Production of acrylamide using alginate-immobilized E. coli expressing Comamonas testosteroni 5-MGAM-4D nitrile hydratase

Article abstract of DOI:10.1002/adsc.200505039

A thermally-stable nitrile hydratase produced by Comamonas testosteroni 5-MGAM-4D has been expressed in Escherichia coli, and the alginate-immobilized transformant evaluated as catalyst for the conversion of acrylonitrile to acrylamide. In batch reactions with catalyst recycle, the catalyst productivity decreased with increasing acrylonitrile concentration or reaction temperature, but was relatively insensitive to acrylamide concentration. A total of 206 consecutive batch reactions with catalyst recycle were run at 5°C and produced 1035 g acrylamide/g dry cell weight and 95 g acrylamide/L; the initial and final volumetric productivities of this series of recycle reactions were 197 and 96 g acrylamide/L/h, respectively. A packed column reactor yielded lower catalyst productivity than consecutive batch recycle reactions, presumably due to contact of the fixed-bed catalyst with a constant high concentration of acrylonitrile.

Full text of DOI:10.1002/adsc.200505039

FULL PAPERS
Production of Acrylamide using Alginate-Immobilized E. coli
Expressing Comamonas testosteroni 5-MGAM-4D Nitrile
Hydratase
Lawrence J. Mersinger, Eugenia C. Hann, Frederick B. Cooling, John E. Gavagan,
Arie Ben-Bassat, Shijun Wu, Kelly L. Petrillo, Mark S. Payne, Robert DiCosimo*
DuPont Central Research and Development, Experimental Station, P. O. Box 80328, Wilmington, DE 19880-0328, U. S. A.
Fax: (þ1)-302-695-8114, e-mail: Robert.DiCosimo@usa.dupont.com
Received: January 23, 2005; Accepted: March 21, 2005
Abstract: A thermally-stable nitrile hydratase pro- weight and 95 g acrylamide/L; the initial and final
duced by Comamonas testosteroni 5-MGAM-4D has volumetric productivities of this series of recycle reac-
been expressed in Escherichia coli, and the alginate- tions were 197 and 96 g acrylamide/L/h, respectively.
immobilized transformant evaluated as catalyst for A packed column reactor yielded lower catalyst pro-
the conversion of acrylonitrile to acrylamide. In batch ductivity than consecutive batch recycle reactions,
reactions with catalyst recycle, the catalyst productiv- presumably due to contact of the fixed-bed catalyst
ity decreased with increasing acrylonitrile concentra- with a constant high concentration of acrylonitrile.
tion or reaction temperature, but was relatively insen-
sitive to acrylamide concentration. A total of 206 con-
secutive batch reactions with catalyst recycle were run Keywords: acrylamide; acrylonitrile; catalyst produc-
at 58C and produced 1035 g acrylamide/g dry cell tivity; nitrile hydratase
Introduction
In addition to temperature instability, both mesophilic
and thermally stable microbial NHase catalysts are sus-
The development of a lab-scale biocatalytic reaction ceptible to inactivation by high concentrations of acryl-
into a commercial process requires the determination onitrile. An acrylonitrile concentration of 1.5–2 wt %
of the contribution of volumetric productivity (g prod- was employed when using Rhodococcus sp. N-774 and
uct/L/h) and catalyst productivity (g product/g biocata- P. chlororaphis B23 catalysts for commercial acrylamide
lyst) to the overall cost of manufacture. For the commer- production, while up to 7 wt % acrylonitrile has been
[
5]
cial process employing biocatalytic conversion of acryl- employed with R. rhodochrous J1. Bacillus sp.
onitrile to acrylamide, successive generations of nitrile BR449 expresses a thermostable NHase that undergoes
hydratase biocatalysts have continuously improved upon rapid inactivation at an acrylonitrile concentration of
[
6]
volumetric and catalyst productivities. A variety of bac- only 2 wt % at 22–508C. The thermophilic NHase of
terial genera are known to possess a diverse spectrum of Bacillus spp. RAPc8 lost activity after a short time
[1]
nitrile hydratase (NHase, EC 4.2.1.84) activities. Many when used for the continuous conversion of 100 mM
mesophilic NHases are remarkably unstable, having acrylonitrile in a column reactor at 408C, where low cat-
very short enzyme activity half-lives in the growth tem- alyst productivity was attributed to exposure of NHase
[
1a,2]
[7]
perature range of 20–358C.
Wild-type cells with to acrylonitrile. A comparison of two Rhodococcus
[3]
NHase activity such as Rhodococcus sp. N-774, Pseu- isolates, one with only a nitrilase activity and one with
[
4]
domonas chlororaphis B23, and Rhodococcus rho- only a combination of NHase and amidase activities,
[5]
dochrous J1 have each been used commercially to con- as catalysts for ammonium acrylate production conclud-
vert acrylonitrile to acrylamide. The NHase activities of ed that the combination of NHase and amidase activities
Rhodococcus sp. N-774 and P. chlororaphis B23 were was less preferred due to the susceptibility of both en-
not stable above 108C, and although the NHase activity zymes to deactivation by acrylonitrile, as well as inhibi-
of R. rhodochrous J1 showed comparatively greater tion of the two enzymes by their respective products.
thermal stability, the reaction temperature for commer- Recombinant organisms expressing exogenous
[8]
cial production of acrylamide using this biocatalyst was NHase have also been demonstrated to be effective cat-
maintained at 108C to improve enzyme stability and cat- alysts for the hydration of nitriles. Expression of nitrile
[
5]
[9]
alyst productivity [g product/g dry cell weight (dcw)].
hydratase genes isolated from C. testosteroni NI1,
Adv. Synth. Catal. 2005, 347, 1125–1131
DOI: 10.1002/adsc.200505039
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1125

FULL PAPERS
Lawrence J. Mersinger et al.
[
10]
[11]
Rhodococcus sp. N-771, Rhodococcus sp. N-774,
[
12]
[13]
R. rhodochrous J1 and P. putida 5B in E. coli has
been reported. The NHase of the moderate thermophile
Pseudonocardia thermophila JCM 3095 has been cloned
and expressed in E. coli for use in the commercial pro-
[
14]
duction of acrylamide.
Comamonas testosteroni 5-
MGAM-4D expresses a thermally stable NHase and
amidase, and has been used for conversion of a variety
[
15]
of nitriles to their corresponding carboxylic acids;
the NHase has recently been cloned and sequenced,
and active NHase has been over-produced in Escheri-
[16]
chia coli SW132. We now report an examination of
this transformant catalyst for the conversion of acryloni-
trile to acrylamide, where the dependence of catalyst
productivity on reaction temperature and concentration
of acrylonitrile and acrylamide has been evaluated.
Figure 2. Time course for hydration of 0.51 M acrylonitrile
using unimmobilized E. coli SW132 cells (0.441 mg dcw/
mL) in 0.1 M potassium phosphate buffer (pH 7.0) at 258C:
acrylonitrile (
), acrylamide (~).
&
Results and Discussion
The K (2.0 mM) and specific activity [31.1 mmol/min/ reactions for hydration of 0.53, 1.06, 2.05 and 3.04 M
m
mg dry cell weight (dcw)] were determined for the acrylonitrile to acrylamide were initially run at 258C us-
NHase activity of E. coli SW132 cells (4.412 mg dcw/ ing a catalyst charge of 20 wt % alginate-immobilized
mL) by measuring the rate of acrylamide production cells in the reaction mixture, and quantitative conver-
over a range of acrylonitrile concentrations of 0.40– sion to acrylamide was observed at each concentration;
8
0 mM (Figure 1). The microbial Km was similar to all reactions additionally contained 2 mM calcium chlor-
that reported for the NHase of R. rhodochrous J1 ide to maintain the integrity of the calcium-cross-linked
1.89 mM), and considerably lower than that of P. putida alginate gel beads. Measurement of reaction rates for
B23 (34.6 mM) and Brevibacterium sp. R312 hydration of 2–3 M acrylonitrile was initially problem-
(
[
5]
(
16.7 mM). Complete conversion of 0.51 M acryloni- atic; the solubility of acrylonitrile in water has been re-
[18]
trile using unimmobilized cells (0.44 mg dcw/mL) pro- ported to be 74 g/L (1.4 M) at 258C, and reactions
duced a quantitative yield of acrylamide, and no acrylic containing 2–3 M acrylonitrile were initially two-phase
acid was detected (Figure 2). E. coli SW132 cells (7.5% mixtures of acrylonitrile and the aqueous catalyst sus-
dcw) were immobilized in 2.75 wt % alginate beads (ca. pension. At these higher concentrations of acrylonitrile,
3
mm diameter), which were subsequently cross-linked it was also noted that a decrease in total volume of the
[17]
with glutaraldehyde and polyethyleneimine.
Batch initially two-phase mixture occurred over the course of
the reaction; when using 3 M acrylonitrile, the final vol-
ume of the reaction mixture was ca. 90% of the initial
volume. Dilution of the final reaction mixture to exactly
one-tenth of the initial reaction volume allowed for the
accurate measurement of the concentration of acryl-
amide present.
The thermal stability of E. coli SW132 NHase has
been previously determined, and the enzyme has a
[
16]
half-life of 3.0 h at 508C. Consecutive batch reactions
with catalyst recycle were initially performed at 358C to
determine the effect of acrylonitrile concentration on
NHase stability at a higher temperature than is typically
employed for commercial production of acrylamide (5–
[
5]
1
08C). In consecutive batch reactions with catalyst re-
cycle, the product mixture was decanted from the cata-
lyst beads, leaving a reaction “heel” containing the cat-
alyst beads and residual product mixture, and the reac-
tor was recharged with acrylonitrile and 2 mM calcium
chloride to produce the desired initial concentration of
Figure 1. Dependence of specific activity on acrylonitrile con-
centration for unimmobilized E. coli SW132 NHase at 258C reactant. The catalyst beads were ca. 90% water by
(
4.412 mg dcw/mL).
weight and contained an acrylamide concentration
1126
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1125–1131

Production of Acrylamide using Alginate-Immobilized E. coli
FULL PAPERS
equivalent to that of the final concentration of acryl-
amide in the product mixture of the previous batch reac-
tion, so the acrylamide remaining in the reaction heel
was carried over into subsequent reactions in a series.
For reactions employing a 20 wt % immobilized cell cat-
alyst charge and acrylonitrile concentrations of 1.06 M,
2
.05 M and 3.04 M, the final concentrations of acryl-
amide produced after ca. five consecutive batch reac-
tions with catalyst recycle remained constant at ca.
1.33 M, 2.56 M and 3.80 M, respectively. For 1.06 M
acrylonitrile recycle reactions at 358C, the initial volu-
metric productivity was 737 g acrylamide/L/h, and a to-
tal catalyst productivity of 126 g acrylamide/g (dcw) cat-
alyst was obtained over the course of 25 catalyst recycles,
with 43% of the initial catalyst activity remaining (Ta-
ble 1). The recovered catalyst activity in recycle reac-
Figure 3. Effect of acrylonitrile concentration on recovered
tions decreased with increasing acrylonitrile concentra- NHase activity of alginate-immobilized E. coli SW132 in con-
tion (Figure 3), where at 3.04 M acrylonitrile only 5% of secutive batch recycle reactions at 358C; acrylonitrile: 1.06 M
the initial catalyst activity remained after four catalyst (*
recycles.
), 2.05 M (~), 3.04 M (
).
&
A significant improvement in catalyst productivity
was achieved by decreasing the reaction temperature. 108C in a parallel series of reactions, there was a de-
The immobilized-cell NHase activity was measured in crease in catalyst productivity (Table 1); after 149 con-
consecutive batch reactions with 1 M acrylonitrile at secutive batch reactions, the recovered NHase activities
5
(
8C and 108C using 20 wt % immobilized-cell catalyst at 58C and 108C were 67% and 31% of their respective
Figure 4). At 58C, a total of 206 consecutive batch reac- initial activities.
tions with catalyst recycle produced 1035 g acrylamide/g The immobilized-cell recycle reactions performed at
dcw) catalyst and 95 g acrylamide/L. The catalyst activ- 58C (described above) were run over a period of four
(
ity in the final reaction was 49% of the initial activity, months, where the catalyst was stored in either product
and the initial and final volumetric productivities of mixture (containing ca. 1.33 M acrylamide) or 2 mM cal-
this series of reactions were 231 and 96 g acrylamide/ cium chloride (containing ca. 0.28 M acrylamide) for pe-
L/h, respectively. Although decreasing the reaction tem- riods of up to seven days at 58C between reactions; no
perature from 358C to 58C resulted in a decrease in re- significant loss of catalyst activity was observed during
action rate (and volumetric productivity), the initial and these storage periods, indicating that concentrations of
final reaction times in this set of recycle reactions were acrylamide of up to 1.33 M did not result in significant
only 40 minutes and 80 minutes, respectively. When loss of NHase activity over the course of these reactions.
the reaction temperature was increased from 58C to The effect of acrylonitrile concentration on recovered
Table 1. Catalyst specific activity, catalyst productivity and volumetric productivity for the conversion of acrylonitrile to acryl-
[a]
amide using alginate-immobilized E. coli SW132 in consecutive batch reactions with catalyst recycle.
Temp.
Acrylo-
nitrile
Initial catalyst
specific
activity [U/g]
Initial
volumetric
productivity
Catalyst
recycles
Final
volumetric
produc-
Remaining
catalyst activity
[% initial
Total catalyst
productivity
[g acrylamide/g
dry cell wt]
[
8C]
[
b]
[
M]
[
g/L/h]
tivity [g/L/h]
activity]
3
3
3
2
2
1
5
5
5
5
5
5
5
0
1.06
2.05
3.04
0.53
1.06
1.06
1.06
3.04
864
1037
1124
447
559
252
737
884
958
381
476
215
197
447
25
10
4
319
225
275
43
25
29
126
97
58
163
206
19
50
96
88
32
49
20
819
1035
274
231
524
[
a]
Reaction conditions: 20-mL reaction volume containing 4.0 g alginate-immobilized E. coli SW132 [3 mm diameter beads,
.5% dry cell weight (dcw), cell lot A] in 2.0 mM calcium chloride.
Specific activity was determined by measuring the rate (mmol/min) of acrylamide production per g dcw E. coli SW132 in the
7
[
b]
reaction mixture.
Adv. Synth. Catal. 2005, 347, 1125–1131
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1127

FULL PAPERS
Lawrence J. Mersinger et al.
Figure 5. Specific activity of alginate-immobilized E. coli
SW132 (1-mm diameter catalyst beads) in consecutive
Figure 4. Specific activity of alginate-immobilized E. coli
SW132 in consecutive 1.06 M acrylonitrile batch reactions
1
5
.06 M acrylonitrile batch reactions with catalyst recycle at
with catalyst recycle: 58C (
&
), 108C (*).
&
~
8C: no added acrylamide ( ), 1.4 M added acrylamide ( ,
trend line), 2.9 M acrylamide (*).
NHase activity in consecutive batch reactions was also
examined at 58C where, for reactions employing beads having a 1-mm diameter were employed to im-
.04 M acrylonitrile, an 80% decrease in catalyst activity prove volumetric productivity at the same catalyst
3
occurred after only 19 recycle reactions; increasing the charge as was previously employed using 3-mm catalyst
concentration of acrylonitrile from 1.06 M to 3.04 M beads. Consecutive 1.06 M acrylonitrile batch reactions
was deleterious to E. coli SW132 NHase activity at ei- with catalyst recycle were performed at 58C in the pres-
ther 358C or 58C.
ence of either 1.4 M or 2.9 M added acrylamide; recycle
When using alginate-immobilized microbial cells, reactions with no added acrylamide were run as a con-
volumetric productivity can often be increased by de- trol. No significant difference in the rate of loss of cata-
creasing the diameter of the catalyst bead. The reaction lyst activity was observed over the course of thirty con-
rate may be limited by the rate of diffusion of reactant secutive recycle reactions that had final concentrations
into the catalyst bead and, if there is product inhibition, of acrylamide of 1.33 M (control), 2.73 M (1.4 M added
[
19]
by the rate of diffusion of product out of the bead. If acrylamide) and 4.23 M (2.9 M added acrylamide), cor-
the reaction rate is significantly lower than the rate of responding 95, 194 and 301 g acrylamide/L, respectively.
diffusion of reactant into the bead, no effect on reaction Based on these results, it should be possible to further in-
rate would be expected; conversely, where the reaction crease catalyst productivity by continuously adding
rate is significantly greater than the rate of diffusion, a acrylonitrile to a reaction to maintain a constant concen-
dependence of rate on diameter is expected. The reac- tration significantly lower than 1 M [but at a concentra-
tion rate and bead diffusion rate could also be independ- tion of at least five times the K (2.0 mM) to maintain
m
ently affected by reaction temperature. The 3.0-mm di- maximum reaction rate] and produce at least 30 wt %
ameter catalyst beads employed in the reactions describ- acrylamide in water.
ed above were produced by dripping the cell/alginate
The catalyst used to measure catalyst productivity at
suspension from a 20 gauge needle into an aqueous sol- 58C (Figure 4) was 3-mm diameter beads prepared us-
ution containing calcium chloride; the diameter of the ing a first lot (lot A) of E. coli SW132 cells, whereas
resulting beads size is typically a function of the cell/al- the recycle reactions depicted in Figure 5 were per-
ginate suspension viscosity and surface tension, not the formed with 1-mm diameter catalyst beads prepared us-
diameter of the needle. Modifying this technique by ing a second lot (lot B) of E. coli SW132 cells. The spe-
placing the needle within the lumen of a second tube cific activity of lot B cells was only ca. 83% of that of
through which air can be metered produced beads hav- lot A cells. Comparison of the catalyst stability of the
[
20]
ing a smaller diameter. E. coli SW132 cells were en- two catalyst bead preparations for reactions run with
capsulated in 1-mm diameter alginate beads using this 1.06 M acrylonitrile at 58C (no added acrylamide) indi-
annular air-flow technique, and at 58C the specific activ- cated that the rate of loss of catalyst activity with the 1-
ity of the 1-mm diameter beads was ca. 1.4-fold greater mm diameter lot B cell catalyst beads was significantly
than that of the corresponding 3-mm diameter beads.
greater than for the 3-mm lot A cell catalyst beads.
The dependence of catalyst productivity on acryl- The dependence of catalyst stability and productivity
amide concentration was examined (Figure 5). Catalyst on bead diameter and cell lot was therefore evaluated
1128
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1125–1131

Production of Acrylamide using Alginate-Immobilized E. coli
FULL PAPERS
using three different catalyst bead preparations: (1) 3-
mm diameter beads, lot A cells, (2) 3-mm diameter
beads, lot B cells, and (3) 1 mm diameter beads, lot B
cells (Figure 6). There was no significant dependence
of catalyst stability on bead diameter, but there was a
difference in catalyst stability between the two lots of
immobilized cells; beads produced with lot B cells lost
their activity at a greater rate than beads produced
with lot A cells. Also, there was an initial increase in
the specific activity of lot B catalyst beads after the first
recycle reaction that was not observed for lot A catalyst
beads; the reason for this initial increase in lot B catalyst
specific activity has not been determined, but could be
due to partial lysis of the immobilized cells under the re-
action conditions. These results indicate that the fer-
mentation of E. coli SW132 requires further optimiza-
tion, and that multiple lots of cells produced by fermen-
tation should routinely be compared to determine the
variability in NHase specific activity and productivity
of microbial cell catalysts.
Figure 7. Time course for conversion of 1.13 M acrylonitrile
in a column reactor containing 3 mm diameter catalyst beads
(
cell lot A) at 58C: acrylamide (
rate adjusted to maintain>90% conversion: 0–21 h at
.5 mL/min, 21–118 h at 0.4 mL/min, 118–143 h at 0.35 mL/
&
), acrylonitrile (*). Flow
0
Immobilized cell catalyst productivity in a column re- min, 143–399 h at 0.3 mL/min. Catalyst productivity from
actor at 58C (Figure 7) was evaluated using the same 0–263 h was 155 g acrylamide/g dcw.
catalyst bead preparation employed for the determina-
tion of catalyst productivity at 58C in batch reactions
with catalyst recycle (3-mm diameter beads, lot A cells; umn reactor when compared to consecutive batch reac-
Figure 3). After an initial adjustment in the feed rate of tions, and this lower productivity may be due at least in
1
9
.13 M acrylonitrile/2 mM calcium chloride to maintain part to the constant exposure of the catalyst beads at the
0–95% conversion of acrylonitrile to acrylamide (run- column inlet to a high concentration of acrylonitrile,
ning at less than 100% conversion allows one to immedi- whereas in batch operation the entire catalyst charge is
ately detect a loss of catalyst activity), the column was initially exposed to this same acrylonitrile concentra-
run continuously for a total of 399 h. At 263 h, the con- tion, but the concentration subsequently decreases
version of acrylonitrile began to decrease markedly at over the course of the reaction.
constant feed rate; the catalyst productivity at this point
was ca. 155 g acrylamide/g (dcw) catalyst. A lower cata-
lyst productivity was observed for the catalyst in a col- Conclusion
A catalyst productivity of 1035 g acrylamide/g dcw for
alginate-immobilized E. coli SW132 is comparable to
the commercial catalyst productivities for polyacryl-
amide-immobilized Rhodococcus sp. 774 and P. putida
[5]
B23 (500 and 850 g acrylamide/g dcw, respectively),
but less than that of the polyacrylamide-immobilized
R. rhodochrous J1 currently employed for commercial
production of acrylamide (>7000 g acrylamide/g
[
5]
dcw). Although the catalyst productivity of thermo-
[
6]
philic NHase catalysts Bacillus sp. BR449 or algi-
[
7]
nate-immobilized Bacillus spp. RAPc8 was not report-
ed for reactions run at 5–108C, comparison of the cata-
lyst productivity of unimmobilized E. coli SW132 at
358C with that of Bacillus sp. BR449 at 30–408C indi-
cates a significantly greater catalyst productivity for E.
coli SW132. It is possible that additional improvement
in E. coli SW132 catalyst productivity might be achieved
by further optimization of the immobilization matrix;
for example, a significant increase in the operational sta-
Figure 6. Comparison of recovered activity of 1-mm and 3-
mm bead diameter immobilized cell catalysts in consecutive
1
5
.06 M acrylonitrile batch reactions with catalyst recycle at
8C: 1 mm diameter, cell lot B (
&
), 3 mm diameter, cell lot bility of immobilized Brevibacterium sp. CH1 NHase for
B (~), 3 mm diameter, cell lot A (*).
hydration of acrylonitrile was observed when cells were
Adv. Synth. Catal. 2005, 347, 1125–1131
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1129

FULL PAPERS
Lawrence J. Mersinger et al.
immobilized in polyacrylamide instead of calcium algi- ODl¼550 was 57.6. The cells were chilled to 5–108C and recov-
[
21]
ered by centrifugation; 490 g (wet cells) were isolated. Based
on SDS-PAGE analysis, NHase protein constituted approxi-
mately 30% of total soluble protein in induced E. coli SW132.
nate.
Despite the fact that E. coli SW132 NHase is relatively
thermally stable compared to the NHase of most meso-
philic bacteria, the catalyst productivity at 358C was sig-
nificantly less than that at 58C. The major determinant
of NHase catalyst productivity at either 58C or 358C
was the concentration of acrylonitrile in the reaction
mixture. This result is in agreement with those of previ-
ous studies with thermally stable microbial NHases,
Hydration of Acrylonitrile to Acrylamide using
Unimmobilized E. coli SW132 cells
A 20-mL reaction vessel equipped with magnetic stirring was
charged with 0.336 mL (0.271 g, 5.10 mmol) of acrylonitrile
where the nucleophilic reaction of protein functional and 9.46 mL of 0.1 M potassium phosphate buffer (pH 7.0).
groups with acrylonitrile has been proposed as a mech- To the reaction vessel was next added 0.200 mL of an aqueous
[
6,7]
anism for enzyme inactivation. Performing the reac- suspension of 22.06 mg dcw/mL of E. coli SW132 cells in
0
.10 M potassium phosphate buffer (pH 7.0), and the resulting
tion at 5–108C decreased this rate of enzyme inactiva-
tion, while at the same time the specific activity of the
microbial catalyst was sufficiently high in this tempera-
ture range to produce acceptable reaction rates (and
volumetric productivities) for biocatalytic conversion
of acrylonitrile to acrylamide.
mixture stirred at 258C in a constant temperature bath. Sam-
ples (0.100 mL) of the reaction mixture were mixed with
0
ous 0.200 M sodium butyrate (HPLC standard). The resulting
mixture was centrifuged, and the supernatant analyzed by
HPLC. All reactions produced only the amide as the hydration
product at 100% conversion of nitrile, with no hydrolysis of the
nitrile to the corresponding carboxylic acid.
.100 mL of water, 0.020 mL of 6 N HCl and 0.200 mL of aque-
Experimental Section
Immobilization of E. coli SW132 Cells in Calcium
Cross-Linked Alginate
General Remarks
Chemicals were obtained from commercial sources and used as
received. Cell paste was stored frozen at ꢀ808C. Wet cell
weight (wcw) of microbial catalysts utilized in reactions or as-
says was obtained from cell pellets prepared by centrifugation
of fermentation broth, dry cell weight (dcw) of cells was deter-
mined by microwave drying of wet cells. Analysis for acryloni-
trile and acrylamide were performed by HPLC using a refrac-
tive index detector and a Bio-Rad HPX-87H column (30 cmꢁ
A 250-mL media bottle (equipped with magnetic stir bar con-
taining 59.7 g of distilled, deionized water at 508C) was slowly
ꢁ
charged with 3.30 g of FMC BioPolymer Protanal LF 10/60 al-
ginate with rapid stirring. The mixture was heated to 75–808C
with rapid stirring until the alginate was completely dissolved,
and the resultingsolution cooled to 258C in a water bath. To the
alginate suspension was added 40.8 g of E. coli SW132 wet cell
paste (22% dry cell weight) and 16.2 mL of distilled water with
stirring. The cell/alginate mixture was added dropwise by sy-
ringe to 640 mL of 0.20 M calcium acetate buffer (pH 7.0) at
7
.8 mm diameter) at 508C; the solvent was 0.001 N sulfuric
acid. The calculated yields of recovered acrylonitrile and acryl-
amide were based on initial acrylonitrile concentrations.
2
58C with stirring. After stirring for 2 h, the buffer was de-
canted and the resulting beads (82 g, 3-mm diameter) were re-
suspended in 200 mL of 0.20 M calcium acetate buffer (pH 7.0)
at 258C. With stirring, 4.10 g of 25 wt % glutaraldehyde (GA)
in water was added and the beads were mixed for 1.0 h at 258C.
To the beadsuspension was thenadded 16.4 gof12.5 wt % poly-
Fermentation of Escherichia coli SW132 Cells
[16]
E. coli strain SW132 was grown in a 500-mL seed culture
flask at 368C, 300 rpm for 10 h to an OD of>2.0 prior to
l¼550
ꢁ
ethyleneimine (PEI, BASF Lupasol PR971L, average mo-
inoculation of the fermentor. The fermentation medium was
prepared in an initial batch of 7.5 L, and contained 32 g
KH PO , 8.0 g MgSO ·7 H O, 8.0 g (NH ) SO , 50 g yeast ex-
lecular weight ca. 750,000) in water, and the beads were mixed
for an additional 18 h at 258C. The GA/PEI-cross-linked beads
were then washed twice with 250 mL of 0.05 M calcium acetate
buffer (pH 7.0) at 258C, and stored in this same buffer at 58C.
2
4
4
2
4
2
4
tract, and 10 mL Mazu DF204 antifoam (BASF). Following
sterilization, 369 g glucose solution (60% w/w), 160 mL trace
element solution (citric acid, 10 g/L; CaCl ·2 H O, 1.5 g/L;
2
2
FeSO ·7 H O, 5.0 g/L; ZnSO ·7 H O, 0.39 g/L; CuSO ·5 H
4
2
4
2
4
2
O, 0.38 g/L; CoCl ·6 H O, 0.20 g/L; MnCl ·4 H O, 0.30 g/L),
2
2
2
2
Hydration of Acrylonitrile to Acrylamide using
Alginate-Immobilized E. coli SW132 Cells in
Consecutive Batch Reactions with Biocatalyst Recycle
and 100 mg/L ampicillin were added. NH OH (40% w/v) and
4
2
0% w/v H SO were used for pH control. The initial set points
2 4
were: agitation, 400 rpm; aeration, 2 L/min; pH, 6.8; pressure,
.5 psig; dissolved oxygen concentration (DO), 25%; tempera-
0
A 50-mL jacketed reaction vessel (equipped with an overhead
stirrer and temperature-controlled at 5, 10 or 358C with a recir-
culating temperature bath) was charged with 4.0 g of GA/PEI-
cross-linked E. coli SW132 cell/alginate beads. To the reaction
vessel was added 0.2 mL of 0.20 M calcium acetate buffer
(pH 7.0, 2.0 mM final calcium ion concentration), 1.40 mL
(1.13 g, 21.3 mmol) of acrylonitrile, and the final volume of
ture, 368C. At culture densities of 20–30 OD additional AMP
was added to 100 mg/L. IPTG was added to 1 mM at culture
densities of 30–35 OD. Glucose feed was started at <5 g/L
and feed rate was reduced if glucose accumulated above 2 g/
L. IPTG (1 mM) was added to the fermentor at 30–35 OD
l¼
and cells were harvested 5 h after IPTG addition; final
550
1130
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 1125–1131

Production of Acrylamide using Alginate-Immobilized E. coli
FULL PAPERS
the reaction mixture adjusted to 20 mL by the addition of dis-
tilled, deionized water. The mixture was stirred at 5, 10 or 358C.
Samples (0.100 mL) of the reaction mixture were mixed with
[4] T. Nagasawa, H. Nanba, K. Ryuno, K. Takeuchi, H. Ya-
mada, Eur. J. Biochem. 1987, 162, 691–698.
[5] T. Nagaswa, H. Shimizu, H. Yamada, Appl. Microbiol.
Biotechnol. 1993, 40, 189–195.
0
.400 mL of water, and then 0.200 mL of the diluted sample
were mixed with 0.200 mL of 0.200 M sodium butyrate
HPLC external standard); the resulting mixture was centri-
[
[
[
[
6] R. Padmakumar, P. Oriel, Appl. Biochem. Biotechnol.
999, 77–79, 671–679.
7] D. Graham, R. Pereira, D. Barfield, D. Cowan, Enz. Mi-
crob. Technol. 2000, 26, 368–373.
8] N. A. Webster, D. K. Ramsden, J. Hughes, Biotechnology
Lett. 2001, 23, 95–101.
9] J. M. Stevens, N. Rao Saroja, M. Jaouen, M. Belghazi, J.-
M. Schmitter, D. Mansuy, I. Artaud, M.-A. Sari, Protein
Expression and Purification 2003, 29, 70–76.
(
1
fuged and the supernatant analyzed by HPLC. At the comple-
tion of the reaction (100% conversion of acrylonitrile), the
product mixture was decanted from the biocatalyst beads,
and additional distilled, deionized water, 0.2 mL of 0.20 M cal-
cium acetate buffer and 1.40 mL (1.13 g, 21.3 mmol) of acrylo-
nitrile mixed with the reaction heel (immobilized-cell catalyst
and remaining product mixture from the previous reaction) at
5, 10 or 358C. At the completion of the second reaction, the
[
10] M. Nojiri, M. Yohda, M. Odaka, Y. Matsushita, M. Tsu-
jimura, T. Yoshida, N. Dohmae, K. Takio, I. Endo, J. Bi-
ochem. (Tokyo) 1999, 125, 696–704.
product mixture was decanted and subsequent reactions per-
formed as before.
[
11] Y. Hashimoto, M. Nishiyama, S. Horinouchi, T. Beppu,
Biosci. Biotechnol. Biochem. 1994, 58, 1859–1865.
12] M. Kobayashi, M. Nishiyama, T. Nagasawa, S. Horinou-
chi, T. Beppu, H. Yamada, Biochim. Biophys. Acta 1991,
Hydration of Acrylonitrile to Acrylamide using
Alginate-Immobilized E. coli SW132 Cells in a Packed
Column Reactor
[
1
129, 23–33.
A 12 cmꢁ2.5 cm diameter stainless steel column was charged
with 34 g of GA/PEI-cross-linked E. coli SW132 cell/alginate
beads (3-mm diameter beads, lot A cells) suspended in
[13] S. Wu, R. D. Fallon, M. S. Payne, Appl. Microbiol. Bio-
technol. 1997, 48, 704–708.
[14] T. Yamaki, T. Oikawa, K. Ito, T. Nakamura, J. Ferment.
Bioeng. 1997, 83, 474–477; A. Miyanaga, S. Fushinobu,
K. Ito, T. Wakagi, Biochem. Biophys. Res. Commun.
0.05 M calcium acetate buffer (pH 7.0). The column was placed
in a controlled-temperature bath at 58C, and an aqueous solu-
tion of 1.13 M acrylonitrile and 2 mM calcium chloride was fed
to the column at an initial flow rate of 0.5 mL/min using a peri-
staltic pump. Samples (0.100 mL) of the column effluent were
mixed with 0.400 mL of water, and then 0.200 mL of the diluted
sample were mixed with 0.200 mL of 0.200 M sodium butyrate
2
001, 288, 1169–1174; A. Miyanaga, S. Fushinobu, K.
Ito, H. Shoun, T. Wakagi, Eur. J. Biochem. 2004, 271,
29–438.
4
[
15] a) J. E. Gavagan, S. Fager, R. D. Fallon, P. W. Folsom,
F. E. Herkes, A. E. Eisenberg, E. C. Hann, R. DiCosimo,
J. Org. Chem. 1998, 63, 4792–4801; b) E. C. Hann, A. E.
Sigmund, S. K. Fager, F. B. Cooling, J. E. Gavagan, A.
Ben-Bassat, S. Chauhan, M. S. Payne, S. M. Hennessey,
R. DiCosimo, Adv. Syn. Catal. 2003, 345, 775–782;
c) DuPont, US Patent 6,670,158 B2, 2003; Chem. Abstr.
(
HPLC external standard) and analyzed by HPLC. The col-
umn flow rate was adjusted during initial operation to achieve
9
0
0
0–95% conversion of acrylonitrile to acrylamide: 0–21 h at
.50 mL/min, 21–118 h at 0.40 mL/min, 118–143 h at
.35 mL/min, 143–399 h at 0.3 mL/min. After continuous op-
eration for 263 h (120 h at 0.30 mL/min), the catalyst produc-
tivity was ca. 155 g acrylamide/g dcw catalyst.
2
003, 139, 148573.
[
[
16] K. L. Petrillo, S. Wu, E. C. Hann, F. B. Cooling, A. Ben-
Bassat, J. E. Gavagan, R. DiCosimo, M. S. Payne. Appl.
Microbiol. Biotechnol. 2005, in press.
17] E. C. Hann, A. E. Sigmund, S. M. Hennessey, J. E. Gav-
agan, D. R. Short, A. Ben-Bassat, S. Chauhan, R. D. Fal-
lon, M. S. Payne, R. DiCosimo, Org. Proc. Res. Dev.
Acknowledgements
The authors thank Susan Fager (DuPont), L. W. Wagner (Du-
Pont) and the staff of the DuPont Fermentation Research Facili-
ty for their technical assistance.
2
002, 6, 492–496.
18] H. S. Davis, O. F. Wiedeman, Ind. Eng. Chem. 1945, 37,
82–485.
19] B. Atkinson, F. Mavituna, Biochemical Engineering and
Biotechnology Handbook, 2nd edn., Macmillan Publish-
ers, Stockton Press, New York, 1991, pp. 582, 764–765.
20] J. Klein, J. Stock, K.-D. Vorlop, Eur. J. Appl. Microbiol.
Biotechnol. 1983, 18, 86–91; T. Rehg, C. Dorger, P. C.
Chau, Biotechnol. Lett. 1986, 8, 111–114; D. B. Seifert,
J. A. Phillips, Biotechnol. Prog. 1997, 13, 562–568.
[
[
4
References
[
1] a) D. A. Cowan, R. A. Cameron, T. L. Tsekoa, Adv.
Appl. Microbiol. 2003, 52, 123–158; b) L. Martinkova,
V. Kren, Biocatal. Biotransform. 2002, 20, 73–93.
2] D. Cowan, R. Cramp, R. Pereira, D. Graham, Q. Alma-
tawah, Extremophiles 1998, 2, 207–216.
[
[
[
[
21] J. S. Hwang, H. N. Chang, Biotechnol. Bioeng. 1989, 34,
3] I. Watanabe, Y. Satoh, K. Enomoto, Agric. Biol. Chem.
3
80–386.
1
987, 51, 3193–3199; I. Watanabe, Y. Satoh, K. Enomoto,
S. Seki, K. Sakashita, Agric. Biol. Chem. 1987, 51, 3201–
206.
3
Adv. Synth. Catal. 2005, 347, 1125–1131
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1131