetic genes and synthesis of PHB in E. colia

Sp act (U/g protein) of:____________________ ____ ____ ____ ____ ____ ____ ____ ____Accumulation of

Strain Medium Acetoacetyl-CoA reductase PHB (% ofPHB synthase P-Ketothiolase NADH NADPH cellular dryNADH NADPH ~~~~~weight)

dependent dependent

JM83 LB-Glu <0.10 <20 20 NDb <2JM83(pUC9-1) LB-Ap-Glu <0.10 <20 10 <10 <2JM83(pUC9-1::PP1) LB-Ap-Glu 1.08 720 10 180 25.7K12 LB-Glu <0.10 <20 15 ND <2K12(pUC9-1::PP1) LB-Ap-Glu 1.50 800 40 ND 30.4S17-1(pSUP202) LB-Tc-Glu <0.10 <20 ND <10 <2S17-1(pSUP202::PP1) LB-Tc-Glu 0.31 170 ND 120 27.4S17-1(pVK101) LB-Tc-Glu <0.10 <20 ND <10 <2S17-1(pVK1O1::PPl) LB-Tc-Glu 0.23 <20 ND 33 18.4

a Cells were grown at 37°C in LB medium, which was supplemented with 0.5% (wt/vol) glucose (Glu) and 50 ptg of ampicillin (Ap) per ml or 12.5 p.g oftetracycline (Tc) per ml as indicated. Cells were harvested in the early stationary growth phase. Activities ofPHB synthase and other enzymes in the crude cellularextract or in the soluble protein fraction were determined as described in Materials and Methods. One unit of enzyme activity is defined as the transformationof 1 p.mol of substrate per min. PHB was determined by the infrared spectroscopic method and was compared with the cellular dry weight. The latter wasdetermined by drying a portion of the cell suspension, which had been sucked on a membrane filter (pore size, 0.2 pum), at 85°C to constant weight.

b ND, Not determined.

for the physical identification of the intact PHB synthasegene. Recombinant strains of E. coli which harbor fragmentPP1 expressed all three genes of the PHB-synthetic path-way. That the A. eutrophus PHB-synthetic pathway isfunctionally active in E. coli was demonstrated by theintracellular accumulation ofPHB granules under conditionsof excess carbon source. It has been reported that cells of E.coli, like those of some other bacteria, accumulate smallamounts ofPHB in the cytoplasmic membrane if the cells aredeprived of certain nutrients (15) and during the develop-ment of competence (45). Nothing is known of the mecha-nism ofPHB synthesis or of the enzymes which are involvedin its synthesis in nonrecombinant strains of E. coli. How-ever, the extent to which PHB accumulates, its deposition incytoplasmic granules, and the concomitant detection of allthree PHB-synthetic enzymes clearly demonstrate that ac-cumulation of PHB in strains of E. coli which harbor PP1 isdue to the heterologous expression of the A. eutrophusgenes. Thiolases I and II and hydroxyacyl-CoA dehydroge-nase (EC 1.1.1.35), which are encoded by the ato or the fadoperon, respectively, are synthesized in E. coli only if fattyacids are used as a carbon source (39). A third thiolase,an NADPH-dependent acetoacetyl-CoA reductase (EC1.1.1.36), or even a PHB synthase has not yet been de-scribed for E. coli.

Like the chromosomally as well as the plasmid-encodedgenes for ribulose bisphosphate carboxylase from strainDSM 428 (24a), expression of the A. eutrophus PHB-syn-thetic genes in E. coli occurred independently from E. colipromoters. In contrast, the promoters of the genes for thechromosomally encoded ribulose bisphosphate carboxylasefrom strain ATCC 17707 (2), the plasmid-encoded gene forphosphoribulokinase (28), or the gene for the fermentativealcohol dehydrogenase (30) from strain DSM 428 are notexpressed in E. coli. First, DNA sequences obtained forgenes ofA. eutrophus have shown that transcription and nottranslation is probably limiting with respect to heterologousexpression in E. coli (1, 25). On the other hand, it has beendemonstrated that catabolic genes of E. coli, like pfkA or tpi,are expressed in A. eutrophus (56).

If we assume that at least a three-step synthetic pathway(Fig. 2) is responsible for the synthesis of PHB in A.eutrophus, the following observations are peculiar: (i) Tn5-induced PHB-negative mutants appeared at an unexpected

low frequency; (ii) none of the PHB-negative mutants wasable to form active PHB synthase, independent of the kind ofmutagenic agent applied; (iii) in all transposon-induced mu-tants, the activities of 13-ketothiolase and NADPH-depen-dent acetoacetyl-CoA reductase were greatly diminished;(iv) the ability to synthesize PHB could be restored in anyPHB-negative mutant examined by complementation intrans with pVK101::PP1; and (v) Tn5-mob mapped within asmall region of approximately 1.2 kbp in PP1 in PHB-negative mutants. Our data suggest that the structural genesfor PHB synthase, P-ketothiolase, and NADPH-dependentacetoacetyl-CoA reductase are clustered in PP1. We assumethat the region in which TnS-mob mapped in PHB-negativemutants encodes the structural gene of PHB synthase orsegments which are essential for the expression of this geneand that the insertion exerts a polar effect on the expressionof both other genes. Mutants which harbor Tn5-mob in theother two genes have not been isolated, because they do notexhibit the PHB-negative phenotype (see below). It has stillto be examined whether all three genes are organized in asingle operon. If this is the case, the gene for PHB synthaseis probably its first gene which is transcribed.

,B-Ketothiolase is needed not only for the synthesis ofPHB but also for the synthesis of P-hydroxy-p-methylglu-taryl-CoA, which is a building block in, e.g., quinones,bactoprenol, and hopanoids (Fig. 2). The latter compoundhas not been detected in A. eutrophus (42; K. Poralla,unpublished results). In addition, P-ketothiolase is alsoinvolved during ,8-oxidation of fatty acids. If only one genefor ,B-ketothiolase is present in A. eutrophus, its insertionalinactivation is lethal to the cell. As shown by Haywood et al.(20) and as also indicated by the presence of residual,B-ketothiolase activity in TnS-induced mutants, isoenzymesfor 1-ketothiolase exist in A. eutrophus; a second ,B-keto-thiolase may fulfill the function of the PHB-synthetic 1B-ketothiolase. In A. eutrophus both an NADH- and anNADPH-dependent acetoacetyl-CoA reductase are present(21, 41). In contrast to the dehydrogenase for 1-hydroxy-butyrate, which is NAD dependent and which is specific forthe D-(-)-stereoisomer (46, 47), the NADP-dependent aceto-acetyl-CoA reductase of A. eutrophus does not displaystereospecificity and is active with the D-(-)- as well as withthe L-(+)- stereoisomer of P-hydroxybutyryl-CoA. If weassume that only these two enzymes are present in A.

J. BACTERIOL.


L(|-BHyroxbuyrte(D

-wL(+)-B-Hydroxybutyryl-CoA _

CoASH

I ~~NAD+ NADH+

l.20

IButyrate Butyryl-CoA -

CoASH I

oFatty acids,

- - -

0D

w |Malonyl-CoA|

-Hydroxy-oB-methyl-glutaryl-CoA

CoASH

NADPH + H +

~l-0D(-)- B-Hydroxybutyryl-CoA 4-|- I D(-)- -Hydroxybutyrate|

-.\ [ CoASH[OAp_ CoASH

PHB.

ICrotonateI

NADP +

FIG. 2. Pathway of PHB synthesis and related reaction steps in A. eutrophus. 1, Acyl-CoA synthetases; 2, 13-ketothiolase; 3,acetoacetyl-CoA reductase (NADPH dependent); 4, acetoacetyl-CoA reductase (NADH dependent); 5, enoyl-CoA hydratase [formingL-(+)-O-hydroxybutyryl-CoA]; 6, enoyl-CoA hydratase [forming D-(-)-P-hydroxybutyryl-CoA); 7, butyryl-CoA dehydrogenase; 8, enzymesinvolved in the 1-oxidation pathway; 9, PHB synthase; 10, hydroxymethylglutaryl-CoA synthase; 11, acetyl-CoA carboxylase.

eutrophus, the NADH-dependent enzyme is only able tocompensate for the NADPH-dependent enzyme upon itsloss if an enoyl-CoA hydratase, which is specific for theD-(-)-stereoisomer of ,B-hydroxybutyryl-CoA, is alsopresent (Fig. 2). An enoyl-CoA hydratase specific for theL-(+)-stereoisomer is involved during 1-oxidation. Underthese circumstances PHB synthesis via a five-step syntheticpathway like that in R. rubrum may occur in addition tosynthesis via the basic three-step synthetic pathway.That the situation is much more complex in A. eutrophus

is also obvious from the fact that substrates like crotonate,butyrate, and valerate (11, 16) are incorporated into thepolymer without cleavage of the carbon skeleton to acetyl-CoA (Fig. 2). It has still to be determined whether thedegradative pathway of these substrates joins the basicthree-step PHB-synthetic pathway at the level of acetoa-cetyl-CoA or D-(-)-13-hydroxybutyrl-CoA.

ACKNOWLEDGMENTS

We thank Horst Priefert and Andreas Pries for technical assis-tance during the preparation of lambda packaging kits and theisolation of mutants, respectively.

This study was supported by grants provided by the Max BuchnerForschungsstiftung and the Bundesministerium fur Forschung undTechnologie.

LITERATURE CITED

1. Andersen, K., and J. Caton. 1987. Sequence analysis of theAlcaligenes eutrophus chromosomally encoded ribulose bis-phosphate carboxylase large and small subunit genes and theirgene products. J. Bacteriol. 169:4547-4558.

2. Andersen, K., and M. Wilke-Douglas. 1987. Genetic and physi-cal mapping of the chromosomally encoded ribulose bisphos-

phate carboxylase genes of Alcaligenes eitrophus. J. Bacteriol.169: 1997-2004.

3. Beisenherz, G., H. J. Bolze, T. Bucher, R. Czok, K. H. Garbade,A. E. Meyer, and G. Pfleiderer. 1953. Diphosphofructo-Aldo-lase, Phosphoglyceraldehyd-Dehydrogenase, Milchsaure-Dehy-drogenase, Glycerophosphat-Dehydrogenase und Pyruvat-De-hydrogenase in einem Arbeitsgang. Z. Naturforsch. Teil B 8:555-557.

4. Berg, C. M., and D. E. Berg. 1987. Uses of transposableelements and maps of known insertions, p. 1071-1109. In F. C.Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M.Schaechter, and H. E. Umbarger (ed.) Escherichia coli andSalmonella typhimurium: cellular and molecular biology. Amer-ican Society for Microbiology, Washington, D.C.

5. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extractionprocedure for screening recombinant plasmid DNA. NucleicAcids Res. 7:1513-1523.

6. Bolivar, F. 1978. Construction and characterization of newcloning vehicles. lIl. Derivatives of plasmid pBR322 carryingunique EcoRI sites for selection of EcoRl generated recombi-nant DNA molecules. Gene 4:121-136.

7. Byrom, D. 1987. Polymer synthesis by microorganisms: tech-nology and economics. Trends Biotechnol. 5:246-250.

8. Davis, J. T., R. N. Moore, B. Imperiali, A. J. Pratt, K.Kobayashi, S. Masamune, A. J. Sinskey, and C. T. Walsh. 1987.Biosynthetic thiolase from Zoogloea ramigera. I. Preliminarycharacterization and analysis of proton transfer reaction. J.Biol. Chem. 262:82-89.

9. Dawes, E. A. 1986. Microbial energetics, 1st ed. Blackie & Son,Ltd., Glasgow, Scotland.

10. Dawes, E. A., and P. J. Senior. 1973. Energy reserve polymersin microorganisms. Arch. Microbiol. Physiol. 14:203-266.

11. Doi, Y., A. Tamaki, M. Kunioka, and K. Soga. 1988. Productionof co-polyesters of 3-hydroxybutyrate from butyric and penta-noic acids. Appl. Microbiol. Biotechnol. 28:330-334.

12. Donald, R. G. K., C. K. Raymond, and R. A. Ludwig. 1985.Vector insertion mutagenesis of Rhizobium sp. strain ORS571:

- -

VOL. 170, 1988

co. (D


direct cloning of mutagenized DNA sequences. J. Bacteriol.162:317-323.

13. Friedrich, B., C. Hogrefe, and H. G. Schlegel. 1981. Naturallyoccurring genetic transfer of hydrogen-oxidizing ability betweenstrains of Alcaligenes eutrophus. J. Bacteriol. 147:198-205.

14. Fukui, T., A. Yoshimoto, M. Matsumoto, S. Hosokawa, T. Saito,H. Nishikawa, and T. Kenkichi. 1976. Enzymatic synthesis ofpoly-3-hydroxybutyrate in Zoogloea ramigera. Arch. Micro-biol. 110:149-156.

15. Gilbert, P., and M. R. W. Brown. 1978. Effect of R-plasmid RP1and nutrient depletion on the cross-cellular composition ofEscherichia coli and its resistance to some uncoupling agents. J.Bacteriol. 133:1062-1065.

16. Gottschalk, G. 1964. Die Biosynthese der Poly-,-hydroxybut-tersaure durch Knallgasbakterien. II. Verwertung organischerSauren. Arch. Mikrobiol. 47:230-235.

17. Griebel, R. J., J. M. Merrick. 1971. Metabolism of poly-,B-hydroxybutyrate: effect of mild alkaline extraction on nativepoly-f3-hydroxybutyrate granules. J. Bacteriol. 108:782-789.

18. Hanahan, D. 1983. Studies on transformation of Escherichia coliwith plasmids. J. Mol. Biol. 166:557-580.

19. Hanna, Z., C. Fregeau, G. Prefontaine, and R. Brousseau. 1984.Construction of a family of universal plasmid vectors. Gene 30:247-250.

20. Haywood, G. W., A. J. Anderson, L. Chu, and E. A. Dawes.1988. Characterization of two 3-ketothiolases possessing dif-fering substrate specifities in the polyhydroxyalkanoate synthe-sizing organism Alcaligenes eutrophus. FEMS Microbiol. Lett.52:91-96.

21. Haywood, G. W., A. J. Anderson, L. Chu, and E. A. Dawes.1988. The role of NADH- and NADPH-linked acetoacetyl-CoAreductases in the poly-3-hydroxybutyrate synthesizing organismAlcaligenes eutrophus. FEMS Microbiol. Lett. 52:259-264.

22. Hohn, B. 1979. In vitro packaging of lambda and cosmid DNA.Methods Enzymol. 68:299-309.

23. Hohn, B., and J. Collins. 1980. A small cosmid for efficientcloning of large DNA fragments. Gene. 11:291-298.

24. Holmes, P. A. 1985. Applications of PHB-a microbially pro-duced biodegradable thermoplastic. Phys. Technol. 16:32-36.

24a.Husemann, M., R. Klintworth, V. Buttcher, J. Salnikow, C.Weissenborn, and B. Bowien. 1988. Chromosomally and plas-mid-encoded gene cluster for CO2 fixation (cfx genes) in Alca-ligenes eutrophus. Mol. Gen. Genet. 214:112-120.

25. Jendrossek, D., A. Steinbuchel, and H. G. Schlegel. 1988. Alco-hol dehydrogenase gene from Alcaligenes eutrophus: subclon-ing, heterologous expression in Escherichia coli, sequencing,and location of TnS insertions. J. Bacteriol. 170:5248-5246.

26. Jorgensen, R. A., S. J. Rothstein, and W. S. Reznikoff. 1979. Arestriction enzyme cleavage map of TnS and location of a regionencoding neomycin resistance. Mol. Gen. Genet. 177:65-72.

27. Juttner, R. R., R. M. Lafferty, and H. J. Knackmuss. 1975. Asimple method for the determination of poly-,-hydroxybutyricacid in microbial biomass. Eur. J. Appl. Microbiol. 1:233-237.

28. Klintworth, R., M. Husemann, C. Weissenborn, and B. Bowien.1988. Expression of the plasmid-encoded phosphoribuloki-nase gene from Alcaligenes eutrophus. FEMS Microbiol. Lett.49:1-6.

29. Knauf, V. C., and E. W. Nester. 1982. Wide host range cloningvectors: a cosmid clone bank of an Agrobacterium Ti plasmid.Plasmid 8:45-54.

30. Kuhn M., D. Jendrossek, C. Frund, A. Steinbuchel, and H. G.Schlegel. 1988. Cloning of the gene for alcohol dehydrogenase ofAlcaligenes eutrophus. J. Bacteriol. 170:685-692.

31. Lemoigne, M. 1926. Produits de deshydration et de polymerisa-tion de l'acide ,-oxybutyrique. Bull. Soc. Chim. Biol. 8:770-782.

32. Loenen, W. A. M., and W. J. Brammar. 1980. A bacteriophagelambda vector for cloning large fragments made with severalrestriction enzymes. Gene. 10:249-259.

33. Lynen, F., and H. Wieland. 1955. 1-Ketoreductase. MethodsEnzymol. 1:566-573.

34. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular

cloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

35. Marmur, J. 1961. A procedure for the isolation of desoxyribo-nucleic acids from microorganisms. J. Mol. Biol. 3:208-218.

36. Merrick, J. M., and M. Doudoroff. 1960. Enzymatic synthesis ofpoly-3-hydroxybutyric acid in bacteria. Nature (London) 189:890-892.

37. Moskowitz, G. J., and J. M. Merrick. 1969. Metabolism ofpoly-3-hydroxybutyrate. II. Enzymatic synthesis of D-(-)-P-hydroxybutyryl coenzyme A by an enoyl hydrase from Rhodo-spirillum rubrum. Biochemistry 8:2748-2755.

38. Nishimura, T., T. Saito, and K. Tomita. 1978. Purification andproperties of 1-ketothiolase from Zoogloea ramigera. Arch.Microbiol. 116:21-27.

39. Nunn, W. D. 1986. A molecular view of fatty acid catabolism inEscherichia coli. Microbiol. Rev. 50:179-192.

40. Ochoa, S. 1955. Crystalline condensing enzyme from pig heart.Methods Enzymol. 1:685-694.

41. Oeding, V., and H. G. Schlegel. 1973. ,-Ketothiolase fromHydrogenomonas eutropha H16 and its significance in theregulation of poly-,3-hydroxybutyrate metabolism. Biochem. J.134:239-248.

42. Ourisson, G., M. Rohmo, and K. Poralla. 1987. Prokaryotichopanoids and other polyterpenoid sterol surrogates. Annu.Rev. Microbiol. 41:301-333.

43. Pedr6s-Alio, C., J. Mas, and R. Guerrero. 1985. The influence ofpoly-p-hydroxybutyrate accumulation on cell volume and buoy-ant density in Alcaligenes eutrophus. Arch. Microbiol. 143:178-184.

44. Ploux, O., S. Masamune, and C. T. Walsh. 1988. The NADPH-linked acetoacetyl-CoA reductase from Zoogloea ramigera.Eur. J. Biochem. 174:177-182.

45. Reusch, R. N., T. W. Hiske, and H. L. Sadoff. 1986. Poly-p-hydroxybutyrate membrane structure and its relationship togenetic transformability in Escherichia coli. J. Bacteriol. 168:553-562.

46. Schindler, J. 1964. Die Synthese der Poly-p-hydroxybuttersauredurch Hydrogenomonas H16: die zu 3-Hydroxybutyryl-Coenzym A fuhrenden Reaktionsschritte. Arch. Mikrobiol. 49:236-255.

47. Schindler, J., and H. G. Schlegel. 1963. D(-)-p-Hydroxybutter-saure-Dehydrogenase aus Hydrogenomonas H16. Biochem. Z.339:309-316.

48. Schlegel, H. G., and G. Gottschalk. 1962. Poly-p-hydroxybut-tersaure, ihre Verbreitung, Funktion und Biosynthese. Angew.Chem. 74:342-347.

49. Schlegel, H. G., H. Kaltwasser, and G. Gottschalk. 1961. EinSubmersverfahren zur Kultur wasserstoffoxidierender Bakte-rien: wachstumsphysiologische Untersuchungen. Arch. Mikro-biol. 38:209-222.

50. Schlegel, H. G., R. Lafferty, and I. Krauss. 1970. The isolation ofmutants not accumulating poly-,B-hydroxybutyric acid. Arch.Mikrobiol. 71:283-294.

51. Simon, E. J., and D. Shemin. 1953. The preparation of S-succinyl-coenzyme A. J. Am. Chem. Soc. 75:2520.

52. Simon, R. 1984. High frequency mobilization of gram-negativebacterial replicons by the in vitro constructed Tn5-mob trans-poson. Mol. Gen. Genet. 196:413-420.

53. Simon, R., U. Priefer, and A. Pukhler. 1983. A broad host rangemobilization system for in vivo genetic engineering: transposonmutagenesis in gram negative bacteria. Bio/Technology 1:784-791.

54. Simon, R., U. Priefer, and A. Puhler. 1983. Vector plasmids forin vivo and in vitro manipulations of gram-negative bacteria, p.98-106. In A. Puhler (ed.), Molecular genetics of the bacteria-plant interaction. Springer Verlag KG, Berlin.

55. Srivastava, S., M. Urban, and B. Friedrich. 1982. Mutagenesisof Alcaligenes eutrophus by insertion of the drug-resistancetransposon Tn5. Arch. Microbiol. 131:203-207.

56. Steinbuchel, A. 1986. Expression of the Escherichia coli pfkAgene in Alcaligenes eutrophus and in other gram-negative bac-teria. J. Bacteriol. 166:319-327.

J. BACTERIOL.

CLONING OF A. EUTROPHUS PHB GENES

57. Tomita, K., T. Saito, and T. Fukui. 1983. Bacterial metabolismof poly-p-hydroxybutyrate, p. 353-366. In D. L. F. Lennon,F. W. Stratman, and R. N. Zahlten (ed.), Biochemistry ofmetabolic processes. Elsevier Science Publishing, Inc., NewYork.

58. Vieira, J., and J. Messing. 1982. The pUC plasmids, an

M13mp7-derived system for insertion mutagenesis and sequenc-ing with universal primers. Gene 19:259-268.

59. Vollbrecht, D., M. A. El Nawawy, and H. G. Schlegel. 1978.Excretion of metabolites by hydrogen bacteria. I. Autotrophicand heterotrophic fermentations. Eur. J. Appl. Microbiol. Bio-technol. 6:145-155.

VOL. 170, 1988 5847

cloning of the alcaligenes eutrophus genes for synthesis of phb

Documents