Lipase chemoselectivity towards alcohol and thiol acyl acceptors in a transacylation reaction

Article abstract of DOI:10.1016/j.molcatb.2010.04.005

The lipase chemoselectivity towards an alcohol and a thiol was investigated for the two lipases Candida antarctica lipase B (CalB) and Rhizomucor miehei lipase (Rml). Hexanol and hexanethiol were used as acyl acceptors in a transacylation reaction with ethyl octanoate in cyclohexane. CalB showed the highest chemoselectivity ratio (k_cat/K_M) _OH/(k_cat/K_M)_SH, of 88,000 while the ratio for Rml was 1200. That could be compared with the ratio, k _OH/k_SH, of 120 for the non-catalyzed reaction. Thus, the enzyme contribution to the chemoselectivity between hexanol and hexanethiol was 730 for CalB and 10 for Rml. High K_M values displayed towards hexanethiol (above 1.8 M) were the largest contribution to the selectivity. No saturation was achieved. The K_M values were more than two orders of magnitude higher than those of hexanol.

Full text of DOI:10.1016/j.molcatb.2010.04.005

Journal of Molecular Catalysis B: Enzymatic 66 (2010) 120–123
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Lipase chemoselectivity towards alcohol and thiol acyl acceptors in a
transacylation reaction
∗
Cecilia Hedfors, Karl Hult, Mats Martinelle
Royal Institute of Technology, School of Biotechnology, Department of Biochemistry, AlbaNova University Center, SE-106 91, Stockholm, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
The lipase chemoselectivity towards an alcohol and a thiol was investigated for the two lipases Candida
antarctica lipase B (CalB) and Rhizomucor miehei lipase (Rml). Hexanol and hexanethiol were used as
acyl acceptors in a transacylation reaction with ethyl octanoate in cyclohexane. CalB showed the highest
chemoselectivity ratio (kcat/KM)OH/(kcat/KM)SH, of 88,000 while the ratio for Rml was 1200. That could be
compared with the ratio, kOH/kSH, of 120 for the non-catalyzed reaction. Thus, the enzyme contribution
to the chemoselectivity between hexanol and hexanethiol was 730 for CalB and 10 for Rml. High KM
values displayed towards hexanethiol (above 1.8 M) were the largest contribution to the selectivity. No
saturation was achieved. The KM values were more than two orders of magnitude higher than those of
hexanol.
Received 26 November 2009
Received in revised form 8 April 2010
Accepted 8 April 2010
Available online 18 May 2010
Keywords:
Chemoselectivity
Acylation
Lipase
Kinetics
© 2010 Elsevier B.V. All rights reserved.
Active site titration
1
. Introduction
The catalytic properties of triacylglycerol lipases have been
functionalized polymers is traditionally performed using a protec-
tion group on the thiol [18,19]. We have previously reported on the
synthesis of thiol end-functionalized polyesters by the chemose-
lective Candida antarctica lipase B in an one-pot reaction without
protection chemistry [20–22]. Mercapto-alcohols were used as ini-
tiators in the ROP of lactones, giving thiol functionalized polymers
in high yields. The selectivity displayed by the lipase favoured the
hydroxyl group compared with the thiol group.
The use of alcohols as acyl acceptors in lipase catalyzed acyl
transfer reactions is well documented. A number of Vmax and KM
values for several lipases have been reported, even though kcat
values based on active-site titration are rare. However, the cor-
responding kinetic data for thiols are very limited [23]. Further,
there are many enantiomeric ratios (E) reported for lipases, while
chemoselectivity data for lipases are not available in literature.
In order to document the chemoselectivity displayed by lipases
towards alcohols and thiols, we here report on the selectivity dis-
played by the two lipases C. antarctica lipase B and Rhizomucor
miehei lipase towards hexanol and hexanethiol as acyl acceptors
in acyl transfer reactions.
extensively explored and they are currently used as efficient cata-
lysts in a number of industrial processes [1–3]. In aqueous solutions
they hydrolyse ester bonds, while in an organic environment ester
bonds are synthesised. Theirability toperform regioselective, enan-
tioselective and chemoselective catalysis is of outermost synthetic
interest. This is used in many areas, like synthesis of chiral drugs
[
4], kinetic resolution of alcohols, acids, esters and amines [5,6]
and regioselective synthesis of sugar esters [7]. In polymer sci-
ence, polyester synthesis using lipases has been documented as
an efficient route of making polyesters by polycondensation or
ring-opening polymerisation (ROP) reactions using various differ-
ent substrates/monomers [8–11]. End-group functionalization of
polymers is of great importance for making further complex poly-
mer structures and architectures such as branched and cross-linked
polymers [12]. Using the inherent selectivity displayed by lipases,
simple one-pot routes are available for making functional and chi-
ral polymers [13]. Thiol as a functional group has many interesting
characteristics; it may undergo oxidation, acylation, alkylation and
form complexes with many heavy metals [14,15]. Further, thiols
can react with a variety of enes in a photo-initiated radical reaction
2. Experimental
(
thiol-ene reaction). In polymer science, this is used as an efficient
2.1. Materials
method for making polymer networks [16,17]. Synthesis of thiol
Immobilized lipases were bought from Sigma–Aldrich; C.
antarctica lipase B (CalB) in form of Novozym 435 and R. miehei
lipase (Rml) in form of Lipozyme. The water activity of the two
enzyme preparations was set to 0.11, using saturated LiCl in a desic-
∗ Corresponding author. Fax: +46 8 5537 8468.
E-mail address: matsm@biotech.kth.se (M. Martinelle).
1
381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.04.005

C. Hedfors et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 120–123
121
Table 1
Active-site titration using methyl 4-methylumbelliferyl hexylphosphonate.
Lipase
Amount active lipase on the carrier
nmol/g carrier
weight (%)
Novozym 435 (CalB)
Lipozyme (Rml)
1,000
35
3.3
0.14
determine the activity difference between the incubations with
and without inhibitor. This was done to determine the fraction of
inhibited enzyme.
2.2.2. Acylation of hexanol and hexanethiol
Typically 2–20 mg immobilized lipase in 3 or 6 ml cyclohex-
ane containing hexanol (0.001–0.1 M) or hexanethiol (0.1–1.8 M)
and decane as internal standard were allowed to stir for 5 min.
The reaction was started by adding 0.5 M ethyl octanoate and was
performed at 20 C. All substrates and solvent used were dried
with molecular sieves (3 Å). Samples were withdrawn and filtered
through glass-wool. Analysis was performed on a Hewlett Packard
Scheme 1. Inhibition of lipases using methyl 4-methylumbelliferyl hexylphospho-
nate. The scheme illustrates the formation of the covalently modified active serine
as a stable phosphonate intermediate. The 4-methylumbelliferonate is formed in a
◦
1
:1 ratio to active enzyme.
5
890 Series II gas chromatograph with a Chrompack CP-SIL 5CB
column (25 m × 0.32 mm). Inlet and detector temperatures were
◦ ◦
set at 250 C. The temperature program started at 40 C for 5 min,
cator. All solvents and substrates were dried over molecular sieves
3 Å) for at least one week before use.
◦ ◦ ◦ ◦
increased by 1 C/min to 55 C, then by 5 C/min to 150 C and finally
(
◦
◦
by 10 C/min to 300 C, where it was kept for 5 min. The reten-
tion times were hexanol 17.5 min, hexanethiol 21.7 min, decane
2
2
.2. Methods
2
7.6 min, ethyl octanoate 34.5 min, hexyl octanoate (alcohol prod-
.2.1. Active-site titration
uct) 43.1, and S-hexyl thiooctanoate (thiol product) 45.4 min.
Immobilized lipase (10–20 mg) was mixed with 1 ml ace-
tonitrile and methyl 4-methylumbelliferyl hexylphosphonate was
added (25 ␮M) which was 3–40 times the lipase concentration.
Acetonitrile was used as solvent due to the insolubility of the active-
site inhibitor, methyl 4-methylumbelliferyl hexylphosphonate, in
cyclohexane. In order to totally inhibit the lipase, an addition of
3. Results and discussion
3.1. Active site titration of immobilized lipases in organic solvents
The inhibition of the immobilized lipase preparations, Novozym
435 (CalB) and Lipozyme (Rml), using the inhibitor methyl 4-
methylumbelliferyl hexylphosphonate is illustrated in Scheme 1
[24,25]. The amount of active lipase differed greatly between the
enzyme preparations, 3.3 wt% on Novozym 435 (CalB) and 0.14 wt%
on Lipozyme (Rml) (Table 1). The inhibition was slow compared to
free CalB in aqueous media (unpublished data). Plateau values were
achieved after four days inhibition. By then >99% of the Novozym
435 (CalB) and 91% of the Lipozyme (Rml) preparations had been
inhibited.
1
% water to the solvent was needed. Controls with only inhibitor
and only lipase were treated in the same way to get the back-
ground hydrolysis of the inhibitor and the fluorescence from the
◦
carrier. All samples were incubated in darkness at 20 C. Aliquots
100 ␮l) from the inhibition reaction (during four days) was added
to a cuvette containing 900 ␮l buffer (100 mM Tris–HCl, pH 8,
mM CaCl ). The active-site concentration was determined from
(
1
2
the fluorescence intensity (Ex. 360 nm, Em. 445 nm) of the leav-
ing group 4-methylumbelliferonate (see Scheme 1). A standard
curve of 4-methylumbelliferone was prepared freshly. All mea-
surements were performed on a Perkin Elmer LS50B fluorescence
spectrophotometer. After fluorescence measurements, the immo-
bilized lipase was filtered off and allowed to dry at 0.11. Test
reactions (0.016 M hexanol and 0.5 M ethyloctanoate in cyclo-
hexane using decane as internal standard) were performed to
3.2. Acyl transfer reactions
The mechanistic route for lipase catalyzed acyl transfer reac-
tions is illustrated in Scheme 2. With hexanol as acyl acceptor both
lipases, CalB and Rml, displayed saturation kinetics. The apparent
Scheme 2. Lipase catalyzed transacylation reactions using hexanol (X = O) or hexanethiol (X = S) as acyl acceptors.

1
22
C. Hedfors et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 120–123
Table 2
ester of vinyl octanoate was used, since the spontaneous reaction
with ethyl octanoate could not be detected within ten days. The
ratio of the rate constants, kOH/kSH, of the uncatalysed reaction with
vinyl octanoate was 120. Thus, the enzymatic contribution to the
chemoselectivity ratio was 10 for Rml and 730 for CalB (Table 3).
For both lipases, the largest contribution to the chemoselectivity
was the higher KM values displayed towards hexanethiol compared
with hexanol. The KM ratio was more than two orders of magnitude.
For CalB the selectivity may entirely be a KM effect, as no satura-
tion was achieved with the thiol. This is in line with results using
a thioester as acyl donor in transacylation reactions giving similar
kcat-value as the corresponding oxy ester but a 15-fold higher KM
Apparent kinetic constants for the acyl transfer using 0.5 M ethyl octanoate as
acyl donor and 0.001–0.1 M hexanol or 0.1–1.8 M hexanethiol as acyl acceptor in
cyclohexane.
(kcat/KM)app (s−1
M
−1
)
k
app
cat
(s )a
−1
K
app
M
(M)a
Lipase
CalB
Hexanol
Hexanethiol
710
14 ± 9
0.019 ± 0.002
>1.8
0.0081b
Rml
Hexanol
Hexanethiol
16,000
130 ± 10
0.0084 ± 0.002
>1.8
b
13
–
a
Non-linear regression of Michaelis–Menten equation.
Calculated from rates as a function of substrate concentrations far below KM.
b
[
31].
Table 3
4. Conclusion
Chemoselectivity between hexanol and hexanethiol in an acyl transfer reaction with
ethyl octanoate.
The chemoselectivity towards hexanol and hexanethiol dis-
Lipase
Chemoselectivity
Ratios relative to
uncatalyzed
played by two immobilized lipase preparations, C. antarctica lipase
B (CalB) in form of Novozym 435 and R. miehei lipase (Rml) in
form of Lipozyme, was investigated. The enzyme contribution
to the chemoselectivity ratio between hexanol and hexanethiol
(kcat/KM)OH/(kcat/KM)SH
CalB
Rml
88,000
1,200
730
10
(
kcat/KM)OH/(kcat/KM)SH, was 10 for Rml and 730 for CalB. Most of the
kOH/kSH
Uncatalyzed
120a
selectivity was a result of high KM values for hexanethiol (>1.8 M),
which were at least two orders of magnitude higher than those for
hexanol.
1
a
kOH/kSH, background reactions with no enzyme using vinyl octanoate as acyl
donor. No product was detected within ten days of reaction using ethyl octanoate
as acyl donor.
Acknowledgement
kinetic parameters kcat and KM were determined (Table 2). In con-
trast, neither of the lipases were saturated when using hexanethiol
as acyl acceptor. For both lipases the reaction rate increased linearly
with hexanethiol concentrations up to 1.8 M and consequently
their KM values towards the thiol were above that concentration.
Concentrations above 1.8 M hexanethiol were not used since it
would alter the reaction conditions severely. Thus, only apparent
kcat/KM values from the slope of the Michaelis–Menten plot could be
The research has been supported by the Swedish Research Coun-
cil.
References
[1] A. Schmid, J.S. Dordick, B. Hauer, A. Kiener, M. Wubbolts, B. Witholt, Nature 409
(
2001) 258–268.
[
[
2] A.J.J. Straathof, S. Panke, A. Schmid, Curr. Opin. Biotechnol. 13 (2002) 548–556.
3] A. Houde, A. Kademi, D. Leblace, Appl. Biochem. Biotechnol. 118 (2004)
155–170.
−
1
−1
M
derived for hexanethiol, and these were found to be 0.0081 s
−1
−1
for CalB and 13 s
M for Rml. Based on the amount of active
[
[
4] V. Gotor-Fernández, R. Brieva, V. Gotor, J. Mol. Catal. B: Enzym 40 (2006)
lipase, Rml was more efficient in the acylation of both hexanol (23
times higher kcat/KM) and hexanethiol (1600 times higher kcat/KM),
compared with CalB.
Both lipases displayed KM-values towards hexanethiol that
were more than two orders of magnitude higher than for hexanol
111–120.
5] V. Gotor-Fernández, E. Busto, V. Gotor, Adv. Synth. Catal. 348 (2006) 797–812.
[6] A. Ghanem, Tetrahedron 63 (2007) 1721–1754.
[
[
[
7] S. Riva, J. Mol. Catal. B: Enzym. 19–20 (2002) 43–54.
8] H. Uyama, S. Kobayashi, Chem. Lett. (1993) 1149–1150.
9] D. Knani, A.L. Gutman, D.H. Kohn, J. Polym. Sci. Part A: Pol. Chem. 31 (1993)
1221–1232.
(
Table 2). Causette et al. reported that Rml displayed an apparent
[10] S. Kobayashi, Macromol. Rapid Commun. 30 (2009) 237–266.
[11] S. Matsumura, Adv. Polym. Sci. 194 (2006) 95–132.
[12] I. Taniguchi, W.A. Kuhlman, A.M. Mayes, L.G. Griffith, Polym. Int. 55 (2006)
1385–1397.
KM for butanthiol in hexane of 1.85 M [23]. Lipases are hydro-
lases that cleave ester bonds in triacylglycerols with use of water.
It has been found in several lipases, i.e. Rml and CalB, that they
have an amphiphilic active-site with a hydrophobic part for lipid
binding and a hydrophilic part, like a water channel, for water
binding [26–28]. We hypothesize that the hydrophilic site around
the active serine in lipases is the reason for the high KM val-
ues found for thiols. Thiols are more hydrophobic than alcohols.
[
[
13] B.A.C. van As, P. Thomassen, B. Kalra, R.A. Gross, E.W. Meijer, A.R.A. Palmans, A.
Heise, Macromolecules 37 (2004) 8973–8977.
14] G.L. Kenyon, T.W. Bruice, Method Enzymol. 47 (1977) 407–430.
[15] P.C. Jocelyn, Biochemistry of the SH Group: The Occurrence, Chemical Proper-
ties, Metabolism and Biological Function of Thiols and Disulphides, Academic
Press Inc., London, 1972.
[
16] C.E. Hoyle, T.L. Lee, T.J. Roper, J. Polym. Sci. Part A: Pol. Chem. 42 (2004)
301–5338.
tr
The transfer free energy (ꢀG ) for thiols when transfered from
5
−
1
cyclohexane to water are typically 15–20 kJ mol higher than for
the corresponding alcohols [29,30]. This means that the partition-
ing from the reaction phase (cyclohexane) to the catalyst phase
[17] N. Simpson, M. Takwa, K. Hult, M. Johansson, M. Martinelle, E. Malmström,
Macromolecules 41 (2008) 3613–3619.
[
18] M. Trollsås, C.J. Hawker, J.L. Hedrick, G. Carrot, J. Hilborn, Macromolecules 31
1998) 5960–5963.
[19] G. Carrot, J. Hilborn, J.L. Hedrick, M. Trollsås, Macromolecules 32 (1999)
171–5173.
20] C. Hedfors, E. Östmark, E. Malmström, K. Hult, M. Martinelle, Macromolecules
8 (2005) 647–649.
(
(
hydrophilic water channel site) will be less favourable for thiols
5
than for alcohols. Based on this simple model 2–3 orders of magni-
tude higher KM-values for thiols as compared with alcohols could
be predicted.
[
3
[21] M. Takwa, N. Simpson, E. Malmström, K. Hult, M. Martinelle, Macromol. Rapid
Commun. 27 (2006) 1932–1936.
[
[
22] M. Takwa, K. Hult, M. Martinelle, Macromolecules 41 (2008) 5230–5236.
23] M. Caussette, A. Marty, D. Combes, J. Chem. Technol. Biotechnol. 68 (1997)
3.3. Chemoselectivity
2
57–262.
24] R. Fujii, Y. Utsunomiya, J. Hiratake, A. Sogabe, K. Sakata, Biochim. Biophys. Acta
631 (2003) 97–205.
25] A.O. Magnusson, J. Rotticci-Mulder, A. Sanagostino, K. Hult, ChemBioChem 6
(2005) 1–7.
[
[
The chemoselectivity ratio (kcat/KM)OH/(kcat/KM)SH was calcu-
lated to be 1200 for Rml and 88,000 for CalB (Table 3). To quantify
the chemoselectivity of the un-catalyzed reaction, the activated
1

C. Hedfors et al. / Journal of Molecular Catalysis B: Enzymatic 66 (2010) 120–123
123
[
[
[
26] M. Norin, F. Haeffner, A. Achour, T. Norin, K. Hult, Protein Sci. 3 (1994)
493–1503.
27] M. Wittrup Larsen, D.F. Zielinska, M. Martinelle, A. Hidalgo, L.J. Jensen, U.T.
Bornscheuer, K. Hult, ChemBioChem 11 (2010) 796–801.
[29] W. Han, C.-K. Wan, Y.D. Wu, J. Chem. Theory Comput. 4 (2008) 1891–1901.
[30] A.V. Marenich, R.M. Olson, C.P. Kelly, C.J. Cramer, D.G. Truhlar, J. Chem. Theory
Comput. 3 (2007) 2011–2033.
1
[31] M. Martinelle, K. Hult, Biochim. Biophys. Acta 1251 (1995) 191–197.
28] J. Uppenberg, M.T. Hansen, S. Patkar, T.A. Jones, Structure 2 (1994) 293–308.