1997, 63(12):4765. Appl. Environ. Microbiol. 

F Wang and S Y Lee recombinant Escherichia coli.fed-batch culture of filamentation-suppressed Production of poly(3-hydroxybutyrate) by

http://aem.asm.org/content/63/12/4765Updated information and services can be found at:

These include:

CONTENT ALERTS more»cite this article),

Receive: RSS Feeds, eTOCs, free email alerts (when new articles

http://journals.asm.org/site/misc/reprints.xhtmlInformation about commercial reprint orders: http://journals.asm.org/site/subscriptions/To subscribe to to another ASM Journal go to:

on October 20, 2013 by guest

http://aem.asm

.org/D

ownloaded from

on O

ctober 20, 2013 by guesthttp://aem

.asm.org/

Dow

nloaded from

on October 20, 2013 by guest

http://aem.asm

.org/D

ownloaded from

on O

ctober 20, 2013 by guesthttp://aem

.asm.org/

Dow

nloaded from

on October 20, 2013 by guest

http://aem.asm

.org/D

ownloaded from

on O

ctober 20, 2013 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPLIED AND ENVIRONMENTAL MICROBIOLOGY,0099-2240/97/$04.0010

Dec. 1997, p. 4765–4769 Vol. 63, No. 12

Copyright © 1997, American Society for Microbiology

Production of Poly(3-Hydroxybutyrate) by Fed-Batch Culture ofFilamentation-Suppressed Recombinant Escherichia coli

FULAI WANG AND SANG YUP LEE*

Department of Chemical Engineering and BioProcess Engineering Research Center,Korea Advanced Institute of Science and Technology, Yusong-gu,

Taejon 305-701, Korea

Received 30 June 1997/Accepted 2 October 1997

Recombinant Escherichia coli XL1-Blue harboring a high-copy-number plasmid containing the Alcaligeneseutrophus polyhydroxyalkanoate synthesis genes could efficiently synthesize poly(3-hydroxybutyrate) (PHB) ina complex medium containing yeast extract and tryptone but not in a defined medium. One of the reasons forthe reduced PHB production in a defined medium was thought to be severe filamentation of cells in thismedium. By overexpressing an essential cell division protein, FtsZ, in recombinant E. coli producing PHB,filamentation could be suppressed and PHB could be efficiently produced in a defined medium. A high PHBconcentration of 149 g/liter, with high productivity of 3.4 g of PHB/liter/h, could be obtained by the pH-statfed-batch culture of the filamentation-suppressed recombinant E. coli in a defined medium. It was also foundthat insufficient oxygen supply at a dissolved oxygen concentration (DOC) of 1 to 3% of air saturation duringactive PHB synthesis phase did not negatively affect PHB production. By growing cells to the concentration of110 g/liter and then controlling the DOC in the range of 1 to 3% of air saturation, a PHB concentration of 157g/liter and PHB productivity of 3.2 g of PHB/liter/h were obtained. For the scale-up studies, fed-batch culturewas carried out in a 50-liter stirred tank fermentor, in which the DOC decreased to zero when cell concen-tration reached 50 g/liter. However, a relatively high PHB concentration of 101 g/liter and PHB productivity of2.8 g of PHB/liter/h could still be obtained, which demonstrated the possibility of industrial production of PHBin a defined medium by employing the filamentation-suppressed recombinant E. coli.

Biodegradable plastics have been drawing much attentiondue to the environmental problems caused by the accumula-tion of nondegradable plastic wastes. Polyhydroxyalkanoates(PHAs), intracellularly accumulated by various microorgan-isms as carbon and energy storage materials, have been con-sidered to be strong candidates for biodegradable polymermaterial (1, 6, 17). A major drawback to the commercializationof PHAs is their much higher production cost compared withpetrochemical-based synthetic plastic materials or other bio-degradable polymers such as poly(lactide) (7). With the aim oflowering the production cost of PHAs, various fermentationprocesses employing several different bacteria have been de-veloped to improve the production of PHAs (7, 13). Alcaligeneseutrophus has been the most widely employed bacterium forthe production of PHAs since it can accumulate a largeamount of PHAs in a simple medium (4). Poly(3-hydroxybu-tyrate) (PHB) is a homopolymer of 3-hydroxybutyrate and isthe most widespread and best-characterized member of thePHAs. In A. eutrophus PHB is synthesized from acetyl coen-zyme A (acetyl-CoA) by three sequential enzymatic reactions(1). Two acetyl-CoA moieties are condensed to acetoacetyl-CoA by b-ketothiolase. An NADPH-dependent reductase con-verts acetoacetyl-CoA to D-(2)-3-hydroxybutyryl-CoA, whichis subsequently added to the growing chain of PHB by the PHAsynthase. The genes coding for these three enzymes werecloned in Escherichia coli and shown to form an operon (16, 18,19). We constructed the stable high-copy-number plasmidpSYL104 containing the A. eutrophus PHA synthesis genes (6,

14). A high concentration of PHB (81 g/liter) could be pro-duced with a high productivity of 2.1 g of PHB/liter/h by thepH-stat fed-batch culture of recombinant E. coli XL1-Blueharboring pSYL104 in a complex medium supplemented witha large amount of yeast extract and tryptone (14). However, itis not desirable to use a large amount of expensive yeast extractand tryptone for the production of PHB, which is to be used asan inexpensive bulk product. With the aim of lowering theproduction cost of PHB, fed-batch cultures were carried out ina chemically defined medium. Unfortunately, only less than25 g of PHB/liter could be produced by fed-batch culture ofXL1-Blue(pSYL104) in a defined medium (11). One of thereasons for the reduced accumulation of PHB in a definedmedium was filamentation of cells accumulating PHB (15). Itwas observed that cells became considerably elongated (up to150 mm in length) in flask and fed-batch cultures (11, 15), andthe extent of filamentation was much more severe in a definedmedium (5). Cell filamentation was due to the inactivation ofan essential cell division protein, FtsZ, and could be sup-pressed by amplifying the activity of FtsZ (5). More interest-ingly, in flask culture, recombinant E. coli with amplified FtsZactivity could more efficiently synthesize PHB in a definedmedium compared with cells that underwent filamentation (5).In this study fed-batch cultures of filamentation-suppressedrecombinant E. coli were carried out for the production ofPHB in a defined medium. A cultivation strategy that alloweda high PHB productivity of greater than 3 g of PHB/liter/h wasdeveloped. It was also demonstrated that PHB could be effi-ciently produced in a pilot-scale fermentor by using this strat-egy.

MATERIALS AND METHODS

Microorganism and plasmid. The E. coli strain used in this study was XL1-Blue (recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac[F9 proAB, lacIqZDM15]

* Corresponding author. Mailing address: Department of ChemicalEngineering and BioProcess Engineering Research Center, Korea Ad-vanced Institute of Science and Technology, 373-1 Kusong-dong,Yusong-gu, Taejon 305-701, Korea. Fax: 42-869-8800. E-mail: leesy@sorak.kaist.ac.kr.

4765

Tn10 [Tetr]). The plasmid pSYL107 containing the A. eutrophus PHA biosyn-thesis genes and the E. coli ftsZ gene has been previously described (5).

Culture conditions. Cells were maintained as a 20% (vol/vol) glycerol stock at280°C after growing in Luria-Bertani medium. Seed and fed-batch cultures werecarried out in a chemically defined MR medium which was modified from thepreviously described R medium (9). The MR medium (pH 6.9) contained thefollowing per liter: KH2PO4, 22 g; (NH4)2HPO4, 3 g; MgSO4 z 7H2O, 0.7 g; citricacid, 0.8 g; trace metal solution, 5 ml. The trace metal solution contained thefollowing per liter of 5 M HCl: FeSO4 z 7H2O, 10.0 g; CaCl2, 2.0 g; ZnSO4 z7H2O, 2.2 g; MnSO4 z 4H2O, 0.5 g; CuSO4 z 5H2O, 1.0 g; (NH4)6Mo7O24 z 4H2O,0.1 g; Na2B4O7 z 10H2O, 0.02 g. Separately sterilized glucose and thiaminesupplemented the medium to final concentrations of 20 g/liter and 10 mg/liter,respectively.

Seed cultures were prepared in flasks and incubated in a rotary shaker over-night at 30°C and 250 rpm. Laboratory-scale fed-batch culture was carried out at30°C in a jar fermentor (6.6 liters) (Bioflo 3000; New Brunswick Scientific Co.,Edison, N.J.) containing 1.2 liters of MR medium. Culture pH was controlled at6.9 by the addition of 28% (vol/vol) ammonia-water. The dissolved oxygenconcentration (DOC) was controlled as desired by automatic control of theagitation speed (up to 1,000 rpm) and the pure oxygen percentage. Antifoam(0.02% [vol/vol]) (antifoam 289; Sigma Chemical Co., St. Louis, Mo.) was addedat the onset of cultivation. The feeding solution used for the fed-batch culturecontained the following per liter: glucose, 700 g; MgSO4 z 7H2O, 15 g; thiamine,250 mg. The pH-stat feeding strategy, which was based on the sharp pH rise uponcarbon source depletion, was employed in fed-batch cultures. When the pH roseto a value greater than its set point (6.9) by 0.1, an appropriate volume of feedingsolution was automatically added to increase the glucose concentration in theculture medium to 20 g/liter. The feeding volume was calculated on-line by usingfermentation software (AFS3.42; New Brunswick Scientific Co.).

For the scale-up studies in a 50-liter stirred tank fermentor (Korea FermentorCompany, Inchon, Korea), seed culture (4 liters) was prepared in a 6.6-literfermentor by batch culture. When glucose was depleted, as indicated by a sharpincrease of the pH, cell broth was immediately transferred to a 50-liter fermentorcontaining 12 liters of MR medium. The culture conditions were similar to thosedescribed above except for some modifications in the feeding strategy and theDOC control scheme to be suitable for the 50-liter fermentor. Since the 50-literfermentor was not interfaced to a computer, a fixed feeding volume (0.5 liters perpulse) was added upon the increase of the pH. The DOC was maintained above20% of air saturation during the first 15 h by varying the air flow rate at the fixedagitation speed (500 rpm, the maximum agitation speed of this fermentor). Whenthe DOC could not be maintained above 20% of air saturation, oxygen-enrichedair (83% O2) generated by an oxygen generator (Korea Energy Industry Co.,Uiwang, Korea) was provided. When the DOC could not be maintained above20% of air saturation even after the oxygen-enriched air supply was added,cultivation was simply continued at this maximum level of oxygen supply.

Analytical procedures. Growth was monitored by measuring the optical den-sity at 600 nm. Cell concentration, defined as the amount (dry weight) of cells perliter of culture broth, was determined by weighing dry cells as described previ-ously (14). PHB concentration was determined by gas chromatography (HP5890;Hewlett-Packard, Wilmington, Del.) with n-butyric acid as an internal standard

(2). Residual cell concentration was defined as cell concentration minus PHBconcentration. PHB content was defined as the percentage of the ratio of theamount of PHB to the amount (dry weight) of cells. The PHB synthesis rate wasdefined as grams of PHB synthesized per gram of residual cells per hour.

RESULTS

Fed-batch culture in a 6.6-liter fermentor. The pH-stat fed-batch culture of XL1-Blue(pSYL107) was first carried out ina laboratory-scale fermentor. The time profiles of cell growth,PHB accumulation, and DOC are shown in Fig. 1. The finalcell concentration, PHB concentration, and PHB content ob-tained in 44 h were 206 g/liter, 149.7 g/liter, and 73%, respec-tively, which resulted in a PHB productivity of 3.4 g of PHB/liter/h. For convenience, the fermentation process could bedivided into two phases, an active growth phase and an activePHB synthesis phase. Before cell concentration reached 60.2g/liter at 15 h, the PHB content was kept nearly constant at ca.16% (active cell growth phase). Afterwards, it steadily in-creased to 73% (active polymer synthesis phase). The DOCcould not be maintained above 12% of air saturation when thecell concentration reached 97 g/liter at 18 h. From 18 h to theend of the cultivation (44 h), the DOC fluctuated in the rangeof 7 to 12% of air saturation. At the end of cultivation, a totalof 4.9 liters of cell broth was harvested. A total amount of 734 gof PHB (without considering the PHB in samples) was pro-duced from 2,692 g of glucose (including the glucose in theinitial medium and seed culture), resulting in a yield of 0.27 gof PHB/g of glucose.

Effect of insufficient oxygen supply on PHB synthesis. Theeffect of insufficient oxygen supply on PHB production wasinvestigated in fed-batch cultures. Oxygen limitation of main-taining the DOC at 1 to 3% of air saturation was appliedduring two different phases of the fermentation as describedearlier. First, oxygen limitation was applied at a low cell con-centration of 8.3 g/liter (active cell growth phase). As shown inFig. 2, cell growth stopped after 1.5 h of oxygen limitation.PHB concentration and PHB content increased only from 1.3to 3.0 g/liter and from 15 to 28%, respectively. Therefore, itwas concluded that oxygen limitation during the active growth

FIG. 1. Time profiles of cell concentration, PHB concentration, residual cellconcentration, PHB content, and DOC during the fed-batch culture of recom-binant E. coli XL1-Blue(pSYL107) in a chemically defined medium. The DOCwas maintained above 7% of air saturation.

FIG. 2. Time profiles of cell concentration, PHB concentration, residual cellconcentration, PHB content, and DOC during the fed-batch culture of recom-binant E. coli XL1-Blue(pSYL107) in a chemically defined medium. The DOCwas controlled in the range of 1 to 3% of air saturation during the active cellgrowth phase.

4766 WANG AND LEE APPL. ENVIRON. MICROBIOL.

phase was detrimental not only to cell growth but also to PHBsynthesis.

Figure 3 shows the results obtained by applying oxygen lim-itation during the active PHB synthesis phase. Active PHBsynthesis began at 21 h when cell concentration reached 75.6g/liter. The DOC was controlled in the range of 1 to 3% of airsaturation after cell concentration reached 109.8 g/liter at 24 h.Cell concentration, PHB concentration, and PHB content ob-tained at 49 h were 204.3 g/liter, 157.1 g/liter, and 77%, re-spectively, resulting in a PHB productivity of 3.2 g of PHB/liter/h. These results suggest that application of oxygenlimitation during the active PHB synthesis phase did not neg-atively affect the production of PHB by this recombinant E. coliin a defined medium.

PHB production in a pilot-scale fermentor. To examine ifthe cultivation strategy described above could be applied to theproduction of PHB in a large-scale fermentor, in which oxygentransfer is generally poor, fed-batch culture of XL1-Blue(pSYL107) was carried out in a 50-liter fermentor (Fig. 4).Oxygen-enriched air generated by the pressure swing adsorp-tion equipment, instead of pure oxygen, was used during thefermentation. In this fermentation, the DOC was not deliber-ately lowered but rather decreased to zero due to the oxygentransfer limitation of the fermentor itself (maximum agitationspeed and oxygen-enriched air flow rate of 500 rpm and 5liters/min, respectively). During the cultivation the switch fromthe active growth phase to the active PHB synthesis phaseoccurred at 12 h (cell concentration of 36.6 g/liter). The DOCdropped to zero when cell concentration reached 54.8 g/liter at15 h. However, a detrimental effect of oxygen limitation wasnot observed. Cell and PHB concentrations obtained in 36 hwere 153.7 and 101.3 g/liter, respectively, resulting in a PHBproductivity of 2.8 g of PHB/liter/h. Therefore, it was con-cluded that the filamentation-suppressed recombinant E. colicould be employed for the efficient production of PHB in adefined medium in a large-scale fermentor. A total of 19.4liters of feeding solution was consumed, and 38 liters of cellbroth was harvested at the end of cultivation. Therefore, the

PHB yield was 0.28 g of PHB/g of glucose (3,849 g of PHBproduced from 13,900 g of glucose).

DISCUSSION

Even though PHAs have been considered to be good can-didates for completely biodegradable plastic or elastomer ma-terial, the high production cost has hampered their use in awide range of applications. Several fermentation processes em-ploying A. eutrophus, Alcaligenes latus, methylotrophs, and re-combinant E. coli have been developed for more economicalproduction of PHB (6, 7, 13). A PHB productivity of 2 to 3 gof PHB/liter/h has been obtained by employing A. eutrophusand recombinant E. coli. Computer-aided process analysis andeconomic evaluation for PHB production by bacterial fermen-tation suggested that obtaining high PHB productivity and highPHB content from inexpensive medium was important for re-ducing the production cost of PHB (3). Among the severalbacteria that have been shown to produce PHB efficiently,recombinant E. coli harboring the A. eutrophus PHA biosyn-thesis genes has been considered to be a strong candidate as aPHB producer due to several advantages such as a wide rangeof utilizable carbon sources, easy recovery of PHB, and nodegradation of PHB during fermentation due to the lack ofintracellular depolymerases (12). However, a high concentra-tion of PHB (greater than 80 g/liter) with a high PHB produc-tivity (greater than 2 g of PHB/liter/h) could be obtained onlyin a complex or semi-defined medium (11, 12, 14). One of theseveral reasons for the reduced synthesis of PHB in a definedmedium was filamentation of cells accumulating PHB (5, 15).We have demonstrated in flask cultures that cell filamentationcould be suppressed by overexpressing FtsZ, an essential celldivision protein (5). Furthermore, this filamentation-sup-pressed recombinant E. coli XL1-Blue(pSYL107) could accu-mulate more PHB in a defined medium in flask cultures (5).Therefore, fed-batch cultures of XL1-Blue(pSYL107) werecarried out to examine if PHB could be efficiently produced ina defined medium. As shown in Fig. 1, an even higher concen-tration of PHB (149 g/liter) and a higher PHB productivity (3.4

FIG. 3. Time profiles of cell concentration, PHB concentration, residual cellconcentration, PHB content, and DOC during the fed-batch culture of recom-binant E. coli XL1-Blue(pSYL107) in a chemically defined medium. The DOCwas controlled in the range of 1 to 3% air saturation during the active PHBsynthesis phase.

FIG. 4. Time profiles of cell concentration, PHB concentration, residual cellconcentration, PHB content, and DOC during the fed-batch culture of recom-binant E. coli XL1-Blue(pSYL107) in a 50-liter stirred tank fermentor in achemically defined medium. The DOC dropped to zero when cell concentrationreached 54.8 g/liter at 15 h.

VOL. 63, 1997 PHB PRODUCTION BY RECOMBINANT ESCHERICHIA COLI 4767

g of PHB/liter/h) could be obtained in a defined medium com-pared with a complex or semi-defined medium. No significantcell filamentation was observed during the entire fermentation.Having achieved the successful production of PHB to a highconcentration with high content and high productivity, we nextexamined the effects of applying oxygen limitation on cellgrowth and PHB production for two purposes. First, it may bepossible to enhance the PHB synthesis rate by increasing theacetyl-CoA flux into the PHB biosynthetic pathway and reduc-ing its flux into the tricarboxylic acid cycle. Second, it is difficultto maintain the DOC above 10% of air saturation at high celldensities without using pure oxygen. Since oxygen transfer iseven worse in a large-scale fermentor and the use of a largeamount of pure oxygen can be economically unfavorable, it willbe beneficial if PHB can be efficiently produced under oxygen-limited conditions.

The fed-batch process of fermentation with recombinantE. coli can be conveniently divided into two phases: (i) anactive growth phase during which PHB content is kept rela-tively constant at a low level (15 to 20%) and (ii) an active PHBsynthesis phase during which PHB is actively accumulated witha concomitant increase of PHB content. When oxygen limita-tion was applied during the active growth phase, cell growthstopped and no apparent enhancement of PHB accumulationwas observed (Fig. 2). However, cell growth and PHB produc-tion were not hampered when oxygen limitation was appliedduring the active PHB synthesis phase (the final PHB concen-tration and PHB productivity were as high as 157.1 g/liter and3.2 g of PHB/liter/h, respectively). A major reason for thisdifference may have been related to acetate accumulation.Oxygen limitation applied during the active growth phase re-sulted in accumulation of acetate to a final concentration of 14g/liter, which is higher than the critical acetate concentrationthat shows growth inhibition (8). The acetate concentrationwas less than 4 g/liter when oxygen limitation was appliedduring the active PHB synthesis phase. The actual reason forthese findings is not clear at this moment. Nevertheless, it isnice from a process point of view to find that oxygen limitationcould enhance PHB production by filamentation-suppressedrecombinant E. coli in a defined medium. To examine if thisstrategy could be employed for the production of PHB in alarge-scale fermentor, fed-batch culture was carried out in a50-liter pilot-scale fermentor. This fermentor is a typicalstirred tank bioreactor equipped with three bottom-drivenRushton turbine impellers with a maximum agitation speed of500 rpm. Instead of using pure oxygen, oxygen-enriched air(83% O2) was supplied when required. Although the DOCdecreased to zero when cell concentration reached 54.8 g/literat the maximum gas flow rate and agitation speed (Fig. 4), ahigh concentration of PHB (101.3 g/liter) was still obtained in36 h, resulting in a relatively high PHB productivity of 2.8 g ofPHB/liter/h. Since an oxygen-limited condition is likely to beencountered in a large-scale fermentor without using pureoxygen, it is encouraging to find that PHB could be efficientlyproduced under this condition.

We next examined whether the enhanced production ofPHB in a defined medium under oxygen limitation was due tothe increased PHB synthesis rate. The maximum PHB synthe-sis rates during the fermentation under oxygen-sufficient (Fig.1) and oxygen-limited (Fig. 3) conditions were 0.11 and 0.12 gof PHB/g of residual cell/h, respectively. Therefore, oxygenlimitation apparently did not enhance the PHB synthesis rate.However, the maximum PHB synthesis rate obtained in a de-fined medium with or without oxygen limitation was lower thanthat obtained in a complex or semi-defined medium (0.15 to0.20 g of PHB/g of residual cell/h) (11, 14, 15). Comparison of

the time profiles of the residual cell concentrations in defined(Fig. 1 and 3), complex (11), and semi-defined (10) mediasuggests that the higher PHB concentration and productivityobtained in a defined medium were due to the higher residualcell concentration before cells entered the active PHB synthe-sis phase. In a complex or semi-defined medium, the typicalresidual cell concentration at the switch from active growthphase to active PHB synthesis phase was about 30 g/liter (10,11), while that in a defined medium was greater than 60 g/liter(Fig. 1 and 3). Since PHB is accumulated in the cytoplasm, theresidual cell concentration determines how much PHA canpotentially be accumulated. However, we cannot increase theresidual cell concentration before the active PHA synthesisphase to too high a value because, as was shown previously,such a concentration results in incomplete PHB synthesis and,subsequently, a low PHB content (7). The residual cell con-centration of 60 g/liter was found to be optimal for the pro-duction of PHB by this recombinant E. coli strain in a definedmedium.

Even though recombinant E. coli has been suggested to haveseveral advantages as a PHB producer, there were two prob-lems to be solved: the requirement for complex nitrogensources and a high oxygen demand. In this study we demon-strated that PHB could be efficiently produced in a definedmedium by the filamentation-suppressed recombinant E. colistrain. We also solved the problem of high oxygen demand bydemonstrating that PHB could be produced to a high concen-tration with high productivity by applying oxygen limitationduring the active PHA synthesis phase. However, there is onenew problem observed, i.e., a slightly lower PHB yield onglucose in a defined medium. The PHB yield obtained in thisstudy was 0.27 to 0.28 g of PHB/g of glucose, which is lowerthan that obtained in a complex medium (0.37 g of PHB/g ofglucose) (12). It seems that more glucose was wasted to carbondioxide in a defined medium. To investigate the possibility ofimproving the PHB yield, we are now carrying out fed-batchcultures under various conditions while monitoring the carbondioxide evolution rate by using an on-line mass spectrometer.

ACKNOWLEDGMENTS

We thank the BioProcess Engineering Research Center for thepartial support of Fulai Wang. This work was also supported by theMinistry of Science and Technology and by LG Chemicals, Ltd.

REFERENCES

1. Anderson, A., and E. A. Dawes. 1990. Occurrence, metabolism, metabolicrole, and industrial uses of bacterial polyhydroxyalkanoates. Microbiol. Rev.54:450–472.

2. Braunegg, G., B. Sonnleitner, and R. M. Lafferty. 1978. A rapid gas chro-matographic method for the determination of poly-3-hydroxybutyric acid inmicrobial biomass. Eur. J. Appl. Microbiol. Biotechnol. 6:29–37.

3. Choi, J. I., and S. Y. Lee. 1997. Process analysis and economic evaluation forpoly(3-hydroxybutyrate) production by fermentation. Bioprocess Eng. 17:335–342.

4. Kim, B. S., S. C. Lee, S. Y. Lee, Y. K. Chang, and S. I. Woo. 1994. Productionof poly(3-hydroxybutyric acid) by fed-batch culture of Alcaligenes eutrophuswith glucose concentration control. Biotechnol. Bioeng. 43:892–898.

5. Lee, S. Y. 1994. Suppression of filamentation in recombinant E. coli byamplified FtsZ activity. Biotechnol. Lett. 16:1247–1252.

6. Lee, S. Y. 1996. Bacterial polyhydroxyalkanoates. Biotechnol. Bioeng. 49:1–14.

7. Lee, S. Y. 1996. Plastic bacteria? Progress and prospects for polyhydroxyal-kanoate production in bacteria. Trends Biotechnol. 14:431–438.

8. Lee, S. Y. 1996. High cell-density culture of Escherichia coli. Trends Bio-technol. 14:98–105.

9. Lee, S. Y., and H. N. Chang. 1993. High cell density cultivation of Escherichiacoli W using sucrose as a carbon source. Biotechnol. Lett. 15:971–974.

10. Lee, S. Y., and H. N. Chang. 1994. Effect of complex nitrogen source on thesynthesis and accumulation of poly(3-hydroxybutyric acid) by recombinantEscherichia coli in flask and fed-batch cultures. J. Environ. Polym. Degrad.2:169–176.

4768 WANG AND LEE APPL. ENVIRON. MICROBIOL.

11. Lee, S. Y., and Y. K. Chang. 1994. Production of poly(3-hydroxybutyric acid)by recombinant Escherichia coli. Ann. N.Y. Acad. Sci. 721:43–53.

12. Lee, S. Y., and H. N. Chang. 1995. Production of poly(3-hydroxybutyric acid)by recombinant Escherichia coli strains: genetic and fermentation studies.Can. J. Microbiol. 41(Suppl. 1):207–215.

13. Lee, S. Y., and H. N. Chang. 1995. Production of poly-(hydroxyalkanoicacid). Adv. Biochem. Eng. Biotechnol. 52:27–58.

14. Lee, S. Y., K. S. Yim, H. N. Chang, and Y. K. Chang. 1994. Construction ofplasmids, estimation of plasmid stability, and use of stable plasmids for theproduction of poly(3-hydroxybutyric acid) in Escherichia coli. J. Biotechnol.32:203–211.

15. Lee, S. Y., K. M. Lee, H. N. Chang, and A. Steinbuchel. 1994. Comparison ofrecombinant Escherichia coli strains for synthesis and accumulation of poly-

(3-hydroxybutyric acid) and morphological changes. Biotechnol. Bioeng. 44:1337–1347.

16. Peoples, O. P., and A. J. Sinskey. 1989. Poly-3-hydroxybutyric biosynthesis inAlcaligenes eutrophus H16. Identification and characterization of the PHBpolymerase gene (phbc). J. Biol. Chem. 264:15298–15303.

17. Poirer, Y., C. Nawrath, and C. Somerville. 1995. Production of poly-hydroxy-alkanoates, a family of biodegradable plastics and elastomers, in bacteria andplants. Bio/Technology 13:142–150.

18. Schubert, P., A. Steinbuchel, and H. G. Schlegel. 1988. Cloning of Alcaligeneseutrophus genes for synthesis of poly-b-hydroxybutyric acid (PHB) and syn-thesis of PHB in Escherichia coli. J. Bacteriol. 170:5837–5847.

19. Slater, S. C., W. H. Voige, and D. E. Dennnis. 1988. Cloning and expressionin Escherichia coli of the Alcaligenes eutrophus H16 poly-b-hydroxybutyratebiosynthetic pathway. J. Bacteriol. 170:4431–4436.

VOL. 63, 1997 PHB PRODUCTION BY RECOMBINANT ESCHERICHIA COLI 4769