Immobilization of rhodococcus AJ270 and use of entrapped biocatalyst for the production of acrylic acid

Article abstract of DOI:10.1007/s007060070094

Rhodococcus AJ270 is adsorbed by Dowex 1 at 15.4 mg dry weight per g resin with maximum amidase specific activity observed at lower loadings. Bacteria form a monolayer on the resin surface, and adsorption is complete within 2 min. AJ270 can be entrapped in agar and agarose gels (optimum loading: 20 mg dry weight bacteria per cm³ gel). Adsorption and entrapment improve amidase thermal stability 3-4 fold, and entrapment shifts the pH optimum from 8 to 7. Adsorbed and free bacteria show similar values for K_m and V_max, but entrapped bacteria have higher K_m values. Compared with bacteria adsorbed to Dowex, the activity per cm³ of matrix of agar-entrapped AJ270 is eight-fold higher. In stirred-tank reactors, exposure to acrylic acid reduces the amidase activity of the biocatalyst in the hydrolysis of acrylamide. In column reactors, entrapped AJ270 suffers little reduction in amidase activity against 0.25 M acrylamide over 22 h continuous operation.

Full text of DOI:10.1007/s007060070094

Monatshefte f uÈ r Chemie 131, 655±666 (2000)
Immobilization of Rhodococcus AJ270 and Use
of Entrapped Biocatalyst for the Production
of Acrylic Acid
Ã
a
John Colby , David Snell , and Gary W. Black
School of Sciences, University of Sunderland, Sunderland SR1 3SD, UK
Summary. Rhodococcus AJ270 is adsorbed by Dowex 1 at 15.4 mg dry weight per g resin with
maximum amidase speci®c activity observed at lower loadings. Bacteria form a monolayer on the
resin surface, and adsorption is complete within 2 min. AJ270 can be entrapped in agar and agarose
3
gels (optimum loading: 20 mg dry weight bacteria per cm gel). Adsorption and entrapment improve
amidase thermal stability 3±4 fold, and entrapment shifts the pH optimum from 8 to 7. Adsorbed and
free bacteria show similar values for Km and Vmax, but entrapped bacteria have higher K values.
m
3
Compared with bacteria adsorbed to Dowex, the activity per cm of matrix of agar-entrapped AJ270
is eight-fold higher. In stirred-tank reactors, exposure to acrylic acid reduces the amidase activity of
the biocatalyst in the hydrolysis of acrylamide. In column reactors, entrapped AJ270 suffers little
reduction in amidase activity against 0.25 M acrylamide over 22 h continuous operation.
Keywords. Acrylamide; Amidase; Acrylate; Biocatalysis; Rhodococcus.
Introduction
When grown on acetamide, Rhodococcus AJ270 expresses high speci®c activities
of nitrile hydratase and amidase [1], enzymes with broad speci®city against a wide
range of nitriles and amides [2,3]. Washed suspensions of AJ270 have been used
for the enantioselective hydrolysis of (R,S)-2-phenylbutyronitrile [2] and of
racemic ibuprofen amide, in the latter case yielding the pharmacologically active
non-steroidal anti-in¯ammatory agent S-(‡)-ibuprofen in 94% e.e. [1]. This
versatile biocatalyst also catalyses the regioselective hydrolysis of certain dinitriles
[2,4]. These applications used the rhodococcal biocatalyst in a `one-shot' batch
mode, whereas immobilization of the bacteria would allow multiple re-use of the
catalyst, possibly in continuous reactors, and would certainly facilitate product
recovery. This paper describes methods for the adsorption and entrapment of
AJ270, reports changes in the properties of the amidase activity resulting from
immobilization, and investigates the use of entrapped bacteria for the hydrolysis of
acrylamide. This latter biotransformation is of commercial signi®cance for the
Ã
Corresponding author
a
Present address: Helena Biosciences, Sunderland Business Park, Sunderland SR5 3XB, UK

656
J. Colby et al.
Table 1. Adsorption of AJ270 to anionic exchange resins. Bacteria (4.2 mg dry weight) were
immobilized to the anionic exchangers (3 g wet weight), and the % of bacteria adsorbed determined
as described in the Experimental section. For the data in the second column, the bacteria and resins
were washed with buffer containing 0.01 M EDTA prior to immobilization
%
of bacteria adsorbed
Resin
� EDTA
‡EDTA
Amberlite IRA 904
Amberlite IRA 400
Amberlite IRA 402
Amberlite IRA 440c
Amberlite IRA 900
Dowex 1X8-400
6
21
2
22
24
7
6
7
4
14
94
50
96
60
Dowex 2X8-400
production of acrylates, commodity chemicals produced globally on a large scale
for the manufacture of a wide variety of polymers [5,6].
Results and Discussion
Adsorption of AJ270 to anionic exchange resins
Since bacteria generally carry an overall negative charge at physiological pH, a
variety of anionic exchange resins were tested as adsorption matrices. Most of
them except Dowex 1 and 2 showed poor adsorption of AJ270, with Dowex 1
having the highest capacity (Table 1). Treatment of the bacteria with EDTA to
remove cations from the bacterial surface improved attachment to some of the
Amberlite resins signi®cantly, but Dowex 1 remained the best adsorption matrix
and was used for all further experiments. Three different Dowex 1 resins were
tested, all of mesh size 200±400 but differing in the degree of cross-linkage (2, 4,
8
%). For each, the dry weight of AJ270 adsorbed increased with the dry weight of
3
bacteria added over the range of 2±8 mg dry weight per cm settled resin, but little
further binding was observed on increasing the amount of bacteria added further.
Dowex 1 Â 4±400 (4% cross-linked) bound slightly more bacteria (7.6Æ0.3 mg dry
3
weight AJ270 per cm settled resin; 15.4 mg dry weight per g dry resin) than the
3
other two resins (7.0Æ0.2 mg/cm for 2% cross-linked and 6.4Æ0.2 mg/ml for 8%
cross-linked), and the data below were obtained with this matrix. The speci®c
amidase activity of the bound AJ270 decreased as the bacterial loading increased
so that resin loaded with 7.6 mg dry weight bacteria per cm had only half the
3
3
speci®c activity of resin carrying 2 mg dry weight per cm (0.9 and 1.6 units per mg
dry weight, respectively).
Under the binding conditions used, the resin reached 80% of capacity within
min. and full capacity after 2 min. Scanning electron microscopy of resin loaded
with 7.6 mg dry weight AJ270 per cm showed the adsorbed biocatalyst to consist
of a monolayer of single, rod-shaped bacteria attached along their longest axis and
1
3

Immobilization of Rhodococcus AJ270
657
ꢀ
Table 2. Thermal stability of amidase at 60 C in AJ270 adsorbed to Dowex 1, entrapped in agar or in
3
free suspension. Samples (1 cm ) of AJ270 in free suspension, adsorbed to Dowex 1X4-400 at 2 mg
3
dry weight bacteria per cm resin or entrapped in agar at 20 mg dry weight per cm gel were quickly
3
ꢀ
heated to 60 C and held at that temperature for timed intervals. They were then rapidly cooled in an
ice-water bath and assayed for amidase activity against acetamide. 100% ˆ 3.1 units per mg dry
weight for free bacteria, 1.7 units per mg dry weight for adsorbed bacteria, or 1.6 units per mg dry
weight for entrapped bacteria
%
of initial amidase activity
Time (min)
Free bacteria
Adsorbed bacteria
Entrapped bacteria
0
5
100
74
35
31
26
19
100
96
61
53
44
34
100
98
70
62
47
38
1
2
4
6
0
0
0
0
evenly covering the surface of the resin beads. This suggests that the groups
involved in attachment are evenly distributed over the bacterial surface whilst
cell-to-cell interactions are not important in attachment. Transmission electron
microscopy of bead cross-sections con®rmed that the bacteria did not penetrate to
the interior of the resin beads.
Effect of salt, temperature, and pH value on the amidase activity
of adsorbed AJ270
Adsorbed amidase activity was not eluted by prolonged washing of the
immobilized biocatalyst in a buffer containing 2.5 M NaCl indicating that factors
other than charge are important in attachment. The temperature pro®le for the
3
amidase activity of immobilized bacteria (tested at 0.2 cm resin or 0.4 mg dry
weight AJ270 per cm buffer) was very similar to that of bacteria in free
3
3
suspension (0.4 mg dry weight per cm buffer), activity increasing steadily to a
ꢀ
ꢀ
maximum at about 55 C with only 60% remaining at 70 C. The amidase activity of
the adsorbed bacteria was signi®cantly more stable to high temperature than that of
bacteria in free suspension (Table 2) with a signi®cant three-fold increase in T0.5
from 7 min to 22 min. The pH pro®les for adsorbed and free bacteria were very
similar at high pH values with clear optima at pH ˆ 8.0, but the adsorbed bacteria
were signi®cantly less active at acid pH values (Fig. 1). This effect was not due to
elution of adsorbed bacteria with decreasing pH since washing the immobilized
biocatalyst with buffers of pH value from 5.5 to 9.0 and then measuring their
amidase activity at pH 8.0 resulted in losses of amidase activity not exceeding 4%.
Entrapment of AJ270 in agar and agarose beads
There are several gels that may be used for the entrapment of bacteria differing in
their properties and method of preparation. In this study, agar was chosen for its

658
J. Colby et al.
Fig. 1. Effect of pH on the amidase activity of Rhodococcus AJ270. AJ270 adsorbed to Dowex 1X4-
4
3
00 beads at 2 mg dry weight per cm of resin entrapped in agar or agarose beads at 20 mg dry
3
weight per cm gel or free in suspension were assayed for amidase activity against acetamide in
buffers of different pH value. 100% ˆ 3.3 units per mg dry weight for free bacteria, 1.7 units per mg
dry weight for adsorbed bacteria, or 1.6 units per mg dry weight for entrapped bacteria
relatively low cost and ease of handling, and compared with agarose, its more
expensive counterpart. When bacteria are entrapped within a carrier gel, the rate of
reaction is dependent not only on the speci®c activity of the bacteria but also upon
mass transfer limitations imposed by the topography of the carrier matrix. If the
mass of bacteria within the gel is increased, there will be a point at which the rate
of reaction becomes diffusion limited, and further increases in bacterial loading
will have little effect. This point occurred above 20 mg dry weight bacteria (per
3
cm gel), equivalent to 1.33 g dry weight (per g dry matrix) with agar and agarose
gels, under the conditions used in this study (1.5% w/v gels; bead size < 1 mm
diameter; 50 mM acetamide; assay conditions as described in the Experimental
section) (Fig. 2). Bacterial loadings in excess of 80 mg dry weight (per cm gel)
3
were readily achieved with agar and agarose gels but with no further increase in
amidase activity per g gel.
Effect of temperature and pH on the amidase activity of entrapped AJ270
The temperature-activity pro®les for the amidase activity of bacteria entrapped in
3
3
agar or agarose (tested at 0.2 cm gel or 4 mg dry weight AJ270 per cm buffer)
were almost identical to those reported above for bacteria in free suspension or for
bacteria adsorbed to Dowex 1. Amidase activity was most stable in agar-entrapped
ꢀ
bacteria with T0.5 at 60 C ˆ 34 min, a more than four-fold increase over the T of
0.5
free bacteria (Table 2). The pH optimum of amidase activity was shifted to lower
pH values in both agar- and agarose-entrapped bacteria (Fig. 1) with highest
activities observed at pH values between 6 and 7, whereas the activity of bacteria in
free suspension was highest at pH 8.0 and was 32% lower at pH 6.

Immobilization of Rhodococcus AJ270
659
Fig. 2. Effect of bacterial loading on the speci®c amidase activity of Rhodococcus AJ270 entrapped
3
in agar beads; 100% ˆ 32 units per cm gel)
Kinetics of amidase activity in free and immobilized AJ270
Apparent K and V values for acetamide and acrylamide hydrolysis by AJ270
washed suspensions and by Dowex-adsorbed, agar- and agarose-entrapped AJ270
m
max
are given in Table 3. The K values for free bacteria and for adsorbed bacteria were
m
Table 3. Kinetic data for amidase activity in free and immobilized AJ270. K and Vmax values were
m
determined by Lineweaver-Burke double reciprocal plots of speci®c amidase activity as a function of
substrate concentration; n.a.: not applicable, n.d.: not done
Immobilisation Matrix
Parameter
Loading
None
n.a.
Dowex 1
20
Agar
20
Agarose
20
3
(
mg dry weight per cm )
Apparent Km for acetamide
mM)
Apparent Vmax for acetamide
U per mg dry wt. bacteria)
Apparent Vmax for acetamide
6.7Æ0.4
3.9Æ0.3
n.a.
7.3Æ0.5
2.4Æ0.3
5
20.4Æ0.8
2.2Æ0.2
48
20.1Æ0.9
2.1Æ0.3
42
(
(
3
(
U per cm matrix)
Apparent Km for acrylamide
mM)
Apparent Vmax for acrylamide
U per mg dry wt. bacteria)
Apparent Vmax for acrylamide
9.2Æ0.3
1.1Æ0.2
n.a.
11.1Æ0.5
0.8Æ0.2
1.6
24.7Æ0.9
0.7Æ0.2
14
n.d.
(
n.d.
(
n.d.
3
(
U per cm matrix)

660
J. Colby et al.
very similar, presumably because the monolayer of bacteria adsorbed to the beads
of anionic exchanger does not signi®cantly affect the free diffusion of substrate to
the bacteria. In contrast, bacteria entrapped in gel beads are impeded in their access
to substrate, and this is re¯ected in the three-fold higher apparent Km value.
Nevertheless, because of the higher bacterial loadings achieved by entrapment, this
process resulted in an immobilized biocatalyst with at least 8-times the activity (per
unit volume) achieved with Dowex-adsorbed bacteria (based on Vmax values).
We conclude that adsorption has little effect on amidase kinetic parameters, but
signi®cantly improves the temperature stability. As might be expected, entrapment
in agar or agarose beads has a greater effect on kinetic parameters producing a
biocatalyst with higher K . However, this disadvantage is outweighed by (i) much
m
higher bacterial loadings, (ii) greater temperature stability, and (iii) a shift in the
pH optimum from 8 to 7. As pointed out by Brennan et al. [5], pH 7 is ideal for
ammonium acrylate production. In addition, the physical characteristics of the gel-
entrapped biocatalysts are better suited to continuous column reactors than Dowex
1
, allowing for higher ¯ow-rates at lower pressure differentials. No advantage was
found in using the more expensive agarose, rather than agar, as the immobilization
matrix.
Biotransformation of acrylamide to acrylic acid in an STR
The kinetic data for acrylamide hydrolysis by agar-entrapped AJ270 (Table 3), in
comparison with the data for acetamide hydrolysis, indicated a somewhat poorer
af®nity for this substrate and a signi®cantly lower reaction rate. Nevertheless, the
data suggest that entrapped AJ270 could be a very useful biocatalyst for acrylic
acid production. In order to test this, the hydrolysis of 250 mM acrylamide to
3
acrylic acid was attempted using 10 cm of agar-entrapped bacteria (20 mg dry
3
3
ꢀ
weight bacteria per cm ) in a 50 cm batch STR at 30 C (Fig. 3). Under the
conditions used (see Experimental), acrylamidase activity was independent of
stirrer speeds between 100 and 500 rpm. A STR operated in batch mode may be
subject to diffusional limitations represented by the stationary boundary layer
between the bulk ¯uid and the support surface. Clearly even the lowest stirrer speed
used in this study was suf®cient to overcome this boundary layer effect. About one
®
fth (2.44 mmol) of the starting acrylamide (12.5 mmol) was converted to acrylic
acid after 30 min (speci®c activity for acrylamide hydrolysis ˆ 0.4 units per mg dry
weight bacteria), but the reaction then slowed down very signi®cantly so that after
2
10 min only 59% of the acrylamide had been hydrolyzed. The values quoted are
the mean values of three separate experiments, and in each case at least 95% of the
acrylamide hydrolyzed was accounted for as acrylic acid.
Inhibition of amidase activity by acrylic acid
The results in Fig. 3 suggest that contact with high concentrations of acrylamide or
acrylic acid inhibits the amidase activity of the immobilized biocatalyst. Exposure
to high acrylamide concentrations was reduced by stepwise addition of the
substrate (Fig. 3). This increased the ®nal concentration of acrylic acid achieved,
but only slightly improved the proportion of acrylamide hydrolysed (from 59 to

Immobilization of Rhodococcus AJ270
661
Fig. 3. Hydrolysis of acrylamide by agar-entrapped Rhodococcus AJ270 in a STR. In the ®rst
3
experiment (E1), agar-entrapped AJ270 (10 cm ; 20 mg dry weight per cm gel) was incubated with
3
3
2.5 mmol acrylamide in a 50 cm STR. In the second experiment (E2), the STR contained 5 mmol
1
acrylamide, and further 5 mmol were added at 60, 120 and 180 min. Acrylamide and acrylic acid
concentrations were determined by GC
6
(
6%), suggesting that acrylic acid is the major inhibitor. This was con®rmed
Table 4) by incubating entrapped bacteria for 20 min with 0±250 mM acrylic acid,
then washing the biocatalyst with buffer, and measuring its amidase activity against
acetamide. Inhibition by acrylic acid in the experiments described above was not
due to a signi®cant fall in pH.
Continuous acrylamide hydrolysis/acrylic acid production in a column reactor
In a continuous column reactor fed with an acrylamide solution, only the
biocatalyst at the outlet end of the column is exposed to high concentrations of
product and therefore subject to inhibition by acrylic acid. The amidase activity of
such a column is unlikely to suffer the dramatic losses observed in the STR.
3
Continuous column reactors containing 40 cm of agar-entrapped AJ270 were
challenged with acrylamide solutions of different concentrations at various ¯ow
Table 4. Inhibition of amidase activity in entrapped AJ270 by acrylic acid. Samples of AJ270
3
ꢀ
entrapped in agar beads at 20 mg dry weight per cm gel were incubated at 4 C for 20 min in buffer
containing various concentrations of acrylic acid. The beads were then spun, washed in buffer, and
3
assayed for amidase activity against acetamide. 100% ˆ 31 units per cm gel.
Acrylic acid (mM)
% Activity remaining
0
100
86
81
74
61
50
2
5
5
0
1
1
2
00
75
50

662
J. Colby et al.
Fig. 4. Effect of ¯ow rate and acrylamide concentration on the amidase activity of agar-entrapped
3
Rhodococcus AJ270 in a column reactor. Agar-entrapped AJ270 (40 cm gel; 20 mg dry weight per
3
cm ) in a column reactor (internal diameter 2.4 cm) was fed with buffers containing 0.05±0.25 M
acrylamide at various ¯ow rates. Acrylamidase activities were computed from the acrylamide
concentration in the input and ef¯uent streams
rates. Speci®c activity increased with both ¯ow rate and acrylamide concentration
until a plateau was reached at about 12.3 units (per cm gel) (Fig. 4). As expected,
3
%
conversion rates followed the reverse trend (Fig. 5). The results suggested that a
3
� 1
¯
ow rate of 1.3±2.0 cm Á min
acrylamide.
In a ®nal set of experiments, similar column reactors packed with agar-
would achieve > 95% conversion of 0.25 M
3
3
entrapped AJ270 (40 cm gel; 20 mg dry weight bacteria per cm of gel) and
ꢀ
maintained at 30 C were fed continuously with 0.25 M acrylamide in 0.1 M tris-
� 1
3
HCl buffer ( pH 7) at 1.5 cm Á min for 30 h. These reactors were operated at
>
97% conversion of the substrate to acrylic acid for 22 h (speci®c acrylamidase
activity ˆ 0.4 units per mg dry weight bacteria), after which the % conversion fell
steadily to 60% at 30 h.
Rhodococci expressing nitrile hydratase and immobilized in polyacrylamide
gel are used for the commercial production of acrylamide from acrylonitrile in a
plant with a capacity of 30000 t/a [7]. In contrast, a commercial process for the
production of acrylate has not been developped, even though this commodity
chemical is widely used in the production of polymeric ¯occulents, thickeners,
adhesives, and adsorbents. Several workers have described the isolation of bacteria
able to grow on acrylamide [8±11], but most work on acrylamide hydrolysis, as in
the present study, has involved bacteria grown on simpler, saturated aliphatic
nitriles and amides.
Rhodococcus AJ270 grown on acetamide expresses high amidase activity
against acrylamide. However, with 0.25 M acrylamide as substrate, exposure to the
acrylic acid product slowly inactivates the amidase activity, eventually limiting the

Immobilization of Rhodococcus AJ270
663
Fig. 5. Effect of ¯ow rate and acrylamide concentration on percent conversion by agar-entrapped
Rhodococcus AJ270 in a column reactor. The conditions were as for Fig. 4; percent conversion
®gures were computed from the acrylamide concentrations in the input and ef¯uent streams
%
conversion of the acrylamide substrate. Inactivation of immobilized acetamide-
grown Corynebacterium nitrilophilus by high concentrations of acrylamide and
acrylic acid during the biocatalytic hydrolysis of acrylamide has been noted
previously [5] and is thought to be due to the effect of these ꢀ-unsaturated
compounds on sulfydryl groups essential for amidase activity [12]. In the present
study, inactivation was minimized by using a ®xed-bed column reactor allowing
continuous production of acrylic acid for 22 h at > 97% conversion of acrylamide.
Improvements on this basic protocol, such as the periodic replacement of the
biocatalyst at the outlet end of the reactor, might enable extended production of
acrylate using this biocatalyst.
Experimental
Chemicals and immobilization matrices
Acrylic acid (99%) and acetamide (99‡%) were obtained from Aldrich (Dorset, UK). Dowex 1X4-
400, puri®ed agar, agarose (type 1-A), and all other chemicals and matrices were purchased from
Sigma (Dorset, UK). Vegetable oil (pure sun¯ower oil) was obtained from a large UK retail outlet
ASDA Stores, Leeds). The ion exchange resins used in the study (see Table 5) were all polystyrene
(
based. Amberlite IRA900 and IRA904 are macroreticular, the others are gels.
Analytical methods
Acrylamide and acrylic acid were determined by gas chromatography as described previously [2].
The concentrations of ammonia in assay samples were determined using Nessler's reagent. The

664
J. Colby et al.
Table 5. Properties of the anionic exchange resins used
Resin
Active group
Mesh size
% Cross linked Capacity
(
Ionic form
meq Á cm�
3
)
�
�
�
Amberlite IRA904
Amberlite IRA400
Amberlite IRA402
quaternary ammonium
quaternary ammonium
quaternary ammonium
20±50
16±50
16±50
16±40
16±50
±
8
6
±
±
8
8
1.0
1.4
1.3
1.0
1.0
1.2
1.2
Cl
Cl
Cl
�
Amberlite IRA440c quaternary ammonium
OH
�
�
Amberlite IRA900
Dowex 1X8-400
Dowex 2X8-400
quaternary ammonium
Cl
Cl
trimethylbenzylammonium 200±400
dimethylethanol-
benzylammonium
�
200±400
Cl
3
3
3
sample (0.025 cm ) was diluted with 0.85 cm of water, and then 0.125 cm of reagent (ammonia
colour reagent from Sigma, Dorset, UK) were added. The colour was allowed to develop for 10 min
before measuring the absorbance at 420 nm. Standard curves were prepared using NH Cl (> 99.5%).
4
Bacterial concentrations of washed suspensions were determined turbidometrically by measuring
their absorbance at 540 nm and converted to dry weights using a standard curve. The data are the
mean values from three separate experiments; in no case the results of single experiments differed by
more than 12%.
Bulk growth of Rhodococcus AJ270
AJ270 was grown on 50 mM acetamide, and washed suspensions were prepared as described
previously [1].
Adsorption on Dowex 1 and other anionic exchangers
Bacteria and resin were both washed three times in 0.1 M tris-HCl buffer, pH 8.0, prior to use.
3
3
Bacteria were re-suspended in the same buffer to 22 mg dry weight (per cm ) and 5 cm of the
suspension were added to a Universal bottle containing 3 g of the washed resin. The mixture was
agitated for 45 min on a Denley Spiramix, the resin allowed to settle for 5 min, and then the
supernatant was removed. The settled resin was washed ®ve times with 0.01 M tris-HCl buffer, pH
8.0, to remove any unadsorbed bacteria. The amount of bacteria adsorbed was determined by
comparing the absorbance (A540nm) of the bacterial suspension before and after adsorption. The basic
method was modi®ed by changing the ratio of bacteria to resin, including EDTA in the washing
buffers, or by using different anionic exchangers.
Entrapment in agar or agarose
The method was modi®ed from Wikstrom et al. [13]. Bacteria were washed three times with 0.1 M
tris-HCl buffer, pH 8.0, resuspended in the same buffer to the desired concentration, and mixed with
ꢀ
an equal volume of 3% (w/v) molten agar or agarose at 48 C. The mixture was quickly added to an
ꢀ
equal volume of vegetable oil also held at 48 C and agitated with a Rushton-style impeller at
250 rpm to form an even emulsion. The temperature of the emulsion was then lowered by immersion
of the vessel in an ice-water bath whilst maintaining the agitation. Once the temperature had fallen
ꢀ
below 20 C, agitation was stopped, and the two phases were allowed to separate. The aqueous phase
consisted of a heterogeneous mix of bead sizes < 1 mm in diameter. The beads were washed with
buffer until no oil was visible on the buffer surface.

Immobilization of Rhodococcus AJ270
665
Bioreactors
3
The stirred-tank reactor (STR) consisted of a glass vessel of working volume 50 cm , ®tted with
baf¯es and agitated using a Rushton-style impeller at 100±500 rpm. Fixed-bed column reactors
3
consisted of jacketed glass columns (2.4 cm internal diameter) packed with 40 cm of agar-entrapped
biocatalyst. Substrate solution was pumped through the columns in an upward direction to prevent
ꢀ
compression of the gel. Both types of reactors were kept at constant temperature (30 C) by
connecting their water jackets to a circulating water bath.
Determination of amidase activity
3
For washed suspensions and Dowex 1-immobilized bacteria, samples (0.1 cm ) and 0.35 cm of
3
ꢀ
0
.1 M tris-HCl buffer, pH 8.0 were placed in an Eppendorf tube and pre-warmed to 30 C in a water
3
bath. Substrate (0.05 cm of 0.5 M sodium acetamide unless stated otherwise) was added, and the
ꢀ
3
tube was incubated at 30 C for 5±30 min before stopping the reaction with 0.025 cm of 0.25 M HCl.
The mixture was then centrifuged for 2 min in a microcentrifuge, and the clear supernatant was
removed for ammonia determination.
3
Gel-entrapped bacteria were drained of all buffer and 5 g removed to a 50cm conical ¯ask. The
3
beads were resuspended in 22.5cm of 0.1M tris-HCl buffer, pH 8.0, and warmed to 30 C in a shaking
ꢀ
3
waterbath. Pre-warmed substrate solution (2.5cm of 0.5 M sodium acetamide unless otherwise stated)
was added, and samples were removed at timed intervals for ammonia determination.
In experiments where amidase activity was determined at different pH values, 0.1 M sodium
phosphate buffers were used for pH 6±7.0 and 0.1 M tris-HCl buffers for pH 7.5±9.5.
Electron microscopy of adsorbed bacteria
ꢀ
Samples were ®xed in 2.5% (w/v) glutaraldehyde in 0.1 M sodium phosphate buffer, pH 7.3, at 4 C.
The ®xed samples were then washed twice with the same buffer before ®xing in 1% (w/v) osmium
tetroxide for 1 h. After washing twice with water, the samples were ®xed in 1% (w/v) uranyl acetate
for 30 min and washed in water for 5 min. After ®xing, samples were dehydrated by washing with
ethanol (50, 70, 90% v/v and then twice with 100% for 15 min each wash).
Samples for TEM were embedded in epoxy resin, sectioned, and the sections mounted on copper
grids. Grids were stained with saturated uranyl acetate in 70% (v/v) ethanol for 10 min, washed with
water, then with alkaline lead citrate for 10 min, and then with water again before drying. Samples
for SEM were mounted on carbon adhesive discs and sputter-coated initially with carbon and then
gold/palladium. Specimens were viewed using a Hitachi H7000 microscope and, for SEM, a H7110
scanning attachment.
Acknowledgements
This work was supported in part by research grant GR/J58299 from BBSRC.
References
[
[
[
[
[
1] Snell D, Colby J (1999) Enz Microbial Technol 24: 160
2] Blakey AJ, Colby J, Williams E, O'Reilly C (1995) FEMS Microbiol Lett 129: 57
3] Meth-Cohn O, Wang M-X (1997) J Chem Soc Perkin Trans 1099
4] Meth-Cohn O, Wang M-X (1997) J Chem Soc Chem Commun 1041
5] Brennan MR, Armitage YC, Mortimer MG, Hughes J, Ramsden DK (1995) Biotechnol Lett 17:
513
[
6] Nagasawa T, Yamada H (1989) Trends Biotechnol 7: 153

666
J. Colby et al.: Immobilization of Rhodococcus AJ270
[7] Yamada H, Kobayashi M (1996) Bioscience Biotechnol Biochem 60: 1391
[8] Nawaz MS, Franklin W, Cerniglia CE (1993) Can J Microbiol 39: 207
[9] Nawaz MS, Franklin W, Cerniglia CE (1994) Environ Sci Technol 28: 1106
[
[
10] Kumar A, Kumar A (1998) J Microbiol Biotechnol 8: 347
11] Nawaz MS, Khan AA, Seng JE, Leakey JE, Siitonen PH, Cerniglia CE (1994) Appl Environ
Microbiol 60: 3343
[
12] Cavins JF, Friedman M (1968) J Biol Chem 143: 3357
13] Wikstrom P, Szwajcer E, Brodelius P, Nilsson K, Mossbach K (1982) Biotechnol Lett 4: 153
[
Received November 18, 1999. Accepted December 14, 1999