Biocatalytic preparation of alkyl esters of citrus flavanone glucoside prunin in organic media

Article abstract of DOI:10.1016/j.procbio.2010.07.022

The biocatalytic preparation of alkyl esters of prunin, a flavanone glucoside derivative of citric flavonoid naringin, was performed in organic media using two commercial immobilized lipases: Novozym 435 (Candida antarctica lipase B immobilized on an acry

Full text of DOI:10.1016/j.procbio.2010.07.022

Process Biochemistry 46 (2011) 94–100
Contents lists available at ScienceDirect
Process Biochemistry
journal homepage: www.elsevier.com/locate/procbio
Biocatalytic preparation of alkyl esters of citrus flavanone glucoside prunin in
organic media
Gustavo Céliz, Mirta Daz∗
Instituto de Investigaciones para la Industria Química (CONICET), Facultad de Ciencias Exactas, Universidad Nacional de Salta, Avenida Bolivia 5150, A4408FVY Salta, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
The biocatalytic preparation of alkyl esters of prunin, a flavanone glucoside derivative of citric flavonoid
naringin, was performed in organic media using two commercial immobilized lipases: Novozym 435
(Candida antarctica lipase B immobilized on an acrylic resin) and Lipozyme RM IM (Rhizomucor miehei
lipase immobilized on an anion exchange resin). The influence of several reaction parameters on the per-
formance and regioselectivity of the biocatalytic process was investigated using four solvents (acetone,
acetonitrile, tetrahydrofurane and t-butanol) and vinyl laurate as a donor. The biocatalytic modification
of prunin was strongly dependant on the solvent, the molar ratio of the substrates, the enzyme and the
chain length of the donor. The acylation was highly regioselective, obtaining 6ꢀꢀ-O-acyl monoesters except
when the donor was vinyl acetate in which case a diester seemed to be formed. Novozym 435 proved
to be a more efficient biocatalyst than Lipozyme RM IM. In acetone and acetonitrile, derivatives with
a 100% conversion yield were synthesized after only 6 h of reaction at 50 ◦C. The modified derivatives
preserved the antiradical activity of the flavanone glucoside and their solubility notably increased in the
hydrophobic medium 1-octanol.
Received 8 January 2010
Received in revised form 16 June 2010
Accepted 22 July 2010
Keywords:
Lipases
Flavonoids
Flavanones
Prunin
Naringin
Citrus
© 2010 Elsevier Ltd. All rights reserved.
1. Introduction
substance in food industry, would leave prunin as a by-product.
Prunin could have applications in the pharmaceutical and food
There is a subclass of flavonoids in citrus fruits, called flavanone
glycosides. Due to their properties and biological activity, these
substances and some of their derivatives have important applica-
tions in the pharmaceutical and food industries [1–4].
The Argentine Northeast region is a citrus producer. In par-
ticular, the province of Salta is the main grapefruit producer
of Argentina. Both in the peal and immature aborted fruits
abundant levels of the flavanone glycoside naringin (4ꢀ,5,7-
trihydroxiflavanone-7-␤-d-␣-l-rhamnosyl(1→2)-␤-d-glucoside)
can be found.
In the food industry, naringin can be used as a flavouring sub-
stance to provide bitterness in foods like marmalades and tonic
water or grapefruit flavoured beverages. Besides these applica-
tions, other valuable derivatives of naringin can be obtained. In this
laboratory a very efficient biocatalytic method was developed to
obtain rhamnose from supersaturated solutions of naringin [5]. The
enzymatic hydrolysis of naringin catalysed by ␣-rhamnosidases
produces rhamnose and prunin (4ꢀ,5,7-trihydroxiflavanone-␤-d-
glucoside). The potential manufacture of rhamnose, a valuable
industries on its own due to its biological activities [6–12]. How-
ever, its application could be limited by its low solubility in
lipophilic media.
A viable alternative is the lipase-catalysed process in order
to synthesize new lipophilic derivatives. By modifying the
hydrophilic/lipophilic character through the acylation of prunin,
new compounds could be obtained that may not only have differ-
ent solubility properties but also probably possess new biological
activities.
Various groups have reported the feasibility of the enzymatic
acylation of flavonoid glycosides catalysed by lipases and proteases
in organic media [13–15], ionic liquids [16,17] and in highly con-
centrated homogeneous solutions [18]. The enzymatic acylation of
naringin in particular, has been studied by various authors [15–30].
However, to our knowledge, enzymatic synthesis of prunin esters
has not been described as yet.
In this work, the enzymatic acylation of prunin was stud-
ied using alkyl vinyl esters with different chain lengths as
donors and Candida antarctica lipase
B (Novozym 435) and
Rhizomucor miehei lipase (Lipozyme RM IM) as biocatalysts. Purifi-
cation processes, antiradical activities and solubility measurements
in water and 1-octanol of the obtained compounds are also
described.
∗
Corresponding author. Tel.: +54 387 4255364; fax: +54 387 4251006.
E-mail address: mdaz@unsa.edu.ar (M. Daz).
1359-5113/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.procbio.2010.07.022

G. Céliz, M. Daz / Process Biochemistry 46 (2011) 94–100
95
2. Materials and methods
11.7 Hz, H-6bꢀꢀ), 2.27 (2H, m, H␣ to C O), 1.52 (2H, m, H␤ to C O), 1,21 (XH, m, CH2),
0.85 (3H, t, terminal CH3).
13C (125.76 MHz): ı 80.0 (d, C-2), 43.9 (t, C-3), 197.7 (s, C-4), 163.4 (s, C-5), 98.8
(d, C-6), 165.4 (s, C-7), 96.9 (d, C-8), 163.0 (s, C-9), 103.9 (s, C-10), 129.1 (s, C-1ꢀ),
128.6 (d, C-2ꢀ, C-6ꢀ), 116.0 (d, C-3ꢀ, C-5ꢀ), 157.7 (s, C-4ꢀ), 100.2 (d, C-1ꢀꢀ), 73.5 (d, C-2ꢀꢀ),
76.7 (d, C-3ꢀꢀ), 70.9 (d, C-4ꢀꢀ), 75.1 (d, C-5ꢀꢀ), 64.3 (t, C-6ꢀꢀ), 35.0 (t, C␣ to C O), 27.7 (t,
C␤ to C O), 174.8 (C O).
2.1. Materials
˚
Molecular sieves (sodium aluminosilicate) with pore diameter 4 A, DPPH and
naringenin (4ꢀ,5,7-trihydroxiflavanone) were from SIGMA (USA). Vinyl acetate, vinyl
butyrate and vinyl laurate were from Fluka (USA). Vinyl decanoate and vinyl stearate
were from Aldrich (Germany). All other reagents were analytical grade. Naringin was
obtained according to the method described by Geronazzo et al. [31] from imma-
ture aborted grapefruits. Prunin was obtained by enzymatic hydrolysis of naringin
according to the method described by Soria and Ellenrieder [32]. Novozym 435 (C.
antarctica lipase B immobilized on an acrylic resin) and Lipozyme RM IM (R. miehei
lipase immobilized on an anion exchange resin) were a gift from Novozymes Latin
America Limited (Brazil).
2.7. Radical scavenging activity
The antioxidant activities of prunin and prunin esters were determined by its
ability to scavenge DPPH [33]. 50 ␮l of the solutions of the samples in DMSO at differ-
ent concentrations were each added to 950 ␮l of DPPH 60 ␮M methanolic solutions.
The blank sample consisted of 50 ␮l of DMSO added to 950 ␮l of 60 ␮M DPPH. The
absorbance was measured at 518 nm as a function of the time taken for the reaction
to reach the steady state. The tests were carried out in triplicate. The results of the
assays are presented as percentages of the remaining DPPH, calculated as follows: %
DPPH = [DPPH]ss/[DPPH]o) × 100, where ([DPPH]ss was the concentration of DPPH
at steady state and [DPPH]o was the initial concentration.
2.2. Synthetic reactions
Water was removed from the solvents and vinyl esters used in the synthesis
by adding molecular sieves (30 mg ml−1), shaking the resulting mixture for 24 h at
room temperature and keeping them in the presence of the adsorbent.
60 ␮mol of prunin or naringin and 200 mg of molecular sieves were placed in
a screw-capped vial with a Teflon seal. 3 ml of an appropriate mixture of vinyl
ester and solvent were added and the resulting reaction medium was incubated
with gentle shaking at the chosen temperature. The reaction was started by adding
50 mg of immobilized enzyme and run with shaking (200 rpm). At regular time inter-
vals, 50 ␮l aliquots were extracted and appropriate dilutions of these samples were
analysed by HPLC. All experiments were carried out in duplicate.
2.8. Water and 1-octanol solubilities
In order to determine the solubility of prunin and prunin esters in water and
1-octanol, saturated solutions were prepared by equilibrating the compounds at
20 ◦C overnight. After centrifugation, the concentration of each compound in the
supernatant was measured by HPLC.
2.9. Data analysis
2.3. Analytical methods
Statistical analysis and graphics were carried out using GraphPad Prism
version 4.00 for Windows, GraphPad Software, San Diego, California, USA,
www.graphpad.com.
Substrate and product concentrations were determined by HPLC using a RP-
8 LiChrospher 100 Merk column (25 cm long, 4 mm internal diameter and 5 mm
particle size). Samples (20 ␮l) containing prunin and prunin esters were eluted iso-
cratically with the following mobile phases and flow-rates: (a) prunin acetates,
50/50 acetonitrile/water at 0.6 ml min−1
; (b) prunin butyrate, 65/35 acetoni-
3. Results and discussion
trile/water at 0.7 ml min−1; (c) prunin decanoate, prunin laurate and prunin stearate,
80/20 acetonitrile/water at 0.8 ml min−1. Samples (20 ␮l) containing naringin and
naringin laurate were eluted with acetonitrile/water 80/20. Elution profiles were
monitored at 280 nm. Calibration curves for prunin, prunin esters, naringin and
naringin laurate were obtained using purified samples. Conversion yields were
expressed by the molar ratio of ester formed to flavanone glucoside (prunin or
naringin) introduced at the beginning of the reaction. Initial rates were calculated
from the slope of the linear portion of plots of the product concentration versus
time. Samples were taking at 10, 20, 30, 40 and 60 min.
The hydrolysis of naringin catalysed by ␣-rhamnosidase pro-
duces rhamnose and prunin (Fig. 1). Rhamnose is a very valuable
sugar in the food industry and with the aim of exploiting this reac-
tion to the maximum, an enzymatic method to produce lipophilic
derivates of prunin was investigated in this work.
From the commercially available immobilized lipases we
selected Novozym 435, C. antarctica lipase B (CALB) and Lipozyme
RM IM, R. miehei lipase (RML). Among the numerous acyl donors
used during transesterification, vinyl esters are the most efficient
since the vinyl alcohol formed during the process tautomerizes to
the low-boiling-point acetaldehyde, thus shifting the equilibrium
towards ester formation [34]. Because of this, vinyl esters were cho-
sen for this work. For studying the influence of different parameters,
vinyl laurate was selected as a donor because of its intermediate
chain length.
2.4. Lipase operational stability
Operational stability of Novozym 435 was tested by reusing the enzyme in
the synthesis of prunin laurate in various solvents. In each cycle of the assay,
20 mM prunin, 200 mM vinyl laurate, 67 mg ml−1 molecular sieves and 17 mg ml−1
Novozym 435 were incubated 6 h at 50 ◦C. After this time, the enzyme was sepa-
rated from the reaction medium, washed with the used solvent and added to a fresh
medium preheated at 50 ◦C. For each cycle, the conversion rate of prunin after 6 h
of reaction was measured.
2.5. Purification of prunin esters
3.1. Effect of the solvent
Prunin acetate, prunin butyrate, prunin decanoate, prunin laurate and prunin
stearate were obtained on a gram scale using Novozym 435 and acetone as the
solvent. In a typical experiment, 190 ml of acetone, 5 g of molecular sieves, 1.75 g
of prunin and the appropriate amount of vinyl ester to achieve a vinyl ester/prunin
molar ratio of 10 were placed into screw-capped recipients. The temperature was
set at 50 ◦C prior to starting the reaction with 1 g of Novozym 435. After achieving a
near 100% yield, the immobilized enzyme and the molecular sieves were separated
from the reaction medium by filtration and the acetone was recovered in a rotating
evaporator at 40 ◦C. The residual mixture was then washed three times with hexane,
centrifugating each time at 2000 rpm for 5 min. Prunin ester was obtained from the
solid residue by redissolving it in DMSO and evaporating the solvent at 50 ◦C. Purity
was checked by HPLC.
The time course of the synthesis of prunin laurate catalysed by
CALB and RML using different solvents at 50 ◦C, is shown in Fig. 2.
Solvents of intermediate hydrophobicity were chosen as reaction
media, such as acetonitrile (log P −0.33), acetone (log P −0.23),
tetrahydrofuran (log P 0.49) and t-butanol (log P 0.89). In all cases,
both prunin and the produced ester were completely solubilized.
The reactions were performed in the presence of molecular sieves
and, except the enzymes, all the reaction components were pre-
viously dried. Chromatograms of HPLC with detection at 280 nm
showed only 2 peaks, prunin and that of the reaction product which
always appeared in the same retention time, indicating that the
same product was obtained for all solvents and both enzymes. The
product was identified as prunin 6ꢀꢀ-O-laurate (Fig. 1; Section 4).
A notable reverse correlation between the solvents hydropho-
bicity, measured by the value of log P and the reaction rates,
was observed for both enzymes. This correlation has been found
in other reactions also catalysed by lipases [35–37]. The best
solvents were acetone and acetonitrile for both enzymes. Other
2.6. Prunin laurate structural identification
1H NMR and 2D NMR experiments were measured on a Bruker Avance DMX-
500 NMR spectrometer operating at 500.13 MHz (1H) and 125.76 MHz (13C) using
standard Bruker software.
1H (500.13 MHz, DMSO-d6): ı 5.33 (1H, m, H-2), 3.10 (1H, m, H-3-ax), 2.75 (1H,
m, H-3-eq), 6.17 (1H, sbr, H-6), 6,18 (1H, sbr, H-8), 7.27 (2H, d, 8.4 Hz, H-2ꢀ, H-6ꢀ), 6.83
(2H, d, 8.4 Hz, H-3ꢀ, H-5ꢀ), 4.88 (1H, dd, 5.0, 7.3 Hz, H-1ꢀꢀ), 3.50 (1H, dd, 7.5, 8.0 Hz,
H-2ꢀꢀ), 3.46 (1H, dd, 8.0, 8.5 Hz, H-3ꢀꢀ), 3.33 (1H, m, H-4ꢀꢀ, superposable to solvent
signal), 3.63 (1H, ddd, 2.0, 8.8, 9.2 Hz, H-5ꢀꢀ), 4.18 (1H, m, H-6aꢀꢀ), 4.42 (1H, dd, 2.0,

96
G. Céliz, M. Daz / Process Biochemistry 46 (2011) 94–100
Fig. 1. Scheme of enzymatic hydrolysis of naringin catalysed by ␣-rhamnosidase and enzymatic synthesis of prunin 6ꢀꢀ-O-acyl ester catalysed by lipase.
authors have also reported that acetone and acetonitrile are very
good reaction solvents in flavonoid glycoside esters synthesis
[21,23,26,28–29,36–38]. In both solvents, the total conversion of
prunin in its ester was achieved in just 6 h, when CALB was used as
the biocatalyst and with a conversion yield of 97,3% in acetone and
87,3% in acetonitrile in 24 h when using RML.
3.3. Effect of the vinyl laurate/prunin molar ratio
In order to find the minimum vinyl laurate/prunin molar ratio
necessary to achieve higher conversions in short reaction times, the
vinyl laurate/prunin molar ratio was tested from 1 to 40 at 50 ◦C
using CALB as the biocatalyst. The conversion yields obtained for
each solvent varying the molar ratio are shown in Fig. 3.
In acetone and acetonitrile, a molar ratio of 10 was enough to
achieve yields of almost 100% in just 6 h of reaction and no sub-
stantial improvement was observed by increasing the molar ratio.
However, in THF a molar ratio of 40 was necessary in order to reach
such a yield and a longer reaction time was needed (24 h). In t-
butanol, a yield of almost 80% was achieved in 24 h of reaction using
a molar ratio of 40.
3.2. Effect of the enzyme
Both CALB and RML were capable of catalysing the prunin
6ꢀꢀ-O-laurate synthesis with appreciable yields and reaction rates
in the four solvents. However CALB proved to be more efficient
than RML and all subsequent studies were performed using the
former.
While most of the authors who studied similar reactions used
CALB to synthesize flavonoid glycoside esters, only a few of them
used RML [16,22], reporting that CALB was a more effective bio-
catalyst than RML, in agreement with observed results in this
paper.
3.4. Structural identification of prunin laurate
The HMBC experiment for the only reaction product obtained
from prunin and vinyl laurate in the four tested solvents showed
Fig. 3. Effect of vinyl laurate/prunin molar ratio on the conversion yields of prunin
6ꢀꢀ-O-laurate synthesis catalysed by CALB (Novozym 435) after 6 h in acetonitrile (ꢀ)
and acetone (ꢁ), and after 24 h in tetrahydrofurane (᭹) and t-butanol (ꢂ). Condi-
tions: 20 mM prunin, 67 mg ml−1 molecular sieves and 17 mg ml−1 enzyme at 50 ◦C.
Data are expressed as the mean of duplicate assays. All standard deviations were
below 5%.
Fig. 2. Time course of prunin 6ꢀꢀ-O-laurate synthesis catalysed by CALB (Novozym
435) (——) and RML (Lipozyme RM IM) (- - - - -) in acetonitrile (ꢀ),acetone (ꢁ),
tetrahydrofurane (᭹) and t-butanol (ꢂ). Conditions: 20 mM prunin, 400 mM vinyl
laurate, 67 mg ml−1 molecular sieves and 17 mg ml−1 enzyme at 50 ◦C. Data are
expressed as the mean of duplicate assays. All standard deviations were below 5%.

G. Céliz, M. Daz / Process Biochemistry 46 (2011) 94–100
97
Fig. 4. Effect of acyl donor chain length on initial rates (white bars) and con-
version yields at 24 h (patterned bars) of prunin esters synthesis catalysed by
CALB (Novozym 435) in acetone. Conditions: 20 mM prunin, 100 mM vinyl ester,
67 mg ml−1 molecular sieves, 3.3, 6.7 and 17 mg ml−1 (white bars) and 17 mg ml−1
(patterned bars) enzyme at 50 ◦C. Data are expressed as the mean of duplicate
assays standard deviation.
Fig. 5. Time course of prunin acetate synthesis catalysed by CALB (Novozym 435)
in acetone. P1 (ꢃ), P2 (ꢁ) and P1+P2 (᭹). Conditions: 20 mM prunin, 100 mM vinyl
acetate and 17 mg ml−1 enzyme at 50 ◦C. Data are expressed as the mean of duplicate
assays. All standard deviations were below 3%.
correlations for carbonyl groups at ıC 174.8 with H-6ꢀꢀa and H-6ꢀb
for both enzymes, clearly indicating that the lauryl was attached
to C-6ꢀꢀ. These data unequivocally lead to the proposal that ester-
ification occurs at the 6ꢀꢀ-hydroxyl group of the sugar moiety
(Fig. 1). The acylation proceeded in a highly regioselective man-
ner.
posed acyl side of the active site pocket is more spacious than
the alcohol side [42]. Pleiss et al. [41] suggested that the lipases
have a long, hydrophobic scissile acyl binding site located inside
the binding pocket (at the wall of a binding funnel) and that
the length and hydrophobicity of this binding site should corre-
late with the acyl chain length profile of the respective lipase.
According to the binding site characteristics described by these
authors, CALB has a high activity on short and medium acyl donors
and a minor one when donors present an alkyl long chain. This
would explain the effect of the chain length observed in this
paper. However, the scissile acyl binding site is not the only
determinant of chain length specificity. The lipase activity and
specificity can be influenced by many factors such as water activ-
ity, solvent, substrates, immobilization support and donor/acceptor
molar ratio. The effect of the acyl donor chain length in synthesis
reactions of flavonoid glycoside esters has been investigated and
discussed in several papers [13–14]. The results obtained by vari-
ous authors showed some contradictions, which could be explained
by the differences in the substrates and the operational conditions
used.
Only one reaction product, which was expected to be prunin
6ꢀꢀ-O-acyl ester, was observed in all cases by HPLC, except in the
chromatograms of prunin acetate synthesis, where two products
eluted with different retention times (P1 and P2). The kinetic
profiles (Fig. 5) show that at the beginning of the reaction, the
concentration corresponding to P1 increases faster than that of P2,
followed by a decrease of the concentration of P1 accompanied by
a rise of P2. It is thought that prunin 6ꢀꢀ-O-acetate is formed first,
and given that the acyl group introduced in the prunin molecule
is small, the obtained derivative is still a substrate for the enzyme
and another hydroxyl of sugar moiety is acetylated. Given the small
size of the binding pocket, the esters with larger acyl groups are not
substrates for the enzyme.
The selectivity of CALB for the C-6ꢀꢀ of glucose from naringin has
been reported using a wide variety of donors [16,22,26]. The bind-
˚
ing pocket of CALB is an elliptical step funnel of 9.5 × 4.5 A wide
˚
and 12 A deep. This narrow and deep substrate binding pocket has,
compared to other lipases, very limited available space in the active
site pocket and so can be expected to exhibit a high degree of selec-
tivity [39–41]. The binding pocket of RML is a shallow bowl with a
˚
˚
˚
long axis of 18 A and a width of 4.5 A at its base and 6 A at the protein
surface [41]. Despite having a more spacious binding pocket, under
these experimental conditions RML has the same regioselectivity
that CALB.
3.5. Effect of acyl donor chain length
Prunin ester synthesis reactions were performed using lin-
ear acyl donors with different chain lengths (vinyl acetate, vinyl
butyrate, vinyl decanoate, vinyl laurate and vinyl stearate) in ace-
tone at 50 ◦C. Assays were performed with 3.3, 6.7 and 17 mg ml−1
of CALB enzyme. Given that when the chain length of the donor
increases its hydrophobicity does the same, a low vinyl ester/prunin
molar ratio was used in the assays to prevent excessive varia-
tions in the hydrophobicity of the media. Prunin and its esters
were solubilized in all the tested reaction mixture. The relation-
ship between initial rates and the amount of enzyme were linear.
The initial rates per mg of enzyme are shown in Fig. 4. In addition,
the conversion yields at 24 h with a quantity of enzyme fixed in
17 mg ml−1, are reported. It is clear that the longer the chain length,
the slower the initial reaction rate is. However, the yields achieved
after 24 h exceeded 80% regardless of the chain length of the donor
used.
The binding pocket of CALB consists of an elliptical and
steep funnel whose walls predominantly constitute hydrophobic
aliphatic amino acid residues. The active site is located at the
bottom of the pocket, which is a more hydrophilic region and
is constituted by the triad Ser 105 Asp 187 His 224. CALB has a
large hydrophobic surface surrounding the entrance of the active
site channel. The active site pocket can be partitioned into two
sides, an acyl side and alcohol side, where the corresponding
parts of the substrate will be located during catalysis. The pro-
These results are consistent with studies of the acylation of iso-
quercitrin, a flavonol glucoside, in experimental conditions similar
to this paper. Danieli et al. [37] studied the transesterification reac-
tion of isoquercitrin using vinyl acetate as a donor and Novozym
435 as the biocatalyst in acetone at 45 ◦C and reported the forma-
tion of 3ꢀꢀ,6ꢀꢀ-O-diacetate. Chebil et al. [36], using a donor/acceptor
molar ratio four times higher than that used by Danieli et al. [37],
obtained 3ꢀꢀ,6ꢀꢀ-O-diacetate and 2ꢀꢀ,3ꢀꢀ,6ꢀꢀ-O-triacetate and did not
observe monoacetate. Other authors have reported the acylation
of isoquercitrin with more bulky donors such as vinyl cinamate
[21–22], vinyl ferulate, vinyl coumarate [22] and several aro-

98
G. Céliz, M. Daz / Process Biochemistry 46 (2011) 94–100
Fig. 6. Time course of prunin 6ꢀꢀ-O-laurate (——) and naringin 6ꢀꢀ-O-laurate (- - - - -)
synthesis catalysed by CALB (Novozym 435) in acetonitrile (ꢀ), acetone (ꢁ), tetrahy-
drofurane (᭹) and t-butanol (ꢂ). Conditions: 20 mM prunin, 200 mM vinyl laurate,
67 mg ml−1 molecular sieves and 17 mg ml−1 enzyme at 50 ◦C. Data are expressed
as the mean of duplicate assays. All standard deviations were below 3%.
Fig. 7. DPPH remaining in the steady state versus compound/DPPH molar ratio
for prunin (ꢂ), naringin (᭹), naringenin (ꢀ), prunin 6ꢀꢀ-O-butyrate (ꢁ), prunin 6ꢀꢀ-
O-decanoate (ꢃ), prunin 6ꢀꢀ-O-laurate (ꢁ) and prunin 6ꢀꢀ-O-stearate (×). Data are
expressed as the mean of duplicate assays. All standard deviations were below 3%.
ing vinyl ester by extracting it with hexane. Once the hexane was
evaporated the only remaining solid was the prunin ester. In the
case the conversion yield would have been lower than 100%, vari-
ous prunin separation processes were tested, looking for solvents
capable of dissolving or the prunin ester or the prunin, but not both.
For example, in the case of prunin and prunin laurate, both ethyl
acetate and diethyl ether were capable of performing the separa-
tion. The separation of prunin estearate and prunin was achieved
using 1-octanol.
matic acid vinyl esters [38] obtaining the corresponding 6ꢀꢀ-O-acyl
monoesters. CALB acylates only position 6 and the more bulky
derivatives are no longer substrates for a further acylation, probably
due to steric hindrance.
3.6. Effect of glycoside flavanone
The acylation of flavanone diglycoside naringin, from which
prunin is derived, has been reported by various authors [15–30]
but it was carried out under different operational conditions. In
order to compare the synthesis reactions of naringin 6ꢀꢀ-O-laurate
and prunin 6ꢀꢀ-O-laurate, this study was performed under the same
conditions using CALB. In Fig. 6 the transesterification reaction pro-
files of prunin and naringin with vinyl laurate in the four solvents
are shown. It was observed that the formation rate of naringin 6ꢀꢀ-
O-laurate was much lower than that of prunin 6ꢀꢀ-O-laurate. This
can be explained by the larger size and decreased hydrophobicity
of naringin, whose sugar moiety is the disaccharide neohesperi-
dose whereas the sugar moiety of prunin is glucose. Gao et al. [22]
also observed that the acylation rate of a flavonoid carrying a disac-
charide moiety was noticeably reduced compared to those with a
monosaccharide aglycon. However, they did not compare flavonoid
glycoside derivatives of the same aglycon. Pedersen et al. [43] sug-
gested that when the chain length of the donor or the size of the
flavonoid is increased, the steric hindrance, that may occur on the
binding pocket, increases too, reducing the probability of formation
of the product.
3.9. Radical scavenging activity
Even though various methods to measure antioxidant activities
have been found [44], the DPPH method is widely used for quan-
tifying the free radical scavenging activity. Experiments under the
same conditions were undertaken in order to measure the radi-
cal scavenging activity, not only of the obtained esters, but also
of prunin, the unmodified flavanone glucoside. Flavonoids have
the ability to scavenge free radicals and many of their biological
properties are associated with this ability. This activity of pheno-
lic compounds depends on the structure, in particular the number
and positions of the hydroxyl groups, and the nature of substitu-
tions on the aromatic rings [45,46]. Because of this, a modification
of the structure might drastically alter its biological properties. This
study examined whether or not the introduction of the alkyl chains
in the prunin molecule alters its antiradical activity.
In Fig. 7 the results of the remaining DPPH percentage versus
compound/DPPH molar ratio are shown. It can be observed that
diglycoside naringin and glucoside prunin possessed an identical
capacity to capture free radicals, even though this is a little below
that of the aglycon naringenin from which they are derived. The
introduction of the hydrophobic chain in prunin not only did not
alter its capacity to capture free radicals, but it seemed to be slightly
increased. No correlation was observed in this experiment between
the chain length and the free radical scavenging activity. This result
suggests that maybe these compounds could be used as antioxi-
dant.
3.7. Lipase operational stability
The operational stability of CALB was evaluated in the synthesis
of prunin 6ꢀꢀ-O-laurate at 50 ◦C in the four solvents employed with a
vinyl laurate/prunin molar ratio of 10. After 12 runs, the conversion
yields remained unchanged in all solvents.
3.8. Purification of prunin esters
The prunin 6ꢀꢀ-O-butyrate, prunin 6ꢀꢀ-O-decanoate, prunin 6ꢀꢀ-O-
laurate and prunin 6ꢀꢀ-O-stearate esters were synthesized on gram
scale in acetone using CALB as the biocatalyst. Considering that in
all cases it was possible to achieve the total conversion of prunin
in a few hours of reaction, it was possible to purify them using a
simple method. The ester purification method consisted basically
of evaporating the acetone first and then eliminating the remain-
3.10. Solubility
The solubility data of prunin and its esters in water and in
1-octanol are shown in Table 1. An increase of the solubility in
1-octanol and a decrease of the solubility in water were clearly
observed when the chain length of the acyl group introduced in
the prunin molecule was increased. The obtained 6ꢀꢀ-O-acyl deriva-

G. Céliz, M. Daz / Process Biochemistry 46 (2011) 94–100
99
Table 1
induction of endogenous antioxidant enzymes. Food Chem Toxicol 2008;46:
3140–6.
[9] Han X, Ren D, Fan P, Shen T, Lou H. Protective effects of naringenin-7-O-
glucoside on doxorubicin-induced apoptosis in H9C2 cells. Eur J Pharmacol
2008;581:47–53.
[10] Imai Y, Tsukaharas S, Asada S, Sugimoto Y. Phytoestrogens/Flavonoids reverse
breast cancer resistance protein/ABCG2-mediated multidrug resistance. Can-
cer Res 2004;64:4346–52.
Water and 1-octanol solubilities of prunin and prunin 6ꢀꢀ-O-acyl esters at 20 ◦C.
nd: not detected. Data are expressed as the mean of duplicate assays. All standard
deviations were below 5%.
Solubility (mM)
Water
1-Octanol
[11] Malterud KE, Bremnes TE, Faegri A, Moe T, Sandanger Dugstad EK. Flavonoids
from the wood of salix caprea as inhibitors of wood-destroying fungi. J Nat Prod
1985;48:559–63.
[12] Yao H, Liao ZX, Wu Q, Lei GQ, Liu ZJ, Chen DF, et al. Antioxidative flavanone
glycosides from the branches and leaves of Viscum coloratum. Chem Pharm
Bull 2006;54:133–5.
Prunin
1.07
0.64
nd
nd
nd
1.11
26.0
66.0
66.7
116
Prunin butyrate
Prunin decanoate
Prunin laurate
Prunin stearate
[13] Chebil L, Humeau C, Falcimaigne A, Engasser JM, Ghoul M. Enzymatic acylation
of flavonoids. Process Biochem 2006;41:2237–51.
ˇ
[14] Viskupicˇová J, Sturdík E, Ondrejovicˇ M. Enzymatic transformation of flavonoids.
Acta Chim Slovaca 2009;2:88–106.
tives were more lipophilic than the parent prunin, what makes
them suitable for taking advantage of its antioxidant properties in
hydrophobic matrices.
[15] Watanabe Y, Nagai M, Yamanaka K, Jose K, Nomura M. Synthesis of lauroyl phe-
nolic glycoside by immobilized lipase in organic solvent and its antioxidative
activity. Biochem Eng J 2009;43:261–5.
[16] Katsoura MH, Polydera AC, Tsironis L, Tselepis AD, Stamatis H. Use of ionic
liquids as media for the biocatalytic preparation of flavonoid derivatives with
antioxidant potency. J Biotechnol 2006;123:491–503.
4. Conclusions
[17] Katsoura MH, Polydera AC, Katapodis P, Kolisis FN, Stamatis H. Effect of different
reaction parameters on the lipase-catalyzed selective acylation of polyhydrox-
ylated natural compounds in ionic liquids. Process Biochem 2007;42:1326–34.
[18] Youn SH, Kim HJ, Kim TH, Shin CS. Lipase-catalyzed acylation of naringin with
palmitic acid in highly concentrated homogeneous solutions. J Mol Catal B
Enzym 2007;46:26–31.
[19] Danieli B, De Bellis P, Carrea G, Riva S. Enzyme-mediated regioselec-
tive acylations of flavonoid disaccharide monoglycosides. Helv Chim Acta
1990;73:1837–44.
[20] Riva S, Danieli B, Luisetti M. A two-step efficient chemoenzymatic synthesis of
flavonoid glycoside malonates. J Nat Prod 1996;59:618–21.
[21] Nakajima N, Ishihara K, Itoh T, Furuya T, Hamada H. Lipase-catalyzed direct and
regioselective acylation of flavonoid glucoside for mechanistic investigation of
stable plant pigments. J Biosci Bioeng 1999;87:105–7.
[22] Gao C, Mayon O, MacManus DA, Vulfson EN. Novel enzymatic approach
to the synthesis of flavonoid glycosides and their esters. Biotechnol Bioeng
2000;71:235–43.
[23] Kontogianni A, Skouridou V, Sereti V, Stamatis H, Kolisis FN. Regioselective
acylation of flavonoids catalyzed by lipase in low toxicity media. Eur J Lipid Sci
Technol 2001;103:655–60.
From the results of this study it can be concluded that hydropho-
bic derivatives of prunin can be easily obtained through lipase
catalysed transesterification under mild reaction conditions. The
best conditions involve the use of a solvent with low toxicity (ace-
tone), a relatively low vinyl ester/prunin molar ratio (10), and an
immobilized enzyme (CALB) which is very stable and can be re-
used many times. The total conversions of prunin are achieved
in a few hours and the products can be easily purified. Prunin
esters preserved the antiradical activity of the flavanone glucoside
parent and their solubility notably increased in the hydrophobic
medium 1-octanol, making them suitable for use in hydrophobic
matrices in the food, cosmetic and pharmaceutical industries. Cer-
tainly, these possible applications are subordinated to toxicological
studies.
[24] Otto RT, Geers B, Weiss A, Petersohn D, Schlotmann K, Schroeder KL. Novel
flavone glycoside derivatives for use in cosmetics, pharmaceuticals and nutri-
tion. WO 0179245 2001.
Acknowledgments
[25] Gayot S, Santarelli X, Coulon D. Modification of flavonoid using lipase in non-
conventional media: effect of the water content. J Biotechnol 2003;101:29–36.
[26] Kontogianni A, Skouridou V, Sereti V, Stamatis H, Kolisis FN. Lipase-catalyzed
esterification of rutin and naringin with fatty acids of medium carbon chain. J
Mol Catal B Enzym 2003;21:59–62.
[27] Passicos E, Santarelli X, Coulon D. Regioselective acylation of flavonoids cat-
alyzed by immobilized Candida antarctica lipase under reduced pressure.
Biotechnol Lett 2004;26:1073–6.
[28] Mellou F, Lazari D, Skaltsa H, Tselepis AD, Kolisis FN, Stamatis H. Biocat-
alytic preparation of acylated derivatives of flavonoid glycosides enhances their
antioxidant and antimicrobial activity. J Biotechnol 2005;116:295–304.
[29] Mellou F, Loutrari H, Stamatis H, Roussos C, Kolisis FN. Enzymatic esterifica-
tion of flavonoids with unsaturated fatty acids: effect of the novel esters on
vascular endothelial growth factor release from K562 cells. Process Biochem
2006;41:2029–34.
[30] Stevenson DE, Wibisono R, Jensen DJ, Stanley RA, Cooney JM. Direct acylation of
flavonoid glycosides with phenolic acids catalysed by Candida antarctica lipase
B (Novozym 435®). Enzyme Microb Technol 2006;39:1236–41.
[31] Geronazzo H, Robin J, Blanco S, Cuevas C, Ellenrieder G. Aprovechamiento inte-
gral de residuos de producción y procesamiento de pomelos. Un proyecto de
innovación tecnólogica. In Actas VIII Congreso Argentino de Ciencia y Tec-
nología de Alimentos. Argentina: Santa Fe. 1999.
[32] Soria F, Ellenrieder G. Thermal inactivation and product inhibition of Aspergillus
terreus CECT 2663 ␣-rhamnosidase and their role on hydrolysis of naringin
solutions. Biosci Biotechnol Biochem 2002;66:1442–9.
[33] Brand-Williams W, Cuvelier ME, Berset C. Use of a free radical method to eval-
uate antioxidant activity. LWT Food Sci Technol 1995;28:25–30.
[34] Degueil-Castaing M, de Jeso B, Drouillard S, Maillard B. Enzymatic reactions
in organic synthesis: 2-ester interchange of vinyl esters. Tetrahedron Lett
1987;28:953–4.
[35] Hazarika S, Goswami P, Dutta NN. Lipase catalysed transesterification of 2-o-
benzylglycerol with vinyl acetate: solvent effect. Chem Eng J 2003;94:1–10.
[36] Chebil L, Anthoni J, Humeau C, Gerardin C, Engasser JM, Ghoul M. Enzymatic
acylation of flavonoids: effect of the nature of the substrate. origin of lipase,
and operating conditions on conversion yield and regioselectivity. J Agric Food
Chem 2007;55:9496–502.
[37] Danieli B, Luisetti M, Sampognaro G, Carrea G, Riva S. Regioselective acylation
of polyhydroxylated natural compounds catalyzed by Candida antarctica lipase
B (Novozym 435) in organic solvents. J Mol Catal B Enzym 1997;3:193–201.
The authors would like to thank Universidad Nacional de Salta
(UNSa), Consejo Nacional de Investigaciones Científicas y Técnicas
(CONICET), Agencia Nacional de Promoción Científica y Tecnológ-
ica (ANPCyT) of Argentina for the financial support that made
this work possible. To Novozymes Latin America Limited for the
enzyme, to Dr Roberto R. Gil (Carnegie Mellon University, USA),
Dr María Laura Uriburu and Dr Juana Rosa de la Fuente for
helping in the creation and interpretation of the NMR spectra.
To Dr Guillermo Ellenrieder and to Ch E Hugo Geronazzo for
providing the naringin. G. Céliz thanks CONICET for his fellow-
ship.
References
[1] Lyseng-Williamson KA, Perry CM. Micronised purified flavonoid fraction: a
review of its use in chronic venous insufficiency, venous ulcers and haemor-
rhoids. Drugs 2003;63:71–100.
[2] Misra MC, Imlitemsu. Drug treatment of haemorrhoids. Drugs
2005;65:1481–91.
[3] Ellenrieder G. Biotransformations of citrus flavanone glycosides. In: Pandey
A, editor. Concise encyclopedia of bioresource technology. New York: The
Haworth press; 2004. p. 189–200.
[4] Benavente-Garcia O, Castillo J. Update on uses and properties of citrus
flavonoids: new findings in anticancer, cardiovascular, and anti-inflammatory
activity. J Agric Food Chem 2008;56:6185–205.
[5] Ellenrieder G, Blanco S, Daz M. Hydrolysis of supersaturated naringin solutions
by free and immobilized naringinase. Biotechnol Tech 1998;12:63–5.
[6] Choi JS, Yokozawa T, Oura H. Antihyperlipidemic effect of flavonoids from
Prunus davidiana. J Nat Prod 1991;54:218–24.
[7] Choi JS, Yokozawa T, Oura H. Improvement of hyperglycemia and hyperlipemia
in streptozotocin-diabetic rats by a methanolic extract of Prunus davidiana
stems and its main component, prunin. Planta Med 1991;57:208–11.
[8] Han X, Pan J, Ren D, Cheng Y, Fan P, Lou H. Naringenin-7-O-glucoside
protects against doxorubicin-induced toxicity in H9c2 cardiomyocytes by

100
G. Céliz, M. Daz / Process Biochemistry 46 (2011) 94–100
[38] Ishihara K, Nakajima N. Structural aspects of acylated plant pigments: stabiliza-
tion of flavonoid glucosides and interpretation of their functions. J Mol Catal B
Enzym 2003;23:411–7.
[39] Uppenberg J, Hansen MT. The sequence, crystal structure determination and
refinement of two crystal forms of lipase B from Candida antarctica. Structure
1994;2:293–308.
[40] Otto RT, Scheib H, Bornscheuer UT, Pleiss J, Syldatk C, Schmid RD. Substrate
specificity of lipase B from Candida antarctica in the synthesis of arylaliphatic
glycolipids. J Mol Catal B Enzym 2000;8:201–11.
[41] Pleiss J, Fischer M, Schmid RD. Anatomy of lipase binding sites: the scissile fatty
acid binding site. Chem Phys Lipids 1998;93:67–80.
reveal a stereospecificity pocket for secondary alcohol. Biochemistry (Wash)
1995;34:16838–51.
[43] Pedersen NR, Wimmer R, Emmersen J, Degn P, Pedersen LH. Effect of fatty acid
chain length on initial reaction rates and regioselectivity of lipase-catalysed
esterification of disaccharides. Carbohydr Res 2002;337:1179–84.
[44] Prior RL, Wu X, Schaich K. Standardized methods for the determination of
antioxidant capacity and phenolics in foods and dietary supplements. J Agric
Food Chem 2005;53:4290–302.
[45] Balasundram N, Sundram K, Samman S. Phenolic compounds in plants and agri-
industrial by-products: antioxidant activity, occurrence, and potential uses.
Food Chem 2006;99:191–203.
[42] Uppenberg J, Öhrner N, Noria M, Hule K, Kleywegt GJ, Patear S, et al. Crystal-
lographic and molecular-modeling studies of lipase B from Candida antarctica
[46] Rice-Evans CA, Miller NJ, Paganga G. Structure–antioxidant activity relation-
ships of flavonoids and phenolic acids. Free Radic Biol Med 1996;20:933–56.