Poly(3-hydroxybutyrate)-depolymerase from Pseudomonas lemoignei: Catalysis of esterifications in organic media

Article abstract of DOI:10.1021/jo000814y

Lipase catalysis in nonaqueous media is recognized as a powerful tool in organic and more recently polymer synthesis. Even though none of the currently known polyhydroxyalkanoate (PHA) depolymerases have lipase activity, they do have a catalytic center that resembles that of lipases. Motivated by the above, the potential of using the poly(3-hydroxybutyrate), PHB, depolymerase from Psuedomonas lemoignei in organic media to catalyze ester-forming reactions was investigated. The effect of different organic solvents (benzene-d₆, cyclohexane-d₁₂, and acetonitrile-d₃) on the activity of the PHB-depolymerase toward propylation of L-lactide was studied. A significant difference in the catalytic rate was observed as a function of solvent polarity. The selectivity of the PHB-depolymerase (P. lemoignei) to catalyze the propylation of a series of different lactones including ε-caprolactone, δ-butyrolactone, γ-butyrolactone, and D, L, meso, and racemic lactides has been studied with the PHB-depolymerase (P. lemoignei) in organic solvents. Important differences in the reactivity of these lactones, as well as selective hydrolysis of stereochemically different linear lactic acid dimers, were observed. Moreover, the ability of the PHB-depolymerase to catalyze the solventless polymerization of ε-caprolactone and trimethylene carbonate was investigated.

Full text of DOI:10.1021/jo000814y

7800
J . Org. Chem. 2000, 65, 7800-7806
P oly(3-h yd r oxybu tyr a te)-d ep olym er a se fr om
P seu d om on a s lem oign ei: Ca ta lysis of Ester ifica tion s in Or ga n ic
Med ia
Ajay Kumar,† R. A. Gross,*,† and D. J endrossek‡
Department of Chemical Engineering, Chemistry and Material Science, Polytechnic University,
Brooklyn, New York, 11201, and Institut fu¨r Mikrobiologie, Universita¨t Stuttgart, Allmandring 31,
70550 Stuttgart, Germany
rgross@poly.edu
Received May 30, 2000
Lipase catalysis in nonaqueous media is recognized as a powerful tool in organic and more recently
polymer synthesis. Even though none of the currently known polyhydroxyalkanoate (PHA)
depolymerases have lipase activity, they do have a catalytic center that resembles that of lipases.
Motivated by the above, the potential of using the poly(3-hydroxybutyrate), PHB, depolymerase
from Psuedomonas lemoignei in organic media to catalyze ester-forming reactions was investigated.
The effect of different organic solvents (benzene-d6, cyclohexane-d12, and acetonitrile-d3) on the
activity of the PHB-depolymerase toward propylation of L-lactide was studied. A significant
difference in the catalytic rate was observed as a function of solvent polarity. The selectivity of the
PHB-depolymerase (P. lemoignei) to catalyze the propylation of a series of different lactones
including ꢀ-caprolactone, δ-butyrolactone, γ-butyrolactone, and D, L, meso, and racemic lactides
has been studied with the PHB-depolymerase (P. lemoignei) in organic solvents. Important
differences in the reactivity of these lactones, as well as selective hydrolysis of stereochemically
different linear lactic acid dimers, were observed. Moreover, the ability of the PHB-depolymerase
to catalyze the solventless polymerization of ꢀ-caprolactone and trimethylene carbonate was
investigated.
In tr od u ction
conserved amino acids, namely, serine, aspartate, and
histidine, constitute the catalytic triad at the active
center of the catalytic domain. The conserved serine is
part of the so-called lipase-box pentapeptide (Gly-Xaal-
Ser-Xaa2-Gly) that has been found in all known serine
hydrolases such as lipases, esterases, and serine-pro-
teases. The hydroxyl of this residue that attacks the ester
bond of an enzyme substrate is increased in nucleophi-
licity through general-base catalysis by the imidazole ring
of histidine. The positive charge on the imidazole ring is
stabilized by the carboxylate group of the aspartate, and
the resulting negatively charged tetrahedral carbon atom
of the transient state is stabilized by a hydrogen bond to
a main chain peptide bond (oxyanion hole). Thus, the
catalytic centers of PHA-depolymerases resemble those
of lipases. However, none of the currently known PHA-
depolymerases have lipase activity. In contrast, some
lipases have PHA-depolymerase activity when the sub-
strates are ω-hydroxyacids, e.g., poly(4-hydroxybutyrate)
(PHB) or poly(6-hydroxyhexanoate).5 Apparently, lipases
and PHA-depolymerases differ in the spatial arrange-
ment of their active sites. Also, lipases generally do not
accept polymeric substrates having side chains along the
carbon backbone.3
Motivations to minimize the environmental impact of
chemical processes have spawned impressive innovations
in biocatalytic reactions using both whole cell and in vitro
catalysis.1 An interesting example has been the microbial
synthesis and enzymatic degradation of polyhydroxy-
alkanoates (PHAs).2 The ability to degrade polyesters in
the PHA family is widely distributed among different
microorganisms and depends on the secretion of specific
hydrolases, which degrade the polymer to water-soluble
products. Many PHA-depolymerases from bacteria and
fungi have been characterized.3,4
PHA-depolymerases have structures that consist of (1)
a large catalytic domain, (2) a linking domain, and (3) a
C-terminal substrate-binding domain. Three strictly
† Polytechnic University.
‡ Universita¨t Stuttgart.
(1) Waldmann, H. Enzyme catalysis in organic synthesis; Drauz, K.,
Waldmann, H., Eds.; VCH: Weinheim, 1995; Vols. I & II. (b) Balles-
teros, A.; Bernabe, M.; Cruzado, C.; Lomas, M. M.; Otero, C. Tetrahe-
dron 1989, 45, 7077-7082. (c) Chenvert, R.; Desjardins, M.; Gagnon,
R. M. Chem. Lett. 1990, 33-34. (d) Moris, F.; Gotor, V. Tetrahedron
1992, 48, 9869-9876. (e) Hiroyuki, A.; Isao, U.; Masako, N. Tetrahe-
dron: Asymmetry 1993, 4, 757-760. (f) Ader, U.; Andersch, P.; Berger,
M.; Georgens, U.; Seemayer, R.; Schneider, M. P. Pure Appl. Chem.
1992, 64, 1165-1170. (g) Kliabanov, A. M. Acc. Chem. Res. 1990, 23,
114-120. (h) Parmar, V. S.; Kumar, A.; Bisht, K. S.; Mukherjee, S.;
Prasad, A. K.; Sharma, S. K.; Wengel, J .; Oleson, C. E. Tetrahedron
1997, 53, 2163-2176.
The utility of hydrolases, most notably lipases and
proteases, for both ester and amide bond formation in
nonaqueous media, continues to increase in technological
importance. This is reflected in the wide participation of
many research groups in this area and the resulting large
(2) (a) Anderson, A. J .; Dawes, E. A. Microbiol. Rev. 1990, 54, 450-
472. (b) Madison, L. L.; Huisman, G. W. Microbiol. Mol. Rev. 1999,
63, 21-53.
(3) J endrossek, D.; Schirmer, A.; Schlegel, H. G. Appl. Microbiol.
Biotechnol. 1996, 46, 451-463.
(5) J aeger, K. E.; Steinbuechel, A.; J endrossek, D. Appl. Environm.
Microbiol. 1995, 61, 3113-3118.
(4) J endrossek, D. Polym. Degrad. Stab. 1998, 59, 317-325.
10.1021/jo000814y CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/21/2000

Catalysis of Esterifications in Organic Media
J . Org. Chem., Vol. 65, No. 23, 2000 7801
purification with acetone ether.10 The weight-average (Mw) and
the polydispersity index (Mw/Mn) of the isolated polyesters
determined by gel permeation chromatography using polysty-
rene standards were 2.3 × 106 g/mol and 2.1, respectively. The
activity of the enzyme was also measured by determination
of the decrease in the optical density (660 nm) of natural PHB
suspensions in TRIS-HCl buffer (50 mM), CaCl2 (1 mM), pH
8.0.10 Activity of the lyophylysate after incubation for 3 h at
70 °C in benzene was measured. The assay was performed by
taking 2 µL of the PHB-depolymerase solution (stock solution
1 mg/0.02 mL) in water and dropping it onto the surface of
the plate. After the drop had diffused into the agar, the Petri
disk was incubated in an oven at about 37 °C. Clearing zone
formation was visible after 1 h, indicating strong activity. For
comparison, the activity of the lyophylysate was determined
after incubating at room temperature for 3 h. It was found
that the clearing zones for both the lyophylysates after
incubating for 3 h at room temperature and 70 °C in benzene
were identical. Hence, thermal inactivation under these condi-
tions was not observed.
number of publications.1 The rationale for employing
enzyme-catalyzed reactions in organic media, as well as
the manipulation of the solvent character to control these
reactions, has been discussed at length elsewhere.6 Since
lipases and proteases are such powerful catalysts for
organic transformations, it seems critical to extend our
knowledge of how other ester hydrolase families may
contribute new attributes to enzyme catalysis in non-
aqueous media. Hence, in this paper, we prepared and
purified a PHB-depolymerase from Pseudomonas lem-
oignei and studied its activity and specificity for the
catalysis of esterification reactions of cyclic esters and
carbonates in organic media. To the best of our knowl-
edge, the use of PHB-depolymerases in nonaqueous
media for bond formation has not yet been investigated.
Ma ter ia l a n d Meth od s
P r ep a r a tion of (R)-P r op yl 3-Hyd r oxybu tr a te (14). The
propanolysis of natural PHB under acid conditions was
Gen er a l Ch em ica ls a n d P r oced u r es. All chemicals and
solvents were of analytical grade and were used as received
unless otherwise noted. ꢀ-Caprolactone donated by Union
Carbide was distilled at 97-98 °C over CaH2 at 10 mmHg.
Lactides donated by Cargill were recrystallized from ethyl
acetate before use. â-Butyrolactone, n-propanol, benzene-d6,
cyclohexane-d12, and acetonitrile-d3 were procured from Al-
drich Chemical Co. and were dried over 4 Å molecular sieves
before use. 1H nuclear magnetic resonance (NMR) spectra were
recorded on a Bruker ARX-250 and -200 spectrometers at 250
performed exactly as described elsewhere.11,12 The product was
25
a colorless liquid: [R] ) -39.98 (1.33, CH3C6H5); IR (neat)
589
1
3465, 1753, cm-1; H NMR (250 MHz, C6D6) δ 4.21 (q, 1 H, J
) 6.3 Hz), 3.97 (t, 2 H, J ) 6.7 Hz), 2.31 (ddd, 2 H, J ) 8.2
and 16.0 Hz, 4.1 and 16.0 Hz), 1.45 (m, 2 H), 1.14 (d, 3 H, J )
6.3 Hz), 0.81 (t, 3 H, J ) 6.7 Hz).
P r op yl (3R/3S)-3-h yd r oxybu ta n oa te (14/15) was ob-
25
589
tained as a colorless oil: [R] ) + 9.81 (0.153, CH3C6H5); IR
1
(neat) 3465, 1751 cm-1; 1H NMR (250 MHz, C6D6) δ 4.21 (q, 1
H, J ) 6.3 Hz), 3.97 (t, 2 H, J ) 6.7 Hz), 2.31 (ddd, 2 H, J )
8.2 and16.0 Hz, 4.1 and 16.0 Hz), 1.45 (m, 2 H, J ) 6.7 Hz),
1.14 (d, 3 H, J ) 6.3 Hz), 0.81 (t, 3 H, J ) 6.7 Hz); 13C NMR
(50 MHz, CDCl3) δ 172.86, 66.18, 64.58, 45.71, 22.81, 22.34,
9.71.
and 200 MHz. H NMR chemical shifts (in parts per million
or ppm) are reported downfield from 0.00 ppm using tetra-
methylsilane (TMS) as the internal reference. The instrumen-
tal parameters were as follows: temperature 300 K, pulse
width 7.8 µs (30°), 32 K data points, 3.178-s acquisition time,
1-s relaxation delay, and 16 transients. 13C NMR spectra were
recorded at 62.9 and 50 MHz on a Bruker ARX-250 spectrom-
eter with chemical shifts in ppm referenced relative to TMS
at 0.00 ppm. Infrared (IR) spectra were recorded by using a
Perkin-Elmer FTIR model 1720X spectrometer. Optical rota-
P r epar ation of P r opyl (2R)-2-Hydr oxypr opan oate (10).
In dry chloroform D-lactide (2) was mixed with NaOC3H7 at
room temperature for 30 min. This mixture was then washed
with water, the chloroform layer was evaporated, and com-
25
589
tion values were measured by using a Perkin-Elmer model 241
pound 10 was isolated as a colorless oil: [R] ) +10.8 (0.25,
25
589
CH3C6H5); IR (neat) 3465, 1756 cm-1 1H NMR (250 MHz,
;
polarimeter and are reported as follows: [R]
) specific
rotation (concentration in g/100 mL of solvent). Gas chroma-
tography (GC) experiments were performed on a Hewlett-
Packard instrument fitted with a γ-cyclodextrin chiral column.
P r ep a r a tion a n d Assa y of P HB-d ep olym er a se. P. lem-
oignei was grown in 10 L of Stinson and Merrick’s mineral
medium7 with 50 mM sodium succinate in a Braun Biostat
fermentor (model ES 10) at 30 °C for 24 h. The conditions of
maximal PHB-depolymerse production by P. lemoignei have
recently been described.8 The cell-free culture fluid was
obtained by centrifugation, and the proteins were concentrated
10-fold by ultrafiltration (exclusion size 30 kDa). The concen-
trate was subjected to ammonium sulfate precipitation (20-
60% saturation), and the pellet of the 60% step was dialyzed
against 5 mM MES-TRIS buffer. The supernatant of the
centrifuged dialysate was lyophylized and stored at -20 °C.
After solubilization of the lyophylized enzyme in water or
2-propanol, the activity of the enzyme was assayed by the
clearing zone formation of opaque PHB granules-containing
agar (100 mM Tris-HCl, pH 8.0, 1 mM CaCl2, 1.5% [wt/vol]
agar) as was described in detail elsewhere.9 The PHB granules
were isolated from sodium gluconate grown Ralstonia eutropha
H16 by digestion with sodium hypochlorite and subsequent
CDCl3) δ 4.25 (q, 1 H, J ) 6.9 Hz), 4.12 (t, 2 H, J ) 6.8 Hz),
1.66 (m, 2 H), 1.39 (d, 3 H, J ) 6.9 Hz), 0.93 (t, 3 H, J ) 6.8
Hz); 13C NMR (50 MHz, CDCl3) δ 176.20, 67.46, 67.12, 22.62,
20.75, 10.58.
Gen er a l P r oced u r e for P HB-d ep olym er a se Ca ta lyzed
Ester ifica tion . The PHB-depolymerase catalyzed esterifica-
tions of ꢀ-caprolactone, â- and γ-butyrolactone, and D-, L-, dl-,
and meso-lactides were performed as follows: A glovebox and
dry argon were used to maintain an inert atmosphere during
transfers. The cyclic ester substrate (44 mg) and the PHB-
depolymerase lyophylysate (2 mg), both dried in a vacuum
desiccator (0.1 mmHg, 25 °C, 24 h), were transferred to a dry
NMR tube. The NMR tube was immediately stoppered (rubber
septum) and purged with argon; deuterated solvent (0.8 mL
dried over 4 Å molecular sieve) and 25 µL of n-propanol (dried
and distilled over CaH2) were added via syringe under argon.
The NMR tube was then placed in a constant temperature oil
bath maintained at 70 °C with agitation. The enzyme lyo-
phylysate was dispersed throughout the reaction medium
although it does settle to the bottom of the NMR tube while
spectra are recorded. A control reaction was set up as described
above except PHB-depolymerase was not added. Also, another
control reaction was performed in which the thermally deac-
tivated PHB-depolymerase (depolymerase boiled in water for
24 h, dried) was added in place of the active form. All the above
(6) (a) Sakurai, T.; Margolin, A. L.; Russel, A. J .; Klibanov, A. M. J .
Am. Chem. Soc. 1988, 110, 7236-7237. (b) Terradas, F.; Teston-Henry,
M.; Fitzpatrick, P. A.; Klibanov, A. M. J . Am. Chem. Soc. 1993, 115,
390-396. (c) Wu, S. H.; Chu, F. Y.; Wang, K. T. Bioorg. Med. Chem.
Lett. 1991, 1, 399-405. (d) Tawaki, S.; Klibanov, A. M. J . Am. Chem.
Soc. 1992, 114, 1882-1884.
(7) Stinson, M. W.; Merrick, J . M. J . Bacteriol. 1974, 119, 152-161.
(8) Terpe, K.; Kerkhoff, K.; Pluta, E.; J endrossek, D. Appl. Envi-
ronm. Microbiol. 1999, 65, 1703-1709.
(9) J endrossek, D.; Mueller, B.; Schlegel, H. G. Eur. J . Biochem.
1993, 218, 701-710.
(10) J endrossek, D.; Knoke, I.; Habibian, R. B.; Steinbuchel, A.;
Schlegel, H. G. J . Environm. Polym. Degrad. 1993, 1, 53-63.
(11) Seebach, D.; Roggo, S.; Zimmermann. In Stereochemistry of
Organic and Bioorganic Transformations; Bartmann, W., Sharpless,
K. B., Eds.; VCH: Weinheim, West Germany, 1987; Vol. 17, pp 7-126.
(12) Le Borgne, A.; Spassky, N. Polym. Prepr. (Am. Chem. Soc., Div.
Polym. Chem.) 1988, 29 (1), 598-599.

7802 J . Org. Chem., Vol. 65, No. 23, 2000
Kumar et al.
Sch em e 1
1
25
589
reactions were monitored from time to time by H NMR. The
reactions were terminated by removal of the enzyme by
filtration under vacuum using a fritted-glass filter (medium-
pore porosity, 4.5-5 µm). The enzyme was washed on the
fritted-glass filter by 1 to 2 washings with chloroform (0.5 mL
portions), the filtrates were combined, and the solvent was
removed in vacuo to give the crude products. When needed,
the crude products were purified by silica gel column chroma-
tography.
tained as a colorless oil; [R] ) + 9.2 (0.403, CH3C6H5); IR
(neat) 3515, 1756 cm-1; 1H NMR (200 MHz, CDCl3) δ 5.19 (q,
1 H, J ) 7.1 Hz), 4.36 (q, 1H, J ) 6.9 Hz), 4.12 (t, 2 H, J ) 6.8
Hz), 1.68 (m, 2 H), 1.54 (d, 3 H, J ) 7.1 Hz), 1.50 (d, 3 H, J )
6.9 Hz), 0.95 (t, 3 H, J ) 6.8 Hz); 13C NMR (50 MHz, CDCl3)
δ 175.48), 170.72, 69.77, 67.50, 67.11, 22.31, 20.82, 17.29,
10.62.
P r op yl h exa n oa te (15) was obtained as a colorless oil: IR
(neat) 3465, 1751 cm-1; 1H NMR (250 MHz, C6D6) δ 4.01 (t, 2
H, J ) 6.7 Hz), 3.62 (t, 2 H, J ) 6.4 Hz), 2.31 (t, 2 H, J ) 7.2
Hz), 1.68-1.57 (m, 6 H), 1.44-1.35 (m, 2 H), 0.93 (t, 3 H, J )
6.7 Hz); 13C NMR (50 MHz, CDCl3) δ 174.34, 66.33, 62.87,
34.63, 32.67, 25.79, 25.07, 22.36, 10.77.
P r ep a r a tion of (2S)-2-h yd r oxyp r op a n oic a cid , 2-p r o-
p yloxy-(1S)-1-m eth yl-2-oxoeth ylester (8): obtained as a
25
589
colorless oil; [R]
) -25.87 (0.143, CH3C6H5); IR (neat)
3495, 1761 cm-1; 1H NMR (200 MHz, CDCl3) δ 5.18 (q, 1 H, J
) 7.1 Hz), 4.37 (q, 1 H, J ) 6.9 Hz), 4.13 (t, 2 H, J ) 6.8 Hz),
1.70 (m, 2 H), 1.54 (d, 3 H, J ) 7.1 Hz), 1.50 (d, 3 H, J ) 6.9
Hz), 0.95 (t, 3 H, J ) 6.8 Hz); 13C NMR (50 MHz, CDCl3) δ
175.57, 170.67, 69.87, 67.55, 67.12, 22.28, 20.91, 17.33, 10.66.
P r ep a r a tion of (2R)-2-h yd r oxyp r op a n oic a cid , 2-p r o-
Resu lts a n d Discu ssion s
Studies were performed to determine the catalytic
activity and specificity in organic media of the PHB-
depolymerase from P. lemoignei. The natural substrate
for this enzyme is enantiopure ([R]-antipode) PHB or
structurally related polyhydroxyalkanoates.3,16 This work
p yloxy-(1R)-1-m eth yl-2-oxoeth yl ester (9): obtained as a
25
589
colorless oil; [R]
) +30 (0.323, CH3C6H5); IR (neat) 3505,
1761 cm-1; 1H NMR (250 MHz, CDCl3) δ 5.18 (q, 1 H, J ) 7.1
Hz), 4.36 (q, 1 H, J ) 6.9 Hz), 4.12 (t, 2 H, J ) 6.8 Hz), 1.67
(m, 2 H), 1.53 (d, 3 H, J ) 7.1 Hz), 1.49 (d, 3 H, J ) 6.9 Hz),
0.95 (t, 3 H, J ) 6.8 Hz); 13C NMR (50 MHz, CDCl3) δ 175.0,
170.65, 69.76, 67.46, 67.10, 22.24, 20.78, 17.25, 10.56.
(13) Names of structures in Table 1 and Scheme 1 are given as 1,
(3S-cis)-3,6-dimethyl-1,4-dioxane-2,5-dione; 2, (3R-cis)-3,6-dimethyl-
1,4-dioxane-2,5-dione; 3, (3R,6S)-3,6-dimethyl-1,4-dioxane-2,5-dione; 4,
(()-3,6-dimethyl-1,4-dioxane-2,5-dione; 5, 4-methyl-2-oxetanone; 6,
P r epar ation of (2R/2S)-2-h ydr oxypr opan oic acid, 2-pr o-
p yloxy-(1S/1R)-1-m et h yl-2-oxoet h yl est er (11/12): ob-

Catalysis of Esterifications in Organic Media
J . Org. Chem., Vol. 65, No. 23, 2000 7803
Ta ble P HB- d ep olym er a se Ca ta lyzed Ester ifica tion s
betw een n -P r op a n ol a n d Cyclic Ester s in Or ga n ic
Solven tsa
substrate reactn time (h)
solvent
product % convn
1
1
1
2
35
65
65
70
benzene-d6
8
8
8
98
90
4
cyclohexane-d12
acetonitrile-d3
benzene-d6
9
38
10
11, 12
13
8
22b
59
3
52
48
benzene-d6
benzene-d6
benzene-d6
benzene-d6
8b
4
major
minor
35
11
21
9
5c
140
14
15
16
17
18
6
6
7
72
48
48
18
30
a
Names of structures (1-18) in Table 1 and Scheme 1 are given
in ref 13. Lactic acid ester unit/2. c The enzyme concentration
b
F igu r e 1. Time versus % conversion for the PHB-depoly-
merase catalyzed propylation of L-lactide (1) in different
solvents. (benzene-d6, cyclohexane-d12, acetonitrile-d3).
was increased by 10-fold for this reaction.
focused on the catalytic activity of the PHB-depolymerase
on a series of nonnatural substrates including lactones,
lactides, and a cyclic carbonate (Scheme 1, 1-7). Lactides
(1-4, Scheme 1) differing in stereochemical configuration
were included. Similar to PHB, lactides have secondary
hydroxyl groups attached to the chiral center that
participate in formation of the ester bonds. Furthermore,
the substituents attached to the chiral centers of both
lactide and PHB are methyl groups. In contrast to PHB
which has hydroxyl groups that are â to the carbonyl,
lactides have R-hydroxyl groups and are low molecular
weight cyclic analogues. The â-methyl-substituted bu-
tyrolactone (Scheme 1, 5) is a small molecule lactone
analogue of PHB. Structures 1-5 all have chiral centers
that permitted the interrogation of PHB-depolymerase
enantiospecificity in organic media. ꢀ-Caprolactone
(Scheme 1, 6) is an analogue of structures 1-5 that is
achiral and does not have a methyl substituent. Six-
membered trimethylene carbonate (Scheme 1, 7) was
studied in bulk as a monomer for PHB-depolymerase
catalyzed ring-opening polymerization.
F igu r e 2. Time versus % conversion for the PHB-depoly-
merase catalyzed propylation of L-, D-, and meso-lactide.
The PHB-depolymerase catalyzed esterification of lac-
tone and lactide stereoisomers by n-propanol (see Scheme
1) was investigated in dry deuterated benzene (C6D6),
cyclohexane (C6D12), and acetonitrile (CD3CN) (Table 1).
Although far from comprehensive, these solvents repre-
sent a broad range of solvent polarity. The extent of
nonenzyme-catalyzed reactions was found to be negligible
based on controls that were conducted without adding
PHB-depolymerase. The plot in Figure 1 shows that
increasing the solvent polarity from benzene-d6 or cyclo-
hexane-d12 to acetonitrile-d3 resulted in a large decrease
in reactivity for L-lactide propylation. An increase in the
medium polarity for lipase-catalyzed esterifications also
generally results in decreased enzyme activity.14
The effect of lactide stereochemistry on lactide pro-
pylation catalyzed by the PHB-depolymerase is shown
in Figure 2. For % conversions above ∼30%, the %
conversion increased in the following order: D-lactide <
meso-lactide < L-lactide. When this reaction was con-
ducted using racemic or dl lactide (a 50/50 mixture of D-
and L-lactide) to 50%-conversion, a preferential ring-
opening of the L-lactide was observed. This resulted in
predominantly L-propyl-lactide that had a 26.6% optical
purity. The extent of nonenzyme catalyzed reactions,
based on controls conducted without adding PHB-depoly-
merase, resulted in <3% products yields for L- and
D-lactides, and <9% for meso-lactide (see Figures 4 and
5).
An important question in lactide ring-opening is
whether propyl lactide reacts further to form two mol-
ecules of propyl lactate. The H NMR of L-lactide has the
hexahydro-2H-oxepin-2-one; 7, trimethylene carbonate; 8, (2S)-2-
hydroxypropanoic acid, 2-propyloxy-(1S)-1-methyl-2-oxoethyl ester; 9,
(2R)-2-hydroxypropanoic acid, 2-propyloxy-(1R)-1-methyl-2-oxoethyl
ester; 10, propyl (2R)-2-hydroxypropanoate; 11/12, (2R/2S)-2-hydroxy-
propanoic acid, 2-propyloxy-(1R/1S)-1-methyl-2-oxoethyl ester; 13,
propyl 2-hydroxypropanoate; 14, propyl (3R)-3-hydroxybutanoate; 15,
propyl (3S)-3-hydroxybutanoate; 16, propylhexanoate; 17, polytri-
methylene carbonate; 18, polycaprolactone.
(14) Dong, H.; Cao, S.; Li, Z.; Han, S.; You, D.; Shen, J . J . Polym.
Sci. A 1999, 37, 1265-1275.
(15) DP-TFA-γ-Cyclodextrin column 0.12 µM (40 m × 0.24 mm),
oven temperature 135 °C, injection temperature 160 °C, detection
temperature 170 °C and at 10 psi.
1
methine protons 2 (2) at δ 4.15 in benzene-d6. As the
PHB-depolymerase catalyzed propylation of L-lactide
1
proceeds, the H NMR spectra showed a decrease in the
intensity of the signal at 4.15 with a concomitant increase
in the intensities of the two quartets at δ 5.18 and 4.37
due to the methine protons Hb (8) and Hc (8), respectively.
However, the methine signals due to propyl lactate were
not observed for reactions conducted for 35 h to 98%
(16) Fitting by two Gaussian peaks in Microcal Origin, Version 5.

7804 J . Org. Chem., Vol. 65, No. 23, 2000
Kumar et al.
F igu r e 3. 1H NMR profile of the CH and OCH2 region of compounds 8, 9, and 10 in benzene-d6.
F igu r e 5. Time versus ester formation for the PHB-depoly-
merase/without depolymerase catalyzed propylation of meso-
lactide (3).
F igu r e 4. Time versus % conversion for the PHB-depoly-
merase/without depolymerase catalyzed propylation of D-
lactide (2).
conversion. Furthermore, the only product of the reaction
at 98% conversion had a 13C NMR spectrum that was
consistent with the formation of L-propyl lactide (see
experimental), with signals at 69.87, 67.55, and 67.12 and
175.57 and 170.67 ppm. Thus, although the n-propylation
reaction with L-lactide was rapid compared to similar
reactions performed with meso and D-lactide, L-propyl
lactide was a poor substrate for further reaction with
n-propanol at the internal ester bond.
17% of 10. In contrast, under similar conditions, n-
propylation of L-lactide gave 98% of 8. To confirm that
n-propylation of D-lactide yielded 9 and 10, the reaction
was repeated for 140 h to 80% total conversion. The
products formed were separated by silica gel chromatog-
raphy (eluent: hexane/ethyl acetate in increasing polar-
ity). Two compounds were isolated in the pure form.
1
Compound 9 showed H NMR signals identical to those
of 8 but had an optical rotation of equal magnitude but
in the opposite direction (see experimental). Thus, 9 is
n-propyl D-lactide or (2R)-2-hydroxypropanoic acid, 2-pro-
pyloxy-(1R)-1-methyl-2-oxoethyl ester. The 1H NMR in
benzene-d6 of the coproduct had a quartet centered at δ
4.25 (CHa), a triplet at 3.95 (OCH2), a doublet at 1.40,
and a triplet at 0.75. Further analysis of this product by
13C NMR further confirmed that the coproduct was
The n-propylation of D- and L-lactide proceeded differ-
ently (see Figure 4). As the propylation of D-lactide
1
progressed, in addition to the H NMR quartets centered
at δ 5.16 and 4.38 due to the methine protons Hb (9) and
Hc (9), quartets at δ 4.28 and 4.15 were also observed
(see above and Figure 3). The presence of these signals
(δ 4.28 and 4.15) suggests that in addition to D-propyl
lactide D-propyl lactate was also formed. After a 35 h
reaction, n-propylation of D-lactide yielded 28% of 9 and
D-propyl lactate 10 (propyl (2R)-2-hydroxypropanoate).
25
589
The specific rotation of 10 is [R]
) +10.5 (0.25,

Catalysis of Esterifications in Organic Media
J . Org. Chem., Vol. 65, No. 23, 2000 7805
F igu r e 6. Time versus ester formation for the PHB-depolymerase/without depolymerase catalyzed propylation of 4-methyl-2-
oxetanone (5).
CH3C6H5). This value is consistent with that of D-propyl
omer is formed selectively. High selectivity would result
in the formation of a product that has high stereochem-
ical purity. If this were the case, it would offer an
explanation for our inability to separate the L-D and D-L
stereoisomers.
25
lactate ([R]
) +10.8 (0.25, CH3C6H5) synthesized by
589
us using another route (see Materials and Methods
section). Thus, the PHB-depolymerase is able to convert
L-lactide to n-propyl L-lactide without further cleavage
of the internal ester. In contrast, the PHB-depolymerase
catalyzes the conversion of D-lactide to n-propyl D-lactide,
but this product is further converted to 2 equiv of
n-propyl D-lactate.
â-Butyrolactone is the single repeat unit cyclic ana-
logue of natural PHB. Hence, the activity of the PHB-
depolymerase for the enantioselective n-propylation of
[R]- and [S]-â-BL seems particularly relevant to the
enzymes natural substrate preferences under aqueous
conditions. The propylation of racemic-â-BL proceeded
so slowly that, for comparison to lactide substrates, it was
necessary to increase the relative enzyme concentration
by a factor of 10 times. Even with this disparity in
enzyme concentrations, n-propylation of racemic-â-BL
still proceeded slower than that observed for any of the
lactide stereoisomers. The reaction was terminated at
∼50% monomer conversion. The enantioselectivity of
â-BL n-propylation was assessed by analyzing the tri-
fluoroacetyl derivatives of 14 and 15 using a chiral
γ-cyclodextrin GC column.15 The resulting GC trace
shows partial resolution of 14 and 15, giving two peaks
at 16.57 and 16.93 min (see Figure 6). On the basis of
the relative peak areas, analyzed by using Peak Fit
software,16 the enantiomeric excess was 23.5%. The
n-propyl-â-HB stereoisomers from the reaction mixture
were purified by column chromatography (silica gel
The potential exists that the reaction of n-propanol
with meso-lactide (D,L-isomer) will occur with preferential
attack at the D- or L-lactide stereocenters. In other words,
depending on the enantioselectivity of the PHB-depoly-
merase, this reaction could give an unequal proportion
of the diastereomers n-propyl-[L-D]-11, or n-propyl-[D-L]-
12 (see Scheme 1). In addition, the rates that 11 and 12
react with n-propanol to form racemic n-propyl lactate
13 may differ. By analysis of the 1H and 13C NMR spectra,
the major reaction products formed by 28 h were 13 (7
mol %) and 48 mol % of n-propyl lactide that, presumably,
consists of some mixture of 11 and 12. The product
mixture of 11, 12, and 13 had a specific optical rotation
25
589
of [R]
) +9.2 (0.403, CH3C6H5). The product 13 was
then removed from the reaction mixture under vacuum.
When the presumed mixture of 11 and 12 was analyzed
by GC chromatography using a chiral γ-cyclodextrin
column,15 a single peak at 14.76 min was found. Variation
of the GC chromatographic conditions (e.g., temperature
and pressure) or derivatization of the ω-hydroxyl groups
by trifluoroacetylation did not assist in the GC resolution
of the presumed mixture. The chiral shift reagent tris-
[3-(heptafluoropropylhydroxymethylene)-(+)camphorato]-
europium(III) was used in varying concentrations relative
column, hexane/ethyl acetate), and the resulting product
25
589
had a specific rotation, [R] , of +9.81 (0.153, CH3C6H5).
To determine the stereochemical configuration of this
product, (R)-propyl â-hydroxybutrate was synthesized
from natural [R]-PHB as described in the Material and
Methods section. The optical rotation of this enantiopure
25
589
1
to 11 and/or 12 and H NMR spectra were recorded. Once
R-isomer is [R] ) -39.98 (1.33, CH3C6H5). Thus, the
again, a resolution of the corresponding NMR signals was
not achieved. The inability of the shift reagent to resolve
such a mixture was not surprising after the identical
experiments conducted with the [D-D]- and [L-L]-lactide
propyl esters did not differentiate these enantiomers.
Generally, for lipase-catalyzed reactions of ester sub-
strates with multiple stereocenters, the stereochemical
configuration of the ester unit in closest proximity to the
serine or cysteine of the active lipase box determines the
stereoselectivity of the reaction.15 One possibility is that,
based on the enantioselectivity of the PHB-depolymerase
for the L-lactide enantiomer, the n-propyl-(L-D) diastere-
PHB-depolymerase reacts preferentially with the (S)-
antipode, and the enantiomeric excess calculated by the
optical rotation measurements was 24.5%. Since the
PHB-depolymerase from P. lemoignei selectively hydro-
lyzes (R)-PHB in aqueous media,17 it appears that a
switch in the stereoselectivity in the organic medium was
observed. A change in enzyme enantioselectivity as a
function of the reaction medium used is not uncommon.6c,d
Since both lactide and â-BL have methyl branches, the
(17) Bachman, B. M.; Seebach, D. Macromolecules 1999, 32, 1777-
1784.

7806 J . Org. Chem., Vol. 65, No. 23, 2000
Kumar et al.
found. Evaluation of control reactions of ꢀ-CL and tri-
methylene carbonate carried out for 48 h at 70 °C without
enzyme showed negligible levels of monomer conversion.
Hence, based on this early study, the PHB-depolymerase
from P. lemoignei appears promising for use in ring-
opening polymerizations.
Su m m a r y of Resu lts
For the first time, a member of the family of PHB
depolymerase enzymes was studied to evaluate its ability
to catalyze bond-forming reactions in organic media. P.
lemoignei was cultured using conditions developed for
high PHB-depolymerase production. The PHB depoly-
merase from P. lemoignei was purified from the cell-free
culture liquid and lyophilized. The lyophilized enzyme
was then studied for its ability to catalyze transesteri-
fication reactions in benzene-d6. The relative rates of
PHB-depolymerase catalyzed n-propylation, from fastest
to slowest, was as follows: L-lactide (SS), meso-lactide
(RS) D-lactide (RR), δ-butyrolactone (dl), ꢀ-caprolactone,
γ-butyrolactone. It was observed that the PHB-depoly-
merase from P. lemoignei reacted differently with L- and
D-lactides. We were able to get exclusive formation of
L-propyl lactide (SS configuration) from L-lactide. In
contrast, the production of D-propyl lactide (RR config-
uration) from D-lactide underwent further reaction with
n-propanol to give propyl (2R)-2-hydroxypropanoate.
Similarly, the selectivity of the PHB-depolymerase re-
sulted in the formation of the S enriched n-propyl-â-HB
from racemic â-butyrolactone. In conclusion, this report
demonstrates the potential of using various PHB-depoly-
merase enzymes as novel catalysts in organic media. The
introduction of new families of enzymes for bond forma-
tion in organic media provides an opportunity to explore
reactions that thus far were not successfully carried out
with existing enzymes. It is noteworthy that no attempts
were made in this study to improve the enzyme activity
of the PHB-depolymerase in organic media by, for
example, using immobilization or surfactant pairing
techniques. Such work is expected to result in substantial
increases in activity of this and other PHB-depolymerases
in organic media.
F igu r e 7. Time versus ester formation for the PHB-depoly-
merase/without depolymerase catalyzed propylation of hexahy-
dro-2H-oxepin-2-one (6).
activity of the PHB-depolymerase for the nonbranched,
achiral, seven-membered ring ꢀ-caprolactone (ꢀ-CL) was
studied (see Scheme 1, compound 6). After a 3-day
reaction time, the % conversion to the corresponding
n-propyl ester 16 reached only 21 mol % and proceeded
very slowly to 63 mol % in 10 days (Figure 7, control %
conversion was 4% and 12%, respectively). Under the
identical reaction conditions, the smaller five-membered
γ-butyrolactone showed no significant PHB-depolymerase
catalyzed n-propylation after a 5-day reaction. Review
of the above shows that the relative rates of PHB-
depolymerase catalyzed n-propylation, from fastest to
slowest, was as follows: L-lactide, meso-lactide, D-lactide,
δ-butyrolactone, ꢀ-caprolactone, γ-butyrolactone.
Recently, there has been increasing interest in using
enzymes as catalysts for polymer synthesis.18 Studies on
lipase-catalyzed polymerizations of lactones18 are par-
ticularly relevant to this work. Therefore, the potential
of using the P. lemoignei PHB-depolymerase for lactone
ring-opening polymerizations was evaluated. The PHB-
depolymerase catalyzed polymerization of ꢀ-CL resulted
in 18% conversion of monomer in 48 h (70 °C) to
oligomers (18, Scheme 1) with an average DP of 4
Ack n ow led gm en t. This work was supported by the
Center for Biocatalysis and Bioprocessing of Macromol-
ecules. One of us (D.J .) thanks the Max-Buchner-
Forschungsstiftung for their kind support. We are also
thankful to Professor Virinder S. Parmar for carrying
out the GC experiments.
1
(determined by H NMR).19 Under the identical solvent-
less polymerization conditions, a 30% conversion of
trimethylene carbonate to oligomers (17, Scheme 1) with
an average DP of 12 (determined by 1H NMR)20 was
(18) Gross, R. A.; Kaplan, D. L.; Swift, G., Eds. Enzymes in Polymer
Synthesis; ACS Symposium Series 684; American Chemical Society:
Washington, DC, 1998.
(19) MacDonald, R. T.; Pulapura, S.; Svirkin, Y. Y.; Gross, R. A.;
Kaplan, D. L.; Akkara, J .; Swift, G. Macromolecules 1995, 28, 73-78.
(20) Bisht, K. S.; Svirkin, Y. Y.; Henderson, L. A.; Gross, R. A.
Macromolecules 1997, 30, 7735-7742.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of compounds 8, 9, 10, (11 and 12), 14, (14 and 15),
and 16. This material is available free of charge via Internet
at http://pubs.acs.org.
J O000814Y