Biocatalytic synthesis of new copolymers from 3-hydroxybutyric acid and a carbohydrate lactone

Article abstract of DOI:10.1016/j.molcatb.2011.03.004

Lipase-catalyzed reaction of 3-hydroxybutyric acid with d-glucono-δ-lactone at 5:1 molar ratio and 80°C yielded a mixture of moderate molecular weight linear and cyclic oligomers. The most efficient biocatalyst, Candida antarctica B lipase (Novozyme 435), allowed the synthesis of new oligomeric compounds with ring-opened gluconolactone units included in the oligomeric chain, without previous derivatization of the sugar, or activation of the acid monomer. The reaction medium nature had an important influence on the product composition. Although the main copolymer amount was synthesized in tert-butanol/dimethylsulfoxide medium, the highest polymerization degrees, up to 9 for the copolymer, and 10 for the 3-hydroxybutyric acid homopolymer co-product, were achieved in solventless conditions.

Full text of DOI:10.1016/j.molcatb.2011.03.004

Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Biocatalytic synthesis of new copolymers from 3-hydroxybutyric acid and a
carbohydrate lactone
Sándor Kakasi-Zsurkaa,c, Anamaria Todeaa, Andrada Buta, Cristina Paula, Carmen G. Boeriub,d
,
Corneliu Davidescua, Lajos Nagyc, Ákos Kukic, Sándor Kékic, Francisc Pétera,∗
a Faculty of Industrial Chemistry and Environmental Engineering, University “Politehnica” of Timisoara, C. Telbisz 6, 30001 Timisoara, Romania
b Wageningen University and Research Centre, Food & Biobased Research, Division of Biobased Products, Wageningen, The Netherlands
c Department of Applied Chemistry, University of Debrecen, H-4010 Debrecen, Hungary
d University “Aurel Vlaicu” of Arad, Department of Chemical and Biological Sciences, Arad, Romania
a r t i c l e i n f o
a b s t r a c t
Lipase-catalyzed reaction of 3-hydroxybutyric acid with d-glucono-␦-lactone at 5:1 molar ratio and 80 ◦C
yielded a mixture of moderate molecular weight linear and cyclic oligomers. The most efficient biocat-
alyst, Candida antarctica B lipase (Novozyme 435), allowed the synthesis of new oligomeric compounds
with ring-opened gluconolactone units included in the oligomeric chain, without previous derivatiza-
tion of the sugar, or activation of the acid monomer. The reaction medium nature had an important
influence on the product composition. Although the main copolymer amount was synthesized in tert-
butanol/dimethylsulfoxide medium, the highest polymerization degrees, up to 9 for the copolymer, and
10 for the 3-hydroxybutyric acid homopolymer co-product, were achieved in solventless conditions.
© 2011 Elsevier B.V. All rights reserved.
Article history:
Received 16 December 2010
Received in revised form 11 March 2011
Accepted 14 March 2011
Available online 21 March 2011
Keywords:
3-Hydroxybutyric acid
Biopolymers
Gluconolactone
Lipase
Biocatalysis
1. Introduction
toxic chemicals as catalysts [2]. Even though the most important
biomacromolecules are nucleic acids, proteins, and polysaccha-
Like many other important compounds, biodegradable poly-
mers based on renewable resources are products of industrial
(white) biotechnology [1]. Three routes are achievable to obtain
polymeric substances using biotechnological procedures: (i) isola-
tion of natural biopolymers produced in plants; (ii) production by
microorganisms; (iii) synthesis from bio-derived monomers by in
vitro enzymatic catalysis. While the first two are based on biosyn-
thetic pathways and basically limited to synthesis of naturally
occurring polymers by in vivo enzymatic catalysis, the third offer
several possibilities to design new synthetic strategies for existing
polymers, or even to obtain new polymeric compounds that are
difficult to be synthesized by conventional chemical catalysis. Enzy-
matic polymerizations also display the well-known advantages of
biocatalytic processes, like mild reaction conditions and avoiding
rides, polyesters have also gained constantly increasing interest.
Polyesters, both aliphatic and aromatic, are already used for manu-
facturing of several industrial and consumer products, but likewise
they are considered emerging biomaterials for medical applica-
tions, or carriers for controlled drug release [3].
The enzymes already investigated for enzymatic synthesis of
polyesters were mainly hydrolases, with lipases demonstrating
special ability to catalyze such reactions. Essentially, they are
two applicable polymerization routes: ring-opening polymeriza-
tion (ROP) of lactones and polyester condensation polymerizations,
both having been intensively investigated during the past decade,
as reviewed by Kobayashi [4], Albertsson and Srivastava [5], and
Gross et al. [6] Recently, another hydrolase, cutinase from Humi-
cola insolens, was successfully used for cell-free polyester synthesis
[7]. In a special case, polymerization of (R)-␤-hydroxyalkanoate
monomers, polyhydroxyalkanoate (PHA) synthases of bacterial ori-
gin have also been demonstrated their capability to work in vitro
[8].
Abbreviations: ROP, ring-opening polymerization; PHA, polyhydroxyalka-
noate; 3-HBA, 3-hydroxybutyric acid; P3HB, poly(3-hydroxybutyrate); GL,
d-glucono-␦-lactone; RO-GL, ring-opened gluconolactone; [Bmim][PF6], 1-butyl-3-
methylimidazolium hexafluorophosphate; t-BuOH, tert-butanol; DMSO, dimethyl-
sulfoxide; MALDI-TOF, matrix-assisted laser desorption/ionization time of flight.
Synthesis of polymeric compounds that contain carbohydrates
had also captured considerable interest, by reason of their new
functional capabilities able to induce specific properties and
applications. Sugar-containing poly(acrylate)-based biocompati-
ble hydrogels suitable for biomedical and membrane applications
∗
Corresponding author. Tel.: +40 256 404216; fax: +40 256 403060.
E-mail address: francisc.peter@chim.upt.ro (F. Péter).
1381-1177/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2011.03.004

S. Kakasi-Zsurka et al. / Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
23
have been obtained by a chemoenzymatic method, but only the 6-
acryloyl sugar monomer was synthesized using a lipase-catalyzed
biocatalytic procedure, the polymerization being performed in tra-
ditional manner [9]. Vinyl sugar esters suitable for polymerization
reactions were also obtained by regioselective transesterification
in the 6 position of several hexoses with vinyl adipate in dimethyl-
formamide, catalyzed by an alkaline protease [10].
chemical catalysis. Tetra-O-acetyl-d-gluconolactone, reacted with
1,4-butanediol in the presence of a metal alcoxide initiator, yielded
a low molecular weight oligoester [20]. A structurally different
d-gluconolactone derivative, subjected to ROP in the presence
of stannous butyrate initiator, yielded polyesters with number-
average molecular weights ranging from 1800 to 7300 Da [21].
After our best knowledge, enzyme-catalyzed ROP of carbohy-
drate lactones or incorporation of non-derivatized carbohydrate
units in polyhydroxybutyrate structures was not yet reported by
other groups. Therefore, the target of this work was to include
gluconolactone-derived units in the enzymatically synthesized
oligomers or polymers of 3HBA.
Enzyme-catalyzed direct inclusion of a sugar-derived moi-
ety in the polymer backbone is not an easy task, taking into
account the difficulty to find and enzyme having specificity for
both monomers, or the very low solubility of non-derivatized sug-
ars in the solvents normally employed as media for biocatalytic
polycondensation or ROP reactions. Good results were reported
with protected sugars, like for lipase-catalyzed acroylation of
a sugar derivative with protected anomeric position, followed
by ROP of ␧-caprolactone initiated by the acryl-sugar interme-
diate, in toluene [11]. ROP of ␧-caprolactone was reported to
be induced by the hydroxyl group in the 6-position of methyl-
glucopyranoside, at 60 ◦C, without solvent, with Novozyme 435
lipase as the catalyst [12]. Sugar-containing polymers were also
synthesized by enzymatic polycondensation reactions, but acti-
vation of the employed ester co-monomers was necessary to
shift the reaction equilibrium toward the polymer product. Poly-
condensation of sucrose with bis(2,2,2-trifluoroethyl)-adipate was
catalyzed by an alkaline protease in anhydrous pyridine, yielding a
polyester that contained ester linkages at 6 and 1ꢀ positions of the
sucrose [13].
(R)-3-Hydroxybutyric acid (3HBA), derived from poly[(R)-3-
hydroxybutyrate] (P3HB), could be a valuable monomer for the
synthesis of new biodegradable oligomers and polymers. Although
microorganisms synthesize polymers and copolymers of 3HBA, the
monomer could be also obtained via an enzymatic process. Produc-
tion of chiral 3-hydroxycarboxylic acids directly in the culture broth
by in vivo depolymerization of the corresponding polyhydroxyalka-
noates was already demonstrated [14]. P3HB can be degraded by
various intracellular and extracellular depolymerases [15]. Using a
thermophilic Streptomyces sp. MG with strong hydrolytic activity
for depolymerization of PHB, the process was considered econom-
ically feasible, even in terms of 3HBA recovery [16]. As a monomer,
3HBA could be a precursor for the synthesis of pure biodegrad-
able P3HB with desired molecular weight, or for the synthesis of
different copolyesters.
Lipase-catalyzed ROP of lactones, mainly ␧-caprolactone,
and polycondensation of hydroxy acids, like lactic or ricinoleic
acid, was subject of numerous investigations in the past decade
[4]. By contrast, literature data about reactions involving 3HBA
and ␤-butyrolactone are scarcer. Lipase-catalyzed ROP of (R)-␤-
butyrolactone was used to obtain optically active linear and cyclic
P3HBs, with molecular weights centered around 2100 Da for the
cyclic and 3200 Da for the linear forms, when 5% porcine pancreatic
lipase was employed as catalyst at 100 ◦C, for 24 h in a solventless
system [17]. Porcine pancreatic lipase was used for polymeriza-
tion of 3HBA in organic solvents, yielding oligomers [18]. Ionic
liquids could be an alternative for organic solvents as reaction
media for the synthesis of polyhydroxyalkanoates by ROP or
polycondensation reactions. Candida antarctica B lipase catalyzed
the ROP of several lactones in 1-butyl-3-methylimidazolium
bis(trifluoromethane) sulfonimide ([Bmim]Tf2N) at 60 ◦C
[19].
Carbohydrates are attractive as highly functionalized renewable
monomers for the synthesis of various polymers by lipase-
catalyzed condensation polymerization reactions, but they require
utilization of protected alditols or activated aldaric esters or
chlorides. Therefore, ROP polymerization of carbohydrate 1,5-
lactones emerged as an easier approach to obtain aliphatic
polyesters, but their polymerization was reported until now only by
2. Experimental
2.1. Materials
(R,S)-3-Hydroxybutyric acid (3HBA), d-glucono-␦-lactone (GL),
tert-butanol (t-BuOH, ∼99% pure), 1-butanol (>99% pure), butyric
acid (>99% pure), isooctane (>99% pure), toluene (>99%), dimethyl-
sulfoxide (DMSO, ∼99.7% pure), 1-butyl-3-methylimidazolium
hexafluorophosphate ([Bmim][PF6]), were purchased from Merck.
Optically active (R)-3-hydroxybutyric acid (R3HBA, >98%) was
a product of Aldrich. Immobilized Candida antarctica lipase B
on acrylic resin (Novozyme 435) and Rhizomucor miehei lipase
(Lipozyme-RM IM) were from Novozymes, Burkholderia cepacia
(Amano PS) and Pseudomonas fluorescens (Amano AK) lipases from
Aldrich, lyophilized Candida antarctica lipase B (CALB-Lecta) from
C-Lecta (Leipzig, Germany), Aspergillus niger and porcine pancreatic
˚
lipase (PPL) from Sigma. Molecular sieves (4 A) were provided by
Acros Organics.
2.2. Methods
2.2.1. Lipase-mediated esterification of 3HBA
Different lipases have been tested to select the suitable enzyme
for the forthcoming polymerization reactions. The reactions were
performed in a 5 mL glass vials containing 3HBA/1-butanol (1:2
molar ratio) in 2 mL isooctane, and 25 mg of enzyme. The mix-
ture was stirred using an orbital shaker (MIR-S100, Sanyo, Japan)
at 300 strokes/min and 40 ◦C (ILW 115 STD incubator, Pol-Eko-
Aparatura, Poland). After 24 h reaction time, samples were drawn
out for gas chromatographic analysis.
2.2.2. Polymerization in organic solvents
3HBA (0.465 mL, 5 mmole) and Candida antarctica lipase B
(50 mg, Novozyme 435) were added to GL (0.178 g, 1 mmole) dis-
solved in 2 mL organic medium. The reactions were performed in
5 mL Micro Reactions Vessels, under argon atmosphere, with acti-
˚
vated 4 A molecular sieves, magnetically stirred at 300 rpm and
80 ◦C. The reactions were stopped by filtration of enzyme. The poly-
mers were isolated by precipitation into methanol (10:1, v/v), and
centrifuged at 4 ◦C and a relative centrifugal force of 5000 × g for
60 min. The supernatant was removed and the resulting precipi-
tate was washed twice with methanol. The precipitated polymer
was dried in vacuum at 60 ◦C, resulting in a white solid.
2.2.3. Polymerization in ionic liquid
The same protocol as for reactions in organic solvents was
applied, excepting the use of 1 mL IL instead of organic solvent.
The products were isolated from the IL by extraction with toluene,
evaporation of toluene, and vacuum-drying at 60 ◦C.
2.2.4. Polymerization in bulk
50 mg Novozyme 435 and 178 mg GL were added to 0.465 mL 3-
HBA placed in a 4 mL vial. The reaction was carried out in the same

24
S. Kakasi-Zsurka et al. / Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
Fig. 1. Biocatalytic route for the synthesis of sugar-containing copolymers of 3-HBA based on renewable resources.
conditions as in organic solvents. After extraction with chloroform,
the reaction product was worked up by removal of the enzyme by
filtration, evaporationof theextractionsolvent, andvacuum-drying
at 60 ◦C.
3. Results and discussion
3HBA obtained from microbial polyhydroxyalkanoates can be
a valuable starting material for biodegradable copolymers with
sugar derivatives, obtained entirely from easily available renew-
able resources by biocatalytic pathways. In the present study, we
used GL as non-derivatized co-monomer, but other derivatives of
sugars, like lactobionic acid, could be also appropriate (Fig. 1).
2.3. Product characterization
2.3.1. Gas chromatography
Gas chromatography (GC) analysis of 3HBA esterification
products was performed by a Dani 86.10 (Dani S.p.A., Italy)
chromatograph equipped with flame ionization detector, and a
15 m × 0.32 mm BPX-5 capillary column (SGE, Australia). The anal-
ysis conditions were set as follows: oven temperature: 70–250 ◦C
with 10 ◦C/min heating rate, injector temperature 300 ◦C, detec-
tor temperature 350 ◦C, carrier gas (helium) flow 1.45 mL/min. The
conversions were calculated based on the internal standard method
(n-dodecane was used as standard) using a calibration curve for 1-
butyl-3-hydroxybutyrate, obtained and purified in our laboratory.
The total conversion of 3HBA was also assayed by GC. A filtered
sample (20 ␮L) from the reaction mixture was derivatized by addi-
tion of 20 ␮L N,O-bis(trimethylsilyl)-trifluoracetamide, incubated
at 80 ◦C for 1 h, and analyzed in the same conditions as for the butyl
esters.
3.1. Screening of lipases as 3-hydroxybutyric acid esterification
catalysts
The catalytic efficiency of different microbial lipases, as well
as porcine pancreatic lipase, was tested in the esterification reac-
tion of 3HBA with 1-butanol, in organic solvent medium (Table 1).
Control reactions were run with n-butyric acid, in the same con-
ditions, only for the most efficient enzymes. The presence of a
3-hydroxy substituent in the carboxylic acid molecule was obvi-
ously detrimental to the enzyme esterification activity. The ester
yields obtained at 24 h reaction time were low for most of the
tested enzymes. The best performing lipase, Novozyme 435, also
showed significantly lower esterification activity for 3HBA than for
the unsubstituted butyric acid. Gorke et al. [19] reported an 11%
faster reaction rate for butyric acid than for 3HBA in the esterifica-
tion reaction with both 1- and 2-butanol, in ionic liquid medium,
catalyzed by Novozyme 435 lipase. They suggested that 3HBA fits
poorly in the acid binding site of lipase, compared to butyric acid.
The slower reaction rate observed for 3HBA could be explained by
the more complicate reaction mechanism than in case of a linear
carboxylic acid, as well. In a previous study of 3HBA esterification
with different alcohols [23], we demonstrated using GC–MS data
that polymerization of the 3HBA monomer and esterification of
the formed acid oligomers occurred competitively with the nor-
mal esterification reaction. Even if the esters of acid dimer and
trimer subsequently undergone transesterification reactions with
the alcohol, yielding finally the same 3HBA ester product, the over-
2.3.2. FT-IR spectroscopy
Fourier Transform Infrared Spectroscopy (FT-IR) spectra were
obtained from a JASCO FT/IR 430 spectrometer (JASCO Corp., Japan),
on 500–4000 cm−1 range at 4 cm−1 spectral resolution. Solid sam-
ples were prepared as KBr pellets and liquid samples as films, using
KBr windows.
2.3.3. MALDI-TOF MS analysis
Matrix-assisted laser desorption/ionization time of flight mass
spectrometry (MALDI-TOF MS) analysis of products was car-
ried out using Bruker BIFLEX III matrix assisted laser desorption
ionization time-of-light mass spectrometry (Bruker Daltonik
GmbH, Germany) at an acceleration voltage of 20 kV using 2,5-
dihydroxybenzoic acid as matrix. Samples (10 mg/mL), the matrix
(20 mg/mL) and the ionization agent NaTFA (5 mg/mL) were indi-
vidually dissolved in organic solvent. 10 ␮L of sample solution,
5 ␮L of ionization agent solution were mixed with 50 ␮L of
matrix solution and an aliquot (0.3 ␮L) was applied to the sample
plate.
Within one MALDI-TOF spectrum, the intensities of all signals
(originated from sodium adduct ions) were added and the total
set as 100%. Separately, a sum of signal intensities was calcu-
lated for every type of oligomer present in the analyzed product.
We calculated the relative composition of the reaction product,
assuming that the ionization efficiency was the same for every
oligomeric type present in the mixture. Such estimations have
been already used to calculate the relative distribution of fructose
oligosaccharide-lauryl ester oligomers [22].
Table 1
Screening of lipases for catalytic efficiency in esterification reactions of racemic and
optically active 3-HBA. The reactions have been carried out at 40 ◦C, with 3-HBA/1-
butanol at 1:2 molar ratio, in isooctane.
Lipase
Acid
Reaction
time (h)
Ester
yield (%)
Candida antarctica B (Novozyme 435)
Candida antarctica B (Novozyme 435)
Candida antarctica B (C-Lecta)
Mucor miehei (Lipozyme-RM IM)
Burkholderia cepacia (Amano PS)
Pseudomonas fluorescens (Amano AK)
Candida cylindracea
Porcine pancreatic lipase
Aspergillus niger
Candida antarctica B (Novozyme 435)
Mucor miehei (Lipozyme-RM IM)
3HBA
R3HBA
3HBA
3HBA
3HBA
3HBA
3HBA
3HBA
3HBA
24
24
24
24
24
24
24
24
24
3
52
27
42
2
<1
<1
9
<1
<1
98
35
Butyric acid
Butyric acid
3

S. Kakasi-Zsurka et al. / Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
25
Fig. 2. Reaction scheme of the enzymatic polymerization reaction. The equations indicate the masses of the corresponding oligomers, m and n meaning the numbers of GL
and HBA repeat units, respectively. 178 and 86 are the masses of the GL and HBA units, respectively, and 18 is the mass of the end-group H2O.
all process was slower. 3-Hydroxybutyric acid is a chiral compound,
and the lipase enantioselectivity could also influence the ester
yields. However, carrying out the same reaction with the best per-
forming enzyme and optically active (R)-3-hydroxybutyric acid as
substrate, the obtained yield was lower than in the case of racemic
acid (Table 1), demonstrating that not enantioselectivity was the
limiting factor of this reaction. Among the tested lipases, Novozyme
435 has proven the highest esterification activity and was further
employed as polymerization catalyst.
the regioselectivity of lipases in the esterification of sugars and their
ability to catalyze ring-opening polymerization of lactones.
An easily available carbohydrate, 1,5-lactone, d-glucono-␦-
lactone (GL), was selected as sugar derivative. The insertion of GL
in the polyester chain of P3HB can be accomplished in two possi-
ble ways, yielding linear (a), or cyclic (b) copolymers, together with
possible homopolymers (c) of 3HBA (Fig. 2).
In the typical experimental protocol, GL was dissolved or sus-
pended in the reaction medium, 3HBA, enzyme, and molecular
sieves added, and the reaction mixture stirred for several days in
argon atmosphere. The product was isolated after removing the
enzyme and molecular sieves by filtration.
3.2. Lipase-catalyzed copolymerization of 3HBA and GL
Infrared spectroscopy was used to monitor the functional group
changes during the polymerization reaction. The FT-IR spectra of
the raw materials and a reaction product are shown in Fig. 3. The
absorption bands in the FT-IR spectrum of the product confirm the
Insertion of sugar-derived polyhydroxylic units in the P3HB
backbone is difficult by traditional chemical methods, but looks
more reliable using a biocatalytic route. We investigated the lipase-
catalyzed synthesis of such oligomeric compounds, to exploit both
Fig. 3. FT-IR spectra of (a) 3HBA-GL reaction product, obtained at 7 days reaction time in [Bmim]PF6 reaction medium, (b) pure 3HBA and (c) pure GL. Inset the expanded
view of 1600–2200 cm−1 region, with the stretching vibration band of C O group.

26
S. Kakasi-Zsurka et al. / Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
Fig. 4. MALDI-TOF MS spectrum of copolymerization reaction product of 3HBA and GL (at 10:1 molar ratio), catalyzed by Novozyme 435 at 80 ◦C in bulk, at 7 days reaction
time. Inset the expanded view of the 710–850 m/z region.
ester formation, by shifting the band assigned to the carbonyl group
stretching vibration from 1724 cm−1 in 3HBA and GL, to 1734 cm−1
in the product. In the special case of reaction accomplished in IL
medium, the FT-IR spectrum also indicates that only the 3HBA
homopolymer was obtained, by the important decrease of the
hydroxyl group absorption band intensity in the 3300–3400 cm−1
region.
The products were identified by MALDI-TOF mass spectrometry.
Fig. 4 shows a typical MALDI-TOF mass spectrum of the reaction
product obtained by copolymerization of 3HBA and GL in bulk,
using Novozyme 453 at 80 ◦C. The spectrum contains a character-
istic series of peaks, which can be assigned to the sodium adduct
ions of different oligomer series ([M+Na]+).
organic solvents that are particularly suitable for lipase-catalyzed
synthetic reactions. Therefore, the copolymerization of 3HBA and
GL should be strongly influenced by the solubility of raw materials
(mainly GL) and synthesized oligomers (Table 2).
Based on the literature data reported for comparable reaction
systems, we investigated several media for the copolymeriza-
tion reactions: toluene, mixtures of t-BuOH and DMSO, the IL
[Bmim]PF6, as well as the solventless system. It is likely to presume
that even a smaller amount of dissolved sugar might be enough to
ensure the progress of reaction, if the formed oligomeric or poly-
meric product is soluble. 3HBA has much better solubility in organic
solvents, therefore only the sugar derivative solubility looks restric-
tive for the reaction.
According to Fig. 4, linear (as shown in Fig. 2a) and cyclic (Fig. 2b)
oligomers containing inserted GL unit, as well as P3HB homopoly-
mers (Fig. 2c) were found in the MALDI-TOF MS spectrum. For
example, the peak at m/z 735.2 corresponds to the sodium adduct
ion of a linear oligomer with n = 6 and m = 1 (i.e., one inserted GL
unit). In addition, P3HB homopolymer series were also found, e.g.
the peak at m/z 729.2 is due to the presence of oligomer with n = 8
and m = 0. Furthermore, the presence of cyclic oligomers with one
inserted GL unit can also be recognized, e.g. the Na adduct of the
cyclic oligomer with n = 6 and m = 1 was visible at m/z 717.2. Inter-
estingly, [M−H + 2Na]+ adducts of linear homopolymers are also
present in the MALDI-TOF MS spectra at a 22 m/z distance from the
P3HB peaks.
All experiments were carried out in the same conditions: 80 ◦C,
argon atmosphere, magnetic stirring, molecular sieves for removal
of water by-product, and 7 days reaction time. The number-average
and weight-average molecular weights were calculated for the
entire reaction product, based on MALDI-TOF MS analysis, as indi-
cated in Section 2. The conversion of 3HBA was calculated based on
unreacted acid, determined by GC analysis. The copolymerization
reaction was not significantly influenced by the reaction medium
in regard of average polymerization degree (Table 1). Although
the 3HBA total conversions (in both copolymer and homopoly-
Table 2
Influence of reaction medium on 3HBA total conversion and average molecular
weight of the product, in the copolymerization reaction with GL catalyzed by
Novozyme 435 lipase.
Based on these results, it can be demonstrated that copolymer-
ization of 3HBA with a nonderivatized sugar lactone is possible
by lipase catalyzed reaction, yielding oligomers with inserted
2,3,4,5,6-pentahydroxy-caproic acid units, derived from the GL.
Reaction medium
3HBA total
Mna
Mwb
PDIc
conversion (%)
Solventless
Toluene
t-BuOH/DMSO, 80/20 (v/v)
[Bmim]PF6
74.6
83.3
>99
451.3
394.5
384.3
469.1
504.7
434.9
422.7
565.8
1.12
1.10
1.10
1.20
3.3. Influence of reaction medium
>99
The reaction medium has an essential role in biocatalytic pro-
cesses. A major drawback of lipase-catalyzed biotransformation
of polyhydroxylic compounds is their low solubility in nonpolar
a
Number average molecular weight.
Weight average molecular weight.
Polydispersity index.
b
c

S. Kakasi-Zsurka et al. / Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
27
Fig. 5. Relative composition 3-hydroxybutyric acid copolymers with ␦-
gluconolactone (10:1 molar ratio), catalyzed by lipase from Candida antarctica B
(Novozyme 435), 7 days, at 80 ◦C, in bulk and in various reaction media.
mer) were high, the highest polymerization degree did not exceed
10, even for the homopolymer. These results agree with findings
of other groups for reactions of 3HBA and ␤-butyrolactone. In
an earlier report concerning enzymatic synthesis of P3HB from
3HBA, Shuai et al. [18] reported number-average molecular weights
of 660 and 720 in hexane and diethyl ether, respectively, but
these solvents are not appropriate for copolymerization with GL.
Studying enzymatic polymerization of several lactones in IL media,
Gorke et al. [19] obtained the lowest DP (up to 5) in case of ␤-
butyrolactone. The explanation of the relatively low DP could be the
not accurate structural conformity between the 3HBA molecule and
the active center of lipase, or the reduced solubility of the higher
DP oligomer in the reaction medium.
Hindrance of the enzyme activity by insufficiency of water could
beanother possible causeoflimitationofmolecular weightincrease
of the product. A more accurate water control throughout the reac-
tion will be realized in the forthcoming studies, in order to increase
the polymerization degree.
Fig. 6. Relative composition of the reaction mixture after incubation of 3HBA and GL
with Novozyme 435 lipase for 7 days, at 80 ◦C, (a) in bulk, and (b) in t-BuOH/DMSO
(80/20, v/v) reaction medium, calculated from MALDI-TOF MS spectral data at dif-
ferent polymerization degrees (DP).
An emerging question is the influence of reaction medium on
the obtained copolymer composition. Such information will help
us to fine tune this composition in the desired direction: a lin-
ear copolymer of 3HBA and RO-GL, P3HB end-functionalized with
a sugar unit, cyclic copolymer, etc. The estimation of the relative
composition of the product was possible based on MALDI-TOF MS
data, as described in Section 2. Between the two possible structures
having the same molecular mass (P3HB end-functionalized with a
GL unit and cyclic copolymer with inserted RO-GL) we assumed
the formation of cyclic oligomers, as cyclization results in lower
polymerization degree, unless the cyclization product itself is a
substrate for the lipase. As only MALDI-TOF data cannot prove this
hypothesis, NMR analysis will be performed in our forthcoming
studies to elucidate the exact structure of all obtained oligomers.
As shown in Fig. 5, the reaction medium had important influ-
ence on product composition. In all reaction media employed, an
important part (at least 80%) of the initial 3HBA was recovered
in the products as P3HB, caused probably by the higher reaction
rate of homopolymer formation and the important 3HBA excess
used.
Toluene could be an appropriate organic medium for lipase-
catalyzed polymerization reactions, as previously demonstrated for
copolymerization of BL with 12-hydroxydodecanoic acid [24]. Uti-
lization of this solvent in the studied process resulted in a very high
relative amount of cyclic product.
Sugars and sugar derivatives are soluble only in highly polar
solvents as DMSO, pyridine, or dimethylformide [25], but these
solvents can lead to an important decrease of lipase activity [26].
To avoid such inactivation, mixtures of DMSO with t-BuOH were
used in the enzymatic synthesis of sugar esters [22]. In the present
study, an organic medium composed of t-BuOH and DMSO in 80/20
(v/v) ratio was the most appropriate reaction medium, yielding
more than 70% relative amount of ring-opened form in the obtained
copolymer.
Another option to increase homogeneity of the reaction system
could be the utilization of ILs with an imidazolium cation, as they
have polarities on the Reichard’s scale in the range of the most polar
organic solvents [27]. Among several ILs investigated, we detected
polymerization products only in [Bmim]PF6 reaction medium, but
almost exclusively homopolymers of 3-HBA, and mainly in the
cyclic form. The explanation is most likely related to the solubility of
GL. Gorke et al. [19] reported that water-immiscible ILs were supe-
rior to water-miscible ILs, as reaction media in the ROP of lactones.
Although [Bmim]PF6 is water-immiscible, its solvation capacity for
GL was probably low and the polymerization reaction was effective
only for 3HBA. An investigation of a much larger spectrum of ILs will
be necessary to exploit the advantages of such reaction media.
In a solventless system, in which excess 3HBA has the role
of reaction medium, the obtained copolymer was formed mainly
from cyclic oligomers, linear oligomers with inserted RO-GL rep-
resenting about one third of the product amount. Formation of
cyclic oligomers as major product was also reported by Mat-
sumura et al. [17] for lipase-catalyzed ROP of (R,S)-␤-butyrolactone,
in bulk.

28
S. Kakasi-Zsurka et al. / Journal of Molecular Catalysis B: Enzymatic 71 (2011) 22–28
Based on the appropriate MALDI-TOF MS spectral data, the
Scientific Research Development, Hungary), and TAMOP 4.2.1./B-
09/1/KONV-2010-0007.
relative composition of the obtained product at different polymer-
ization degrees was also calculated, as showed in Fig. 6. In bulk
(Fig. 6a), they were no significant differences between the relative
ratio of copolymer and homopolymer at different DP. Oligomers
with DP in the range 4–5 have been detected at highest content in
the product, in accordance with the calculated average molecular
weight. In t-BuOH/DMSO (Fig. 6b), the majority of oligomers with
RO-GL were found at DP3, and the overall polymerization process
showed a slower increase of the obtained molecular mass than in
the bulk system. In both cases, the increase of the copolymer DP was
achieved mostly by addition of a 3HBA repetitive unit to the pre-
vious oligomer, as we identified only small amounts of oligomers
with more than one RO-GL included.
References
[1] W. Soetaert, E. Vandamme, Biotechnol. J. 1 (2006) 756–769.
[2] S. Kobayashi, H. Uyama, S. Kimura, Chem. Rev. 101 (2001) 3793–3818.
[3] S. Kobayashi, Proc. Jpn. Acad. Ser. B 86 (2010) 338–365.
[4] S. Kobayashi, Macromol. Rapid Commun. 30 (2009) 237–266.
[5] A.-C. Albertsson, R.K. Srivastava, Adv. Drug Deliv. Rev. 60 (2008) 1077–1093.
[6] R.A. Gross, M. Ganesh, W. Lu, Trends Biotechnol. 28 (2010) 435–443.
[7] D. Feder, R.A. Gross, Biomacromolecules 11 (2010) 690–697.
[8] S. Sato, M. Minato, Y. Kikkawa, H. Abe, T. Tsuge, J. Chem. Technol. Biotechnol.
85 (2010) 779–782.
[9] B.D. Martin, S.A. Ampofo, R.J. Linhardt, J.S. Dordick, Macromolecules 25 (1992)
7081–7085.
[10] M. Kitagawa, H. Fan, T. Raku, S. Shibatani, Y. Maekawa, Y. Hiraguri, R. Kurane,
Y. Tokiwa, Biotechnol. Lett. 21 (1999) 335–339.
[11] R. Kumar, R.A. Gross, J. Am. Chem. Soc. 124 (2002) 1850–1851.
[12] A. Cordova, T. Iversen, K. Hult, Macromolecules 31 (1998) 1040–1045.
[13] D.R. Patil, D.G. Rethwisch, J.S. Dordick, Biotechnol. Bioeng. 37 (1991) 639–646.
[14] Q. Ren, K. Ruth, L. Thöny-Meyer, M. Zinn, Macromol. Rapid Commun. 28 (2007)
2131–2136.
[15] Y. Tokiwa, B. Calabia, Biotechnol. Lett. 26 (2004) 1181–1189.
[16] Y. Tokiwa, C.U. Ugwu, J. Biotechnol. 132 (2007) 264–272.
[17] S. Matsumura, Y. Suzuki, K. Tsukada, K. Toshima, Macromolecules 31 (1998)
6444–6449.
[18] X. Shuai, Z. Jedlinski, M. Kowalczuk, J. Rydz, H. Tan, Eur. Polym. J. 35 (1999)
721–725.
[19] J.T. Gorke, K. Okrasa, A. Louwagie, R.J. Kazlauskas, F. Srienc, J. Biotechnol. 132
(2007) 306–313.
[20] A.F. Haider, C.K. Williams, J. Polym. Sci. A: Polym. Chem. 46 (2008) 2891–2896.
[21] M. Tang, A.J.P. White, M.M. Stevens, C.K. Williams, Chem. Commun. (2009)
941–943.
[22] R. ter Haar, H.A. Schols, L.A.M. van den Broek, D. Saglam, A.E. Frissen, C.G. Boeriu,
H. Gruppen, J. Mol. Catal. B: Enzym. 62 (2010) 183–189.
[23] S. Kakasi-Zsurka, C. Zarcula, L. Trasca, F. Peter, Ann. West Univ. Timisoara Ser.
Chem. 17 (2008) 67–72.
4. Conclusions
Insertion of a sugar-derived monomeric unit in the P3HB chain
induces new functionalities and could be a route to bio-based poly-
mers with possible industrial or medical applications.
For the first time, copolymerization of 3HBA with GL was
achieved in a lipase-catalyzed reaction, yielding oligomers con-
taining inserted 2,3,4,5,6-pentahydroxy-caproate units. Novozyme
435 lipase demonstrated the highest esterification activity for the
3HBA substrate, while the highest selectivity toward formation of
copolymers with inserted RO-GL was obtained in a t-BuOH/DMSO
(4/1, v/v) reaction medium. Extension of this novel polyesterifica-
tion reaction to copolymers with inserted lactobionic acid units is
currently in study.
[24] Z. Jedlinsky, M. Kowalczuk, G. Adamus, W. Sikorska, J. Rydz, Int. J. Biol. Macro-
mol. 25 (1999) 247–253.
Acknowledgements
[25] J.F. Kennedy, H. Kumar, P.S. Panesar, S.S. Marwaha, R. Goyal, A. Parmar, S. Kaur,
J. Chem. Technol. Biotechnol. 81 (2006) 866–876.
[26] S. Park, R.J. Kazlauskas, J. Org. Chem. 66 (2001) 8395–8401.
[27] C. Reichardt, Green Chem. 7 (2005) 339–351.
Financial support of this work was provided in Romania by
CNCSIS-UEFISCDI through PNII-IDEI grant No. 368/2007, and in
Hungary by grants K-72524 given by OTKA (National Found for