JOURNALOFBIOSCIENCEAND BIOENGINEERING

Vol. 93, No. 2, 125-129.2002

Influence of External Mass Transfer Limitation on Apparent

Kinetic Parameters of Penicillin G Acylase Immobilized
on NonPorous Ultrafine Silica Particles
AZADEH KHEIROLOMOOM,‘* FARHAD KHORASHEH,’
AND HOSSEIN FAZELINIA’

Sharif Universityof Technology,Department ofchemical Engineering,

PO. Box 113654’639, Azadi Ave, Tehran. Iran’

Received 2 I June 2001/Accepted 24 October 200 I

Immobilization of enzymes on nonporous supports provides a suitable model for investigating

the effect of external mass transfer limitation on the reaction rate in the absence of internal dif-
fusional resistance. In this study, deacylation of penicillin G was investigated using penicillin acy-
lase immobilized on ultrafine silica particles. Kinetic studies were performed within the low-sub-
strate-concentration region, where the external mass transfer limitation becomes significant. To
predict the apparent kinetic parameters and the overall effectiveness factor, knowledge of the ex-
ternal mass transfer coefficient, kLu, is necessary. Although various correlations exist for estima-
tion of &Q, in this study, an optimization scheme was utilized to obtain this coefficient. Using the
optimum values of k,a, the initial reaction rates were predicted and found to be in good agree-
ment with the experimental data.

[Key words: penicillin G acylase, ultrafine silica particles, apparent kinetic parameters, external mass transfer
limitation]

Immobilized enzymes are of great technological and sci-
entific importance. To develop reactor design for immobi-
lized enzyme catalytic systems, it is important to examine in
some detail the kinetic properties of immobilized enzymes.
In general, the maximum reaction rate and the Michaelis
constant determined for an immobilized enzyme are differ-
ent from the kinetic parameters evaluated from its free state.
These kinetic parameters may be affected by a number of
factors including interaction between enzyme and support,
steric hindrance, and diffusive and electrostatic phenomena,
which serve to alter the concentrations of substrates and
other species at the site of enzyme action from bulk solution
values ( l-6).
The kinetic parameters defined for immobilized enzyme
systems are called intrinsic parameters in the absence of
mass transfer limitation. Apparent kinetic parameters are
those observed in the presence of mass transfer limitations.
Unfortunately, much less work on obtaining intrinsic kinetic
parameters for immobilized enzyme systems has been car-
ried out. On the other hand, because of the denaturation of
enzyme protein and the interaction between the enzyme and
the support, it is very difficult to compare the kinetic param-
eters of an immobilized enzyme system with those of its
free state ( 1,2, 5-7).
The combined effects of diffusive resistance, electrostatic
field and substrate dependence on the rate of the reaction
catalyzed by an enzyme immobilized on porous and nonpo-
* Corresponding author. e-mail: akheirolomoom@,hotmail.com
phone: +98-2 l-60058 19 fax: +98-2 I-6022853
rous supports have been reported (8). Furthermore, equa-
tions for apparent kinetic parameters have also been ob-
tained for an enzyme immobilized on a porous ceramic sup-
port (9).
In the present work, the applicability of the proposed
equations for apparent kinetic parameters was investigated
using the experimental data obtained for penicillin G acy-
lase immobilized on nonporous ultratine silica particles. For
this purpose, the initial reaction rates were measured for a
wide range of initial substrate concentrations. The veritica-
tion of the proposed equations was performed using the ex-
perimental data at sufficiently low substrate concentrations,
where the Michaelis-Menten equation can be used (1 O-13).
THEORETICAL BACKGROUND
Penicillin G acylase is generally recognized to be inhib-
ited by high substrate concentration and since the product
formation is negligible at the initial stage of the reaction, the
following model has been suggested for its kinetic behavior
(14):
KS,
V= ^. (1)
K,,,+S, l+?
i ,)
where S, is the bulk concentration of substrate, V,,, is the
maximum reaction rate, K, is the Michaelis constant and K,
is substrate inhibition constant. When the enzyme is immo-
bilized on a fixed surface, the concentration of the substrate
at the surface may be different from that in the bulk fluid

125
126 KHEIROLOMOOM ET AL. J. Broscl. BIOENF..

phase due to external and internal mass transfer resistance,

and electrostatic effects. In the absence of these limitations, Ef and f are employed to calculate the apparent kinetic
the reaction rate can be expressed by Eq. 1. However, in the b
presence of diffusional and electrostatic effects, introducing parameters.
an effectiveness factor can modify the actual reaction rate as In the presence of diffusional resistances and electrostatic
follows:
effect, the relationship between
Sb and S,, becomes non-
v
Ksb
“=Ef
,+,(1+~) linear (l-3). For simplicity, Vim and K? are defined as the
apparent kinetic parameters and Eq. 4 can then be rewritten
Here, the kinetic constants V,, and Km are called intrinsic ki- as Eq. 9.
netic parameters. The above equation can be rearranged to
the following form: (9)

(3) These apparent kinetic parameters can be calculated from

It has been shown that a high concentration of penicillin tangent lines to the nonlinear plot of $ -Sb (8, 9, 15).
G is a mild inhibitor of penicillin G acylase and the value of
the substrate inhibition constant is two orders of magnitude (10)

higher than S,, so that the $ term in the above equation can
I
be neglected at sufficiently low substrate concentration (1 O-
Km E,+P,(P,+ ll$ 1 (11)
Ef-(pb+
I)$
b
13). By neglecting the substrate inhibition term, the above K”m=Km~,=
equation reduces to:
b

Here, y, and y2 are dimensionless parameters representing

the deviations of the apparent values of the maximum re-
action rate and the Michaelis constant from their intrinsic
which is generally known as the Michaelis-Menten equa-
values. Substitutions of Eq. 5 and 8 into Eqs. 10 and 11 re-
tion.
sult in Eqs. 12 and 13, respectively.
The combined effects of external diffusive resistance and
electrostatic field on the rate of reaction catalyzed by an en- J?m&(P
-Jr>’ (12)
zyme immobilized on a nonporous surface, which follows
“= 2~E{(~+pb)(p-h/r)-2pbcp,}
irreversible Michaelis-Menten kinetics, have been investi-
gated and the following expression was derived for the cal-
culation of the effectiveness factor (5,6, 8, 15): (13)
m (‘k++Pb)(fhh2Pb(PE

(5) In the present work, the application of the above equa-

tions for surface-immobilized enzyme in the absence of the
where electrostatic effects was investigated.
u=Pb+cpE+e* (6)
5=+4&p, MATERIALS AND METHODS
(7)
Chemicals Penicillin G potassium (PGK) was purchased
Equation 5 shows that for a surface-immobilized enzyme,
from Jaber Iben-e-Hayan Pharmaceutical Co. (Tehran, Iran). Non-
the effectiveness factor is a function of the dimensionless porous ultratine silica particles (Snowtex 30,30.4 wt%) were a gift
sb
from Kobe University (Kobe) and had an average diameter of
substrate concentration &,= K, dimensionless surface po- 15 nm (16, 17). All chemicals used were of analytical grade.
rn Production of penicillin G acylase and its partial purification
Kl Escherichia coli (E. cob) ATCC 1I 105 was grown in a culture
tential h and the Thiele modulus (pE= Mk K , where Mis
La m medium consisting of the following: phenylacetic acid 0.2% (w/v),
veast extract 0.5% (w/v). NHXI 0. I % (w/v\. K,HPO. 0.1% (w/v>.
the electrostatic modifier and kLa is the external mass trans- i(H,PO, 0.01% (wiv), ‘hgSi),.7H,O 6.026 (k/v). ‘The pfi w&
fer coefficient. adjusted to 7.00 with NaOH. Aliquots (100 ml) of the culture me-
Differentiation of Eq. 5 with respect to p,, leads to follow- dium were dispensed into 500-ml conical flasks with lint caps and
ing equation: sterilized at 12 1°C for 20 min. The flasks were inoculated with E.
coli and incubated on a rotary shaker (220 rpm) at 24°C for 46 h.
dE, _ Cells were harvested by centrifugation, washed and incubated
--$--[h@b++y)-V+&] (8) in 10 mM sodium phosphate buffer (pH=7.80). Subsequently, cells
dab
were ruptured using an ultrasonicator Bondlin-HD200 at cycle 90
for 90 s. Cell debris was removed by centrifugation (8OOOxg, 30
VOL.. 93,2002 MASS TRANSFER EFFECT ON SURFACE-IMMOBILIZED PENICILLIN G ACYLASE 127

1 ammonia k

z 0 5 10 15 20 25
lime (min)
1 prow I 1 buffer 1
FIG. 2. Time-course of ammonia titration at a substrate concentra-
tion of 0.2 mM.
FIG. I Schematic diagram of the reactor used for the initial veloc-
ity measurements.

min, 04°C). The enzyme solution was then partially purified us-
ing an ammonium sulfate salt, and the fraction precipitated be-
tween 30% and 60% of ammonium sulfate saturation was collected
by centrifugation (14,000 xg, 40 min, 0_4”C), dissolved in 10 mM
sodium phosphate buffer (pH=7.80), and dialyzed against the
same buffer.
No penicillinase activity was detected in the partially purified
enzyme solution.
Enzyme immobilization An enzyme solution (4 ml) of the
desired concentration (0.446mg/ml, which was optimized in an
other investigation) was mixed with an equal volume of suspen- L ob -~ - -- --_ ~ ~-_ ~~~~
5 10 15 20 25 30
sion of ultrafine silica particles, and 0.1 ml of glutaraldehyde solu-
SbO (mM)
tion (1%) was added to this mixture (I 6, 17). The immobilization
was carried out, in the presence of phenylacetic acid as a com- FIG. 3. Nonlinear plot of $ versus S, in the deacylation of peni-
petitive inhibitor of penicillin G acylase to protect the active site,
for 2 h at 30°C (IO-I 2). The ultrafine particles covalently cross- cillin C potassium by penicillin G acylase immobilized on ultrafine
linked to the enzyme were then recovered by centrifugation silica particles. Solid line represents the calculated results in the ab-
(14,000 xg, 40 min, 04OC). These immobilized enzyme molecules sence of external mass transfer limitation.
were washed and resuspended in 10 mM sodium phosphate buffer
(pH=7.50) and used for kinetic studies. In all experiments, the
surface area of the particles used per unit volume of the reaction observed (9).
mixture was 1.Ol m’/mi. Based on the initial substrate concentration S,, and ne-
Reactor conditions and determination of the reaction rate glecting the substrate inhibition term at low substrate con-
A schematic diagram of the reactor used for the initial velocity
centration, the initial reaction rate can be expressed by:
measurement is presented in Fig. 1. Aliquots of 20 ml of aqueous
solutions of PGK in 10 mM sodium phosphate buffer (pH=8.00) at
different concentrations were prepared and the reactions were
started by the addition of an appropriate amount of the immobi-
lized enzyme, to achieve a linear time course for PGK consump- Since for the system investigated in this study neither the
tion during the first 15 min of the reaction. The agitation speed for support nor the substrate is electrically charged, the electro-
all experiments was 170 rpm, which was found to be the optimum
static modifier (I#) was set to unity. Figure 3 indicates that
agitation speed in a previous investigation.
Initial enzymatic reaction rates were measured by a pH-stat s
method. This method is based on the titration of phenylacetic acid the relationship between $ versus S,,, becomes linear at
formed during PGK hydrolysis by ammonia. The initial reaction
rate was calculated from the slope of the time course of ammonia high substrate concentrations. This suggests that external
consumption at the time of reaction (IO, 1I). All reported data diffusional limitation becomes negligible under such condi-
were mean values obtained from three repeated experiments. tions, leading to an effectiveness factor value of unity. The
intrinsic kinetic parameters are determined from the tangen-
tial line to the plot in this high substrate concentration range.
RESULTS AND DISCUSSION
Therefore, V,,, and K,,, can be determined from the slope
Figure 2 shows the time course of ammonia titration at a and intercept of this line. The calculated values for V, and
substrate concentration of 0.2 mM. As stated above, the ini-
mm01
tial reaction rates were obtained at each substrate concentra- K,, were found to be 0.150 x 10m3s and 9.15 -
lit ’
tion. Figure 3 shows a plot of the relationship of y, against respectively. The deviations from this line become more
pronounced at lower substrate concentrations, indicating
S,, for the surface immobilized enzyme, which in this study significant external diffUsiona limitations.
is convex downward, whereas in the presence of internal Tangent lines to this nonlinear plot at lower substrate con-
diffusion for porous supports, an upward convex would be centrations were obtained using a cubic-spline fit. From the
128 KHEIROLOMOOM ET AL. J. BIOSCI. BIOENG.,

slopes and intercepts of these tangent lines, the values for Assign a large value for
apparent kinetic parameters ViWand K? were obtained and objective fmction,OFN
are represented by solid squares in Figs. 5 and 6, respec-
4
tively. These plots indicate that at low substrate concentra- Set iteration counter I=1
tions, the values of Viw and KF are both high and approach Set iteration counter J=l
their intrinsic values at high substrate concentrations; the
J
higher the diffusional resistance, the higher the values of Specify initial search
Vi* and K”,“. region and parameter
To investigate the accuracy of the proposed equations (S), estimate3

Eqs. 12 and 13 were rewritten in terms of the dimensionless

initial substrate concentration (&J:
parametervalues in the search region
K&P-&Y
(15)
V”,“= 2~e~(~+PbO)(P-Jr)-2Pbo(PE}

KmPbo*&~++~~)
K:y = (16) SolwEqs. lSand16andthe~Eq 17
a+ PM)(P 4) -2PbO(PE to obtain inkal rates for different
iraitialsubstrate concentrations
Using the intrinsic kinetic parameters, which were deter-
mined in the high substrate concentration range and high 4
external mass transfer coefficient, apparent kinetic parame- Detemune the objective funchcm,OFN
ters were predicted. In this investigation, optimization was
4
employed to obtain the optimum value of ,$a. The optimi- OFIWOF
zation scheme was based on a direct search method with i I
systematic reduction of the search region. The flow chart of
the algorithm is presented in Fig. 4. Selected values of kLa
were randomly chosen within the search region to minimize
the following objective function (18):
Objective function = ~(VR..,- VW,~~)*
i=l
where N is the number of experiments, vO,_ is the experi-
E$!isl Store parameter values as optimumand set OF=OFN

mental reaction rate and v,, theois the theoretical reaction rate,
which is calculated from Fq. 17.

Vo=
V”,“&
(17)
K: + S,,
The optimized value of kLu was found to be 1.99x 1O-*&. YES
The (Pi was consequently calculated to be 0.82. STOP
The predicted values of Vzp”and KT were obtained using
Eqs. 15 and 16, and are represented by the solid lines in FIG. 4. Flow chart for the optimization algorithm for parameter
Figs. 5 and 6, respectively. The theoretical curves for V”,” estimation.
and K”,” are in good agreement with those obtained experi-
mentally at low substrate concentration. At high substrate
concentrations, however, deviations between experimental
and theoretical values become more significant. These devi-
ations are probably related to the enzyme inhibition by high
penicillin G concentration. The exclusion of the substrate
inhibition term in Eq. 3 represents l-7% deviation depend-
ing on the substrate concentration.
Conclusions The apparent kinetic parameters of peni-
cillin G acylase immobilized on the surface of nonporous
2 c
ml00 --~
E ,
______
ultrafine silica particles were measured. Immobilization of - -~-.~.___._ _ _- _~

the enzyme on the surface of these particles made it possible

to study the effect of external mass transfer limitation in the 5 10 15 20 25 30
absence of internal mass transfer limitation. Dependency of L SbO (mM)
the apparent kinetic parameters on the magnitude of the
FIG. 5. Relationship between apparent maximum reaction rate and
substrate concentrations was then theoretically and experi- initial substrate concentration. Solid line represents results calculated
mentally shown. To estimate the apparent kinetic parame- using Eq. 15. Solid squares show the values calculated from the slopes
ters, the external mass transfer coefficient was optimized and intercepts of tangent lines drawn to the nonlinear plot in Fig. 3.
via an optimization scheme based on a direct search method
VOL. 93. 2002 MASS TRANSFER
01 - 5 10
0 15 20 2;. 30
SbO (mM)
FIG. 6. Relationship between the apparent Michaelis constant and
initial substrate concentration. Solid line shows the results calculated
using Eq. 16. Solid squares indicate the values calculated from the
slopes and intercepts of tangent lines drawn to the nonlinear plot in
Fig. 3.
with systematic reduction of the search region. The appar-
ent kinetic parameters predicted theoretically were in good
agreement with the experimental data.
ACKNOWLEDGMENTS
The authors wish to thank the Biochemical and Bioenvironmen-
tal Research Center (BBRC) of Sharif University of Technology
(Tehran, Iran) for technical assistance. This work was supported
financially by The OffIce of the Vice Chancellor of Sharif Univer-
sity of Technology (Tehran, Iran) in Research Affairs and the sup-
port is gratefully acknowledged.
REFERENCES
Moo-Young, M. and Kobayashi, T.: Effectiveness factor for
immobilized-enzyme reactions. Can. J. Chem. Eng., 50, l62-
167 (1972).
Kasche, V.: Correlation of experimental and theoretical data
for artificial and natural systems with immobilized biocata-
lysts. Enzyme Microb. Technol., 5, 2-13 (I 983).
Lee, G. K., Lesch, R. A., and Reilly, P. J.: Estimation of
intrinsic kinetic constants for pore diffusion-limited immo-
bilized enzyme reactions. Biotechnol. Bioeng., 23, 487-497
(1981).
Engasser, J. M. and Horvath, C.: Effect of internal diffusion
in heterogeneous enzyme system: evaluation of true kinetic
parameters and substrate diffusivity. J. Theor. Biol., 42, l37-
EFFECI ON SURFACE-IMMOBILIZEDPENlCILLlNG ACYLASE 129
155 (1973).
5. Shuler, M. L., Aris, R., and Tsuehiya, H. M.: Diffusive and
electrostatic effects with insolubilized enzymes. J. Theor.
Biol., 35, 67-76 (1972).
6. Hamilton, B. K., Stockmeyer, L. J., and Colto, C. K.: Com-
ments on diffusive and electrostatic effects with immobilized
enzymes. J. Theor. Biol., 41, 547-560 (1973).
7. Chibata, I. (ed.): Immobilized enzymes: Research and de-
velopment. Kodansha, Tokyo; J. Wiley & Sons, New York
(I 978).
8. Shiraishi, F.: Substrate-concentration dependences of the
apparent, maximum-reaction rate and Michaelis-Menten con-
stant in immobilized enzyme reactions. Inter. Chem. Eng., 32,
140-147 (1992).
9. Shiraishi, F.: Experimental evaluation of the usefulness of
equations describing the apparent maximum reaction rate and
apparent Michaelis constant of an immobilized enzyme reac-
tion. Enzyme Microb. Technol., 15, 150-154 (I 993).
IO. Erarslan, A. and Giiray, A.: Kinetic investigation of peni-
cillin G acylase from a mutant strain of Escherichia co/i
ATCC I1 105 immobilized on oxirane-acrylic beads. J. Chem.
Tech. Biotechnol., 51, 181-195 (1991).
Il. Braun, J., Le Chanu, P., and Le Goffc, F.: The immobiliza-
tion of penicillin G acylase on chitosan. Biotechnol. Bioeng.,
33,242&246 (I 989).
12. Fonseca, L. P., Cardoso, J. P., and Cabral, J. M. S.: Immo-
bilization studies of an industrial penicillin acylase prepara-
tion a silica carrier. J. Chem. Tech. Biotechnol., 58, 27-37
(1993).
13. Park, J. M., Choi, C.Y., Seong, B. L., and Han, M. H.:
Effect of mass transfer in recirculation batch reactor system
for immobilized penicillin amidase. Biotechnol. Bioeng., 24,
2215-2226(1982).
14. Ospina, S. S., Lopez-Munguia, A., Gonzalez, R. L., and
Quintero, R.: Characterization and use of a penicillin acylase
biocatalyst. J. Chem. Tech. Biotechnol., 53, 205-214 (1992).
15. Shiriahi, F.: Diffiusional and electrostatic effects on apparent
kinetic parameters of reactions catalyzed by enzyme immobi-
lized on the external surface of a support. J. Ferment. Bioeng.,
79,373-377 (1995).
16. Kondo, A., Urabe, T., and Higashitani, K.: Bioconversions
in an aqueous two-phase system using enzymes immobilized
on ultratine silica particles. J. Ferment. Bioeng., 77, 700-703
(1994).
17. Kondo, A. and Urabe, T.: Relationships between molecular
states (conformation and orientation) and activities of a-
amylase adsorbed on ultrafine silica particles. Appl. Microb.
Biotechnol., 43, 801-807 (1995).
18. Luus, R. and Jaakola, T. H. I.: Optimization by direct search
and systematic reduction of the size search region. AIChE,
19, 760-766 (1973).