Anderson Et Al. (2011) Produção de Bioplásticos

8/11/2019 Anderson Et Al. (2011) Produo de Bioplsticos

1/15

Synthesis of Short-Chain-Length/Medium-Chain Length

Polyhydroxyalkanoate (PHA) Copolymers in Peroxisomes

of Transgenic Sugarcane Plants

David J. Anderson &Annathurai Gnanasambandam &

Edwina Mills &Michael G. OShea &Lars K. Nielsen &

Stevens M. Brumbley

Received: 5 May 2011 /Accepted: 15 June 2011 /Published online: 16 August 2011# Springer Science+Business Media, LLC 2011

Abstract Metabolic engineering of crops is a potential routeto economically viable production of polyhydroxyalkanoates(PHAs), biodegradable and renewable alternatives to conven-tional plastics. In particular, short-chain-length (SCL)/medi-um-chain-length (MCL) PHA copolymers have attractedcommercial interest for their wide range of potential applica-tions. To date, examples of SCL/MCL PHA copolymer

production in plant peroxisomes have involved single trans-gene approaches in transgenic Arabidopsis. We attempted to

produce SCL/MCL PHA copolymers using a multigenestrategy in peroxisomes of the high biomass food andindustrial crop, sugarcane (Saccharum hybrids). Our ap-

proach involved peroxisomal targeting of a 3-ketothiolase,acetoacetyl-CoA reductase, enoyl-CoA hydratase and PHAsynthase, as well as plastid targeting of a acyl-ACPthioesterase and 3-ketoacyl-ACP synthase to increase perox-isomal -oxidation flux. Of 143 transgenic sugarcane linesgenerated by co-bombardment with the six transgenes, sixwere identified with PHA copolymers at up to 0.015% leafdry mass, consisting mainly of saturated C4C16 3-hydroxyalkanoic acids. One line with high acetoacetyl-CoAreductase and low 3-ketothiolase transcript levels hadincreased 3-hydroxybutyrate content, and acyl-ACP thioes-terase and 3-ketoacyl-ACP synthase expression were associ-

Communicated by: Robert Birch

D. J. Anderson : A. Gnanasambandam :E. Mills : M. G. OShea :

S. M. BrumbleyBSES Limited,50 Meiers Road,Indooroopilly, QLD 4068, Australia

D. J. Andersone-mail: [email protected]

A. Gnanasambandame-mail: [email protected]

E. Millse-mail: [email protected]

M. G. OSheae-mail: [email protected]

D. J. Anderson : A. Gnanasambandam :E. Mills : M. G. OShea :

L. K. Nielsen :S. M. BrumbleyCooperative Research Centre for Sugar Industry Innovationthrough Biotechnology, The University of Queensland,Brisbane, QLD 4072, Australia

L. K. Nielsene-mail: [email protected]

L. K. Nielsen :S. M. BrumbleyAustralian Institute for Bioengineering and Nanotechnology,The University of Queensland,Brisbane, QLD 4072, Australia

Present Address:

S. M. Brumbley (*)Department of Biological Science, The University of North Texas,1155 Union Circle, # 305220,Denton, TX 76203-5017, USA

e-mail: [email protected]

Present Address:

D. J. AndersonSchool of Agricultural and Food Sciences,The University of Queensland,Brisbane, QLD 4072, Australia

Present Address:

A. GnanasambandamDepartment of Primary Industries,Horsham, VIC 3400, Australia

Tropical Plant Biol. (2011) 4:170184

DOI 10.1007/s12042-011-9080-7


2/15

ated with altered MCL monomer profiles. SCL/MCL PHAcopolymer from the highest-yielding line showed a weight-average molecular weight of 111 KDa and polydispersityindex of 1.2. Transmission electron microscopy of leafsections from this line indicated the presence of PHA granulesin peroxisomes. This work demonstrates SCL/MCL PHAcopolymer biosynthesis in sugarcane peroxisomes and pro-

vides a basis for further development of mechanisms forcontrolling PHA composition in transgenic crop plants.

Keywords Polyhydroxyalkanoate . copolymer.

Saccharum . Transgenic sugarcane . Peroxisome

Abbreviations

ACP Acyl carrier proteinDM Dry massHPLC High performance liquid chromatographyPDI Polydispersity indexPHA polyhydroxyalkanoate

PHB polyhydroxybutyrateSCL short-chain-lengthMCL medium-chain-lengthGC-MS gas chromatographymass spectrometryGPC gel permeation chromatography

Introduction

Polyhydroxyalkanoates (PHAs) are polyesters of hydrox-yacids synthesised by over 100 genera of bacteria as a

carbon and energy reserve (Steinbchel and Hein 2001;Lenz and Marchessault 2005). They are biodegradable,renewable, and exhibit a wide range of physical properties,making them a potential alternative to petrochemical-derived plastics. (Steinbchel and Valentin 1995; Hazerand Steinbchel2007). PHAs are most commonly composedofR-3-hydroxyalkanoic acids, and are classified by monomerchain length as short-chain-length (SCL, three to five carbons),or medium-chain-length (MCL, six to 16 carbons). SCL PHAssuch as polyhydroxybutyrate (PHB) are rigid and brittle withlimited practical uses, while MCL PHAs have elastomeric

properties but lack the mechanical strength required for many

consumer or biomedical products. SCL/MCL PHA copoly-mers, containing both SCL and MCL monomers, can have arange of properties to suit a wide variety of applications(Satkowski et al. 2001; Philip et al. 2007). In the typified

pathway ofRalstonia eutropha, PHB is synthesised from twomolecules of acetyl-CoA by 3-ketothiolase (PhbA),acetoacetyl-CoA reductase (PhbB) and PHA synthase (PhbC)activities. Bacteria produce MCL PHAs either by diverting R-3-hydroxyacyl-ACP intermediates from fatty acid biosynthe-sis, or production ofR-3-hydroxyacyl-CoAs from fatty acid -

oxidation intermediates (Steinbchel 2001). A number ofnaturally occurring bacterial strains produce SCL/MCL PHAcopolymers, includingAeromonas caviae(Doi et al.1995),A.hydrophila (Lee et al. 2000), and various Pseudomonasspecies (Lee et al. 1995; Kato et al.1996).

Metabolic engineering of bacterial strains to produceSCL/MCL PHA copolymers has focused on optimising

yield and monomer composition for particular applications(Zou and Chen 2007). The family of PHA copolymersdeveloped by Proctor & Gamble and Kaneka under thetrade name Nodax are an example (Noda et al. 2005b).Their high 3-hydroxybutyrate (H4:0) content (typically~90 mol%) provides crystallinity for rigidity and strength,while the smaller MCL proportion decreases meltingtemperature and increases ductility for improved process-ing. Extensive materials testing has demonstrated a range of

potential applications including flexible packaging, agricul-tural films, moulded articles, disposable sanitary products,synthetic paper and medical devices (Satkowski et al.2001;

Noda et al. 2005b). Nonetheless, the higher cost of PHAproduction by bacterial fermentation compared to chemicalsynthesis of conventional plastics is a major barrier forlarge-scale commercial production of PHAs. Engineering oftransgenic crops for PHA production offers a potentialsolution, particularly if implemented as a value-added

product (Poirier and Gruys 2005; Bohlmann 2006; vanBeilen and Poirier 2008). However, there are significantchallenges for commercialisation of transgenic crops

producing PHAs, such as achieving economically viableyields without affecting agronomic performance, regulatoryand market issues associated with transgenic crops, and the

development of more efficient recovery technologies(Bohlmann2006; Philip et al. 2007). For SCL/MCL PHAcopolymers, an additional challenge is to control monomercomposition for the required functionality. Attaining theflexibility and control of composition that is currently

possible with bacterial fermentation in transgenic crops islikely to be difficult and may restrict the number of viableend product plastic types (van Beilen and Poirier2008).

PHB has been produced in a number of plant specieswith yields up to 40% dry mass (DM) when accumulated in

plastids (Bohmert et al. 2000), although yields aboveapproximately 4% DM have typically been associated with

adverse phenotypic effects (Poirier and Gruys2005). MCLPHA was first produced in plants using a peroxisomal-targeted PHA synthase in transgenic Arabidopsis, relyingon supply of R-3-hydroxyacyl-CoAs from -oxidationintermediates via an unknown mechanism, possibly involv-ing an enoyl-CoA hydratase II and/or an R-3-hydroxyacyl-CoA epimerase (Mittendorf et al.1998). A small number ofstudies have demonstrated SCL/MCL PHA copolymer

production in transgenic Arabidopsis with limited controlof monomer composition. In all examples of peroxisomal

Tropical Plant Biol. (2011) 4:170184 171


3/15

SCL/MCL PHA production, monomer composition hasbeen determined soley by the substrate preferences of thePHA synthase, either natural or engineered (Mittendorf etal. 1998; Arai et al . 2002; Matsumoto et al. 2005;Matsumoto et al. 2006). SCL/MCL PHA copolymers withsmall MCL contents below 1 mol% have also been

produced in plastids of transgenic Arabidopsis using an

engineered 3-ketoacyl-ACP synthase III (Matsumoto et al.2009). Other control mechanisms for monomer composi-tion have been demonstrated with MCL PHA production,either providing or supplementing MCL monomers fromfatty acid biosynthesis intermediates in plastids (Romano etal.2005; Wang et al.2005) or-oxidation intermediates in

peroxisomes (Mittendorf et al.1999).Sugarcane is a high biomass crop that currently

accounts for more than 87% of world sugar and 35%of world bioethanol production, and is expected tosupply similar proportions of these commodities as their

production increases over the next decade (OECD/FAO,

2010). PHA might potentially be used as a value-addingco-product in sugarcane either by production in leaves,which are not usually harvested, or by production in theculm, where it would need to be extracted along withsugar. Our group investigated the potential of this cropfor producing PHA by demonstrating and characterisingaccumulation of PHB at levels up to 2.5% leaf DM inleaves of transgenic sugarcane plants using a plastid-

based strategy (Petrasovits et al. 2007; Purnell et al.2007). More recently, we demonstrated peroxisomalaccumulation of PHB in sugarcane leaves at levels up to1.6% DM (Tilbrook et al. 2011). Here we investigate the

feasibility of producing SCL/MCL PHA copolymerin sugarcane. Our transgenic approach consisted of

peroxisome-targeted PHA synthesis enzymes to accessand polymerise substrates from fatty acid -oxidation, and

plastid-targeted enzymes for production of MCL fattyacids to increase flux through -oxidation (Fig.1). For co-

polymerisation of SCL and MCL monomers we selectedPhaC2, a PHA synthase with broad substrate specificityfrom Pseudomonas fluorescens strain GK13 (Liebergesellet al. 2002). When expressed in R. eutropha, PhaC2

produces PHA cop oly mers of var iable com pos itiondepending on fatty acid co-feed and concentration, with

H4:0 contents ranging from 21 to 94 mol% (Noda et al.2005a). To increase H4:0 content, we used R. eutrophaPhbA and PhbB to produce R-3-hydroxybutyryl-CoAmonomers from the peroxisomal acetyl-CoA pool. Finally,we attempted to increase the availability of MCL

R-3-hydroxyacyl-CoA monomers and PHA yield using anR-specific enoyl Co-A hydratase (PhaJ), which in bacteriasupplies monomers for MCL PHA synthesis from fattyacid -oxidation intermediates (Fig. 1). PhaJ2 from

Pse udomonas aer uginos a was selected for its high

stereo-specificity for R-enantiomers and high specificactivities for C6 C12 enoyl-CoAs (Tsuge et al. 2003).To increase carbon flux through fatty acid -oxidation andsupplement MCL content, we used two enzymes thattogether are responsible for the high caprate and lauratecontent in seed storage lipids ofCuphea wrightii(Leonard

et al. 1998). FatB2 (Fig. 1) is a plastidal acyl-ACPthioesterase that increases 16:0 seed oil content and causesunusual accumulation of 10:0, 12:0 and 14:0 whenexpressed in Arabidopsis (Leonard et al. 1998). When asimilar plastidial thioesterase was co-expressed with a

peroxisomal PHA synthase in developing seeds ofArabi-dopsis, the unusual MCL fatty acids generated weredegraded and incorporated as PHA, indicating that plantssense free fatty acids and upregulate -oxidation inresponse (Poirier et al. 1999). KasA1 (Fig. 1) i s a3-ketoacyl-ACP synthase that catalyses extension of 6:0-and 8:0-ACP, thereby enriching steady state levels of 8:0-

and 10:0-ACP in fatty acid biosynthesis. Co-expression ofKasA1 with FatB2 in Arabidopsis shifts the seed oil fattyacid profile towards shorter chain lengths, but alsoincreases production of 12:0 (Leonard et al.1998). In thisstudy we demonstrate synthesis of SCL/MCL PHAcopolymers in transgenic sugarcane, and compare mono-mer contents with expression profiles of the six trans-genes. We also characterise the molecular weightdistribution for a selected SCL/MCL PHA copolymer,and investigate the subcellular site of PHA accumulation.

Fig. 1 Schematic diagram of strategy for SCL/MCL PHA copolymerproduction in peroxisomes. FAB, fatty acid biosynthesis; FAD, fattyacid degradation (-oxidation). Transgenic enzymes are indicated ingrey lettering

172 Tropical Plant Biol. (2011) 4:170184


4/15

Results

Production and Selection of PHA-Accumulating TransgenicSugarcane Plants

ThephbA, phbB,phaC2and phaJ2genes were modified toinclude the C-terminal peroxisomal type 1 targeting

sequence, RAVARL, which efficiently targets heterologousproteins to peroxisomes in tobacco (Volokita 1991) andmaize (Hahn et al. 1999). This sequence has been used totarget R. eutropha PhbA and PhbC to sugarcane perox-isomes, while ARL alone is sufficient for PhbB (Tilbrook etal. 2010; Tilbrook et al. 2011). The FatB2 and KasA1coding sequences contain native putative plastid-transit

peptides (Leonard et al. 1997; Slabaugh et al. 1998). Alltransgenes were placed under the control of the maize Ubi-1promoter (Christensen et al.1992) and nopaline synthaseterminator in direct gene transfer vectors. To facilitate therecovery of transgenic lines expressing phaJ2, th e 3-

aminoglycoside transferase II (nptII) selectable marker andphaJ2 cassettes were combined in tandem on a singlevector, while expression cassettes for all other transgeneswere contained in separate vectors. Sugarcane embryogeniccallus was co-bombarded with a total of six vectors. Leaf

blade samples from glasshouse-grown sugarcane plants for143 independent transgenic lines were screened by gaschromatographymass spectrometry (GC-MS). Ethanoltrans-esterification was used for the screening process dueto problems with unwanted derivatives and loss of the morevolatile esters produced by methanol trans-esterificationduring initial method establishment. GC-MS analysis of the

ethanol-transesterified chloroform extracts revealed sixlines with multiple ethyl 3-hydroxyester peaks (~4% oftotal lines).

Synthesis of SCL/MCL PHA Copolymers

To enable conclusive identification and quantitative analysisof monomer content using a full range of known standards,the six lines were re-analysed using methanol trans-esterification. GC-MS analysis of derivatised extractsrevealed a number of peaks containing the characteristicm/z 103 ion corresponding to the methyl 3-hydroxypropionic

acid fragment, which is common to all methyl estersof saturated 3-hydroxyalkanoic acids and unsaturated3-hydroxyalkanoic acids with double bonds beyond thethird carbon. The six lines produced low yields of PHAcopolymers that consisted primarily of saturated 3-hydroxyacid monomers with even-numbered carbonchains ranging from C4C16 (as shown for line J41 inFig.2a). Small amounts of H5:0 (Fig.2a) and traces of 3-hydroxyoctadecanoic acid (H18:0) were also present insome samples (Fig.2b).

No methyl 3-hydroxyesters were detected in samples fromleaves of wild type or UKN transformed control plants(transformed with the vector pUKN, containing only theselectable marker cassette, kindly provided by Dr Priya Joyce),with the exception of methyl 3-hydroxybutyrate, which wasdetected at very low levels (equivalent to a mean value of1.4g/g DM PHB; Table 1). Ethanolysis of wild type and

UKN transformed control leaf samples produced similarlevels of the corresponding ethyl ester (data not shown),indicating that the 3-hydroxyesters originated from the trans-esterification reaction. Only line J41 had a H4:0 yield that wasclearly higher than wild type and UKN transformed controllevels (Table 1). Hence, low level H4:0 background is

probably present in all lines, and may comprise a substantialproportion of the H4:0 content in lines J2, J40, J72 and J142.

The composition of PHA copolymers obtained wasconsistent among replicate samples and broadly similaracross the six lines, with H8:0 comprising the largest molar

proportion in all cases. Conversely, H5:0 and H16:0

comprised the smallest and second smallest molar propor-tions in most lines. Mean total PHA yields ranged from17 g/g DM (J40) to 87 g/g DM (J41), while themaximum total PHA yield for an individual sample was155 g/g DM (0.0155%) for line J41 (Table 1). Since themean recovery of PHB standard applied to the groundsample matrix through the entire extraction and analysis

process was 16.5% (Table1), the measured yields are likelyto be ~six-fold underestimates of actual PHA contents,assuming similar recoveries for each monomer species.

The PHA copolymer compositions contrast with the MCLPHA copolymers obtained by Mittendorf et al. (1998) from

Arabidopsis seedlings expressing P. aeruginosa PhaC1,which contained substantial proportions of unsaturated 3-hydroxyalkanoic acids, as well as trace amounts of somesaturated 3-hydroxyalkanoic acids with uneven numbers ofcarbons. However, close inspection of the GC-MS chromato-grams revealed a number of minor peaks for the m/z 103 ionthat were not present in wild type or UKN transformedcontrols (Fig.2b). Due to the small size of the peaks relativeto background, identification of the mass spectra with totalion chromatograms was not possible. Based on the m/z valueand comparison with the elution order under similarconditions presented by Mittendorf et al. (1998), we

putatively assigned these as methyl esters of unsaturated 3-hydroxyalkanoic acids or saturated 3-hydroxyalkanoic acidswith uneven numbers of carbons (Fig.2b).

Comparison of Transgene Expression and PHAComposition

Expression levels of the six transgenes for the peroxisomestrategy were determined in cDNA populations of all PHA-

producing lines except J171, which was lost under



5/15

glasshouse conditions. We selected competitive PCR andMassARRAY technology for transgene expression anal-ysis for high sensitivity and amenability for multiplexing(Ding and Cantor2003; Oeth et al.2004). This methodol-ogy quantifies endogenous transcript using a syntheticoligonucleotide differing from the target amplicon at one

base position as an internal control (the competitor).Competitor and target PCR products are subjected to a

primer extension reaction and detected by matrix-assistedlaser desorption ionization time-of-flight mass spectrometry

(MALDI TOF MS). Since known quantities of thecompetitor are used, absolute quantities of a transcript can

be determined within a sample without reference to anexternal standard. From expression data for actin, 18S RNAand glyceraldehyde 3-phosphate dehydrogenase (GAPDH)genes, the latter two were selected and used for normal-isation with the geNORM algorithm (Vandesompele et al.2002) to obtain final transcript levels (Table 2). Notranscripts for any of the six transgenes were detected inwild type or UKN transformed control lines. As expected,

Fig. 2 Comparison of GCMSchromatograms for methanol-transesterified chloroformextracts from a UKN trans-formed control line (inverted)and line J41. a Major saturatedmethyl 3-hydroxyester peaks.Chromatograms shown werecollected in selective ion

monitoring mode and consistof ions with m/z ratios of 71, 74and 103, except for the internalstandard region from 11.56-11.8 min, which shows ions withm/z ratios of 77, 105, and 136.bMinor putative unsaturatedmethyl 3-hydroxyester saturatedmethyl 3-hydroxyesters withodd numbers of carbons. i-std,internal standard (methyl

benzoate). Chromatogramsshow only the m/z ratio 103 ion.Methyl esters are indicated bytheir corresponding 3-

hydroxyalkanoic acid label



6/15

phaC2 transcripts were present in all five PHA-producinglines analysed, all at similar levels ranging from 3.0 to 6.7fM. Despite inclusion of the nptIIselectable marker on thesame vector, phaJ2 expression was only detected in linesJ72 and J142 at very low levels. Transcripts forphbA and

phbB were absent in lines J2 and J40, respectively, but

present in all others. However, all phbA levels were at orbelow 2.7 fM, whereas phbB transcripts were expressed atlevels up to 13.4 fM. Line J41, the only line with a H4:0yield clearly above wild-type levels, had both the lowestlevel of phbA transcripts and the highest level of phbBtranscripts of any line that co-expressed phbA.

Four lines showed FatB2 transcript expression, with thehighest level in J142, while all five lines tested showedvarying levels of KasA1 transcript expression (Table 2).When expressed in transgenic Arabidopsis, FatB2 elevateswild type 16:0 and generates novel 14:0, 12:0 and 10:0(Leonard et al.1998). Superimposed on the wild type profile

of beta-oxidation intermediates, FatB2 could be expected toincrease H10:0-H16:0 proportions. Line J2 functions as aFatB2-negative control, with its KasA1 expression expectedto have little or no effect on the fatty acid profile in theabsence of FatB2 (Leonard et al. 1998). The closestapproximation to a FatB2-positive, KasA1-negative lineobtained was J72, which had the second highest FatB2andlowestKasA1transcript levels. Molar proportions of H10:0,H12:0, H14:0 and H16:0 were all greater in J72 compared to

J2 (Fig.3a). As expected, the FatB2-expressing lines J40 andJ142 also had larger combined H10:0-H16:0 proportionsthan J2 (Fig.3a). However, line J41 did not, possibly due todilution by large H4:0 and H8:0 proportions. Given theresults of Leonard et al. (1998), co-expression ofFatB2and

KasA1would be expected to increase H10:0 and H12:0 at

the expense of H14:0 and H16:0 when compared asproportions of total H10:0-H16:0 content. Accordingly, alllines co-expressingFatB2and KasA1had molar proportionsshifted towards H10:0 and H12:0 relative to J2 (Fig. 3b),which lacked FatB2expression.

J41 SCL/MCL PHA Copolymer has a Moderate MolecularWeight and Low Polydispersity

The molecular weight distributions of copolymers producedby lines J2, J40, J41, J72 and J142 were investigated by gelpermeation chromatography (GPC) using scaled-up PHA

extractions. Chloroform extracts of line J41 showed a singlepeak that was not present in wild type or UKN transformedcontrol samples (data not shown). Although a chloroformextract of line J2 also contained a single peak at the sameretention time, it was not sufficiently large for analysis. No

peaks were observed for lines J40, J72 or J142, probablydue to their very low PHA yields. Based on chloroformsolubility and absence from wild type and UKN trans-formed control extracts, we concluded that the peaksidentified for lines J2 and J41 represented PHA. The PHAcopolymer in line J41 showed a weight-average molecularweight (Mw) of 111 KDa and a polydispersity index (PDI)

of 1.2 (Table3). Standards of bacterial origin for PHB andpoly[3-hydroxybutyrate-co-3-hydroxyhexanoate] (PHB-PHHx) had larger molecular weights of 2.37105 and6.71105, respectively, as well as larger PDI values of 3.0and 3.4, respectively. To assess the effect of the extraction

process on molecular weight distribution, 5 g of the PHBstandard was extracted following application onto theground wild type sugarcane leaf matrix and analysed. Theextraction process resulted in a 20.7% reduction in Mwcompared to the non-extracted value (Table3). It is known

Table 1 Yields of chloroform-extractable PHA from wild type, spiked wild type and mature transgenic leaf samples

Line (no. replicate extractions) WT (2), UKN (1) WT+PHB (5) J2 (3) J40 (4) J41 (5) J72 (5) J142 (5) J171 (4)

Total PHA (g/g DM) Max. 2.1 7.2 33.8 18.5 154.8 30.0 121.0 113.3

Mean 1.4 6.7 27.0 16.7 86.7 22.5 67.5 80.4

SE 0.4 0.2 3.5 0.6 17.1 2.3 20.9 18.2

H4:0 (g/g DM) Mean 1.4 6.7 3.2 1.6 15.7 1.6 1.9 3.2

SE 0.4 0.2 0.6 0.2 1.6 0.1 0.1 0.3

DMdry mass; WTwild type;UKNUKN transformed control; WT+PHBwild type spiked with 5 g PHB. The number of independent replicateextractions is indicated in brackets next to the line label. Samples consisted of pooled blade tissue from the six oldest, non-senescent leaves fromone plant, and thus extractions from this material are analytical replicates

Table 2 Absolute transcript expression levels for five PHA-producing

lines determined using competitive PCR and MassARRAY

technologyLine phbA phbB phaC2 phaJ2 FatB2 KasA1

J2 13.4 3.9 0.8

J40 1.5 4.2 6.3 1.4

J41 0.4 11.7 3.0 5.6 0.6

J72 2.7 3.2 4.0 0.1 7.3 0.5

J142 1.1 2.2 6.7 0.1 28.5 5.1

The fM values shown represent the mean of two replicate analyses ofthe same cDNA sample. R2 values for all standard curves were >0.9



7/15

that the molecular weight distribution of PHB extractedfrom bacteria can be influenced by the extraction processused (Holmes 1988). Hence the data for line J41 are

probably an underestimate of the actual Mw.Compared to PHA copolymer from line J41, PHB extracted

from leaves of the transgenic sugarcane line TA4 (Petrasovits etal.2007) and analysed in this study showed a ~7-fold highermolecular weight of 8.45105 and a ~4-fold higher PDI of 5.0(Table 5). Poirier et al. (1995a) reported a similarly highmolecular weight but larger PDI value for PHB produced by

transgenic Arabidopsis suspension cells (Table 3). In bothcases, PHB biosynthesis involved R. eutropha PhbC. Using

P. aeruginosa PhaC1, Mittendorf et al. (1998) reported alower Mwand more polydisperse distribution for MCL PHAcopolymers produced in transgenicArabidopsisthan for lineJ41 PHA copolymer, while Nakashita et al. (1999) reported alower Mw but similar polydispersity for PHB produced in

transgenic tobacco using A. caviae PhaC (Table3).

Accumulation of PHA Copolymer in Peroxisomes

To investigate the subcellular localisation of the PHA copoly-mers, we examined leaf sections of the PHA-producing lines bytransmission electron microscopy. Characteristically, peroxi-somes are (1) bounded by a single membrane, (2) spheroid inshape, ranging from 0.21.7m in diameter, and (3) contain acoarsely granular or fibrillar matrix, occasionally with amor-

phous or paracrystalline inclusions (Huang et al.1983; Fig.4a).While most peroxisomes from line J41 appeared normal, a

small proportion from various leaf cell types containedelectron-lucent, granular inclusions approximately 0.10.2 m in size (Fig. 4bd) surrounded by a thin, electron-dense layer. These features are typical of the appearance ofPHA granules produced in bacteria as well as various plantorganelles (Poirier and Gruys2005), which reflects a structureconsisting of a PHA core surrounded by a phospholipidmembrane embedded with PHA synthase and other proteinmolecules (Rehm2003). The granules observed here are alsohighly similar to PHB granules produced in sugarcane leaf

peroxisomes (Tilbrook et al.2011). No inclusions were foundin UKN transformed control (Fig. 4a) or wild type leaf

peroxisomes (data not shown). Based on the cellular location,absence from wild type and UKN transformed control samples,and close resemblance to typical bacterial and plant PHAgranules, we conclude that the granules consist of PHAcopolymer. No inclusions that were clearly identifiable as PHAgranules were found in leaf cell peroxisomes of lines J2, J40,J72 or J142 (data not shown). It may be that PHA inclusionswere more rare or difficult to identify in these lines due theirlower PHA yields and/or lower SCL monomer contents.

Discussion

Evaluation of Mechanisms for Regulating PHA CopolymerComposition

We aimed to produce PHA copolymers with regulated mono-mer content in sugarcane peroxisomes using three mecha-nisms. The first mechanism drew on peroxisomal acetyl Co-Ato increase H4:0 content using 3-ketothiolase and acetoacetyl-CoA reductase activities. A useful comparison to the results ofthis study is provided by the recent demonstration by our

Fig. 3 Compositions of chloroform-extractable PHA copolymersfrom transgenic lines as determined by GC-MS. a Total monomer

proportions. b H10:0 H16:0 monomer proportions only. Meanmol% values are shown within bars for each monomer series, and thenumber of independent sample extractions is indicated in bracketsnext to the line label. Samples consisted of pooled blade tissue of thesix oldest, non-senescent leaves from one plant, and thus extractionsfrom this material are analytical replicates. Error bars show SE



8/15

Table 3 Comparison of molecular weight distributions of PHAs from lines J41 and TA4 to standards, and PHAs previously produced intransgenic plants

Sample Mw103 Mn10

3 Mw/Mn PHA source PHA synthase Reference

Line J41 111.4 94.0 1.2 Sugarcane leaves(peroxisomes)

P. fluorescensPhaC2 This study

Line TA4(PHB producer)

845.1 168.0 5.0 Sugarcane leaves(plastids)

R. eutrophaPhbC (Petrasovits et al.2007);this study

PHB 237.3 79.2 3.0 Standard fromR. eutropha

R. eutrophaPhbC This study

Wild type sugarcaneleaves+PHB

188.2 81.5 2.3 Standard fromR. eutropha, appliedonto ground leaf matrix

R. eutrophaPhbC This study

PHB-PHHx 671.5 196.2 3.4 Standard from A. hydrophila A. hydrophilaPhaC This study

PHB 615.0 58.7 10.5 Arabidopsissuspensioncells cytoplasm

R. eutrophaPhbC (Poirier et al.1995b)

MCL PHAcopolymers

23.7 5.5 4.3 Arabidopsisseedlings(peroxisomes)

P. aeruginosa PhaC1 (Mittendorf et al.1998)

PHB 57.6 32.0 1.8 Tobacco leaves(cytoplasm)

A. caviaePhaC (Nakashita et al.1999)

PHB-PHHx poly[3-hydroxybutyrate-co-3-hydroxyhexanoate]

Fig. 4 Transmission electronmicrographs of leaf sectionsfrom a UKN transformedcontrol line and PHA-producingline J41.a Peroxisomes within aUKN transformed control line

bundle sheath cell (image kindlyprovided by Ms. KimberleyTilbrook); (b-d) PHA inclusionswithin peroxisomes of line

J41 mestome sheath (b), phloemcompanion (c) and bundlesheath (d) cells. Labels: c,chloroplast; cw, cell wall; i,inclusion; m, mitochondrion; n,nucleus; p, peroxisome; v,vacuole; ve, vesicle. Scale bars:(a) 500 nm; (b-d) 250 nm



9/15

group of PHB production in sugarcane peroxisomes. Tilbrooket al. (2011) observed PHB at levels up to 1.6% DM inleaves of transgenic sugarcane plants containing the same

peroxisome-targeted R. eutropha PhbA and PhbB enzymesin combination with R. eutropha PhbC. This indicates thatPhaC2 was a key limiting factor for yield in this study, whichcould be due to either incomplete silencing of the phaC2

transgene, low PhaC2 activity in the sugarcane peroxisomecontext, or both. While the modest (~1.5-fold) increase inH4:0 content in line J41 compared to the other lines (Fig.3a,Table 4) holds some promise that PhbA and PhbB may beuseful for boosting the H4:0 content of PHA copolymers

produced in peroxisomes, analysis of a larger number oflines would be required to demonstrate an effect. Evaluationof the effect of PhaA and PhaB would be facilitated byhigher PHA copolymer yields, which might be achieved viaoptimised PHA synthase expression and/or the use ofalternate PHA synthases with broad substrate specificities.

The second mechanism used FatB2 and KasA1 to supplyunusual MCL fatty acids from plastid fatty acid biosynthesis.The PHA composition and transgene expression data providedsome evidence that FatB2 may have increased H10:0-H16:0contents, and in combination with KasA1, shifted thedistribution within H10:0-H16:0 towards H12:0 and H10:0(Fig.3, Table 2). However, no conclusions about the effect

can be made with the small number of transgenic linesobtained. Since carbon flux through fatty acid biosynthesis insugarcane leaves is expected to be low compared to tissuessuch as developing Arabidopsis seeds, any effects of FatB2and KasA1 are likely to be subtle and may require a largenumber of lines to establish.

Finally, the third mechanism attempted to enhance diversionof MCL -oxidation intermediates with an R-specific enoyl-CoA hydratase. No conclusions could be made about thismechanism due to the lack of any substantial phaJ2expression(Table2). This was unexpected since phaJ2was contained on

Table 4 Sources and primers used for amplification of transgenes

Vector Gene(s) Source

organism

Genbank

accession

Forward primer

(5'-3')

Reverse primer

(5'-3')

Amplicon

size (base

pairs)

pATS phbA R. eutropha J04987 TGAGGATCCAT

GACTGACGTTG

TCATCGTATCCG

ACTGAGCTCTTAT

AATCTGGCAACA

GCACGTTTGCGCT

CGACTGCCAGCG

1218

pBTS phbB R. eutropha J04987 TGAGGATCCAT

GACTCAGCGCA

TTGCGTATG

ACTGAGCTCTTAT

AATCTGGCAACA

GCACGGCCCATAT

GCAGGCCGCCG

777

pC2TS phaC2 P. fluorescens AX105569 CAGTGATCAAT

GCGAGAGAAAC

AGGTGTCG

ACTGAGCTCTTAT

AATCTGGCAACA

GCACGGCGCACGT

GCACGTAGGTGC

1719

pFB2 FatB2 C. wrightii U56104 AAAGGATCCAA

ACATGGTGGTG

GCTGC

TCGGAGCTCTTTCA

TGTTGATATCGCC

1251

pKA1 KasA1 C. wrightii U67316 GGCAGATCTTT

GGTGTTTCAAT

GGCGG

TGGGAGCTCGGCA

TTAAGCTACTAAC

G

1689

pJ2K phaJ2 P. aeruginosa AB040026 CGCGGATCCAT

GGCGCTCGATC

CTGAG

ACTGAGCTCTTAT

AATCTGGCAACA

GCACGGTCCGGCC

GCTCTGGCGG

908

pJ2K nptII E. coli E00425 CCGAAGCTTGA

ATACGAATTCC

CGATC

CCGAAGCTTGAAT

ACGAATTCCCGAT

C

3312

Restrict ion enzyme sites:BamHI,BclI, BglII, SacI,Hind II I. Bases encoding the peroxisome targeting signal RAVARL are indicated in bold

type.



10/15

the same vector as the selectable marker, and all five othergenes used for the strategy were expressed at 10-fold or greaterlevels in at least some lines. One possible explanation is that thelack of phaJ2 expression was caused by efficient transgenesilencing, which is a known problem in sugarcane (Mudge etal. 2009; Birch et al. 2010). Another possibility is thatexpression of PhaJ2 is harmful, and that only lines with low

or no expression were recovered through the somatic embryo-genesis and selection process. Some Arabidopsis doublemutants for genes encoding core -oxidation enzymes haveknown embryo-lethal phenotypes (Goepfert and Poirier2007).

Consistent with the small proportion of natural lipids withodd-numbered carbon chains, only trace amounts of putativeH7:0, H9:0, H11:0 and H13:0 were detected (Fig. 2b).However, higher levels of H5:0 were present, as noted

previously for PHA copolymers produced in plant perox-isomes (Arai et al.2002; Matsumoto et al.2006; Tilbrook etal.2011). It has been suggested that this phenomenon may

be due to an unknown metabolic flux (Arai et al. 2002),

possibly via condensation of propanoyl-CoA and acetyl-CoA to yield 3-ketopentanoyl-CoA, and subsequent reduc-tion to 3-hydroxypentanoyl-CoA (Matsumoto et al.2006). Itis also interesting that H8:0 was the most abundant monomer

by molar proportion in all lines (Fig. 3a). A similarabundance of H8:0 and 3-hydroxyoctenoic acid (H8:1) wasobserved by Mittendorf et al. (1998) in MCL PHAcopolymers produced by Arabidopsis seedlings expressing

peroxisomal P. aeruginosaPhaC1 PHA synthase. H8:0 wasalso the dominant saturated monomer in PHA accumulatedin Arabidopsis seeds using the same peroxisome-targetedPHA synthase (Poirier et al. 1999). Mittendorf et al. (1998)

concluded that their results were consistent with trienoic anddienoic fatty acids with cis double bonds at an even carbonundergoing -oxidation via an epimerase pathway, whichinvolves direct production ofR-3-hydroxyoctenoyl-CoA and

R-3-hydroxyoctanoyl-CoA intermediates, respectively.In contrast, we did not detect any H8:1 and putatively

identified only trace peaks for other unsaturated PHA mono-mers (Fig. 2b), probably reflecting the differences in fatty acidmetabolism between sugarcane leaves and Arabidopsisseed-lings. The lack of H8:1 and scarcity of other unsaturatedmonomers rules out-oxidation via an epimerase pathway asan explanation for the high molar proportion of H8:0 in our

case. It is also unlikely that the substrate specificities ofPhaC2 are responsible, since the enzyme is capable of

producing PHAs with a broad range of MCL contents (Nodaet al.2005b). An alternative explanation is that eight-carbon-oxidation intermediates are present at higher steady statelevels compared to intermediates of other chain lengths. Thismight be caused by differing chain-length specificities ofenzymes catalysing any of the four core -oxidationreactions. For example, plants contain a family of acyl-CoA oxidases with partially overlapping substrate specific-

ities that catalyse the first step of the -oxidation cycle(Arent et al.2008). Interestingly, the three acyl CoA oxidaseswith known SCL or MCL activities all have comparativelylow specific activities for octanoyl-CoA, which forms theoverlap point for their specific activity profiles (Froman et al.2000).

Physical Properties of SCL/MCL PHA CopolymerProduced

The substantial MCL monomer content of the PHAcopolymer produced by line J41 (~65 mol%; Fig. 3a)would be expected to exhibit very elastomeric properties,and considerably higher H4:0 content would probably berequired for commercial applications. This might be achievedin two ways. The first is to increase the availability ofR-3-hydroxybutyryl-CoA, as we have attempted in this studyusing phbAandphbB. The second is to use an alternate PHAsynthase with substrate specificity characteristics that will

produce PHA with the required H4:0 content from theavailable R-3-hydroxyacyl-CoAs in the peroxisome. Forexample, Pseudomonas sp. 613 PhaC1 expressed in

Arabidopsis peroxisomes produced PHA copolymers withan average of 40 mol% H4:0 (Matsumoto et al. 2006).

PHA copolymer from line J41 showed a low molecularweight and polydispersity compared to bacterial PHB andPHB-PHHx standards, but similar to some previousexamples of PHAs produced in plants. Low polydispersityvalues such as that determined for J41 PHA copolymer arenot uncommon for biological polymers. For example, P.sp.613 PhaC1 expressed in E. coliproduced SCL/MCL PHA

copolymer with a PDI of 1.5 (Takase, 2004). Highermolecular weights of 500,000-700,000 are typically re-quired for commercial applications (Noda et al. 2005b).This might be achieved using PHA synthases engineered for

production of higher molecular weight PHA, of which therehave been several examples (Nomura and Taguchi2007).

Opportunities for Increasing PHA Yield in PlantPeroxisomes

The maximum PHA yield of 0.015% DM obtained is lowerbut comparable to several previous examples of PHA

copolymer production in Arabidopsis peroxisomes in vege-tative leaves using other PHA synthases. It is approximately3-fold less than forA. caviaePhaC (Arai et al.2002); three-quarters of the yield obtained with P. aeruginosa PhaC1(Mittendorf, 1998); and two thirds that forP. sp. 61-3PhaC1(Matsumoto et al.2006). A higher level of 0.4% DM MCLPHA was achieved by expression ofP. aeruginosaPhaC1 in7-day-old germinating seedlings, in which -oxidation isstrongly induced for mobilisation of seed lipid reserves(Mittendorf et al. 1999). Nonetheless, PHB yields of 1.6%



11/15

DM in sugarcane leaf peroxisomes (Tilbrook et al.2011) andapproximately 2.0% DM in maize suspension cells (Hahn etal. 1999) indicate that peroxisomes have the potential forgreater PHA production. As already noted, the PHA synthaseused may be a key determinant of PHA yield, and mayexplain the large yield differences between studies pro-ducing PHA copolymers with MCL content and those

producing PHB. It has been suggested that the high PHAyields that occur in bacteria from -oxidation may be enabled

by a 3-hydroxyacyl-CoA epimerase or other activity that is notpresent in plants, and that metabolic channeling of intermedi-ates in plant peroxisomes may also limit yields (van Beilenand Poirier2008). If this is the case, R-specific enoyl-CoAhydratases may hold potential for yield improvement ifexpression can be achieved. PHA yield might also beimproved by increasing the activities of the PHA synthaseand other enzymes, either through optimisation of transgeneexpression or the use of engineered enzymes.

Methods

Gene Constructs

All transformation vectors were based on the pU3Z-mcs-nosvector (McQualter et al. 2005) which is a modification of

pAHC20 (Christensen and Quail 1996) that contains themaize ubi-1 promoter and nos terminator. Transgenes wereamplified using the primers listed in Table4 from plasmidscontaining the genes. The plasmids were kindly provided by:Prof. Yves Poirier, University of Lausanne, Switzerland

(phbAand phbB); Dr. Phil Green, The Proctor and GambleCompany, Cincinnati, Ohio, USA (phaC2); Dr. Priya Joyce,BSES Limited (nptII); Dr. Mary Slabaugh, Oregon StateUniversity, Corvallis, Oregon, USA (FatB2 and KasA1);Prof. Y. Doi, RIKEN Institute, Saitama, Japan (phaJ2). Theamplification products were digested with BamHI/SacI,except for the phaC2 and KasA1 products, which weredigested with BclI/SacI and BglII/SacI, respectively. Thedigested transgene amplification products were cloned into

pU3Z-mcs-nos digested with BamHI/SacI. To constructpJ2K, phaJ2 was inserted into pU3Z-mcs-nos to create anintermediate vector, pJ2. pJ2 was linearized with HindIII and

ligated to the amplifiedUbi-1promoter::nptII::nosterminatorcassette with compatible ends produced by digestion with

HindIII. A clone with the phaJ2and nptIIcassettes orientedin the same direction to each other and the ampR gene wasselected as the transformation vector pJ2K.

Sugarcane Transformation

Embryogenic callus cultures of commercial sugarcanecultivar Q117 were initiated and maintained as described

by Bower et al. (1996). Essentially, 4 days followingsubculture, nodular embryogenic callus pieces of 3 to 5 mmdiameter were arranged to cover a circle of approximately3 cm diameter on MSC3medium supplemented with 0.2 Mmannitol and 0.2 M sorbitol as an osmotic treatment for 4 h

prior to bombardment and 4 h after bombardment. Calliwere bombarded with 1 m DNA-coated gold micro-

projectiles (Bio-Rad Laboratories, Hercules, CA, USA)using the Bio-Rad PDS-1000 system (Bio-Rad Laborato-ries) at 1200 psi. Microprojectile preparation and bombard-ment were carried out according to the manufacturersinstructions. Following bombardment, embryogenic calliwere cultured on MSC3 medium in darkness withoutselection for 3 days. They were then transferred to MSC3medium containing 50 mg/L G418 (Geneticin, LifeTechnologies Corporation, Carlsbad, California) in darknessand subcultured every 2 weeks to provide escape-freeselection. After 810 weeks, actively growing calli were

placed on MSC0 medium (MSC3 medium without 2,4-D)

containing 50 mg/L G418. Regeneration of plants from callusoccurred 812 weeks after transfer to MSC0. Only one shootwas recovered from each antibiotic-resistant callus clump toensure that each transgenic line was derived from anindependent transformation event. Regenerating plants weremaintained at 28C under fluorescent lights until ready forestablishment in pots in a containment glasshouse.

PHA Extractions and GC-MS Analysis

Samples for PHA extractions were taken from pooled leafblade tissue of the six oldest, non-senescing leaves of

primary stalks of 1314 month-old sugarcane plants. Theexcised tissue was freeze dried for at least 18 h beforestorage at20C until use. The PHA extraction method forinitial screening of transgenic lines was adapted from Araiet al. (2002). Approximately 100 mg freeze dried leaf bladetissue was pulverised for 20 min at 30 Hz in a RetschMM300 ball mill (Retsch GmbH, Haan, Germany). Ground

powder was transferred to glass centrifuge tubes (Corning#8142-10 with #9998 phenolic/PTFE seal cap, Corning,

NY, USA, supplemented with custom-made 1 mm thickPTFE seal) and weighed. To remove lipids and othercontaminants, the powder was extracted with 8 mL

n-hexanes at 55C, centrifuged at 3000g, and the superna-tant discarded. This extraction was repeated six times over24 h, followed by an identical extraction protocol withmethanol, then evaporated to dryness. PHA was extractedfrom the dried powder with 4 mL chloroform overnight at55C. All extractions were performed at 55C with constantmixing in a Hybaid rotary hybridisation oven (Thermo-Electron Corp. Waltham, MA, USA). The PHA-containingchloroform was extracted twice with 4 mL water to removesolids, then evaporated to a volume of 0.5 mL. The



12/15

chloroform extract was subjected to ethanolysis by adding1.7 mL ethanol, 0.2 mL concentrated HCl, and incubating at100C for 4 h. Following ethanolysis, 2 g of methyl 3-hydroxypentanoate was added as an extraction standard forthe remaining steps. The chloroform phase was recovered byextraction with 7 mL 0.9 M NaCl and neutralised byextraction with 2 mL saturated Na2CO3. Methyl 3-

hydroxybutyrate was added as an internal standard and thepurified chloroform phase analysed on an Agilent 6890 gaschromatograph (Agilent HP-5MS 30 m column, 250 minternal diameter, 0.25 m film) coupled to an Agilent 5973mass spectrometer (Agilent Technologies, Santa Clara, CA,USA). Methyl 3-hydroxybutyrate and the target standardsethyl-3-hydroxybutyrate and ethyl-3-hexanoate were pur-

Table 5 Primer and competitor sequences used for transgene expression analysis.

Gene Amplicon size(base pairs)

Forward PCR primer Reverse PCR primer Extension primer Target sequence [ALLELE/competitor]

18S RNA 90 ACGTTGGATGTCCGCATAGCTAGTTAGCAG

ACGTTGGATGTTGGTGGAGCGATTTGTCTG

TTTGTCTGGTTAATTCCGTTAA

ATGGGTGCATCTTTGCTTGGGGCAGAGATAACAACCTTCTTG[C/a]CACCACCCTTCAGATGCGCAGCAGCCTTGTCCTTGTCAGTGAA

GAPDH 106 ACGTTGGATGATGGGTGCATCTTTGCTTGG

ACGTTGGATGTTCACTGACAAGGACAAGGC

GCATCTGAAGGGTGGTGC

ATGGGTGCATCTTTGCTTGGGGCAGAGATAACAACC

TTCTTG[C/a]CACCACCCTTCAGATGCGCAGCAGCCTTGTCCTTGTCAGTGAA

Actin 85 ACGTTGGATGAAAGGCCAACAGGGAGAAGA

ACGTTGGATGCGTACATGGCAGGAACATTG

ACATTGAAAGTCTCGAACATAATCC

CGTACATGGCAGGAACATTGAAAGTCTCGAACATAATC[A/c]GGGTCATCTTCTCCCTGTTGGCCTTT

phbA-TS 72 ACGTTGGATGAAATCCACCCGTCGGCACCT

ACGTTGGATGGGATACGATGACAACGTCAG

GACAACGTCAGTCATGG

AAATCCACCCGTCGGCACCTCCGCTTCAAGGTCGACTCTAGAGGA[T/a]CCATGACTGACGTTGTCATCGTATCC

phbB-TS 117 ACGTTGGATGAATGGCGGTTCCGATACCAC

ACGTTGGATGTCGGCACCTCCGCTTCAAG

CTCTAGAGGATCCATGAC

AATGGCGGTTCCGATACCACCCATGCCGCCGGTCACATACGCAATGCGCTG[A/g]GTCATGGATCCTCTAGAGTCGACGCTAGACAAGTCAGATTCTC

phaC2-TS 120 ACGTTGGATGTGCGCGTTCATGTAGTTAGC


CTCCCGACACCTGTTTC

TGCGCGTTCATGTAGTTAGCGGGGACCGGCAAGGCTCCCGACACCTGTTTC[T/c]CTCGCATTGATCCTCTAGAGTCGACCTTGAAGCGGAGGTGCCGA

phaJ2-TS 94 ACGTTGGATGGTAGTTCAGGAGCACCTCAG


CCCCAGCACCTCAGGATCGAG

GTAGTTCAGGAGCACCTCAGGATCGAG[C/t]GCCATGGATCCTCTAGAGTCGACCTTGAAGCGGAGGTGCCGA

FatB2 118 ACGTTGGATGCTAGGTGCTGGAACAGGGAA


TGGAACAGGGAAGAATGC

CTAGGTGCTGGAACAGGGAAGAATGC[A/t]GAACTTGCTGCAGCAGCCACCACCATGTTTGGATCCTCTAGAGTCGACCTTGAAGCGGAGGTGCCGA

KasA1 113 ACGTTGGATGGTACAGAATGGGGACGCAAC


GACGCAACCATGGAAGC

GTACAGAATGGGGACGCAACCATGGAAGC[G/c]GCGGCCGCCATTGAAACACCAAAGATCCTCTAGAGTCGACCTTGAAGCGGAGGTGCCGA

The universal 5PCR primer tag is indicated in bold type



13/15

chased from Sigma Aldrich (St Louis, Missouri); methyl 3-hydroxypentanoate from Fluka AG (Buchs, Switzerland).

A modified version of this method was used for finalanalyses of PHA-producing lines, with the followingmodifications: (1) Methyl 3-hydroxybutyrate and methyl3-hydroxypentanoate were replaced by methyl benzoate,which was added prior to methanolysis, acting as an

extraction standard for subsequent steps and an internalstandard for GC-MS analysis; (2) The chloroform extractwas subjected to methanolysis rather than ethanolysisusing the same procedure, by replacing ethanol with thesame quantity of methanol. This was superior to themethanolysis method used initially, which used methanolcontaining 3% v/v sulphuric acid and produced unwantedderivatives that interfered with analysis. The problem ofloss of volatile methyl 3-hydroxyesters was resolved inthe improved method by using the glass centrifuge tubesand custom-made PTFE seals described above for themethanolysis reaction. Methyl benzoate and the target

standard methyl 3-hydroxyhexanoate were purchasedfrom Sigma Aldrich. Additional target standards methyl3-hydroxyoctanoate, methyl 3-hydroxydecanoate, methyl3-hydroxydodecanoate, methyl 3-hydroxytetradecanoateand methyl 3-hydroxyoctadecanoate were purchased fromLarodan Fine Chemicals AB (Malm,Sweden). Quantita-tion of targets was performed in selective ion monitoringmode using ions with m/z ratios 71, 74, and 103 for methyl3-hydroxyesters and 77, 105, and 136 for methyl benzoate.

Gel Permeation Chromatography

PHA was extracted from approximately 2.5 g freezedried leaf blade tissue using the same method as for GC-MS samples, but without the derivatisation and purifi-cation steps. The chloroform extract was concentrated toa final volume of 300 L, and 100 L used for injection. Separations were performed on a Shimadzu10A HPLC equipped with four columns in series:Phenogel guard, Phenogel Linear-2 mixed bedcolumn (10010,000 KDa), Phenogel 104 (5500KDa), Phenogel 103 (175 KDa) (all 5 m bore,300 7.8 mm; Phenomenex, Torrance, CA, USA; order aslisted; chloroform mobile phase at 1 mL/min). Peaks were

observed with a refractive index detector. ReadyCalpolystyrene standards (Fluka AG, Buchs, Switzerland)were used for Mw calibration.

Transgene Expression Analysis

Leaf blade samples for RNA extractions were taken fromdeveloping leaves of young secondary stalks of 2526month-old sugarcane plants. The excised tissue was immedi-ately frozen in liquid nitrogen before storage at80C until use.

Total RNAwas extracted from 100 mg of leaf blade tissue usingan RNeasy kit (QIAGEN GmbH, Hilden, Germany) includingoptional on-column DNase treatment according to manufac-turers instructions. Reverse transcription was performed using2g total RNA with an Omniscript kit and random hexamers(QIAGEN GmbH, Hilden, Germany) according to manufac-turers instructions. Competitive PCR and Mass ArrayTM

(Sequenom Inc., San Diego, CA) was carried out by theAustralian Genome Research Facility (The University ofQueensland, QLD, Australia) according to the methodologyof Ding and Cantor (2003). Primers used are listed in Table 5.

Transmission Electron Microscopy

Samples were prepared according to Bohmert et al. (2000),except that leaves were fixed in 3% glutaraldehyde.Electron microscopy was performed with a JEOL 1010electron microscope (JEOL, Tokyo, Japan) equipped with aVeleta TEM digital camera and iTEM imaging software

(Olympus Soft Imaging Systems, GMBH, Mnster, Germany).

Acknowledgements This work was funded jointly by the AustralianGovernment through the Co-operative Research Centre for SugarcaneIndustry Innovation through Biotechnology (CRCSIIB) and BSESLimited. AG was a recipient of a Smart State Fellowship awarded bythe Department of State Development, Trade and Innovation of theQueensland Government. We wish to thank Ms. Liz Burns (BSESLimited) for assistance with initial GC-MS screening; Mr. Niall Masel,BSES Limited, for assistance with GPC analysis; Dr. Deb Stenzel,Queensland University of Technology, and Ms. Kimberley Tilbrook,The University of Queensland, for assistance with transmissionelectron microscopy.

References

Arai Y, Nakashita H, Suzuki Y et al (2002) Synthesis of a novel classof polyhydroxyalkanoates in Arabidopsisperoxisomes, and theiruse in monitoring short-chain-length intermediates of beta-oxidation. Plant Cell Physiol 43:555562

Arent S, Pye VE, Henriksen A (2008) Structure and function of plantacyl-CoA oxidases. Plant Physiol Biochem 46:292301

Birch RG, Bower RS, Elliott AR (2010) Highly efficient, 5 -sequence-specific transgene silencing in a complex polyploid. Trop PlantBiol 3:8897

Bohlmann GM (2006) Polyhydroxyalkanoate production in crops. In:Bozell JJ (ed) Feedstocks for the future: renewables for theproduction of chemicals and materials. American ChemicalSociety Symposium Series 921, vol 921. Oxford UniversityPress, Oxford, pp 253270

Bohmert K, Balbo I, Kopka J et al (2000) Transgenic Arabidopsisplants can accumulate polyhydroxybutyrate to up to 4% of theirfresh weight. Planta 211:841845

Bower R, Elliott AR, Potier BAM et al (1996) High-efficiency,microprojectile-mediated cotransformation of sugarcane, usingvisible or selectable markers. Mol Breed 2:239249

Christensen AH, Sharrock RA, Quail PH (1992) Maize polyubiquitingenesstructure, thermal perturbation of expression and tran-



14/15

script splicing, and promoter activity following transfer toprotoplasts by electroporation. Plant Mol Biol 18:675689

Christensen AH, Quail PH (1996) Ubiquitin promoter-based vectorsfor high-level expression of selectable and/or screenable markergenes in monocotyledonous plants. Transgenic Res 5:213218

Ding CM, Cantor CR (2003) A high-throughput gene expressionanalysis technique using competitive PCR and matrix-assistedlaser desorption ionization time-of-flight MS. Proc Natl Acad SciUSA 100:30593064

Doi Y, Kitamura S, Abe H (1995) Microbial synthesis and character-ization of poly(3-hydroxybutyrate-co-3-hydroxyhexanoate).Macromol 28:48224828

Froman BE, Edwards PC, Bursch AG et al (2000) ACX3, a novelmedium-chain acyl-coenzyme A oxidase from Arabidopsis. PlantPhysiol 123:733741

Goepfert S, Poirier Y (2007) Beta-oxidation in fatty acid degradationand beyond. Curr Opin Plant Biol 10:245251

Hahn JJ, Eschenlauer AC, Sleytr UB et al (1999) Peroxisomes as sitesfor synthesis of polyhydroxyalkanoates in transgenic plants.Biotechnol Prog 15:10531057

Hazer B, Steinbchel A (2007) Increased diversification of polyhy-droxyalkanoates by modification reactions for industrial andmedical applications. App Microbiol Biotechnol 74:112

Holmes PA (1988) Biologically produced PHA polymers andcopolymers. In: Bassett DC (ed) Developments in CrystallinePolymers2. Elsevier Applied Science Publ, London/NewYork, pp 165

Huang AHC, Trelease RN, Moore TS (1983) Plant peroxisomes.Academic, New York

Kato M, Bao HJ, Kang CK et al (1996) Production of a novelcopolyester of 3-hydroxybutyric acid and medium chain length3-hydroxyalkanoic acids by Pseudomonassp. 613 from sugars.App Microbiol Biotechnol 45:363370

Lee EY, Jendrossek D, Schirmer A et al (1995) Biosynthesis ofcopolyesters consisting of 3-hydroxybutyric acid and medium-chain length 3-hydroxyalkanoic acids from 1,3-butanediol orfrom 3-hydroxybutyrate by Pseudomonas sp. A33. App Micro-

biol Biotechnol 42:901909

Lee SH, Oh DH, Ahn WS et al (2000) Production of poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) by high-cell-densitycultivation of Aeromonas hydrophila. Biotechnol Bioeng67:240244

Lenz RW, Marchessault RH (2005) Bacterial polyesters: Biosynthesis,biodegradable plastics and biotechnology. Biomacromol 6:18

Leonard JM, Slabaugh MB, Knapp SJ (1997) Cuphea wrightiithioesterases have unexpected broad specificities on saturatedfatty acids. Plant Mol Biol 34:669679

Leonard JM, Knapp SJ, Slabaugh MB (1998) A Cuphea beta-ketoacyl-ACP synthase shifts the synthesis of fatty acids towardsshorter chains in Arabidopsis seeds expressing Cuphea FatBthioesterases. Plant J 13:621628

Liebergesell M, Fallis P, Dong J et al (2002) Polyhydroxyalkanoatesynthase genes. US Patent 6,475,734,

Matsumoto K, Nagao R, Murata T et al (2005) Enhancement of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) production in thetransgenic Arabidopsis thaliana by the in vitro evolved highlyactive mutants of polyhydroxyalkanoate (PHA) synthase from

Aeromonas caviae. Biomacromol 6:21262130Matsumoto K, Arai Y, Nagao R et al (2006) Synthesis of short-chain-

length/medium-chain-length polyhydroxyalkanoate (PHA)copolymers in peroxisome of the transgenic Arabidopsis thalianaharboring the PHA synthase gene from Pseudomonassp 613. JPolymers Env 14:369374

Matsumoto K, Murata T, Nagao R et al (2009) Production of short-chain-length/medium-chain-length polyhydroxyalkanoate (PHA)copolymer in the plastid of Arabidopsis thaliana using an

engineered 3-ketoacyl-acyl carrier protein synthase III. Bioma-cromol 10:686690

McQualter RB, Fong Chong B, Baker A et al (2005) Initial evaluationof sugarcane as a production platform for p-hydroxybenzoic acid.Plant Biotechnol J 3:2941

Mittendorf V, Robertson EJ, Leech RM et al (1998) Synthesis ofmedium-chain-length polyhydroxyalkanoates in Arabidopsisthaliana using intermediates of peroxisomal fatty acid beta-oxidation. Proc Natl Acad Sci USA 95:1339713402

Mittendorf V, Bongcam V, Allenbach L et al (1999) Polyhydroxyalka-noate synthesis in transgenic plants as a new tool to study carbonflow through beta-oxidation. Plant J 20:4555

Mudge SR, Osabe K, Casu RE et al (2009) Efficient silencing ofreporter transgenes coupled to known functional promoters insugarcane, a highly polyploid crop species. Planta 229:549558

Nakashita H, Arai Y, Yoshioka K et al (1999) Production ofbiodegradable polyester by a transgenic tobacco. Biosci Bio-technol Biochem 63:870874

Noda I, Bond EB, Green PR et al (2005a) Preparation, properties, andutilization of biobased biodegradable NodaxTM copolymers. In:Cheng HN, Gross RA (eds) Polymer Biocatalysis and Biomate-rials. American Chemical Society Symposium Series, vol 900.Oxford University Press, Oxford, pp 280291

Noda I, Green PR, Satkowski MM et al (2005b) Preparation and

properties of a novel class of polyhydroxyalkanoate copolymers.Biomacromol 6:580586

Nomura CT, Taguchi S (2007) PHA synthase engineering towardsuperbiocatalysts for custom-made biopolymers. App MicrobiolBiotechnol 73:969979

Oeth P, Correll D, Jurinke C (2004) Multiplexed gene expressionanalysis using competitive PCR and MassARRAYTM. In:Sequenom Application Notes. Sequenom Inc. http://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdfCited 30 June 2008

Petrasovits LA, Purnell MP, Nielsen LK et al (2007) Production ofpolyhydroxybutyrate in sugarcane. Plant Biotechnol J 5:162172

Philip S, Keshavarz T, Roy I (2007) Polyhydroxyalkanoates: biode-gradable polymers with a range of applications. J ChemicalTechnol Biotechnol 82:233247

Poirier Y, Nawrath C, Somerville C (1995a) Production of polyhy-droxyalkanoates, a family of biodegradable plastics and elasto-mers, in bacteria and plants. Biotechnol 13:142150

Poirier Y, Somerville C, Schechtman LA et al (1995b) Synthesis of high-molecular-weight poly([R]-()-3-hydroxybutyrate) in transgenic

Arabidopsis thalianaplant cells. Int J Biol Macromol 17:712Poirier Y, Ventre G, Caldelari D (1999) Increased flow of fatty acids

toward beta-oxidation in developing seeds of Arabidopsisdeficient in diacylglycerol acyltransferase activity or synthesizingmedium-chain-length fatty acids. Plant Physiol 121:13591366

Poirier Y, Gruys KJ (2005) Production of polyhydroxyalkanoates intransgenic plants. In: Doi Y, Steinbchel A (eds) Polyesters I:

biological systems and biotechnological production. Biopoly-mers, vol 3a. Wiley-VCH, Weinheim, pp 281316

Purnell MP, Petrasovits LA, Nielsen LK et al (2007) Spatio-temporalcharacterization of polyhydroxybutyrate accumulation in sugar-cane. Plant Biotechnol J 5:173184

Rehm BHA (2003) Polyester synthases: natural catalysts for plastics.Biochem J 376:1533

Romano A, van der Plas LHW, Witholt B et al (2005) Expression ofpoly-3-(R)-hydroxyalkanoate (PHA) polymerase and acyl-CoA-transacylase in plastids of transgenic potato leads to the synthesisof a hydrophobic polymer, presumably medium-chain-lengthPHAs. Planta 220:455464

Satkowski MM, Melik DH, Autran J-P et al (2001) Physical andprocessing properties of polyhydroxyalkanoate copolymers. In:Doi Y, Steinbchel A (eds) Polyesters II: properties and chemical

http://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdfhttp://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdfhttp://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdfhttp://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdfhttp://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdfhttp://www.sequenom.com/Assets/pdfs/appnotes/Multiplexing_for_Gene_Expression_Analysis.pdf


15/15

synthesis. Biopolymers, vol 3b. Wiley VCH, Weinheim, pp 231264

Slabaugh MB, Leonard JM, Knapp SJ (1998) Condensing enzymesfrom Cuphea wrightii associated with medium chain fatty acid

biosynthesis. Plant J 13:611620Steinbchel A, Valentin HE (1995) Diversity of bacterial polyhydrox-

yalkanoic acids. FEMS Micro Lett 128:219228Steinbchel A (2001) Perspectives for biotechnological production

and utilization of biopolymers: metabolic engineering of poly-

hydroxyalkanoate biosynthesis pathways as a successful example.Macromol Biosci 1:124

Steinbchel A, Hein S (2001) Biochemical and molecular basis ofmicrobial synthesis of polyhydroxyalkanoates in microorgan-isms. In: Scheper T (ed) Advances in biochemical engineering/

biotechnology 71. Springer, Heidelberg, pp 81123Tilbrook K, Gnanasambandam A, Schenk PM et al (2010) Efficient

targeting of polyhydroxybutyrate biosynthetic enzymes to plantperoxisomes requires more than three amino acids in thecarboxyl-terminal signal. J Plant Physiol 167:329332

Tilbrook K, Gebbie L, Schenk PM et al (2011) Peroxisomalpolyhydroxyalkanoate biosynthesis is a promising strategy for

bioplastic production in high biomass crops. Plant Biotechnol J(In press). doi:10.1111/j.1467-7652.2011.00600.x

Tsuge T, Taguchi K, Taguchi S et al (2003) Molecular characterizationand properties of (R)-specific enoyl-CoA hydratases from

Pseudomonas aeruginosa: metabolic tools for synthesis ofpolyhydroxyalkanoates via fatty acid beta-oxidation. Int J BiolMacromol 31:195205

van Beilen JB, Poirier Y (2008) Production of renewable polymersfrom crop plants. Plant J 54:684701

Vandesompele J, De Preter K, Pattyn F et al (2002) Accurate normalizationof real-time quantitative RT-PCR data by geometric averaging ofmultiple internal control genes. Genome Biol 3:112

Volokita M (1991) The carboxy-terminal end of glycolate oxidase directsa foreign protein into tobacco leaf peroxisomes. Plant J 1:361366

Wang YH, Wu ZY, Zhang XH et al (2005) Synthesis of medium-chain-length-polyhydroxyalkanoates in tobacco via chloroplast geneticengineering. Chinese Sci Bull 50:11131120

Zou XH, Chen GQ (2007) Metabolic engineering for microbialproduction and applications of copolyesters consisting of 3-hydroxybutyrate and medium-chain-length 3-hydroxyalkanoates.Macromol Biosci 7:174182

http://dx.doi.org/10.1111/j.1467-7652.2011.00600.xhttp://dx.doi.org/10.1111/j.1467-7652.2011.00600.x

Documents

Anderson Et Al. (2011) Produção de Bioplásticos