Professional Documents
Culture Documents
Enzyme Immobilization On Poly (Ethylene-Co-Acrylic Acid) Films Studied by Quartz Crystal Microbalance With Dissipation Monitoring
Enzyme Immobilization On Poly (Ethylene-Co-Acrylic Acid) Films Studied by Quartz Crystal Microbalance With Dissipation Monitoring
www.elsevier.com/locate/jcis
Abstract
In this study, we use the quartz crystal microbalance with dissipation monitoring (QCM-D) to study the immobilization of the enzyme
horseradish peroxidase (HRP) on poly(ethylene-co-acrylic acid) (PEAA) films. The surface polarity of spin-coated PEAA films was varied
by heat treatments in air or in a 30% NaOH aqueous solution leading to COOH-depleted or COOH-enriched surfaces, respectively. Two
reaction schemes, direct adsorption and amine coupling, were employed for HRP immobilization on the two surfaces. The shifts in frequency
and dissipation, f and D, measured by QCM-D and the ratio D/f were used to evaluate the binding amount and the conformation
of the adsorbed enzyme. It is found that HRP immobilized via covalent linkages forms rigid and little dissipative films. In contrast, directly
adsorbed HRP films exhibit a highly dissipative structure. HRP-catalyzed oxidation of the 4-chloro-1-naphthol in the presence of H2 O2 was
used to characterize the catalytic activity of the HRP films. The results show that the enzymatic activity of the covalently immobilized HRP
tends to be higher.
2005 Elsevier Inc. All rights reserved.
1. Introduction behavior, such as for layers of proteins [3] or DNA [4], sub-
stantial deviations from the Sauerbrey equation can occur,
Quartz crystal microbalance (QCM) is a well-established depending on how the oscillatory motion of the crystal prop-
noninvasive technique suitable for the investigation of ad- agates into and through the adsorbed viscoelastic films [5].
sorption/desorption processes on solid surfaces [1]. The The damping behavior of the QCM crystal is related to the
merits of the QCM technique are the simplicity and sensitiv- conformational or structural properties of the viscous lay-
ity, by which a wide range of interfacial adsorption reactions ers.
can be monitored, on a variety of supports, in real time. For One way to study the damping effect on a quartz sensor
rigid, ultrathin, and evenly distributed adsorbed layers, the is to measure the dissipation D (inversely proportional to
Sauerbrey equation [2] describes successfully the propor- the Q factor) of a nondriven (freely oscillating) crystal [6].
tional relationship between the adsorbed mass (m) and the The so-called QCM-D is such an instrument that measures
shift of the QCM crystals’ resonance frequency (f ). On the changes in frequency (f ) and dissipation (D) simulta-
other hand, if the adsorbed material exhibits a viscoelastic neously. This is achieved by periodically switching off the
driving power and by recording the decay of the damped
* Corresponding author. Fax: +65-68720785. oscillation. The time constant of the decay is inversely pro-
E-mail address: xd-su@imre.a-star.edu.sg (X. Su). portional to D and the period of the decaying signal gives
0021-9797/$ – see front matter 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2005.01.089
36 X. Su et al. / Journal of Colloid and Interface Science 287 (2005) 35–42
surface was first treated with a 0.4 M EDC and 0.1 M NHS the substrate and the primary amine of the protein to be im-
aqueous solution (1:1 mixture) for 7 min. After this treat- mobilized [26]. The first part of Fig. 2 shows the QCM-D
ment, the surface was equilibrated with PBS buffer and the response for the EDC/NHS activation of a COOH-depleted
HRP solution (0.5 mg/mL in PBS) was then applied. After surface. The application of EDC/NHS (2 min) after cali-
binding equilibrium was reached, the cell was rinsed with bration of the initial baseline in PBS results in significant
PBS buffer to remove loosely attached enzyme. In the other shifts in frequency and dissipation, which are almost entirely
scheme, the HRP solution was directly applied on the PEAA reversed after replacing the EDC/NHS solution with PBS
surface calibrated in PBS buffer, without EDC/NHS activa- buffer. This response likely reflects the difference in viscos-
tion. ity and density between PBS buffer and EDC/NHS aqueous
solution and is thus not surface related. The changes in the
2.5. Catalytic activity of the bound HRP adsorbed mass associated to the formation of N -hydroxy-
succiniimide ester are expected to be small [27], which
HRP-catalyzed oxidation of 4-CN in the presence of is indeed reflected by the minor net changes in f and D
H2 O2 was used to study the catalytic properties of the immo- measured after rinsing in buffer solution. The successive ap-
bilized HRP. Surfaces with immobilized HRP were exposed plication of HRP on this activated surface (Fig. 2, 13 min)
to a 4-CN substrate solution in the absence of H2 O2 to cal- resulted in a steady frequency decrease accompanied with
ibrate the baseline. After equilibrium, the same substrate a dissipation increase over time, which reflects the adsorp-
solution containing 7.5 × 10−3 mM of H2 O2 was injected tion of the enzyme at the sensor surface. f and D at
to react. The precipitation of oxidized 4-CN on the sensor saturation were −22 Hz and 0.8 × 10−6 , respectively. The
surfaces was followed by QCM-D. resulting ratio |D/f | (the induced energy loss per cou-
pled unit mass) is 0.04 × 10−6 Hz−1 .
The same immobilization reaction was repeated on the
COOH-enriched surface. The EDC/NHS activation results in
3. Results
only minor net changes in f and D (data not shown) as in the
COOH-depleted case. Successive exposure to HRP (Fig. 3)
3.1. Surface preparation and HRP immobilization results in f = −17 Hz and D = 0.7 × 10−6 that leads
to |D/f | = 0.04 × 10−6 Hz−1 . These values are similar
PEAA is an organic polymer with a nonpolar scaffold to those obtained on the COOH-depleted surface, indicating
containing a small amount of polar carboxyl groups. Heat- that the amount of HRP binding and the protein conforma-
ing the polymer in air or in aqueous sodium hydroxide tion on both types of PEAA surfaces are similar.
will change the surface polarity due to the diffusion of In another reaction arrangement, HRP was directly ap-
the functionality between the surface and the bulk of the plied onto PEAA surfaces that were not activated by EDC/
polymer [24]. Fig. 1 depicts the process of the change in sur- NHS. Figs. 4 and 5 show the QCM-D responses for the
face composition of spin-coated PEAA films on the sensor COOH-depleted and COOH-enriched surfaces, respectively.
crystal. Upon heating in air, the polymer surface becomes We note that both the frequency shifts (−36 Hz versus
depleted in COOH groups, while heating in 30% NaOH −25 Hz) and the |D/f | values (0.16 × 10−6 Hz−1 ver-
leads to a COOH-enriched surface. Using the TBO dying sus 0.08 × 10−6 Hz−1 ) at saturation differ for these two
method, we confirmed that the surface COOH density for the surfaces. This suggests a profound influence of the COOH
COOH enriched surfaces is 2–3 times of that of the COOH- density on the protein adsorption, in contrast to our observa-
depleted surfaces. The corresponding polarity indicated by tions on the EDC/NHS-activated surfaces.
the advancing contact angles is 92.0 and 68.5◦ , respectively. If comparing between immobilization with and without
EDC/NHS-based amine coupling is a frequently used EDC/NHS activation, adsorption proceeds generally slower
method for the covalent tethering of proteins onto COOH- on nonactivated surfaces, but leads to higher frequency shifts
containing substrates. EDC/NHS activates the COOH groups and higher dissipation changes. The |D/f | values are
to form N -hydroxysucciniimide esters. The latter promote always higher for the physically adsorbed HRP on each in-
the formation of amide bonds between the COOH groups on dividual surface. Table 1 summarizes the QCM-D signals
Fig. 1. Schematic illustration of the diffusion of carboxylic acid groups between the surface and the bulk of the PEAA films spin-coated on the sensor crystals
upon heating treatments.
38 X. Su et al. / Journal of Colloid and Interface Science 287 (2005) 35–42
Fig. 3. HRP immobilization on EDC/NHS-activated COOH-enriched surface (EDC/NHS activation is not shown).
(f, D, and |D/f |) at saturation for HRP bound to thickness and viscosity of the deposited HRP films through
the two surfaces using the two reaction arrangements. fitting using a Voigt-based model [4]. The “effective” values
In addition to the qualitative observations of the HRP describe the entire film composed of water and HRP. The fit-
immobilization behavior, we have quantified the effective ted parameters are based on a HRP density of 1.15 g/cm3 (an
X. Su et al. / Journal of Colloid and Interface Science 287 (2005) 35–42 39
Table 1
Summary of the QCM-D data (f and D at saturation) for HRP immobilization on COOH-enriched and COOH-depleted PEAA surfaces under different
reaction schemes
COOH-enriched surface COOH-depleted surface
+ EDC/NHS No EDC/NHS + EDC/NHS No EDC/NHS
f (Hz) 17 25 22 36
D (×10−6 ) 0.7 2.1 0.8 5.7
D/f (×10−6 ) Hz−1 0.04 0.08 0.04 0.16
Thickness (nm) 3.7 5.1 4.1 6.2
Mass (ng/cm2 ) 426 587 472 713
Viscosity (×10−3 N s/m2 ) 6 3.6 6 2.9
average of 1.0 g/cm3 for water and 1.3 g/cm3 for the pro- immobilization is achieved using direct adsorption on a
tein) [7–9].1 The fitted results (Table 1) show that the HRP COOH-depleted surface. When the QCM-D sensor carrying
films, formed on the EDC/NHS-activated surfaces, have a the HRP layer was exposed to the HRP substrate solu-
higher viscosity. The resulting effective thicknesses, d ≈ tion (4-CN in Tris/diethylene glycol buffer in the absence
4.1 nm (COOH-depleted surface) and d ≈ 3.7 nm (COOH- of H2 O2 ), the change in viscosity of the liquid medium
enriched surface), are similar to a previously reported HRP resulted in a drop in f and an increase in D (data not
thickness of 3.6 ± 1.9 nm (tapping mode AFM data), where shown). Subsequent coincubation of 4-CN and H2 O2 in the
the HRP molecules were trapped on a dithiobis-N -suc- same buffer led to substantial changes in f and D (Fig. 6),
cinimidyl propionate-treated gold surface using a similar which are indicative of the formation and deposition of oxi-
covalent attachment chemistry [28]. dized 4-CN. At saturation, f = 264 Hz, and |D/f | =
0.067 × 10−6 Hz−1 are reached. The initial rate of the sub-
3.2. The catalytic activity of immobilized HRP strate deposition is 2.7 Hz/s.
Qualitatively similar responses are obtained for all condi-
HRP is known to catalyze the oxidation of 4-CN in the
tions of HRP immobilization, confirming that the deposited
presence of H2 O2 leading to the precipitation of the oxide
proteins are enzymatically active. For covalently bound HRP
in diethylene glycol containing solution. In this study we
on EDC/NHS-activated, COOH-enriched PEAA, the final
use this precipitation reaction to evaluate the catalytic ac-
values of f = 342 Hz and D/f = 0.063 × 10−6 Hz−1
tivity of surface-bound HRP by exposing the surface to a
mixture of 4-CN and H2 O2 . As an example, Fig. 6 shows are obtained. The initial rate of the substrate deposition
the corresponding QCM-D measurement, where the HRP is 4.0 Hz/s, being higher than that with the directly ad-
sorbed HRP. The comparison of the initial rates of pre-
cipitate deposition [29] reveals that the covalently bound
1 The effect of the choice of the HRP density on the fitted viscosity and
HRP has a stronger activity. In addition, it is notable that
the mass is negligible. The fitted thickness varies reversely proportional
with changes in the input density [3]. The error is, however, expected to
the |D/f | values are similar for all employed HRP-im-
be smaller than ±15% as the physically reasonable range for the density mobilization protocols. This indicates that the precipitate
(1.0 to 1.3 g/cm3 ) is narrow. layers exhibited identical viscoelastic properties. The differ-
40 X. Su et al. / Journal of Colloid and Interface Science 287 (2005) 35–42
Fig. 6. Enzyme-catalyzed oxidation of 4-CN in the presence of H2 O2 . The enzyme-immobilized COOH-depleted surface (via direct adsorption) is exposed to
the HRP substrate solution (4-CN in Tris buffer with a small amount of diethylene glycol) to create a response baseline. Once this response is stabilized, the
substrate solution is replaced by the same solution containing H2 O2 .
ences in f observed for the precipitation reaction represent enriched surfaces have no further effect on the binding
thus true differences in the amount of adsorbed precipitate. behavior.
Compared to the immobilization on EDC/NHS-activated
surfaces, direct adsorption is observed to proceed more
4. Discussion slowly. It appears that in order for physisorption to take
place, an activation barrier must be overcome. Once ad-
We have investigated the binding behavior of HRP on sorbed, however, the enzyme is immobilized irreversibly. In
PEAA surfaces via different immobilization strategies. As addition, directly adsorbed proteins exhibited an adsorption
evidenced by Table 1 and Figs. 2–5, the HRP adsorption behavior with rather high dissipation, indicating the forma-
on the PEAA surfaces using different immobilization strate- tion of a water-rich and soft film structure. The fitting results
gies exhibits remarkable differences in the binding kinetics (Table 1) reveal that these HRP films are thicker and exhibit
and saturation behavior (f , D, and |D/f | ratio). lower viscosity than the HRP films adsorbed on EDC/NHS-
These differences are believed to be related to the amounts activated surfaces. Unlike the case of EDC/NHS-activated
of bound protein, the binding mechanisms, and the structure surfaces, the surface polarity significantly influences the vis-
of the formed protein films and will be discussed below. coelastic state (|D/f |) and the total adsorbed mass.
According to the general understanding of the behavior It is known that a native protein is usually covered
of a macromolecular film, e.g., adsorbed proteins, as sensed with hydrophilic residues (e.g., carboxyl, amino, and thiol
by QCM-D, water is known to be trapped and to be sensed groups). At a pH value near the pI value, the protein tends to
as additional mass [7–10]. The amount of trapped water can retain its native structure without exposing the hydrophobic
vary significantly depending on the structure of the formed residues [32]. This is exactly the case for the HRP in PBS
protein film [3], and an increased water content is usually buffer (pH 7.2) in this study (pI of the HRP is 7.2 [19]).
correlated with increased flexibility, i.e., a larger |D/f | Thus the direct adsorption of HRP to the PEAA surfaces
ratio [3,30]. Based on this understanding, we interpret our (Figs. 4 and 5) is believed to be dominated by polar–polar
observation of the smaller |D/f | values for the HRP pro- interactions (some electrostatic interactions may also be in-
teins adsorbed to EDC/NHS-activated surfaces as a strong volved as the surface COOH groups are partially ionized at
indication that the adsorbed films are more rigid (contain- pH 7.2 [22]). We suggest that the increased polarity of the
ing less water) than directly adsorbed films. In agreement, COOH-enriched surface leads to more extensive polar–polar
the adsorption on EDC/NHS-activated surfaces results in interactions between the protein and the surface, thus result-
protein films of higher viscosity and lower thickness (Ta- ing in a more efficient HRP adhesion on the surface, being
ble 1). The COOH density did not significantly influence the detected as a layer with a smaller flexibility (|D/f | =
film structure (D/f ) and the adsorbed amount (f ) of 0.08 × 10−6 Hz−1 ) if compared to the COOH-depleted case
these surfaces. Considering the size of the HRP molecule (|D/f | = 0.16 × 10−6 Hz−1 ).
(3.5 × 6.0 × 7.5 nm) [31], it appears that the number of sites From Figs. 2–5, the adsorption-induced dissipation shifts
that is available on surfaces with the lower COOH density seem, in both reaction schemes, to follow the frequency
is sufficient for tight attachment of a complete monolayer shifts (cf., mass uptake), but not exactly. As examples,
of HRP. As such, additional binding sites on the COOH- Figs. 7 and 8 show the plot of D versus f for the immo-
X. Su et al. / Journal of Colloid and Interface Science 287 (2005) 35–42 41
5. Conclusion
[15] S.M. Notley, S. Biggs, V.S.J. Craig, L. Wagberg, Phys. Chem. Chem. [26] L. Fägerstam, Techniques in Protein Chemistry II, Academic Press,
Phys. 6 (2004) 2379. San Diego, 1991, pp. 71–95.
[16] M. Debreczeny, V. Ball, F. Boulmedais, B. Szalontai, J.C. Voegel, P. [27] E. Huang, F.M. Zhou, L. Deng, Langmuir 16 (2000) 3272.
Schaaf, J. Phys. Chem. B 107 (2003) 12734. [28] D. Hobara, Y. Uno, T. Kakiuchi, Bunseki Kagaku 51 (2002) 455.
[17] E.T. Vandenberg, R.S. Brown, U.J. Krull, Immobilized Biosystems: [29] J.M. Abad, F. Pariente, L. Hernandez, H.D. Abruna, E. Lorenzo, Anal.
Theory and Practical Applications, Blackie Academic & Professional, Chem. 70 (1998) 2848.
London, 1994. [30] J. Vörös, Biophys. J. 87 (2004) 553.
[18] E.A. Kulik, K. Kato, M.I. Ikada, Biomaterials 14 (1993) 763. [31] M. Gajhede, D.L. Schuller, A. Henriksen, A.T. Smith, T.L. Poulos,
[19] E.T. Kang, K. Tan, K. Kato, Y. Uyama, Y. Ikada, Macromolecules 29 Nature Struct. Biol. 4 (1997) 1032.
(1996) 6872. [32] J.R. Crowther, ELISA Theory and Practice, Institute for Animal
Health/Humana Press, Working, UK/Totowa, NJ, 1995, p. 63.
[20] Z.F. Li, E.T. Kang, K.G. Neoh, K.L. Tan, Biomaterials 19 (1998) 45.
[33] F. Höök, M. Rodahl, B. Casemo, P. Brzezinski, Proc. Natl. Acad. Sci.
[21] E.P. Ivanova, M. Papiernik, A. Oliveira, I. Sbarski, T. Smekal, P.
USA 95 (1998) 12271.
Grodzinski, D.V. Nicolau, Smart Mater. Struct. 11 (2002) 783.
[34] R.C. Ebersole, M.D. Ward, J. Am. Chem. Soc. 110 (1998) 8623.
[22] P. Zhang, C. Fawcett, J.A. Evans, T. Hurt, K.G. Harvey, R.C. Craven,
[35] X. Su, S.F.Y. Li, Anal. Chim. Acta 429 (2001) 27.
Anal. Biochem. 282 (2000) 218.
[36] X. Su, S.J. O’Shea, Anal. Biochem. 299 (2001) 241.
[23] K.T. Chong, X. Su, E.D.J. Lee, S.J. O’Shea, Langmuir 18 (2002) 9932. [37] S.M. Reddy, J.P. Jones, T.J. Lewis, P.M. Vadgama, Anal. Chim.
[24] M. Rodahl, F. Höök, A. Krozer, P. Brzezinski, B. Kasemo, Rev. Sci. Acta 363 (1998) 203.
Instrum. 66 (1995) 3924. [38] H. Speijer, R.H. Laterveer-Vreeswijk, J.F.C. Glatz, W. Nieuwen-
[25] D.R. Gagnon, T.J. McCarthy, J. Appl. Polym. Sci. 29 (1984) 4335. huizen, W.T. Hermens, Anal. Biochem. 326 (2004) 257.