Biosynthesis and structural characterization of medium-chain-length poly(3-hydroxyalkanoates) produced by Pseudomonas aeruginosa from fatty acids

International Journal of Biological Macromolecules 29 (2001) 107– 114 www.elsevier.com/locate/ijbiomac

Biosynthesis and structural characterization of medium-chain-length poly(3-hydroxyalkanoates) produced by Pseudomonas aeruginosa from fatty acids Alberto Ballistreri a, Mario Giuffrida b, Salvatore P.P. Guglielmino c, Santina Carnazza c, Annamaria Ferreri c, Giuseppe Impallomeni b,* b

a Dipartimento di Scienze Chimiche, Uni6ersita` di Catania, 95125 Catania, Italy Istituto per la Chimica e la Tecnologia dei Materiali Polimerici, Consiglio Nazionale delle Ricerche, Viale A. Doria 6, 95125 Catania, Italy c Dipartimento di Scienze Microbiologiche, Genetiche e Molecolari, Facolta` di Scienze, Uni6ersita` di Messina, 98166 Messina, Italy

Received 6 February 2001; received in revised form 24 May 2001; accepted 24 May 2001

Abstract In this study, we investigated the ability of Pseudomonas aeruginosa ATCC 27853 to grow and synthesize poly(3-hydroxyalkanoates) (PHAs) from saturated fatty acids with an even number of carbon atoms, from eight to 22, and from oleic acid. In a non-limiting medium, all carbon sources but docosanoic acid supported cell growth and PHA production, with eicosanoic acid giving the highest yield. In magnesium-limiting conditions, higher yields were obtained from sources with up to 16 carbon atoms. Composition was estimated by gas chromatography of methanolyzed samples and 13C nuclear magnetic resonance. The 3-hydroxyalkanoate units extended from hexanoate to tetradecanoate or tetradecenoate, with octanoate and decanoate as the predominant components. Weight average molecular weights ranged from 78 000 to 316 000. Fast atom bombardment mass spectrometry of partially pyrolyzed samples, coupled to statistical analysis, showed that these PHAs are random copolymers. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Pseudomonas aeruginosa; Poly(hydroxyalkanoate) structure; Biosynthesis; Nuclear magnetic resonance; Copolymer sequence; Fast atom bombardment mass spectrometry

1. Introduction A wide variety of bacteria are able to synthesize poly(3-hydroxyalkanoates) (PHAs) as intracellular granules in nutrient-limiting conditions and in the presence of carbon source excess [1 – 4]. This class of polyesters has the following general structure: R (OCHCH2CO)n  R =side-chain substituent

* Corresponding author. Tel.: + 39-095-339-926; fax: + 39-095221-541. E-mail address: [email protected] (G. Impallomeni).

where, due to the stereospecificity of biosynthetic enzymes, the chiral centres have the (R) stereochemical configuration, so that they are isotactic and optically active [5,6]. These biopolymers are receiving much attention because of their potential use as renewable and biodegradable plastics [7,8]. Alcaligenes eutrophus has the ability to synthesize short-chain-length (scl) polyesters in which RCH3 or C2H5, whereas Pseudomonas oleo6orans produces medium-chain-length (mcl) polyesters in which RC3H7 to C9H19 [9]. Moreover, PHAs containing functional groups, such as olefin [10], phenyl [11,12], chloride [13], bromide [14], fluoride [15,16], phenoxy [17] and cyanophenoxy [18] have been obtained from P. oleo6orans or Pseudomonas putida grown on functional substrates. Monomer composition of PHAs produced by pseudomonads belonging to rRNA homology group I is

0141-8130/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 4 1 - 8 1 3 0 ( 0 1 ) 0 0 1 5 4 - 4

108

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114

determined by the specificity of the PHA biosynthetic system, by the nature of the substrate added as a carbon source to the growth media and by the metabolic pathways leading to PHA synthesis [19]. The comonomeric composition of these PHAs, and its direct relationship with the structure of the growth substrate, suggest that the mcl-PHA biosynthetic route is a direct branch of the fatty acid oxidation pathway [20]. When the carbon source has from six to 12 carbon atoms, the monomers in the PHA are of the same length as the carbon source, or have been shortened by two, four or six carbon atoms. When C13 –C18 fatty acids are given, the composition of the PHA is no longer substrate-dependent, because the longest monomer inserted has 11 or 12 carbon atoms, according to the odd or even number of carbon atoms in the source [21]. When P. oleo6orans is fed with heptadecanoic or octadecanoic acids, however, PHA is not detected, because these carbon sources do not allow it to grow. The mcl-PHA synthesis is now considered not only pertinent to P. oleo6orans, but common to all fluorescent pseudomonads [21]. The mcl-PHAs show elastomeric properties and can be classified as thermoplastic elastomers of interest for several uses. Fluorescent pseudomonads can produce PHAs on a large variety of carbon sources of agricultural origin [19,21]. Among these, the fatty acids derived from vegetable oils seem the most interesting [21– 23]. Lauric, myristic, palmitic, stearic and oleic acids are the most widespread fatty acids and are relatively cheap. Their mixtures can be found as waste products in vegetable oil production, so they are economically more attractive. Furthermore, many efforts have been made to lower the cost of PHA production by the isolation of bacterial strains more efficient in fermentation processes. The aim of this study was to investigate the ability of Pseudomonas aeruginosa ATCC 27853, a clinic isolate able to utilize several carbon sources, to grow and synthesize PHAs from even-carbon-atoms n-alkanoic acids (from octanoic to docosanoic) and from cis-9-octadecenoic acid (oleic acid) as an unsaturated carbon source. The PHAs produced were characterized by gas chromatography (GC), gel permeation chromatography (GPC), 1H and 13C nuclear magnetic resonance (NMR) spectroscopy. Moreover, since in the field of materials science it is important to know if a homopolymer, a copolymer or a mixture of homo- or copolymers has been obtained from a fermentation process, fast atom bombardment mass spectrometry (FAB MS) analysis was applied to partial pyrolysis products of PHAs, following a procedure previously described [24,25], to determine the sequence distribution of repeating units.

2. Materials and methods

2.1. Media and growth conditions P. aeruginosa ATCC 27853, cultured in Luria Bertani (LB) broth at 30 °C, was used throughout the experiments. For PHA production, cells were cultured in an E* medium [26] containing the following: 1.1 g/l (NH4)2HPO4, 5.8 g/l K2HPO4, 3.7 g/l KH2PO4, 10 ml/l MgSO4 0.1 M, supplemented with 1 ml/l of a microelement solution [20]. MT stock contained the following: 2.78 g/l FeSO4·7H2O, 1.98 g/l MnCl2·4H2O, 2.81 g/l CoSO4·7H2O, 1.67 g/l CaCl2·2H2O, 0.17 g/l CuCl2·2H2O, 0.29 g/l ZnSO4·7H2O. Magnesium-limiting conditions were created by depriving the medium of MgSO4. n-Octanoic, n-decanoic, n-dodecanoic, ntetradecanoic, n-hexadecanoic, n-octadecanoic, n-eicosanoic, n-docosanoic and cis-9-octadecenoic acids (Aldrich) were added as carbon sources to give a final concentration of 10 mM. The pH was adjusted to 7.0, and the medium was autoclaved. Exponential-phase cells, precultured in 500 ml LB broth cultures, were inoculated in 2 l E* medium batch cultures. Experiments were performed under aerobic conditions in a temperature-controlled shaker set at 250 rpm and at a constant temperature of 30 °C. After 72 h cells were harvested by centrifugation, washed and lyophilized.

2.2. Extraction of poly(3 -hydroxyalkanoate) PHAs were extracted from lyophilized cells into chloroform using a Soxhlet extractor. After a reflux period of 6 h, the chloroform solution was concentrated with a rotary evaporator. The crude polymer product was dissolved in chloroform at a ratio of 5:1 (volume of chloroform (ml) to weight of crude extract (g)), and precipitated into ten volumes of rapidly stirred ethanol. After 30 min of stirring, the ethanol– chloroform mixture was decanted o.n., and the precipitated polymer was separated by centrifugation (Beckman J2-21; JA-20 rotor; 20 °C; 9000×g), washed twice with ethanol, and dried in vacuo (1 mmHg) at room temperature. The product was weighed and the precipitation was repeated twice more to obtain the final product.

2.3. Gas chromatography analysis To determine the copolymer composition, 2–5 mg of PHAs were subjected to methanolysis by heating for 140 min at 100 °C in a mixture of 1 ml chloroform and 1 ml methanol containing 15% sulfuric acid [9]. The reaction mixture was washed with 1 ml water and the chloroform layer was analyzed by GC. A Perkin–Elmer 8420 GC system was used, equipped

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114

with an AT-50 Alltech capillary column (30 m× 0.25 mm; carrier gas He, 1 ml/min) and with a flame ionization detector; the temperature program was 80 °C for 5 min, then the temperature was increased by 8 °C/min to 280 °C and held for 10 min. 0.2 ml of the organic phase was analyzed after injection (split ratio 10:1). The methyl esters obtained by methanolysis were identified by comparing their retention times with those of standard methyl-3-hydroxyalkanoates (Sigma).

2.4. Molecular weight measurements Average molecular weights were determined by GPC using a Waters Model 6000A solvent delivery system with five Styragel HR columns connected in series (in the order: HR5, HR4, HR3, HR2 and HR1) and a Model 401 refractive index detector. Chloroform was used as an eluent at a flow rate of 1.0 ml/min; 200 ml of a 5 mg/ml solution were injected for each sample. A molecular weight calibration curve was generated with polystyrene standards of low polydispersity (Polymer Laboratories) using GPC Caliber acquisition and processing software (Polymer Laboratories).

2.5. Nuclear magnetic resonance spectroscopy Two hundred Mega Hertz 1H NMR spectra of the samples were recorded at room temperature in a CDCl3 solution on a Bruker AC 200 spectrometer with a 4 s pulse repetition, a 2000 Hz spectral width, 16K data points and 256 scans accumulation. 1H decoupled 50 MHz 13C NMR spectra were recorded on the same samples with 1.6 s acquisition time, 10 000 Hz spectral width, 32K data points and 30 000 scans accumulation.

109

Inverse gated 13C NMR spectra were acquired using the same spectral width, but with 0.8 s acquisition time and an additional 5 s relaxation delay.

2.6. Partial pyrolysis Partial pyrolyses of microbial polyesters were performed with a Perkin–Elmer TGS/2 thermogravimetric (TG) apparatus under a nitrogen atmosphere (60 ml/min) at a heating rate of 10 °C/min. The temperature was increased until weight loss was 20%. After cooling rapidly to room temperature (100 °C/min), the residue was recovered. This residue was completely soluble in chloroform.

2.7. Fast atom bombardment mass spectrometry A double-focusing Kratos MS 50S was used, equipped with a standard FAB source and with Maspec2 data acquisition and processing system (Mass Spectrometry Services Ltd.). The cesium ion gun was operated at 20 kV and the instrument was scanned from m/z 3000 to 60 at a scan rate of 10 s per decade and at an acceleration potential of 8 kV in negative ion mode. Cesium and rubidium iodides (50:50 by weight) were used for computer calibration. Samples obtained from partial pyrolysis of the polymers in TG experiments were placed on the target of the direct insertion probe and mixed with 3-nitrobenzyl alcohol as a matrix. The experimental peak intensities corresponding to monomers were excluded from calculations, because, as previously shown [25], the volatility of these species makes copolymer compositions estimated by these means unreliable. The peak intensity values reported represent the average of three samples.

Table 1 Production of PHAs by P. aeruginosa cultured on various fatty acids Fatty acid carbon source

Cellular dry weight (mg/l)

PHA yield (% cellular dry weight)

PHA yield (mg/l)

Octanoic Octanoic−Mg Decanoic Decanoic−Mg Dodecanoic Dodecanoic−Mg Tetradecanoic Tetradecanoic−Mg Hexadecanoic Hexadecanoic−Mg Octadecanoic Octadecanoic−Mg 9-Octadecenoic 9-Octadecenoic−Mg Eicosanoic Eicosanoic−Mg Docosanoic Docosanoic−Mg

989 380 1170 628 803 558 1395 628 1500 1065 2580 907 2346 1092 2518 990 3510 330

Trace 7.9 Trace 2.7 0.2 1.1 0.4 18.9 0.7 2.7 0.9 Trace 0.4 6.1 7.1 Trace – –

Trace 30 Trace 17 2 6 6 119 10 29 22 Trace 9 67 180 Trace – –

Cultures were grown in a complete E* medium and in a Mg-deprived medium (−Mg).

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114

110

The spectrum from each sample was averaged over at least five scans. 3. Results and discussion PHAs were isolated from P. aeruginosa ATCC 27853 after its growth on different n-alkanoic acids with an even number of carbon atoms ranging from octanoic to docosanoic, including also oleic acid as a carbon source with unsaturated bonds. All the biosyntheses were carried out, both in a complete E* medium and in magnesium-limiting conditions, for each carbon source. Yields in biomass and in PHA are reported in Table 1. No PHA was detected when acids with less than eight carbons were supplied, and no nutrient limitation was required to induce PHA synthesis when fatty acids from C8 to C20 were used, although better yields were obtained when the Table 2 Molecular weights and polydispersities of PHAs obtained from various fatty acid carbon sources Fatty acid

Mw 10−3

Mn 10−3

Mw/Mn

Octanoic Octanoic−Mg Decanoic Decanoic−Mg Dodecanoic Dodecanoic−Mg Tetradecanoic Tetradecanoic−Mg Hexadecanoic Hexadecanoic−Mg Octadecanoic Octadecanoic−Mg 9-Octadecenoic 9-Octadecenoic−Mg Eicosanoic Eicosanoic−Mg

n.d. 316 n.d. 251 274 216 204 255 92 241 200 n.d. 78 182 273 n.d.

n.d. 191 n.d. 121 195 127 118 148 60 113 103 n.d. 49 106 120 n.d.

n.d. 1.7 n.d. 2.1 1.4 1.7 1.7 1.7 1.5 2.1 1.9 n.d. 1.6 1.7 2.3 n.d.

Cultures were grown in a complete E* medium and in a Mg-deprived medium (−Mg). Mw, weight— average molecular weight; Mn, number— average molecular weight; n.d., not determined. Table 3 Monomer composition (mol%) of PHAs obtained from various fatty acid carbon sources as determined by GC Fatty acid

C

O

D

D

Octanoic Decanoic Dodecanoic Tetradecanoic Hexadecanoic Octadecanoic 9-Octadecenoic Eicosanoic

10 5 4 4 4 4 4 5

86 66 62 61 54 44 55 38

4 28 26 26 29 38 27 38

1 7 7 10 11 8 15

T

Te

1 2 3 3 6 4

The PHAs analyzed are those obtained in the culture medium giving the highest yield from each carbon source (Table 1); for symbols used see Fig. 1.

Fig. 1. Chemical structure of the PHAs obtained from P. aeruginosa ATCC 27853 cultured on various fatty acids (Table 1).

medium was magnesium deprived, with the exception of octadecanoic and eicosanoic acids. Cell growth was observed from docosanoic acid, but not PHA production. Interestingly, the highest PHA yield (180 mg/l) was observed with eicosanoic acid in a complete E* medium. This was an unexpected result, because PHA production by pseudomonads from fatty acids higher than C18 has never been reported. The average molecular weights measured by GPC are listed in Table 2. Molecular weights ranged from 78 000 to 274 000 when the PHAs were obtained from the complete medium, and from 182 000 to 316 000 when they were obtained from the Mg-deprived medium. The composition of the polyesters produced in the highest yield from each carbon source, as determined by GC of the 3-hydroxyalkanoate methyl esters prepared by methanolysis, is shown in Table 3. The PHA from octanoic acid consisted of three different repeating units, the PHA from decanoic acid of four, and the PHAs from dodecanoic to eicosanoic acids of five. The most predominant units were 3-hydroxyoctanoate and 3-hydroxydecanoate, in agreement with the results of Huisman et al. [21], who concluded that the PHA polymerizing enzyme system shows a preference for monomers with a carbon chain length between eight and ten. The trend observed is that higher monomer units are more abundant when a carbon source with a longer carbon chain is used. The medium used (complete or Mg-deprived) did not substantially alter the composition of the polyesters synthesized from each carbon source. The proposed chemical structure of the PHAs obtained from saturated carboxylic acids is shown in Fig. 1. Fig. 2(a)–(c) shows the 13 C NMR spectra of the PHAs isolated from P. aerugi-

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114

Fig. 2. Fifty Mega Hertz 13C NMR spectra of the PHAs isolated from P. aeruginosa ATCC 27853 cultured on eicosanoic (a), tetradecanoic (b), and octanoic (c) acids. For symbols used see Fig. 1.

Fig. 3. Fifty Mega Hertz 13C NMR spectrum of the PHA isolated from P. aeruginosa ATCC 27853 cultured on cis-9-octadecenoic acid. For symbols used see Fig. 1.

Table 4 Monomer composition (mol%) of PHAs obtained from various fatty acid carbon sources as determined by GC, 13C NMR, and FAB MS D

T

26 25 25

7 11a 7

2

55 52 55

27 35a 22

8

38 39 37

38 37 38

15 19a 10

Fatty acid

Technique

C

O

D

Octanoic

GC NMR MS

10 9 8

86 87 90

4 4 2

Tetradecanoic

GC NMR MS

4 5 6

61 59 60

9-Octadecenoic

GC NMR MS

4 6 9

GC NMR MS

5 6 10

Eicosanoic

Te

2 6 7 6

8 4 5

The PHAs analyzed are those obtained in the culture medium giving the highest yield from each carbon source (Table 1); for symbols used see Fig. 1. a Mol% of D and T units or D and D units are given as their sum as their signals are not resolved.

111

nosa grown on eicosanoic, tetradecanoic and octanoic acids, respectively, chosen as representative of the saturated series studied. Assignments were made following Gross et al. [9] for the carbons of 3-hydroxyhexanoate (C), 3-hydroxyoctanoate (O) or 3-hydroxydecanoate (D) units; for the carbons belonging to 3-hydroxydodecanoate (D) and 3-hydroxytetradecanoate (T) units, resonances were assigned using the known additive shift parameters for hydrocarbons [27]. For brevity 13C NMR spectra of the PHAs obtained from decanoic, dodecanoic, hexadecanoic and octadecanoic acids are not reported here. They are similar to those shown in Fig. 2(a)–(c), differing only in the relative intensity of the 13C signals. Fig. 3 shows the 13C NMR spectrum of the PHA isolated with oleic acid as the carbon source. A magnesium-deprived medium was used to obtain higher yields (67 mg/l against 9 mg/l, Table 1). With respect to the 13C NMR spectra of the PHAs containing only saturated 3-hydroxyalkanoates, in the spectrum of the PHA from oleic acid additional signals were found at 123 and 134 ppm, indicating the presence of a 3-hydroxy-cis-5-tetradecenoate (Te) repeating unit. The assignments are based on those reported previously for the PHA obtained from P. putida [23] and Pseudomonas resino6orans [28] grown on the same carbon source. The proposed chemical structure is shown in Fig. 1 and its composition measured by GC is reported in Table 3. The PHA produced by P. aeruginosa from oleic acid contained a high concentration of 3-hydroxyoctanoate and 3-hydroxydecanoate units, and the whole composition differs slightly from that reported for P. putida and P. resino6orans. This suggests a common b-oxidation pathway in these species involving the isomerization of the cis-D3 double bonds of 3-cis-dodecanoyl CoA to trans-D2 double bonds by cis-D3-enoyl-CoA isomerase. The composition of the polyesters from octanoic, tetradecanoic, eicosanoic and oleic acid was also determined from peak areas of 13C NMR signals in inversegated decoupling experiments. This was possible as the GC analysis and FAB MS of partial pyrolyzates (vide infra) allowed us to exclude the presence of monomers other than C, O, D, D, T and Te. The results are reported in Table 4. Nevertheless it is not possible to establish by GC or 13 C NMR whether a single copolymer, or a mixture of copolymers or of homopolymers, has been obtained. In fact, the chemical shifts of carbon atoms in the units higher than 3-hydroxyhexanoate are essentially identical [9] and no dyads or triads analysis may be applied. We therefore analyzed some of the PHAs by investigating the structure of their oligomers formed by partial pyrolysis, by means of FAB MS, following a procedure previously described [24,25]. Assuming a Bernoullian (random) distribution of repeating units in these copolymers, the probability of finding a given Cx Oy …Tz sequence

112

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114

(Px,y …z ) can be calculated by the Leibnitz formula as follows: Px,y,…z =

(x+ y+···z)! x y P CP O…P zT. x!y!…z!

Table 5 Experimental and calculated relative amounts of the partial pyrolysis products of some PHAs produced by P. aeruginosa from fatty acids

(1)

In this equation PC, PO and PT are the molar fractions of the C, O and T units in the copolymer. The experimental intensities of the FAB MS peaks of a partially pyrolyzed sample are compared to those computed starting from an arbitrary set of monomer abundances. A measure of the fit of the calculated oligomer intensities to the experimental ones is given by the agreement factor (AF) [25]. The best match between the experimentally observed oligomer distribution and the calculated statistical abundances gives the copolymer composition (Table 4). If this composition is compared to that measured by GC or NMR and the two sets of data are very similar, the statistical model used to generate the theoretical oligomer intensities is validated. Fig. 4 shows the negative ion FAB mass spectrum of the partial pyrolyzate of the PHA from tetradecanoic acid, chosen as an example of the whole saturated PHAs series studied. The spectrum essentially consists of the pseudomolecular ions corresponding to the oligomers whose structure is shown in Fig. 5, where R may be a n-propyl, n-pentyl, n-heptyl, n-nonyl or n-undecanyl group. Monomers are identified at m/z 113 (C), m/z 141 (O), m/z 169 (D), m/z 197 (D), and m/z 225 (T). The dimers and trimers identified in Fig. 4 are listed in Table 5 together with their experimental and calculated intensities. The change in pattern of the FAB spectra as a function of PHA composition can be observed in Fig. 6(a)– (c), where the negative ion FAB mass spectra of the partial pyrolyzate of the PHAs from octanoic, tetradecanoic and eicosanoic acid are shown. For clarity the trimers zone has been cut out.

Fig. 4. Negative ion FAB mass spectrum of the residue from partial pyrolysis of the PHA isolated from P. aeruginosa ATCC 27853 cultured on tetradecanoic acid.

m/z (MH)−

FAB MSa

Calculatedb

PHA from octanoic acidc C/O/D (8/90/2)d Dimers C2 CO O2, CD OD D2

227 255 283 311 339

2 12 81 4 1

0.6 14 82 3 0.4 AFe 0.034

Trimersf C20 C22 C24 C26

369 397 425 453

3 19 74 4

2 20 74 4 AFe 0.022

PHA from tetradecanoic acidc C/O/D/D/T (6/60/25/7/2)d Dimers CO CD, O2 CD, OD CT, OD, D2 OT, DD DT

255 283 311 339 367 395

7 40 32 14 6 1

7 39 32 15 6 1 AFe 0.024

Trimersf C24 C26 C28 C30 C32

425 453 481 509 537

29 31 24 11 5

30 33 22 11 4 AFe 0.059

PHA from eicosanoic acidg C/O/D/D/T (10/37/38/10/5)d Dimers CO CD, O2 CD, OD CT, OD, D2 OT, DD DT, D2

255 283 311 339 367 395

8 21 32 23 12 4

8 22 31 23 11 5 AFe 0.045

Trimersf C24 C26 C28 C30 C32 C34

425 453 481 509 537 565

18 24 26 17 10 5

16 25 26 19 10 4 AFe 0.064

10 35 28

10 36 27

PHA from 9-octadecenoic acidc C/O/D/D/Te (9/55/22/8/6)d

Fig. 5. Chemical structure of the PHA oligomers produced by partial pyrolysis.

Dimers CO CD, O2 CD, OD

255 283 311

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114 Table 5 (Continued) −

a

b

m/z (MH)

FAB MS

Calculated

CTe OD, D2 OTe DD DTe

337 339 365 367 393

3 13 5 4 2

2 14 6 3 2 AFe 0.060

Trimersf C24 C26 C28 C30 C32

425 453 481 507 509

32 30 22 8 8

31 32 21 7 9 AFe 0.045

a

Relative intensity of (MH)− ions of the partial pyrolysis products in the FAB mass spectrum. b Relative intensity of the partial pyrolysis products as calculated by Eq. (1). c Obtained from a Mg-deprived medium. d Composition giving the lowest agreement factor. For symbols used see Fig. 1. e Agreement factor. f All possible isobaric trimers containing n carbon atoms are indicated as Cn. g Obtained from a complete medium.

Fig. 6. Negative ion FAB mass spectra of the residue from partial pyrolysis of the PHAs isolated from P. aeruginosa ATCC 27853 cultured on octanoic (a), tetradecanoic (b), and eicosanoic (c) acids.

113

The lowest AF for the PHA from octanoic acid was obtained for the composition C/O/D of 8/90/2, very close to 10/86/4 and 9/87/4 obtained from GC and NMR data (Table 4), and the calculated dimer and trimer intensities were in good agreement with the experimental values. It is therefore possible to conclude that the sequence of monomeric units in this polyester is random. Similar agreements were found for the PHAs from tetradecanoic and eicosanoic acids, as shown in Table 5. Fig. 7 shows the negative ion FAB mass spectrum of the partial pyrolyzate of the PHA from oleic acid. In this case R in Fig. 5 may be an n-propyl, n-pentyl, n-heptyl, n-nonyl or n-undecenyl group. Monomers are identified at m/z 113 (C), m/z 141 (O), m/z 169 (D), m/z 197 (D) and m/z 223 (Te). The dimers and trimers identified in Fig. 7 are listed in Table 5. The lowest AF was given by the composition C/O/D/D/Te of 9/55/22/8/6 very close to 4/55/27/8/6 and 6/52/35 (D+ D)/7 (Table 4) obtained by GC and NMR, establishing a random sequence also for this copolymer. 4. Conclusions P. aeruginosa produced PHAs when was grown on even carbon atom fatty acids from octanoic to eicosanoic. The last carbon source gave the best PHA yield (180 mg/l). This is an interesting result, because PHA production by pseudomonads from fatty acids higher than C18 has never been reported. The compositions of these polyesters were substrate dependent: higher monomer units were more abundant when carbon sources with longer carbon chains were used. Octanoate and decanoate, the predominant components, varied their respective percent molar ratio from 86/4 in the PHA from octanoic acid to 38/38 in that obtained from eicosanoic acid. Compositional analysis data obtained by three different techniques (GC, 13C NMR and FAB MS) were in good agreement. Finally, FAB MS of partially pyrolyzed samples coupled to statistical analysis showed that these polyesters are random copolymers, confirming that this technique remains the only one at present to supply chemical evidence of the copolymeric nature of these mcl-PHAs.

Fig. 7. Negative ion FAB mass spectrum of the residue from partial pyrolysis of the PHA isolated from P. aeruginosa ATCC 27853 cultured on cis-9-octadecenoic acid.

114

A. Ballistreri et al. / International Journal of Biological Macromolecules 29 (2001) 107–114

Acknowledgements Many thanks are due to Mr Roberto Rapisardi for his continuous and skilful technical assistance with FAB MS. Partial financial support from the Consiglio Nazionale delle Ricerche (CNR, Rome), the Ministero dell’Universita` e della Ricerca Scientifica e Tecnologica (MURST, Rome), and the European Economic Community (contribution no. 94.05.09.013 ARINCO no. 94.IT.16028) is gratefully acknowledged. References [1] Doi Y. Microbiol. Microbial Polyesters. New York: VCH Publishers, 1990. [2] Anderson AJ, Dawes EA. Microbiol Rev 1990;54:450. [3] Steinbu¨ chel A, Valentin HE. FEMS Microbiol Lett 1995;128:219. [4] Madison LL, Huisman GW. Microbiol Mol Biol Rev 1999;63:21. [5] Akita S, Einega Y, Miyaki Y, Fiyita H. Macromolecules 1976;9:774. [6] Alper R, Lundgren DG, Marchessault RH, Cote WA. Biopolymers 1963;1:545. [7] Hocking PJ, Marchessault RH. In: Griffin GJL, editor. Chemistry and Technology of Biodegradable Polymers, Biopolyesters. New York: Blackie Academic & Professional, 1994. [8] Holmes PA. Phys Technol 1985;16:32. [9] Gross RA, DeMello C, Lenz RW, Brandl H, Fuller RC. Macromolecules 1989;22:1106.

[10] Preusting H, Nijenhuis A, Witholt B. Macromolecules 1990;23:4220. [11] Fritzsche K, Lenz RW, Fuller RC. Makromol Chem 1990;191:1957. [12] Hazer B, Lenz RW, Fuller RC. Polymer 1996;37:5951. [13] Doi Y, Abe C. Macromolecules 1990;23:3705. [14] Kim YB, Lenz RW, Fuller RC. Macromolecules 1992;25:1852. [15] Abe C, Taima Y, Nakamura Y, Doi Y. Polym Commun 1990;31:404. [16] Kim O, Gross RA, Hammar WJ, Newmark RA. Macromolecules 1996;29:4572. [17] Kim YB, Rhee YH, Han SH, Heo GS, Kim JS. Macromolecules 1996;29:3432. [18] Kim O, Gross RA, Rutherford DR. Can J Microbiol 1995;41:32. [19] Eggink G, de Waard P, Huijberts GNM. FEMS Microbiol Rev 1992;103:159. [20] Lageveen RG, Huisman GW, Preusting H, Ketelaar P, Eggink G, Witholt B. Appl Environ Microbiol 1988;54:2924. [21] Huisman GW, de Leeuw O, Eggink G, Witholt B. Appl Environ Microbiol 1989;55:1949. [22] Eggink G, de Waard P, Huijberts GNM. Can J Microbiol 1995;41:14. [23] de Waard P, van der Wal H, Huijberts GNM, Eggink G. J Biol Chem 1993;268:315. [24] Montaudo MS, Ballistreri A, Montaudo G. Macromolecules 1991;24:5051. [25] Ballistreri A, Giuffrida M, Impallomeni G, Lenz RW, Fuller RC. Int J Biol Macromol 1999;26:201. [26] Volgen HJ, Bonner DM. J Biol Chem 1956;218:97. [27] Silverstein RM, Bassler GC, Morrill TC. Spectrometric Identification of Organic Compounds, 5th ed. New York: Wiley, 1991:234. [28] Ashby RD, Foglia TA. Appl Microbiol Biotechnol 1998;49:431.