Biochemical Engineering Journal 16 (2003) 135–143

Evidence of medium-chain-length polyhydroxyoctanoate accumulation intransgenic potato lines expressing thePseudomonas oleovoransPha-C1

polymerase in the cytoplasm

Andrea Romanoa,1, Dick Vreugdenhilb, Diaan Jamarb, Linus H.W. van der Plasb,Guy de Rooc, Bernard Witholtc, Gerrit Egginka, Hans Mooibroeka,∗

a Wageningen University and Research Centre, ATO B.V., 6700 AA, Postbus 17, Wageningen, The Netherlandsb Institute of Biotechnology, ETH Hönggerberg, CH-8093, Zürich, Switzerland

c Laboratory of Plant Physiology, Wageningen University and Research Centre, Arboretumlaan 4, 6703 BD, Wageningen, The Netherlands

Received 5 July 2002; accepted after revision 23 September 2002

Abstract

ThephaC1 gene fromPseudomonas oleovorans, coding for the Pha-C1 polymerase, was introduced into the potato genome. Transgeniccallus and plant lines which transcribed and translated the transgene were selected and cell suspension cultures from the wild typeand transgenic lines were established. The substrate for the Pha-C1 polymerase, 3-(R)-hydroxyoctanoate, was provided to the growthmedium. In the transgenic lines, but not in the wild type or in transgenic cell suspension cultures without Pha-C1 expression, evidence ofmedium-chain-length polyhydroxyalkanoate accumulation ranging from 0.02 to 9.7 mg of polymer per gram of dry weight was observedafter feeding in the growth medium the substrate.© 2003 Elsevier Science B.V. All rights reserved.

Keywords: Solanum tuberosum; Medium-chain-length polyhydroxyalkanoate; Plant cell suspension culture; Recombinant DNA; Biosynthesis; Bioconversion

1. Introduction

Fluorescent pseudomonads belonging to the rRNA groupI accumulate granules that contain medium-chain-lengthpolyhydroxyalkanoate (mcl-PHA), consisting of 3-(R)-hydroxyalkanoic acids with carbon chains ranging from C6to C14 [1,2]. The PHA-polymerase is the key enzyme forthe synthesis of mcl-PHA using 3-(R)-hydroxy-acyl-CoAderived from fatty acid metabolism. In these pseudomon-ads, thepha-operon[3,4] comprises thephaC1 andphaC2genes coding for the Pha-C1 and Pha-C2 polymerases,respectively. These PHA-polymerases are classified astype II and are specifically active on mcl-alkanoic acids.The two PHA-polymerases share 50–54% identity at theprotein sequence level and have a molecular weight of62–64 kDa. Several additional genes are present[4–7] in

Abbreviations: Mcl, medium-chain-length; PHA, polyhydroxyalka-noate; 35-S, cauliflower mosaic virus 35-S; E-35-S, enhanced 35-S pro-moter; nos, nopaline synthase; 2-4-D, 2,4-dichlorophenoxyacetic acid

∗ Corresponding author. Tel.:+31-317-475316; fax:+31-317-475347.E-mail address:a.mooibroek@ato.wag-ur.nl (H. Mooibroek).

1 Department of Obstetrics and Gynecology, Maastricht UniversityHospital, Debyelaan 25, 6202 AZ Maastricht.

the operon which are involved in PHA degradation or instructural aspects of the PHA granules[8,9]. A high degreeof identity (69–80%) at the amino acid level is observedbetween corresponding mcl-PHA-polymerases from differ-ent species and 40% identity is observed with theRalstoniaeutropha polyhydroxybutyrate (PHB)-polymerase,[3,4],which is active on short-chain-length alkanoates (C4 andC5 carbon units). PHB-polymerase (type I) and type IIPHA-polymerase contain a lipase box[10] and it has beenpostulated that the catalytic mechanism involves residuesCys-296 and Cys-430 (Pseudomonas oleovoransnumbering[3,10]).

Both Pha-C1 and Pha-C2 polymerases are functionalproteins which are able to catalyse mcl-PHA formationindependently from each other when expressed in heterol-ogous hosts[11,12] including plants[13]. Overproductionof the mcl-PHA-polymerase inE. coli did not result in acorresponding increase of PHA yield, and most of the pro-tein was recovered as inactive protein in inclusion bodies[14]. However, solubilisation of the trapped protein re-sulted in active polymerase, and this enzyme was capableof in vitro synthesis of mcl-PHA without any additionalcomponent.

1369-703X/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S1369-703X(03)00030-5

136 A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143

In this work, the Pha-C1 polymerase ofP. oleovoranshasbeen expressed in the cytoplasm of transgenic potato lines.Although plant cytoplasm does not contain any PHA precur-sors due to the compartmentation of fatty acid metabolismsin plants (�-oxidation occurs mainly in peroxisomes andfatty acid biosynthesis (FAB) in the plastids), the abilityof potato to express the Pha-C1 polymerase and to accu-mulate mcl-PHA was tested using a feeding approach, ex-ogenously providing the substrate (3-(R)-hydroxyoctanoate)to cell suspensions of the selected transgenic lines. Evi-dence for mcl-PHA accumulation in transgenic lines ex-pressing thephaC1 gene, but not in the wild type line nor intransgenic lines not expressing the Pha-C1 polymerase, wasobtained.

2. Material and methods

2.1. Plant material and transformation

In vitro grown Solanum tuberosum, genotype 1024-2(amf, diploid; [15]) was used as starting material for parti-cle bombardment-mediated co-transformation as describedin Romano et al.[16–18]. Transgenic cell suspension lineswere derived from both regenerated plants resistant tokanamycin and calli resistant to kanamycin as describedfurther.

2.2. DNA constructs

Plasmid 35-S-Kan containing thenptII selectable markerunder the control of the 35-S promoter and terminator,coding for the neomycin-phospho-transferase, conferringresistance to kanamycin was kindly provided by the JohnInnes Centre (Norwich, UK). Plasmid pAPP63 containedthe phaC1 gene fromP. oleovoransunder the control ofthe enhanced 35-S promoter (E-35S) andnos-terminatorand was constructed as follows (Fig. 1a): Pwo (Pyro-coccus woesei) DNA polymerase (Eurogentec) was used,as described by the manufacturer, for PCR amplificationusing plasmid pET101 (phaC1 gene. [19]) as a tem-plate. Primer extension was used in order to introduce theNcoI and BamHI restriction sites at the 5′ and 3′ endsof the gene. In order to avoid the possible occurrence ofPCR mistakes in the gene (1680 bp long), upstream and

�

Fig. 1. Construction of plasmid pAPP63 (b) containing thephaC1 gene fromP. oleovoransunder the control of the E-35S andnos terminator. ThephaC1 upstream linker (phaC1-UP-LINKER, 357 bp, includingNcoI and SacII restriction sites at the 5′ and the 3′ ends, respectively) and thephaC1downstream linker (phaC1-DW-LINKER, 466 bp, includingXmaIII and BamHI restriction sites at the 5′ and the 3′ ends, respectively) were amplifiedby PCR and cloned in pGEM-T-easy vectors giving rise to the pAPP52 (upstream linker) and pAPP53 (downstream linker). TheSacII/XmaIII 857 bpwild-type fragment of thephaC1 was cut off from pET101 and cloned intoSacII/XmaIII of pBlueKSp (Promega) giving rise to pAPP54. Plasmid pAPP55was created by cloning the 466 bpXmaIII/ BamHI fragment from pAPP53 into pAPP54 digested withXmaIII/ BamHI restriction endonucleases. The 393bp EcoRI fragment from pAPP52, comprising the upstream linker, was cloned into theEcoRI site of pUC18 (Pharmacia) giving plasmid pAPP56.Eventually, the 1323 bpSacII/BamHI fragment from pAPP55 was cloned intoSacII/BamHI sites of pAPP56, thus yielding pAPP57 the complete syntheticphaC1 gene which was subsequently cloned into plasmid pAPP23 (seeSection 2).

downstream linkers were created by PCR using primers5′-CGCGGATCCACCATGGGTAACAAGAACA-3′ and5′-TGACGAACTGGCCGCGGCTGATG, annealing at po-sition 351–373 (which includes the restriction siteSacIIat position 357) and at the 5’ of thephaC1 gene and us-ing primers 5′-CCCTGCCGGCCGCCTTCCACG-3′ and5′-TCCGGATCCTCAACGCTCGTGAACGTA-3′, anneal-ing at position 1209–1228 (including a siteXmaIII atposition 1214) and at the 3′ of the phaC1 gene. All PCRproducts were cloned into pGEM-T-easy vectors (Promega)and sequenced. The completephaC1 gene was re-assembledin plasmid pAPP57 using the two linkers and the 857 bpSacII/XmaIII fragment from pET101 containing the centralpart of the wild-type gene as described in detail inFig. 1a.The syntheticphaC1 gene was extracted byNcoI/BamHIrestriction from plasmid pAPP57 and cloned in pAMV-1[20] giving plasmid pAPP62. TheBglII/BamHI fragmentfrom pAPP62 was finally cloned intoBamHI site of pAPP23[16], giving rise to pAPP63 (Fig. 1b).

2.3. PCR and Southern blot analysis

All molecular techniques were performed using standardprotocols[21]. PCR analyses on genomic DNA was per-formed using DNA isolated with the Sigma Gene-Elute KITaccording to the manufacturer’s recommendations and us-ing the upstream and downstream linker primers to monitorthe integration into the plant genome of thephaC1 gene.For Southern blot analyses, genomic DNA was isolatedusing the CTAB protocol[22] from leaves of plants root-ing on kanamycin-containing medium or cell suspensions,and equal amounts of DNA were blotted on a positivelycharged nylon membrane using the TurboblotterTM sys-tem (Schleicher & Schuell) and using standard techniques[21]. The filters were hybridised with digoxigenin-labelledprobe as described by the manufacturer (Boehringer,Mannheim, DE).

2.4. RT-PCR and Northern blot

RNA was extracted from young leaves and cell suspen-sions as follows: plant material was ground in liquid ni-trogen and RNA was extracted in 1 vol. of RNA-extractionbuffer (0.2 M Na acetate pH 5.0, 10 mM EDTA, 1% SDS)and 1 vol. of equilibrated-phenol at 65◦C. One volume of

A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143 137

138 A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143

chloroform was added and centrifuged for 10 min at 4◦C.The water phase was re-extracted with chloroform and RNAwas precipitated overnight at 4◦C by adding 1/3 vol. of LiCl8 M. RNA was washed with LiCl 2 M and with ethanol 80%and finally re-suspended in water. C-DNA was synthesisedusing the SuperscriptTM system (GibcoBRL®) according tothe manufacturer’s recommendations and 6�l of cDNA so-lution were used for PCR amplification.

For Northern blot analysis, 15�g of total RNA were sepa-rated on a 2% agarose gel and blotted on a positively chargednylon membrane using the TurboblotterTM system (Schle-icher & Schuell). The filters were hybridised with32P la-belled probe using standard techniques[21].

2.5. Sodiumdodecylsulfate polyacrylamide gelelectrophoresis (SDS-PAGE) and immunoblotting

Fresh plant/cell suspension material was homogenised in100 mM Tris buffer, pH 7, containing the CompleteTM miniprotease inhibitor cocktail (Roche) and proteins were sep-arated by SDS-PAGE[23] in a Mini PROTEAN II system(Bio-Rad) under reducing denaturing conditions. Proteinswere stained with Comassie Brilliant Blue (CBB R-350,Pharmacia). For Western blot analysis, proteins were blot-ted electrophoretically onto a PVDF filter (Millipore) usinga Mini Trans-Blot Cell (Bio-Rad) in CAPS buffer (2.2 g/l,pH 11) and 10% ethanol. Filters were blocked overnightwith 5% skim milk powder in TBST (0.1 M Tris–HCl, pH7.5; 1.5 M NaCl; 0.1% Tween-20) at 4◦C. Subsequently,filters were incubated for 2 h with anti-Pha-C1 antibodies(1:10000.[24]) and with alkaline-phosphatase-conjugatedanti-rabbit secondary antibodies (Sigma) in TBST con-taining 1% skim milk powder. The filters were washedin TBST and in detection buffer (0.1 M Tris–HCl, pH9.5; 0.1 M NaCl) and antibodies–protein complexes weredetected by incubating the filters in detection buffer con-taining 5-bromo-4-chloro-3-indoylphosphate 0.5 mM andnitro-blue-tetrazolium 0.1 mM (Amersham PharmaciaBiotech).

2.6. Plant cell suspension culture

Crumbling calli were induced on MS[25] salts plus vi-tamins (Duchefa), 0.8% agar, supplemented with sucrose2.5%, casein hydrolysed 0.1%, myo-inositol 0.1%, 2-4-D3 mg/l, kinetine 0.2 mg/l, thiamine pyrophosphate 0.5 mg/l,pyridoxine 0.5 mg/l, nicotinic acid 0.5 mg/l, glycine 2 mg/l,folic acid 0.5 mg/l, biotine 0.05 mg/l, pH 5.7 from in vitrogrown microtubers or from calli resistant to kanamycin thatemerged from bombarded internodes. Cell suspension cul-tures were started from crumbling calli on MS salts plus vi-tamins liquid medium supplemented with sucrose 3%, and2–4 D, 2 mg/l and maintained by subculturing 10 ml of sus-pension in 40 ml of the same medium every 10 days. Micro-tubers were induced from nodes of in vitro grown plantletsas described in Romano et al.[16].

2.7. Viability test using the fluorescein diacetate (FDA)

To check the percentage of living cells, the FDA stainingmethod was performed as described by Hoeberichts et al.[26].

2.8. Feeding experiment

Pure 3-(R)-hydroxyoctanoate was produced as describedby de Roo et al.[27]. After solubilisation in buffer phosphate50 mM, pH 7, at a concentration of 100 mM, the substratewas added to the suspension, 5 days after subculturing, toa final concentration of 1 mM. The substrate feeding wasrepeated every 2 days.

2.9. PHA isolation, gas chromatography (GC) andGC–MS (mass spectrometry) analyses

Cell suspensions (fed with the substrate or not fed) werewashed three times with tap water, ground in liquid nitro-gen and freeze dried. The resulting powder was washedwith ethanol at 55◦C for 48 h to remove fatty acids, andthe content of the de-fatted powder was analysed by GC af-ter methanolysis as described by Lageveen[28]. In short,50–100 mg of powder was stirred at 110◦C for 5 h in 2 ml ofmethanol supplemented with sulphuric acid 15% and 2 mlof chloroform. Methanol/chloroform phases were separatedby adding 1 ml of water, and the methylesters in the chloro-form phase were analysed by GC on a Carlo-Erba GC6000(Carlo Erba) equipped with a 25 m CP-Sil 5CB capillary col-umn (Chrompack). GC–MS spectra were obtained using aCarlo-Erba HRGC/MS gas chromatograph equipped with aCP-Sil 5CB column attached by a direct interface to a CarloErba QMD 1000 mass spectrometer.

3. Results

3.1. Transformation and selection of transgenic lines

Plasmids 35-S-Kan and pAPP63, containing thenptIIselectable marker and theP. oleovorans phaC1 gene, re-spectively, were simultaneously introduced in potato cellsby particle bombardment mediated co-transformation asdescribed in Romano et al.[16–18]. In short, plasmids35-S-Kan and pAPP63 were co-precipitated on gold parti-cles at a molar ratio of 1:2 and were delivered using theHelium Inflow Gun device to in vitro grown potato intern-odes. Transgenic plants were regenerated and 30 lines wererooted in the presence on kanamycin. Thirty-eight calligrowing on kanamycin containing medium but which didnot accomplish regeneration, were maintained and propa-gated on callus medium containing kanamycin (100 mg/l).Genomic DNA was extracted from plant or callus lines re-sistant to kanamycin and screened for the integration of thephaC1 gene by Southern blot or PCR analyses. Finally, 12

A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143 139

Fig. 2. Western blot analyses using antibodies raised against the Pha-C1polymerase ofP. oleovorans. The Pha-C1 polymerase band is indicatedby an arrow. Lane+: purified Pha-C1; lane WT: wild type line, lanes 1and 2: two independent transgenic lines.

co-transformed lines (6 callus and 6 plant lines) containingthe phaC1 gene were selected and 10 of these showed, byNorthern blot or RT-PCR analyses, that the transgene wastranscribed. Western blot analysis confirmed that the Pha-C1polymerase was expressed in 5 of these 10 lines (Fig. 2).

3.2. Feeding experiment

Plant cell suspensions cultures were established from thewild type line and seven transgenic lines, four lines whichshowed and three lines which did not show the expression ofthe Pha-C1 polymerase by Western blot analyses (Table 1).Feeding with 3-(R)-hydroxyoctanoate was started after theinitial early lag phase of the growth curve had finished andthe substrate was supplemented to the growing medium ata final concentration of 1 mM every 2 days. Too early addi-tion of the substrate resulted in the death of the cell suspen-sion. Preliminarily, different parameters (biomass, pH, cellviability) were monitored during feeding, to check to whichextent the treatment affected growth and survival of the sus-pensions. Furthermore, depletion of the substrate from thegrowing medium was monitored in the wild type and thetransgenic lines. The substrate was detectable in the grow-ing medium by GC analysis after its addition, and it wascompletely depleted after 2 days. The absence of substratedepletion in a negative control without cells (medium plussubstrate) suggested that it was indeed taken up into the cell,

Table 1List of plant and callus lines selected and used in the feeding experiments

Cell suspensionlines

Derived from Pha-C1Western blot

mcl-PHA(mg)/CDW (g)

Wild type Plant – –PB002.1A1 Callus line – –PB002.1E1 Callus line – –PB002.1F1 Callus line + 2.2 × 10−2

PB002.1L1 Plant + 9.7PB002.2A1 Plant + 2.0PB002.3.4 Callus line + 0.9PB002.3.5 Plant – –

Results of Western blot and GC analyses are indicated. CDW: cell dryweight.

as previously observed in bacteria (de Roo, personal com-munication) and for other forms of fatty acids[29]. No sig-nificant differences were observed between cell suspensionsthat were or were not exposed to the substrate.

3.3. PHA analysis

The PHA content of the cell suspensions was analysed byGC. In the extraction procedure, lyophilised material wasfirst washed with warm ethanol. Ethanol solubilised lipidsand chlorophyll, while PHA, soluble in chlorinated solventand not in ethanol, was retained in the powder. The remain-ing content of the powder (including PHA) was subsequentlymethanolysed and analysed by GC.

In the four transgenic lines expressing the Pha-C1 poly-merase (Table 1), only after 3-(R)-hydroxyoctanoate feed-ing, a peak was observed at the expected retention timecorresponding to the pure substrate used in the experiment,(Fig. 3a), but no peak was evident when the same lineswere not exposed to the substrate. Moreover, in the wildtype and transgenic lines in which the polymerase was notexpressed, no peak, at the same expected retention time,was observed, irrespective of whether the substrate hadbeen supplemented or not (Fig. 3a). The amount of poly-hydroxyoctanoate detected ranged between 0.022–9.7 mgof polyhydroxyoctaonate per gram of dry weight. Con-sidering the total amounts of polyhydroxyoctanoate pro-duced in separate batch cultures and the amount of3-(R)-hydroxyoctanoate fed during the experiment the ratiodepleted-substrate/polymerised-substrate ranged between150 and 1800. GC–MS analysis (Fig. 3b) was subsequentlyperformed. The spectra of the methyl-esters extracted fromthe lines were analysed and were compared with the spec-tra of the substrate. This confirmed the GC data, i.e. thepeak observed in the transgenic lines expressing the Pha-C1corresponded to 3-(R)-hydroxyoctanoate. The ethanol usedfor the washing step was also analysed by GC, but onlytraces of 3-(R)-hydroxyoctanoate were observed. These re-sults suggest that the substrate provided from outside thecell, upon entrance into the cytoplasm, was either degraded(wild type and Pha-C1 negative lines) or in part degradedand in part polymerised to polyhydroxyoctanoate (Pha-C1positive lines).

4. Discussion

This paper describes, for the first time, the successfulexpression of the Pha-C1 polymerase fromP. oleovoransand the accumulation of mcl-PHA in transgenic potato cellsuspension cultures. Without substrate feeding, expressionof the Pha-C1 polymerase in the cytoplasm did not lead tothe accumulation of mcl-PHA in vivo, presumably due tothe lack of the substrate in this cell compartment. Thus, thefeeding approach in cell suspensions (Fig. 4a) was attemptedfor two main reasons: on the one hand, to check whether

140 A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143

Fig. 3. (a) GC analysis of the methyl esters extracted from the wild type (fed or not fed with 3-(R)-hydroxyoctanoate) and two of the transgenic linesproducing mcl-PHA after substrate feeding. The arrows indicate the expected retention time for 3-(R)-hydroxyoctanoate. (b) GC–MS analyses to confirmthe nature of the methyl esters observed by GC. Top: MS spectra of 3-(R)-hydroxyoctanoate. Bottom: MS spectra of the compound observed in thetransgenic lines expressing the Pha-C1 polymerase eluted at the same retention time of 3-(R)-hydroxyoctanoate.

theP. oleovorans phaC1 gene would be expressed in potato,on the other hand to study the minimum gene set requiredfor mcl-PHA accumulation in potato. Low expression ofprokaryotic genes in plants has been reported before[30,31].The cytoplasm as target for the expression has the advantagethat bacterial genes can be expressed with no further protein

modification. The expression of thephaC1 andphaC2 genesin Arabidopsis thalianahas been recently published[13] butthe source of the genes in this case wasP. aeruginosa.

Seven transgenic potato lines, 4 showing expression and3 showing no expression of the Pha-C1 polymerase, and awild type line were selected for further analyses. Mcl-PHA

A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143 141

Fig. 4. (a) Scheme of the model implemented in this work to producemcl-PHA. The Pha-C1 polymerase is expressed in the cytoplasm of trans-genic potato lines and the substrate (3-OH-octanoate) is exogenouslyprovided. See text for explanations. (b) Proposed model to explain theobserved data. Upon entrance into the cell of lines showing no Pha-C1polymerase expression, the substrate is degraded through the�-oxidation(top; wild type and transgenic lines not expressing the Pha-C1 poly-merase). In those lines in which the Pha-C1 polymerase is expressed,part of the substrate is converted into a polymer (bottom). See text forexplanation.

contents were analysed in corresponding cell suspensionscultures, after they were provided with the substrate byfeeding. Controls in which the same cell suspensions werenot supplemented with substrate were included. None of thelines analysed showed mcl-PHA accumulation if the sub-strate was not supplemented. However, only the four linesshowingphaC1 gene expression, but not the wild type nortrangenic lines with nophaC1 expression, showed mcl-PHAaccumulation after 3-(R)-hydroxyoctanoate feeding exper-iment. The kind of analyses performed in this work doesnot demonstrate the presence of high-molecular-weight

polymer (GC is based on monomer analysis after methanol-ysis). However, the correlation between presence of thePha-C1 polymerase (Western blot) and the recovery of3-(R)-hydroxyoctanoate-methyl esters by GC, and the ab-sence of the same compound in the wild type or in thetransgenic lines not expressing the polymerase, was consid-ered proof for mcl-PHA synthesis.

A model is proposed to explain these data (Fig. 4b).When supplemented in the medium, the substrate is takenup by the cell and eventually activated with CoA. Acyl-CoAsynthase activity has been shown to be present in the plantcytoplasm (reviewed by Ohlrogge et al.[32]). The pres-ence of fatty acids which diverge from the normal pool offatty acids present in the plant (i.e. palmitic, stearic, oleic,linoleic acids) activated the�-oxidation pathway[33,34],thus, no 3-(R)-hydroxyoctanoate, although supplemented,could be detected in the transgenic lines not expressingthe phaC1 gene and in the wild type (Fig. 4b, top). Inthe transgenic lines expressing the Pha-C1 polymerase,3-(R)-hydroxyoctanoate was converted to an inert polymeras terminal product and could subsequently be detectedbecause potato has no bacterial enzymatic activity capableof degrading the polymer (Fig. 4b, bottom). Thus, in thetransgenic lines expressing the polymerase,�-oxidation andmcl-PHA biosynthesis may have competed for the utilisa-tion of the substrate. Assuming that all the substrate enteredthe cells and was directed into�-oxidation or mcl-PHA(Fig. 4b, bottom), �-oxidation appeared to be 150–1800times more efficient than the mcl-PHA biosynthesis.

In conclusion, this work demonstrates that thephaC1 genefrom P. oleovoransis expressed, leading to the synthesis ofactive Pha-C1 polymerase and accumulation of mcl-PHA intransgenic potato. The feeding approach in plant cell sus-pension cultures attempted during the course of these ex-periments provided an excellent tool to study heterologousgene expression in vivo.

Accumulation of mcl-PHA in plants has been reportedonly for A. thaliana. Moreover, accumulation of PHAs ingeneral (mcl-PHA and PHB), has been reported predomi-nantly in oil crops[35–38], and only very recently the useof transgenic maize[39] and potato[40] for the productionof PHB has been published. Due to the generally larger pro-ductivity of starch crops, compared to oil crops, the formersare considered more suitable hosts for PHA accumulation[41]. Therefore this work opens new promising prospectsfor mcl-PHA biosynthesis in plants[42].

Acknowledgements

We gratefully acknowledge Pieter van de Meer forGC–MS analyses. We also want to thank Dr. Krit Raemak-ers and Prof. Richard Visser for discussions and criticalreading of the manuscript. This study has been funded by aMarie Curie Fellowship (contract number FAIRCT98-5036)and by an ATO-IAC grant.

142 A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143

References

[1] G. Huisman, O. Leeuw de, G. Eggink, B. Witholt, Synthesisof poly-3-hydroxyalkanoates is a common feature of fluorescentpseudomonads, Appl. Environ. Microbiol. 55 (1989) 1949–1954.

[2] A. Timm, A. Steinbüchel, Formation of polyesters consisting ofmedium-chain-length 3-hydroxyalkanoic acids from gluconate inPseudomonas aeruginosaand other fluorescent presudomonads,Appl. Environ. Microbiol. 56 (1990) 3360–3367.

[3] G. Huisman, E. Wonink, R. Meima, B. Kazemier, P. Terpstra,B. Witholt, Metabolism of poly(3-hydroxyalkanoates) (PHAs) byPseudomonas oleovorans. Identification and sequences of genes andfunction of the encoded proteins in the synthesis and degradation ofPHA, J. Biol. Chem. 266 (1991) 2191–2198.

[4] A. Timm, A. Steinbüchel, Cloning and molecular analysis of thepoly(3-hydroxyalkanoic acid) gene locus ofPseudomonas aeruginosaPAO1, Eur. J. Biochem. 209 (1992) 15–30.

[5] A. Steinbüchel, E. Hustede, M. Liegebergesesell, U. Pieper, A. Timm,H. Valentin, Molecular basis for biosynthesis and accumulation ofpolyhydroxyalkanoic acids in bacteria, FEMS Microbiol. Rev. 103(1992) 217–230.

[6] M. Prieto, M. Kellerhals, G. Bozzato, D. Radnovich, B. Witholt, B.Kessler, Engineering of stable recombinant bacteria for productionof chiral medium-chain-length poly-3-hydroxyalkanoates, Appl.Environ. Microbiol. 65 (1999) 3265–3271.

[7] S. Klinke, Q. Ren, B. Witholt, B. Kessler, Production ofmedium chain length poly(3-hydroxyalkanoates) from gluconate byrecombinantEscherichia coli, Appl. Environ. Microbiol. 65 (1999)540–548.

[8] A. Steinbüchel, S. Hein, Biochemical and molecular basis ofmicrobial synthesis of polyhydroxyalkanoates in microorganisms,Adv. Biochem. Eng. Biotechnol. 71 (2001) 81–123.

[9] B. Kessler, B. Witholt, Factors involved in the regulatory network ofpolyhydroxyalkanoate metabolism, J. Biotechnol. 86 (2001) 97–104.

[10] G.W. Huisman, E. Wonink, G.D. Koning, H. Preusting, B. Witholt,G. De Koning, Synthesis of poly(3-hydroxyalkanoates) by mutantand recombinantPseudomonasstrains, Appl. Microbiol. Biotechnol.38 (1992) 1–5.

[11] S. Langenbach, B.H.A. Rehm, A. Steinbüchel, Functional expressionof the PHA synthase gene phaC1 fromPseudomonas aeruginosain Escherichia coli results in poly(3-hydroxyalkanoate) synthesis,FEMS Microbiol. Lett. 150 (1997) 303–309.

[12] Q. Qi, B. Rehm, A. Steinbüchel, Synthesis of poly(3-hy-droxyalkanoates) inEscherichia coliexpressing the PHA synthasegene phaC2 fromPseudomonas aeruginosa: comparison of PhaC1and PhaC2, FEMS Microbiol. Lett. 157 (1997) 155–162.

[13] V. Mittendorf, E.J. Robertson, R.M. Leech, N. Kruger,A. Steinbuchel, Y. Poirier, Synthesis of medium-chain-lengthpolyhydroxyalkanoates inArabidopsis thalianausing intermediatesof peroxisomal fatty acid�-oxidation, Proc. Natl. Acad. Sci. U.S.A.95 (1998) 13397–13402.

[14] Q. Ren, N. Sierro, M. Kellerhals, B. Kessler, Properties of engineeredpoly-3-hydroxyalkanoates produced in recombinantEscherichia colistrain, Appl. Environ. Microbiol. 66 (2000) 1311–1320.

[15] E. Jacobsen, J. Hovenkamp-Hermelink, H. Krijgsheld, H. Nijmad,L. Pijnacker, B. Witholt, W. Feenstra, Phenotypic and genotypiccharacterization of an amylose-free mutant of the potato, Euphytica44 (1989) 43–48.

[16] A. Romano, K. Raemakers, R. Visser, H. Mooibroek, Transformationof potato (Solanum tuberosum) using particle bombardment, PlantCell Rep. 20 (2001) 198–204.

[17] A. Romano J.-P. Vincken, K. Raemakers, H. Mooibroek, R.Visser, Potato genetic transformation and its application in polymermodification, in: R. Singh, P. Jaiwal (Eds.), Plant GeneticEngineering, Sci-Tech Publishing Company, Houston, USA, 2002,in press.

[18] A. Romano, K. Raemakers, J. Bernardi, R. Visser, H. Mooibroek,Transgenes organisation in potato after particle bombardment-mediated (co-)transformation using plasmids and gene cassettes,Transgenis Res., 2003, in press.

[19] Q. Ren, Biosynthesis of medium-chain-length poly-3-hydroxy-alkanoates: fromPseudomonasto Escherichia coli, 1997, Ph.D.thesis, dissertation. ETH No. 12459, Swiss Federal Institute forTechnology, Zurich, 2002, pp. 47–62.

[20] G. Rouwendal, O. Mendes, E. Wolbert, A. De Boer, Enhancedexpression in tobacco of the gene encoding green fluorescent proteinby modification of its codon usage, Plant Mol. Biol. 33 (1997) 989–999.

[21] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: aLaboratory Manual, Cold Spring Harbor Laboratory Press, ColdSpring Harbor, NY, 1989.

[22] S. Rogers, A. Benich, Extraction of total DNA from plants, algaeand fungi, in: S.A. Gelvin SB, S.B. Gelvin, R.A. Schilperoort,D.P.S. Verma (Eds.), Plant Molecular Biology Manual, 1994, KluwerAcademic Publisher, Dordrecht, The Netherlands, pp. 1–8.

[23] U. Laemmli, Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4, Nature 227 (1970) 680–685.

[24] M. Kraak, T. Smits, B. Kessler, B. Witholt, Polymerase C1 levelsand poly(R-hydroxyalkanoate)synthesis in wild-type and recombinantPseudomonasstrains, J. Bacteriol. 179 (1997) 4985–4991.

[25] T. Murashige, F. Skoog, A revised medium for rapid growth andbioassay with tobacco cultures, Physiol. Plant. 15 (1962) 473–479.

[26] F. Hoeberichts, D. Orzaez, L.H.W. van der Plas, E. Woltering,Changes in gene expression during programme cell death in tomatocell suspensions, Plant Mol. Biol. 45 (2001) 641–654.

[27] G. de Roo, M. Kellerhals, Q. Ren, B. Kessler, B. Witholt,Production of R-3-hydroxy fatty acids and R-3-hydroxy fatty acidsmethyl esters via hydrolytic degradation of mcl-PHA synthesized byPseudomonads, Biotechnol. Bioeng. 77 (2002) 717–722.

[28] R. Lageveen, G. Huisman, H. Preusting, P. Ketelaar, G. Eggink, B.Witholt, Formation of polyesters byPseudomonas oleovorans: effectof substrate on formation and composition of poly-(R)-3-hydroxy-alkanoates and poly-(R)-3-hydroxyalkenoates, Appl. Environ.Microbiol. 54 (1988) 2924–2932.

[29] L. Allenbach, Y. Poirier, Analysis of the alternative pathways forthe �-oxidation of unsaturated fatty acids using transgenic plantssynthesizing polyhydroxyalkanoates in peroxisomes, Plant Physiol.124 (2000) 1159–1168.

[30] E. De Rocher, T. Vargo-Cogola, S. Diehn, P. Green, Direct evidencefor rapid degradation ofBacillus thuringensistoxin mRNA as a causeof poor expression in plants, Plant Physiol. 117 (1998) 1445–1461.

[31] S. Diehn, W.-L. Chiu, E. De Rocher, P. Green, Prematurepolyadenilation at multiple sites within aBacillus thuringensistoxingene-coding region, Plant Physiol. 117 (1998) 1433–1443.

[32] J. Ohlrogge, J. Browse, Lipid biosynthesis, Plant Cell 7 (1995) 957–970.

[33] V. Mittendorf, V. Bongcam, L. Allenbach, G. Coullerez, N. Martini,Y. Poirier, Polyhydroxyalkanoates synthesis in transgenic plants asa new tool to study carbon flow through�-oxidation, Plant J. 20(1999) 45–55.

[34] V.S. Eccleston, A.M. Cranmer, T.A. Voelker, J. Ohlrogge, Medium-chain fatty acid biosynthesis and utilization inBrassica napusplantsexpressing lauroyl-acyl carrier protein thioesterase, Planta 198 (1996)46–53.

[35] C. Nawrath, Y. Poirier, C. Somerville, Targeting of the poly-hydroxybutyrate biosynthetic pathway to the plastids ofArabidopsisthaliana results in high levels of polymer accumulation, Proc. Natl.Acad. Sci. USA 91 (1994) 12760–12764.

[36] C. Nawrath, Y. Poirier, C. Somerville, Plant polymers forbiodegradable plastics: cellulose, starch and polyhydroxyalkanoates,Mol. Breeding 1 (1995) 105–122.

[37] K.L. Houmiel, S. Slater, D. Broyles, L. Casagrande, S. Colburn, K.Gonzalez, T.A. Mitsky, S.E. Reiser, D. Shah, N.B. Taylor, M. Tran,

A. Romano et al. / Biochemical Engineering Journal 16 (2003) 135–143 143

H.E. Valentin, K.J. Gruys, Poly(�-hydroxybutyrate) production inoilseed leukoplasts ofBrassica napus, Planta 209 (1999) 547–550.

[38] S. Slater, T.A. Mitsky, K.L. Houmiel, M. Hao, S.E. Reiser, N.B.Taylor, T. Minhtien, H.E. Valentin, D.J. Rodriguez, D.A. Stone,S.R. Padgette, K. Ganesh, K.J. Gruys, M. Hao, G. Kishore,Metabolic engineering of Arabidopsis and Brassica for poly(3-hydroxybutyrate-co-3-hydroxyvalerate) copolymer production, Nat.Biotechnol. 17 (1999) 1011–1016.

[39] J.J. Hahn, A.C. Eschenlauer, U.B. Sleytr, D.A. Somers, F. Srienc,Peroxisomes as sites for synthesis of polyhydroxyalkanoates intransgenic plants, Biotechnol. Prog. 15 (1999) 1053–1057.

[40] K. Bohmert, I. Balbo, A. Steinbüchel, G. Tischendorf, L. Wilmitzer,Constitutive expression of the�-ketothiolase gene in transgenicplants. A major obstacle for obtaining polyhydroxybutyrate-producing plants, Plant Physiol. 128 (2002) 1282–1290.

[41] F. van der Leij, B. Witholt, Strategies for the sustainable production ofnew biodegradable polyesters in plants: a review, Can. J. Microbiol.41 (1995) 222–238.

[42] A. Romano, Production of polyhydroxyalkanoates (PHAs) intransgenic potato, Ph.D. thesis, Wageningen University, TheNetherlands, ISBN 90-5808-726-3.