Glycerol acyl-transfer kinetics of a circular permutated Candida antarctica lipase B

Article abstract of DOI:10.1016/j.molcatb.2011.06.002

Triacylglycerols containing a high abundance of unusual fatty acids, such as γ-linolenic acid, or novel arylaliphatic acids, such as ferulic acid, are useful in pharmaceutical and cosmeceutical applications. Candida antarctica lipase B (CALB) is quite often used for non-aqueous synthesis, although the wild-type enzyme can be rather slow with bulky and sterically hindered acyl donor substrates. The catalytic performance of a circularly permutated variant of CALB, cp283, with various acyl donors and glycerol was examined. In comparison to wild-type CALB, butyl oleate and ethyl γ-linolenate glycerolysis rates were 2.2- and 4.0-fold greater, respectively. Cp283 showed substrate inhibition by glycerol, which was not the case with the wild-type version. With either ethyl ferulate or vinyl ferulate acyl donors, cp283 matched the performance of wild-type CALB. Changes in active site accessibility resulting from circular permutation led to increased catalytic rates for bulky fatty acid esters but did not overcome the steric hindrance or energetic limitations experienced by arylaliphatic esters.

Full text of DOI:10.1016/j.molcatb.2011.06.002

Journal of Molecular Catalysis B: Enzymatic 72 (2011) 175–180
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Glycerol acyl-transfer kinetics of a circular permutated Candida antarctica
lipase Bଝ
Joseph A. Laszloa,∗, Ying Yub, Stefan Lutzb, David L. Comptona
a Renewable Product Technology Research, USDA-Agricultural Research Service, National Center for Agricultural Utilization Research, 1815 N, University St., Peoria, IL 61604, USA
b Department of Chemistry, Emory University, 1515 Dickey Drive, Atlanta, GA 30322, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Triacylglycerols containing a high abundance of unusual fatty acids, such as ␥-linolenic acid, or novel
arylaliphatic acids, such as ferulic acid, are useful in pharmaceutical and cosmeceutical applications.
Candida antarctica lipase B (CALB) is quite often used for non-aqueous synthesis, although the wild-
type enzyme can be rather slow with bulky and sterically hindered acyl donor substrates. The catalytic
performance of a circularly permutated variant of CALB, cp283, with various acyl donors and glycerol
was examined. In comparison to wild-type CALB, butyl oleate and ethyl ␥-linolenate glycerolysis rates
were 2.2- and 4.0-fold greater, respectively. Cp283 showed substrate inhibition by glycerol, which was
not the case with the wild-type version. With either ethyl ferulate or vinyl ferulate acyl donors, cp283
matched the performance of wild-type CALB. Changes in active site accessibility resulting from circular
permutation led to increased catalytic rates for bulky fatty acid esters but did not overcome the steric
hindrance or energetic limitations experienced by arylaliphatic esters.
Received 11 January 2011
Received in revised form 31 May 2011
Accepted 2 June 2011
Available online 13 June 2011
Keywords:
Glycerolysis
Transesterification
Specific activity
Enzyme engineering
Circular permutation
Published by Elsevier B.V.
1. Introduction
tions 283 and 282 of the wild-type CALB sequence, resulted in
an increased hydrolytic activity for p-nitrophenol butyrate (11-
Lipase B of Candida antarctica (CALB) is an exceedingly versatile
biocatalyst for applications involving non-aqueous synthesis due
to its regio- and stereoselectivity, high activity and thermostabil-
ity [1]. Site-directed and random mutagenesis has been applied to
the CALB sequence in an effort to enhance various attributes [2–5].
Rational protein engineering approaches have produced CALB vari-
ant proteins with improved thermostability, activity and altered
enantioselectivity [6,7]. DNA family shuffling was used to create
mutants with enhanced stability and hydrolysis activity toward a
prochiral diester substrate [8]. These studies indicate that improve-
ments can be made to the enzyme to suit specific applications and
substrates.
Circular permutation (cp) of the CALB sequence, which intro-
duces new C- and N-termini into the polypeptide sequence,
greatly increased the catalytic efficiency of the catalyst for select
substrates, particularly esters with bulky leaving groups [9,10].
Variant cp283, which has its N- and C-termini located at posi-
fold) and 6,8-difluoro-4-methylumbelliferyl octanoate (175-fold)
[9]. The engineered lipase further demonstrated enhanced triacyl-
glycerol deacylation activity (5–7-fold) [11], which indicated that
it may be useful for biodiesel production [12–14]. Biodiesel rep-
resents a high-volume, low-value product, for which there exists
a substantial installed base of conventional catalysts that perform
well [15,16]. In contrast, there are many triacylglycerol high-value
transformations for which biocatalysts are ideally suited.
Structured lipids are acylglycerols refashioned by chemical or
enzymatic processes to alter their fatty acid composition and/or
the stereochemical positions of fatty acids on glycerol. These
changes are targeted for specific metabolic effects, nutritive or ther-
apeutic purposes, and improved physicochemical properties [17].
Structured lipids composed of polyunsaturated fatty acids such as
conjugated linolenic acid, ␥-linolenic acid (6,9,12-octadecatrienoic
acid), and the ␻-3 eicosapentaenoic acid and docosahexaenoic
acids are sought in concentrated form for dietary supplementation.
Beyond substitution of one fatty acid for another, a broad range
of carboxyl-containing molecules can be usefully attached to the
glycerol backbone. For instance, novel esters of glycerol have been
investigated for their potential as lipid-soluble antioxidants [18].
Lipid derivatives of the natural dietary components ferulic and caf-
feic acids show strong promise in this regard [19]. Lipase-catalyzed
transesterification is the preferred route for production of struc-
tured lipids containing easily oxidized/isomerized or otherwise
ଝ
Names are necessary to report factually on available data; however, the USDA
neither guarantees nor warrants the standard of the product, and the use of the
name by the USDA implies no approval of the product to the exclusion of others
that may also be suitable. USDA is an equal opportunity provider and employer.
∗
Corresponding author. Tel.: +1 309 681 6322; fax: +1 309 681 6686.
E-mail address: Joe.Laszlo@ars.usda.gov (J.A. Laszlo).
1381-1177/$ – see front matter. Published by Elsevier B.V.
doi:10.1016/j.molcatb.2011.06.002

176
J.A. Laszlo et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 175–180
readily altered substituents. However, the enzymatic processes can
suffer from low reaction rates and yields when working with these
unconventional fatty acid substitutes.
Haute, IN, USA). Each reaction was conducted in triplicate. After
reaction initiation by addition of enzyme, sampling (20 ␮L) was
conducted at specified intervals. Samples were diluted into ace-
tone, filtered (Whatman Anotop-10 0.2-␮m inorganic filters), and
stored at −20 ◦C until analyzed. Product formation was determined
by HPLC using a system detailed previously [21]. A C8 reverse phase
column (5 ␮m, 250 mm × 4.6 mm; Phenomenex, Torrance, CA, USA)
was employed for separations. The column was developed isocrat-
ically (1.5 mL min−1) with solvent consisting of 40% (v/v) acetone
(containing 1% (v/v) acetic acid) and 60% (v/v) acetonitrile. Evap-
orative light scattering detector responses (70 ◦C drift tube, 44 ◦C
exhaust, 1.8 SLPM N2) for lipid products were determined from
authentic standards. The concentration of feruloyl glycerol was
ascertained by UV detection at 325 nm, using ethyl ferulate as the
calibration standard. Initial reaction rates were determined by lin-
ear regression of production concentration as a function of time for
samples in which there was less than 10% substrate consumption.
Triolein butanolysis: The reaction of triolein with 1-butanol
(Fig. 1) was conducted without solvent by combining 1.0 mmol
of triolein with 3.0 mmol of 1-butanol and 10 mg of catalyst (total
weight of resin and support immobilized wild-type or cp283 CALB).
For the reaction performed in 2-methyl-2-butanol, the triolein and
1-butanol concentrations were 0.4 and 1.2 M, respectively, and
10 mg of catalyst was dispersed in 1.0 mL of reaction medium.
Butyl oleate glycerolysis: The interesterification of butyl oleate
with glycerol (Fig. 1) was performed in 2-methyl-2-butanol at 23 ◦C.
Catalyst (5–10 mg) was dispersed in 1.0 mL of reaction medium.
The butyl oleate concentration varied from 0.1 M to 1.5 M, and the
glycerol concentration varied from 0.25 M to 1.0 M.
The present work compares the catalytic activity of wild-type
CALB and cp283 in the synthesis of a variety of acyl glycerols
(Figs. 1 and 2). Previous work demonstrated the superiority of
cp283 for alcoholysis of vegetable oils [11]. Herein a reaction emu-
lating the reverse course was examined to determine whether
cp283 retains its performance edge in such reactions. In addition,
glycerolysis reactions with ␥-linolenic and ferulic acid esters, as
representatives of novel or bulky acyl donors, were studied to ascer-
tain the extent to which the molecular changes wrought by circular
permutation of CALB can deliver catalytic improvements for a broad
range of substrates.
2. Experimental
2.1. Reagents and materials
Novozym 435 (acrylic resin immobilized CALB) and Lewatit
VP OC 1600 poly(methyl methacrylate) resin were obtained from
Novozymes North America. Immobilized cp283 was prepared as
previously described [9]. Methyl 4-methylumbelliferyl hexylphos-
phonate was prepared using a procedure modified from Qian
et al. [9,20]. Ethyl ␥-linolenate and mono-␥-linolenin glycerol were
obtained from Nu-Chek-Prep (Elysian, MN, USA). Ethyl ferulate was
provided by Ash Ingredients (Glen Rock, NJ, USA). Ferulic acid and
spectroscopic grade glycerol were purchased from Sigma–Aldrich
˚
and stored in a dry box. Activated molecular sieves (3 A) were stored
Ethyl ꢀ-linolenate glycerolysis: The interesterification of ethyl
␥-linolenate with glycerol (Fig. 1) was performed in 2-methyl-2-
butanol at 23 ◦C. Catalyst (5–10 mg) was dispersed in 1.0 mL of
reaction medium. The ethyl ␥-linolenate concentration varied from
0.05 M to 1.5 M, and the glycerol concentration was 1.0 M.
under nitrogen. Vinyl acetate was purchased from Sigma–Aldrich
(St. Louis, MO, USA) and stored under nitrogen at 4 ◦C. Mercury
acetate, ethyl acetate, tetrahydrofuran (THF), hexane, and conc.
sulfuric acid were purchased from Sigma–Aldrich and used as
received. All other reagents were obtained from Sigma–Aldrich.
Ethyl and vinyl ferulate glycerolysis: The interesterification of
ethyl or vinyl ferulate with glycerol (Fig. 2) was performed in 2-
methyl-2-butanol at 55 ◦C. Catalyst (100 mg) was combined with
1.0 mL of reaction medium. The ethyl or vinyl ferulate concentra-
tion varied from 0.25 M to 1.5 M, and the glycerol concentration
was 1.0 M.
2.2. Methods
2.2.1. Active-site titration
Catalyst activity is reported as specific activity (kcat) based on the
amount of active enzyme determined by active site titration. The
determination was conducted using methyl 4-methylumbelliferyl
hexylphosphonate (4-MUHP) and a procedure adopted from Qian
et al. [9,20]. Catalyst (5–20 mg) was solvated with 1 mL of ace-
tonitrile containing 1% (v/v) water and 60 ␮M 4-MUHP and then
reacted for one week at 23 ◦C. At the end of the reaction period,
the solvent was sampled for fluorometric determination of 4-
methylumbelliferone (4-MU) released by the single turnover of
each active CALB molecule. Acetonitrile extracts and reaction
solvent were combined. Aliquots were diluted into 10% acetoni-
trile/90% 0.1 M ammonia buffer (pH 9.5). Sample excitation was
at 365 nm and fluorescence emission was measured at 448 nm (2-
nm slit widths). A 4-MU linear calibration curve was constructed
(15–90 nM). The concentration of 4-MU released from the lipase
treatments was determined by regression. Control 4-MUHP solu-
tions containing bare resin (Lewatit VP OC 1600) were used to
correct for background hydrolysis. Analyses were performed on
catalysts in triplicate, for which mean and standard deviation val-
ues are reported.
2.2.3. Vinyl ferulate synthesis
The vinyl ferulate synthesis protocol was modified from Gao
et al. [22] and conducted using standard Schlenk line techniques.
Ferulic acid (9.07 g, 46.7 mmol) and mercury acetate (330 mg,
1.0 mmol) were combined in a 250-mL Schlenk flask equipped with
a stirbar, and then evacuated and flushed with nitrogen three times.
THF (50 mL) was added via syringe with stirring followed by an
excess of vinyl acetate (75 mL, 813.9 mmol). The cream colored
slurry slowly dissolved into a light yellow solution upon stirring for
30 min at ambient temperature. Concentrated sulfuric acid (10 ␮L,
0.2 mmol) was added with a Gilson Pipetman against a nitrogen
flow and the solution was heated to 40 ◦C and stirred for 24 h. The
amber colored solution was cooled and 150 ␮L more conc. sulfuric
acid (1.5 mmol) was added and the solution stirred at 40 ◦C for an
additional 24 h. The reaction was terminated with the addition of
sodium acetate (2.5 g, 30.8 mmol) and stirring for 1 h. The slurry
was filtered through a 100-mL medium frit containing a bed of sil-
ica gel (0.05–0.2 nm, 70–325 mesh). The bed was washed with 0.5 L
of 10% (v/v) ethyl acetate/hexane. The solvent was removed from
the filtrate using a rotovap to produce ∼12 g of a crude, yellow,
crystalline solid.
2.2.2. Enzyme reactions
Reactions were conducted using immobilized wild-type CALB
(Novozym 435) and circularly-permutated CALB (cp283), solvents
(when used), and substrates equilibrated to a water activity of 0.11
using a saturated LiCl solution at 23 ◦C in a closed container for at
least one week. Reactions were conducted in 1.5-mL polypropy-
lene microcentrifuge tubes attached to a rotator (Glas-Col, Terre
The crude vinyl ferulate was purified in ∼3.0 g portions on a
40 cm × 2 cm silica gel (0.05–0.2 nm, 70–325 mesh) column. The
column was conditioned with 0.5 L of hexane followed by 0.1 L of 5%
(v/v) ethyl acetate/hexane. The crude samples was dissolved in 10%

J.A. Laszlo et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 175–180
177
Fig. 1. Alcoholysis reactions with fatty acid esters.
(v/v) ethyl acetate/hexane and loaded onto the column. The column
was developed sequentially with 1.0 L volumes of 5% (v/v) ethyl
acetate/hexane, 10% (v/v) ethyl acetate/hexane, and ethyl acetate.
Fractions containing the vinyl ferulate were combined, and the sol-
vent was removed via rotovap. The resultant white powder was
dried at 30 ◦C overnight in vacuo resulting in 6.08 g of vinyl ferulate
with >98% purity based on 1H NMR and GC–MS.
rates were determined such that initial condition changes were
minimal.
3.1. Active-site titration of immobilized wild-type and cp283
CALB
Based on the amount of 4-MU released by 4-MUHP treat-
ment of the enzyme, the active CALB content was 1.00 0.03
and 0.16 0.01 mmol g−1 of catalyst for Novozym 435 and cp283,
respectively. These values were used to normalize reaction rates
between the two catalysts based on active enzyme amounts.
Yield: 55.3% (based on ferulic acid). 1H NMR (d6-acetone,
500 MHz):
ı
8.22 (0.89 H, bs, –OH), 7.74 (d, J = 16.0 Hz,
1.00 H, Ar–CH CH–), 7.41 (overlap of m, 2.01 H, Ar–H and
–C(O)–O–CH CH2), 7.21 (dd, J = 2.0 Hz, JJ = 8.1 Hz, 1.08 H,
Ar–H), 6.90 (d, J = 8.1 Hz, 1.00 H, Ar–H), 6.46 (d, J = 15.7 Hz,
1.05 H, Ar–CH CH–), 4.92 (dd, J = 1.4 Hz, JJ = 14.0 Hz, 1.04 H,
–C(O)–O–CH CH2), 4.62 (dd, J = 1.5 Hz, JJ = 6.4 Hz, 1.05 H,
–C(O)–O–CH CH2), 3.94 (s, 3.19 H, Ar–O–CH3).
3.2. Triolein butanolysis
Yu and Lutz [11] demonstrated that the initial transesterification
rate of various vegetable oils with 1-butanol by cp283 was 4- to 7-
fold faster than wild-type CALB. In the present work, butanolysis of
triolein in the absence of cosolvent and at room temperature was
found to be 5.3-fold faster with cp283 in comparison to wild-type
CALB (Table 1), which is consistent with the prior findings. Rais-
ing the reaction temperature to 55 ◦C doubled the reaction rate for
both enzymes, so the enzymes are similarly responsive to altered
reaction temperature. Adding a cosolvent (2-methyl-2-butanol) to
the reaction moderately increased (60%) the butanolysis rate of the
3. Results
A comparison was made of the catalytic efficacy of wild-type
CALB and cp283 with a variety of substrates in which glyc-
erides serve as acyl donors, or with glycerol as an acyl acceptor
(Figs. 1 and 2). Reactions were selected to elucidate differences in
the kinetic attributes of the two enzymes, apart from the influence
of reaction conditions (temperature, solvent). Product formation
Fig. 2. Glycerolysis reaction with ethyl and vinyl ferulates.

178
J.A. Laszlo et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 175–180
Table 1
Butanolysis rates for wild-type and cp283 CALB.
Conditions
Initial rate (min−1
Wild type
)
Relative rate
cp283
No solvent, 23 ◦
C
63
120 10
100 10
5
330 10
610 80
300 40
5.3
5.0
3.0
No solvent, 55 ◦C
2M2B solvent, 23 ◦
C
Relative rate: (cp283 rate)/(wild type) rate. 2M2B is 2-methyl-2-butanol. Cp283 is a
circular permutated CALB variant whose N terminus starts at amino acid 283 of the
wild-type sequence.
wild-type enzyme but did not significantly affect the cp283 rate,
which is reflected in a lower relative reaction rate for cp283 in sol-
vent (Table 1). The enzymes thus displayed a small difference in
their response to the presence of solvent in the reaction medium.
Note that addition of solvent served to lower substrate concen-
trations, which may have augmented or masked differences in the
enzymes’ response to the presence of solvent (e.g., altered substrate
affinities). These data suggest however that 2-methyl-2-butanol
does not adversely impact the activity of either enzyme.
3.3. Butyl oleate glycerolysis
The reaction of butyl oleate with glycerol is ostensibly the
reverse of triolein butanolysis (Fig. 1); more so in the presence of
solvent, which kept the substrates miscible. Butyl oleate glycerol-
ysis rates were 5.5- and 2.4-fold greater for wild-type CALB and
cp283, respectively, than their corresponding triolein butanolysis
rates (with 1 M substrates at 23 ◦C; cf. Tables 1 and 2). The higher
rates presumably reflect the case that butyl oleate is smaller than
triolein as an acyl donor and thus more readily oriented in the active
site to form the first enzyme-substrate activated intermediate of
the enzyme’s ping-pong bi–bi reaction pathway [23].
A more detailed study of the butyl oleate glycerolysis reaction
revealed the impact of the structural changes resulting from the
circular permutation of wild type CALB to cp283 (Fig. 3). Both
enzymes demonstrated substrate inhibition at 1.5 M butyl oleate
(data not shown). Substrate inhibition by butyl oleate observed at
1.5 M may not be due to an impact on enzyme’s active site but rather
result from changes to the solvent/substrate conditions at this high
concentration. The butyl oleate may have been starting to form a
separate immiscible liquid phase in the presence of 1 M glycerol
and 2-methyl-2-butanol. In order to simplify the kinetic analysis,
only butyl oleate concentrations ≤1.0 M were used for estimating
kinetic constants (Table 3). The kinetic constants for the wild-type
CALB (Table 3) were adequately determined by non-linear regres-
sion fitting of the experimental data to Eq. (1) for the initial rate of
the reaction,
Fig. 3. Initial reaction rates of butyl linolenate glycerolysis. Wild-type CALB, panel A;
cp283, panel B. Lines were drawn from Eq. (1) (wild-type) and Eq. (2) (cp283) based
on the parameters given in Table 3. Reactions were conducted in 2-methyl-2-butanol
at 23 ◦C.
where v0 is the initial rate, [ET] is the total active enzyme concentra-
tion, kcat is the first-order rate constant, [A0] and [B0] are the initial
substrate concentrations of butyl oleate and glycerol, respectively,
and KmA and KmB are the steady-state Michaelis constants for butyl
v0
kcat
A
[
B
0] [ ]
0
[
=
]
(1)
KmB
A
0
+ KA
B
]
+ A
[
B
0] [ ]
0
E
[
T
[
]
0
m
Table 2
Table 3
Reaction rates for wild-type and cp283 CALB catalyzed acyl transfer reactions with
glycerol in 2-methyl-2-butanol.
Kinetic constants for wild-type and cp283 CALB catalyzed acyl transfer reactions
with lycerol in 2-methyl-2-butanol.
Reaction acyl donor/
temperature
Initial rate (min−1
)
Relative rate
Acyl donor
kcat (min−1
)
KmA (M)
KmB (M)
Wild
type
cp283
Wild
type
cp283
Wild
type
cp283
Wild type
Cp283
Butyl oleate, 23 ◦
C
350 40
960 140
790 40
3800 230
2.2
4.0
Butyl oleate
Ethyl ␥-linolenate
670
1900
1800
5600
0.65
0.65
0.30
0.11
0.43
–
0.02
–
Ethyl ␥-linolenate,
23 ◦C
Reactions were conducted at 23 ◦C. The KiB (inhibition constant for glycerol with
cp283) was determined to be 0.4 M from the butyl oleate + glycerol reaction. Kinetic
constants for the butyl oleate + glycerol reaction were derived from fits to the data
shown in Fig. 3. Kinetic constants for the ethyl ␥-linolenate + glycerol reaction were
derived from fits to the data shown in Fig. 4.
Ethyl ferulate, 55 ◦
Vinyl ferulate, 55 ◦
C
C
0.67 0.06
6.2 0.6
0.90 0.10
6.0 0.1
1.3
1.0
Substrate (acyl donor and glycerol) concentrations were 1.0 M. Relative rate: (cp283
rate)/(wild type) rate.

J.A. Laszlo et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 175–180
179
were not re-examined for the reaction with ethyl ␥-linolenate as
the acyl donor. Eq. (4) was applied to KmApp to calculate KmA for cp283.
KmApp
KmA
=
(4)
1 + I /KB
[ ]
i
From the determined kinetic constants (Table 3) it can be
concluded that wild-type CALB did not display an affinity (Km) dif-
ference between the butyl oleate and ethyl ␥-linolenate substrates,
while cp283 increased its affinity (lowered Km) by two-thirds. Fur-
thermore, the ␥-linolenate selectivity constant (kcat/KmA ) of cp283
was 18-fold greater than what was observed with wild-type CALB.
In comparison to wild-type CALB, circular permutation cp283 thus
showed substantially improved catalytic performance with the
polyunsaturated ␥-linolenate substrate.
3.5. Ethyl and vinyl ferulate glycerolysis
Fig. 4. Initial reaction rates of ethyl ␥-linolenate glycerolysis. Wild-type CALB, filled
circles; cp283, open circles. Lines (wild-type, solid; cp283, dashed) were drawn from
Eq. (3) based on the parameters given in Table 3. Reactions were conducted in 2-
methyl-2-butanol at 23 ◦C with 1 M glycerol.
The reaction of ethyl ferulate with glycerol (Fig. 2) is four orders
of magnitude slower than that of the comparable reaction of butyl
oleate with glycerol for wild-type CALB and cp283 (Table 2). There
was a minimal difference in the initial rates between the two
enzymes with this substrate. As expected [26], use of the acti-
vated ester vinyl ferulate lead to substantially higher initial reaction
rates with both enzymes (∼7–9-fold), but there was no preferential
increase by either enzyme. Thus, the rate enhancement observed
with cp283 when compared to the wild type enzyme for fatty
acid alcoholysis reactions was not extended to reactions involving
unsaturated arylaliphatic esters.
oleate and glycerol, respectively. Cp283 demonstrated substrate
inhibition by glycerol (Fig. 3, panel A), therefore Eq. (2) was applied
to determine its kinetic constants, including KiB for the glycerol
inhibition constant (Table 3).
v0
kcat
A
B
]
0
[
0] [
ꢀ
ꢁ
ꢂꢃ
=
]
(2)
KmB
A
0
+ KA
B
1 + B0/KB + A
B
0] [ ]
0
E
[
T
[
]
[
]
[
0
m
i
4. Discussion
The kcat value of cp283 was 2.7-fold higher compared to wild-
type CALB for the reaction. In addition the KmA (butyl oleate) was
somewhat lower for cp283. Most notably, cp283 displayed dead-
end substrate inhibition by glycerol (KiB = 0.4 M), while wild-type
CALB did not do so under these reaction conditions. Cp283 dis-
played a KmB (glycerol) substantially lower than that of the wild-type
enzyme, reflective of cp283s greater affinity for this substrate.
In the present study, wild-type CALB and its circular permu-
tation variant cp283 were examined to determine whether the
circular permutant displayed greater activity than the wild-type
enzyme for a range of fatty acid ester and arylaliphatic ester sub-
strates. Glycerol was selected as the acyl acceptor for reaction
with these substrates. Glycerides commonly serve as active agent
derivatives for pharmaceutical and cosmeceutical applications. The
reaction of butyl oleate with glycerol (Fig. 1) was selected for study
because prior work had demonstrated that cp283 was particularly
efficacious for butanolysis of triglycerides [11], and thus the butyl
oleate reaction represents the first step in the reversal of this pro-
cess. Glycerolysis of the specialty fatty acid ester ethyl ␥-linolenate
(Fig. 1) and arylaliphatic esters of ferulic acid (Fig. 2) represent reac-
tions of potential commercial utility. Steric interactions of these
acyl donors in most lipase active sites can be unfavorable, leading
to very low reaction rates [27,28]. Variants better able to accommo-
date bulky acyl groups would provide utilitarian value via increased
bioprocessing throughput. Relocation of the N- and C-termini to
within the vicinity of the active site region is believed to lessen
steric constraints in catalytic activity through an increase in chain
flexibility and active site access.
In comparison to wild-type CALB, butyl oleate and ethyl
␥-linolenate glycerolysis rates were 2.2- and 4.0-fold greater,
respectively, with the cp283 variant. These catalytic rates are con-
sistent with previous findings that the permutated enzyme had 2.6-
to 9-fold greater relative catalytic efficiency than wild-type CALB
for transesterification of triglycerides [11]. Greater active site acces-
sibility in cp283 compared to wild-type CALB is likely to result in
faster substrate binding and product release for these lipids [29].
Cp283 showed substrate inhibition by glycerol, which was not
evident in wild-type CALB. Adsorption of glycerol to immobilized
lipase can have an adverse impact on transesterification reactions
by creating a hydrophilic diffusional barrier for hydrophobic sub-
strates [30]. Glycerol adsorption to the support material Lewatit
3.4. Ethyl ꢀ-linolenate glycerolysis
Transesterification of the polyunsaturated ␥-linolenic acid
(18:3, n-6) with glycerol proceeded quite rapidly with either wild-
type CALB or cp283. Neither enzyme displayed substrate inhibition
by ethyl ␥-linolenate, unlike that which was observed with butyl
oleate at high concentration. Ethanol generated during the reaction
would be expected to act as an inhibitor, but only at concentra-
tions greater than that reached during initial rate determinations
(>0.2 M) [24]. The reaction rate was 2.7-fold faster than the butyl
oleate reaction for the wild-type enzyme (Table 2), reflecting the
known preference of this lipase for polyunsaturated C18 fatty acid
substrates [25]. Cp283 also demonstrated an increased reaction
rate with ␥-linolenate as the acyl donor such that its rate was 4-fold
faster than that of wild-type CALB (Table 2).
A simple Michaelis–Menten formulation (Eq. (3)) was applied
to the data (Fig. 4) to determine kinetic constants by non-linear
regression (SigmaPlot 10.0):
v0
kcat A
[ ]
0
=
]
(3)
KmApp + A
E
T
[
[
]
0
Glycerol was held constant (1 M) and the acyl donor concen-
tration was varied. KmApp is the apparent value for the acyl donor.
This reaction would be expected to have the same KmB for glycerol, as
well as the same degree of glycerol inhibition for cp283 (KiB), as was
observed for butyl oleate glycerolysis. Therefore, these parameters

180
J.A. Laszlo et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 175–180
VP OC 1600 used to immobilize wild-type CALB (Novozym 435)
and cp283 is known [31]. There is disagreement in the literature
as to whether use of a moderate polarity solvent such as tert-
butanol overcomes this effect [31,32]. However, this influence is
expected to be the same for either enzyme as they are immobilized
on the same support material. The substrate inhibition demon-
strated by cp283 (Fig. 3) thus likely reflects an altered sensitivity
of the enzyme to glycerol rather than due to the imposition of a
diffusion barrier, which would be the same for both enzymes.
The cp283 variant of CALB introduces highly polar termini
into an otherwise very nonpolar ␣-helix (helix 17), which opens
up the substrate tunnel to the active site [29]. The increased
polarity of this sequence segment may account for cp283’s sus-
ceptibility to substrate inhibition by glycerol, and could also
contribute to the variant’s lower glycerol KmB (Table 3). However, the
commercially-sourced wild-type CALB (Novozym 435) and cp283
were not immobilized under identical conditions (i.e., differences
in co-adsorbates and surface coverage by lipase), which cannot be
precluded as contributing factors governing the kinetic responses
of the lipases to glycerol.
It could be a direct result of a change in the polar environment
near the enzyme’s active site, or may be related to the significantly
lower enzyme immobilization load for the CALB variant. Optimiza-
tion of the immobilization procedure and the support surface could
eliminate the undesirable effect. In future experiments, a targeted
screening of circular permutation libraries of CALB holds promise
for the identification of candidates with improved activity for ster-
ically challenged substrates and esters with unfavorable electronic
effects.
Acknowledgments
Leslie Smith (NCAUR) provided valuable technical support for
this research. This work was supported in part by funding from
the Biotechnology Research and Development Corporation (to J.A.L.
and D.L.C.), US National Science Foundation (CBET-0730312 to S.L.),
and the Petroleum Research Fund by the American Chemical Soci-
ety (PRF 47135-AC1 to S.L.). These organizations did not contribute
to the study design, data collection, analysis and interpretation, or
writing of the report.
With either ethyl ferulate or vinyl ferulate acyl donors, cp283
did not perform better than wild-type CALB. Changes in active
site accessibility resulting from the termini relocation which led
to increased catalytic rates for bulky fatty acid esters hence did not
overcome the steric hindrance or energetic effects experienced by
arylaliphatic acids and esters [33]. Guyot et al. [27] first pointed out
the electron donating effect in cinnamic acid ester synthesis. The
electron donating effects intrinsically deactivate the electrophilic
carbon center of the carboxylic group for nucleophilic attack of
the alcohol group of the active site serine. The use of activated
esters (ethyl or vinyl leaving groups) partially counter this effect.
The greater active site accessibility in cp283 compared to wild-type
CALB which resulted in faster substrate binding and product release
for the fatty acids failed to do so for ferulic acid esters. Given the
ping-pong mechanism of the reaction, a possible reason for this
lack of improvement may be that there is a difference the in rate-
limiting step for the two types of substrates; the deacylation step
for fast substrates such as fatty acid esters, and active site (serine)
acylation for slow reacting ferulate esters [23].
References
[1] E.M. Anderson, K.M. Larsson, O. Kirk, Biocatal. Biotransform. 16 (1998) 181–204.
[2] R. Kourist, H. Brundiek, U.T. Bornscheuer, Eur. J. Lipid Sci. Technol. 112 (2010)
64–74.
[3] M. Skjøt, L. De Maria, R. Chatterjee, A. Svendsen, S.A. Patkar, P.R. Østergaard, J.
Brask, ChemBioChem 10 (2009) 520–527.
[4] S. Lutz, Tetrahedron: Asymmetr. 15 (2004) 2743–2748.
[5] P.B. Juhl, K. Doderer, F. Hollmann, O. Thum, J. Pleiss, J. Biotechnol. 150 (2010)
474–480.
[6] S. Patkar, J. Vind, E. Kelstrup, M.W. Christensen, A. Svendsen, K. Borch, O. Kirk,
Chem. Phys. Lipids 93 (1998) 95–101.
[7] D. Liu, P. Trodler, S. Eiben, K. Koschorreck, M. Müller, J. Pleiss, S.C. Maurer, C.
Branneby, R.D. Schmid, B. Hauer, ChemBioChem 11 (2010) 789–795.
[8] W.C. Suen, N. Zhang, L. Xiao, V. Madison, A. Zaks, Protein Eng. Des. Sel. 17 (2004)
133–140.
[9] Z. Qian, C.J. Fields, S. Lutz, ChemBioChem 8 (2007) 1989–1996.
[10] Z. Qian, S. Lutz, J. Am. Chem. Soc. 127 (2005) 13466–13467.
[11] Y. Yu, S. Lutz, Biotechnol. Bioeng. 105 (2010) 44–50.
[12] Y. Shimada, Y. Watanabe, A. Sugihara, Y. Tominaga, J. Mol. Catal. B: Enzym. 17
(2002) 133–142.
[13] C.C. Akoh, S.-W. Chang, G.-C. Lee, J.-F. Shaw, J. Agric. Food. Chem. 55 (2007)
8995–9005.
[14] H. Fukuda, A. Kondo, H. Noda, J. Biosci. Bioeng. 92 (2001) 405–416.
[15] K.R. Jegannathan, C. Eng-Seng, P. Ravindra, Renew. Sust. Energ. Rev. 15 (2011)
745–751.
5. Conclusions
[16] S. Jain, M.P. Sharma, Renew. Sustain. Energ. Rev. 14 (2010) 3140–3147.
[17] H.T. Osborn, C.C. Akoh, Compr. Rev. Food Sci. Food Saf. 1 (2002) 110–120.
[18] H. Houte, V. Partali, H.R. Sliwka, E.G.K. Quartey, Chem. Phys. Lipids 105 (2000)
105–113.
[19] J.A. Laszlo, K.O. Evans, K.E. Vermillion, M. Appell, J. Agric. Food. Chem. 58 (2010)
5842–5850.
[20] J.A. Laszlo, M. Jackson, R.M. Blanco, J. Mol. Catal. B: Enzym. 69 (2011) 60–65.
[21] D.L. Compton, J.A. Laszlo, M.A. Berhow, J. Am. Oil Chem. Soc. 83 (2006) 753–758.
[22] C. Gao, P. Mayon, D.A. MacManus, E.N. Vulfson, Biotechnol. Bioeng. 71 (2000)
235–243.
[23] M. Martinelle, K. Hult, Biochim. Biophys. Acta 1251 (1995) 191–197.
[24] J.-W. Chen, W.-T. Wu, J. Biosci. Bioeng. 95 (2003) 466–469.
[25] I. Karabulut, G. Durmaz, A. Hayaloglu, J. Am. Oil Chem. Soc. 87 (2010)
1301–1307.
[26] Y. Yu, Y. Zheng, J. Quan, C.-Y. Wu, Y.-J. Wang, C. Branford-White, L.-M. Zhu, J.
Am. Oil Chem. Soc. 87 (2010) 1443–1449.
[27] B. Guyot, B. Bosquette, M. Pina, J. Graille, Biotechnol. Lett. 19 (1997) 529–532.
[28] M. Ardhaoui, A. Falcimaigne, J.M. Engasser, P. Moussou, G. Pauly, M. Ghoul,
Biocatal. Biotransform. 22 (2004) 253–259.
[29] Z. Qian, J.R. Horton, X. Cheng, S. Lutz, J. Mol. Biol. 393 (2009) 191–201.
[30] A. Marty, V. Dossat, J.-S. Condoret, Biotechnol. Bioeng. 56 (1997) 232–237.
[31] E. Séverac, O. Galy, F. Turon, C.A. Pantel, J.-S. Condoret, P. Monsan, A. Marty,
Enzyme Microb. Technol. 48 (2011) 61–70.
[32] Z.-Q. Duan, W. Du, D.-H. Liu, Process Biochem. 45 (2010) 1923–1927.
[33] R.T. Otto, H. Scheib, U.T. Bornscheuer, J. Pleiss, C. Syldatk, R.D. Schmid, J. Mol.
Catal. B: Enzym. 8 (2000) 201–211.
A detailed kinetic analysis demonstrated that the cp283 vari-
ant, which demonstrated very high catalytic performance relative
to wild-type CALB in hydrolytic reactions involving bulky leaving
groups [10], also retains this advantage for reactions of fatty acid
esters with glycerol. Cp283 displayed a sensitivity to substrate inhi-
bition by glycerol that was not evident in wild-type CALB. While
circular permutation has proven quite effective in identifying a
CALB variant with enhanced rates of ester and triglyceride hydrol-
ysis and synthesis, the same variant performed largely at levels
similar to wild type enzyme on sterically hindered arylaliphatic
esters such as ferulates. Whether these functional discrepancies
result from biases in the screening of circular permuted CALB for
tributyrin hydrolysis or whether these difference are a reflection of
fundamental aspects related to this protein engineering strategy is
not clear. Nevertheless, cp283 shows good activity for ␥-linolenate
which warrants further development of the engineered biocat-
alyst. Separately, the cause of the inhibitory effects of glycerol
observed for cp283 is unclear. Spectroscopic analysis indicated no
direct interaction between substrates and the termini of CP283 [9].