A new study of the degradation of hyaluronic acid by hyaluronidase using quartz crystal impedance technique

Talanta 53 (2001) 1021 – 1029 www.elsevier.com/locate/talanta

A new study of the degradation of hyaluronic acid by hyaluronidase using quartz crystal impedance technique Deliang He, Anhong Zhou, Wanzhi Wei, Lihua Nie, Shouzhuo Yao * College of Chemistry and Chemical Engineering, Hunan Uni6ersity, Changsha 410082, People’s Republic of China Received 9 February 2000; received in revised form 14 September 2000; accepted 14 September 2000

Abstract A new quartz crystal impedance sensing technique for the assay of hyaluronidase (HAse) activity is presented. It is based on the changes in viscosity and density during the enzymatic hydrolysis of hyaluronic acid (HA) by HAse. The variations of equivalent circuit parameters of piezoelectric quartz crystal (PQC) during the enzymatic degradation are discussed. The motional resistance shift curves indicate that the viscosity of the test solutions decreases during the hydrolysis process. The initial hydrolysis rates of HA are obtained from changes in viscosity and density as a function of incubation time. Kinetic parameters (the Michaelis constant Km and the maximum hydrolysis rate Vmax) of the degradation reaction are estimated by using a linear Lineweaver – Burk plot in this work. The Km was 0.44 90.03 mg·ml − 1 and the Vmax was −(5.2990.36)×10 − 3 kg·m − 2·s − 1/2·min − 1. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Quartz crystal impedance technique; Hyaluronidase; Hyaluronic acid; Kinetic parameter; Enzymatic hydrolysis

1. Introduction Hyaluronidase (HAse, EC 3.2.1.35), is an endoglycosidase which randomly cleaves internal b-Nacetyl-hexosamine [1 – 4] glucosidic linkages in hyaluronic acid (HA) and chondroitin-4- and -6sulfate and their desulfated derivatives to liberate oligosaccharides containing equimolar glucuronic acid residues at their reducing terminals and Nacetylglucosamine (NAG) at the non-reducing ends [1]. The enzyme has been demonstrated in a wide * Corresponding author. Tel.: + 86-731-8822286; fax: +86731-8824525. E-mail address: [email protected] (S. Yao).

range of mammalian tissues, such as synovial fluid, serum, alveolar macrophages, brain, skin, kidney, liver, spleen, and lung [2]. These tissue-associated HAses function in such diverse biological processes as wound healing, embryological development, angiogenesis, tumorgenesis, and in the acceleration of absorption and diffusion of venoms [3,4]. Northup et al. postulated that HAse may play a role in cancer invasiveness [5]. Stern et al. described elevated levels of HAse in urine of children with Wilm’s tumor and suggested that HAse may be used as a tumor marker [6]. Clinically, the enzyme is used in chemotherapy to enhance the antineoplastic activity of cytostatics and in the treatment of acute myocardial infarction [7,8].

0039-9140/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 0 0 ) 0 0 5 9 9 - 3

1022

D. He et al. / Talanta 53 (2001) 1021–1029

Hyaluronic acid, a negatively charged highmolecular-weight acid mucopolysaccharide, is member of the general class of glycosaminoglycans with the lowest charge density. It is a linear polysaccharide built from repeated disaccharide units with a …[D-glucuronic acid (1-b-3) N-acetylstructure [9]. glucosammine (1-b-4)]n … Hyaluronic acid is widely distributed in soft connective tissues. Its biological functions include the maintenance of such mechanical properties as compressive stiffness and swelling in connective tissues, control of tissue hydration, a backbone for proteoglycan aggregation, providing lubricating properties in synovial fluid and heart valves, and binding to cell surfaces in cell – cell adhesion. The HA substrate for HAse is becoming increasingly prominent in biology [10]. Assays for HAse activity using HA substrate can be divided into three groups as follows. (1) Change in viscosity; rheological analysis has proven to be a useful tool for the assay of HAse [11]. Indeed, viscosity is drastically decreased when HAse is used during the course of its endoglycosidase cleavage reaction of high molecular weight HA. Wyer measured the decrease in viscosity during the enzymatic reaction of HAse [12]. Cobbin and Dicker used a pre-cooled Poiseuille-type viscometer of a capacity of 0.3 ml to study the activity of serum HAse [13]. Salegui et al. described the activity determination of HAse by measuring changes in viscosity of the HA substrate [14] and compared the activity of serum and testicular HAse at 30°C [15]. (2)

Fig. 1. Schematic representation and equivalent circuit of the quartz crystal impedance system. Cs, static capacitance; Lm, motional inductance; Cm, motional capacity; Rm, motional resistance.

Loss of turbidity or decrease in HA; the HAse activity was studied by the loss of ability to form insoluble [16] or colored [17] salts. Quantitative assays of the activity of HAse are performed by using time-dependent changes in intensity of light scattering [18]. (3) Generation of N-acetylamino groups; Gacesa et al. studied HAse activity by spectrophotometry of the liberated hexosamine endgroups [19]. Bonner and Cantey originally developed the assay for serum HAse activity based on the liberation of saccharides with NAG endgroups from HA [20]. The NAG is quantitated by heating with alkaline tetraborate to form an intermediate that reacts with p-dimethylaminobenzaldehyde in acidic medium to form a colored product. C.R. Wilkinson et al. optimized this assay to study the HAse activity of 70 normal sera [21]. However, there are many other assays for HAse that have been reported in the literature. For example, Tung et al. use a microplate assay for HAse and HAse inhibitors at 37°C [22]. Frost and Stern proposed a sensitive, rapid micro-titerbased assay for HAse activity that does not require highly specialized biological reagents [23]. Although many methods are used to assay HAse activity, these assays are either insensitive or lack specificity or generally are tedious and time-consuming and are reliant on unique reagents not generally available. In addition to insensitivity and inconvenience, these single-point reductimetric assays depend on the choice of an appropriate standard and the structure of the reducing sugar. It is thereby very important to develop quantitative and sensitive high-throughput HAse assay methods using generally available reagents. The quartz crystal impedance technique (QCIT), a network analysis method [24], provides much information assessing changes in some physical and/or chemical properties of the tested system [25,26]. Generally, an electrical equivalent circuit as shown in Fig. 1 is used to analyze quartz crystal resonator, in which motional resistance Rm corresponds to energy losses in the environment, motional inductance Lm corresponds positively to crystal mass, motional capacity Cm corresponds to flexibility, and Cs is the static capacitance which is the capacitance deter-

D. He et al. / Talanta 53 (2001) 1021–1029

mined at a frequency far from the crystal resonance [27,28]. The QCIT is a powerful tool for the qualitative or quantitative process-investigation. The QCIT can be used to study the enzymatic degradation of HA by HAse as a sensitive viscosity and density detector, like the technique presented here. The response of quartz crystal impedance system facilitates monitoring of the enzymatic degradation process as well as estimations of the kinetic parameters. Some methods have been used to measure the kinetic parameters of HAse when HA acts as the substrate. Reed et al. determined the Henri–Michaelis – Menten coefficients by monitoring the time-dependence of the scattered light from solutions [18], Km =0.18 90.03 mg·ml − 1 and Vmax/ET =9.6 mM bonds cleaved per h U − 1·ml − 1. They also obtained the kinetic coefficients by reductimetric assay of HAse activity as described by Keleti and Lederer [29], Km is 0.46 9 0.05 mg·ml − 1 and Vmax/ET is 4.3 90.2 mM bonds cleaved per h U − 1·ml − 1. A gel permeation chromatography (GPC) system for the analysis of HA was set up and used for the kinetic assay of HAse [30], Km is 0.44 9 0.04 mg·ml − 1 and Vmax is 1.8 × 10 − 12 90.3×10 − 12 Katals·ml − 1. Friszer measured them in 0.15 M NaCl, 0.05 M potassium phosphate, 0.25 M acetate buffer (pH 6.0, 37°C), Km is 0.09 mg·ml − 1 [31]. Houck and Pearce obtained the kinetic parameters in 0.1 M acetate, 0.15 M NaCl buffer (pH 5.0, 27°C) with a Km of 0.3 mg·ml − 1 [32]. The result from Jones is that Km is 0.6 mg·ml − 1 (pH 5.0, 37°C, 20 mM MES, 0.1 M NaCl) [33]. The application of QCIT in studying the enzymatic hydrolysis of HA by HAse has not been reported previously. HAse from sheep testes and HA from human umbilical cord were used as enzyme source and substrate source, respectively, in this paper. The variations of Rm were recorded during the enzymatic hydrolysis using optimized conditions. The initial degradation rate of HA could deviate from the rate of change rate of DRm. The Michaelis constant Km and the maximum hydrolysis rate Vmax of HAse were obtained by fitting the Lineweaver– Burk equation [34]. The results gained in this work confirm values reported previously.

1023

2. Experimental

2.1. Apparatus As shown in Fig. 1, the quartz crystal impedance system used included a HP 4395A Network/Spectrum/Impedance analyzer, an IBM P233MMX personal computer with Intel cards for data sharing. Impedance and admittance of the piezoelectric quartz crystal (PQC) resonator were measured on the HP 4395A equipped with a HP 43961A impedance test adapter and a HP 16092A test fixture. A user program written in Visual Basic (VB) 6.0 was used to control the HP 4395A and to acquire conductance (G) and susceptance (B) data synchronously via a HP 82341C high-performance HP–IB interface card for Windows 3.1/NT/95. On-line impedance analysis was achieved by a non-linear least square-fitting program for the simultaneous fits of G and B data. And equivalent circuit parameters were acquired during experiments at a frequency of ca. 0.5 Hz. An AT-cut 9 MHz PQC of ca. 12.5 mm diameter with gold electrode (6 mm diameter) on both sides were used. The crystal was sealed to the test tube with one side facing the tested solution. Another gold electrode on the opposite face of the PQC faced air and served as the connection electrode to the non-ground terminal of the HP16092A test fixture. The test solutions were stirred with a magnetic stirrer (Shanghai Electrocommunication Instrumentation Factory). A biochemical incubator was used to control the reaction temperature.

2.2. Reagents and chemicals HAse (from sheep testes, E.C.3.2.1.35, enzyme activity is 845 U mg − 1 solid) was purchased from Sigma Chemical Corporation. One unit is defined as the change in absorbance at 600 nm (change in turbidity) of a USP reference standard HAse which is assayed concurrently with each lot of product. HA (from human umbilical cord, sodium salt) also purchased from Sigma Chemical Corporation. Other chemicals were of analytical grade. Double distilled water was used throughout.

D. He et al. / Talanta 53 (2001) 1021–1029

1024

Sodium phosphate solutions were used as buffers at pH 5.5, 6.0, 6.4, 7.0 and 7.5. 5 mg·ml − 1 HAse solutions in sodium phosphate buffer were prepared each day. Hyaluronic acid was also dissolved in sodium phosphate buffers. The concentrations of HA stock solutions employed ranged from 0.1 to 5 mg·ml − 1.

ters during fitting [35]. The frequency giving maximum G, Gmax, was automatically found by the program and used as the initial value of fs, and 1/Gmax was used as the initial value of Rm during fitting iterations. The fitting iterations could be complete when the sum of the residual square, q, and the relative sum of the residual square, qr, as given in (Eq. (2)), had become minimum.

2.3. Quartz crystal impedance measuring of enzymatic hydrolysis of HA by HAse

N

qr = A 2 ml aliquot of HA solution was removed from the stock and put into the detection cell. Then the PQC was immersed into the solution, and the quartz crystal impedance measurement was carried out. After several minutes, 0.1 of 5 mg·ml − 1 HAse solution was added to the detection cell, and the equivalent circuit parameters and frequency shift of PQC were recorded during the enzymatic hydrolysis process. All experiments were carried out in the biochemical incubator and the reaction temperature controlled at 379 0.1°C. The range of concentration of HA solutions was from 0.1191 to 3.8096 mg·ml − 1. Changes of Rm in these HA solutions were measured by QCIT during the degradation processes, respectively. A series of sucrose solutions with different weight percent were prepared to obtain different viscosity and density. The Rm of PQC sensor used in this work was determined by QCIT, respectively. Experiments were performed at room temperature (2090.1°C).

2.4. Non-linear fitting of admittance spectra Admittance (Y), conductance (G), and susceptance (B) for the modified Butterworth–Van Dyke equivalent electrical circuit as shown in Fig. 1 can be expressed as Y= G+ jB=





Rm U +j vCs − 2 2 R +U R m +U 2 2 m

N

%(Gfit − Gexp)2 + %(Bfit − Bexp)2 1

1 N

1

1

% G 2exp + % B 2exp q

=N N % G 2exp + % B 2exp 1

(2)

1

where the subscripts, fit and exp, denote fitted and experimental results and N is the number of frequency points in experiments.

3. Results and discussion

3.1. Study of the enzymatic degradation of HA by QCIT The equivalent circuit of quartz crystal sensor has been derived by Cady and Bottom [36]. The four circuit elements with parameters Rm, Lm, Cm, and Cs correspond to one mode of vibration of the quartz crystal. The spectra of G and B were measured while the frequency span covered the complete resonant region. Then the equivalent circuit parameters were calculated by non-linear regression analysis. The circuit elements are related to the properties of the quartz crystal as follows [24]: Rm =

e 3r 8Ao 2

Lm =

e 3r 8Ao 2

Cm =

8Ao 2 p 2ec

(1)

where U= vLm −1/(vCm). We measured G and B synchronously in experiments; therefore, G and B data acquired in one frequency scan were fitted simultaneously by using the above-mentioned non-linear least-squares fitting program. Rm, Cs, 1/Cm, and fs were selected as estimation parame-

N

Cs =

ko0A e

(3)

D. He et al. / Talanta 53 (2001) 1021–1029

1025

is needed to define the actual process. As DF could be effected by many factors during the quartz crystal impedance measurement, such as mass change, surface roughness, reactant depletion within the diffusion layer, the viscosity and density of the solution, the acoustoelectric effect, the shape of the curves of DF and Rm are not the same. Changes of CS during the hydrolysis process are shown in curve c of Fig. 2. The value of Cs increased while the enzyme solution was added, while little change in Cs was observed during the degradation process. Fig. 2. Changes of the parameters of PQC during the hydrolysis of HA. The HA used is 0.9524 mg·ml − 1 HA solution and the HAse used is about 201.19 U·ml − 1. The perpendicular arrows indicate the addition of HAse. (a) Curve of Rm; (b) curve of and DF; (c) curve of Cs.

in which e and A are the thickness and area of the quartz crystal, o the piezoelectric stress constant, r the dissipation coefficient, r the density, c the elasticity constant, k the dielectric constant of quartz and o0 is the permittivity of free space. The QCIT was utilized to monitor the hydrolysis of HA by catalysis of HAse. The changes of Rm during the degradation process are shown in curve a of Fig. 2. As pointed out by Muramatsu et al. [37], for a PQC oscillator in contact with a liquid, the motional resistance is proportional to (rLhL)1/2 where rL and hL are the liquid density and viscosity. So with the change in Rm, we can assess the change in viscosity and density of liquid. After the addition of HAse, the viscosity and density of the test solution decreases obviously because of the degradation of HA by HAse. The frequency shift (DF) during the degradation of HA by HAse is shown in curve b of Fig. 2. It can be seen that DF decreases within 10 min and increases during 10 – 30 min, then decreases in the following degradation process. The decrease in viscosity and density will result in increase in frequency. The decrease in frequency may also result from the adsorption of the hydrolytic product of HA by HAse to the electrode of PQC, since an increase of the mass of the sensor resulted in decrease of the frequency. Further study

3.2. The relationship between Rm and the 6iscosity and density of solution The motional resistance Rm can be expressed as follows [37]: Rm = RQ + RV

(4)

where RQ is the unperturbed mounted crystal resistance, RV reflects the contribution due to viscoelastic losses by liquid. RV =

(pfsrLhL)1/2A K2

(5)

In Eq. (5) fs is the series resonant frequency, rL the density of liquid, hL the viscosity of liquid, A the area of Au-electrode of PQC and K is the electromechanical coupling factor. It can be found that Rv is proportional to (hLrL)1/2 of the tested solutions. A series of sucrose solutions with different weight percent were prepared to obtain different viscosity and density. The Rm of PQC sensor used in the following process was determined by QCIT, respectively. Experiments were performed at room temperature (20°C) and the values of density and viscosity for these solutions were obtained from standard table [38]. Fig. 3 shows the dependence of Rm on the viscosity and density of solution ((hLrL)1/2). It can be found that Rm is linear to (hLrL)1/2 within the whole tested concentration range. The relationship between them can be obtained by linear fitting and it is:

1026

Rm = 45.13+157.67(hLrL)1/2

D. He et al. / Talanta 53 (2001) 1021–1029

(r = 0.991, n= 9) (6)

where Rm is in V, h in kg·m − 1·s − 1, r in kg·m − 3.

3.3. Change in 6iscosity and density under different pH conditions A series of HA solutions with different pH were hydrolyzed by HAse. The change of Rm was recorded by QCIT during the degradation process. The term (hLrL)1/2 was calculated from Rm according to Eq. (6). Changes in viscosity and density (D[(hLrL)1/2)] are obtained as follows: D[(hLrL)1/2]= [(hLrL)1/2]T − [(hLrL)1/2]1

Fig. 3. Responses of a PQC in sucrose aqueous solutions. Lines show the best fits of the corresponding data.

(7)

where [(hLrL)1/2]T is the viscosity and density when HA solution is degraded by HAse; [(hLrL)1/ 2]1 is the viscosity and density when HAse was added into HA solution. Typical response curves of D[(hLrL)1/2] and incubation time are shown in Fig. 4. It can be seen that (hLrL)1/2 decreases during this process, but D[(hLrL)1/2] decreases more quickly under pH 6.4 than other pH values. It can be found that HAse has maximal activity to hydrolyze HA when pH is 6.4. It is identical to the result determined by Cobbin and Dicker [13]. So the following experiments are carried out under pH 6.4 condition.

3.4. Change in 6iscosity and density under different HA concentrations Fig. 4. The curves of D[(hLrL)1/2] with incubation time under different pH conditions as follows, (1) 5.5 (2) 6.0 (3) 6.4 (4) 7.0 (5) 7.5.

Fig. 5. Typical response curves of D[(hLrL)1/2] vs. incubation time at different concentrations (mg·ml − 1) of HA. The concentrations of HA are as follows (37°C, pH 6.4), (a) 0.1191 (b) 0.2381 (c) 0.3572 (d) 0.4762 (e) 0.7143 (f) 0.9524 (g) 1.4286 (h) 1.9048 (i) 2.8572 (j) 3.8096.

A series of HA solutions with different concentrations were hydrolyzed by HAse. The change of Rm was recorded by QCIT at each time. The term (hLrL)1/2 was calculated from Rm according to Eq. (6). Changes in viscosity and density (D[(hLrL)1/2]) are obtained as mentioned above. Typical response curves of D[(hLrL)1/2] and incubation time under different HA concentrations are shown in Fig. 5. It can be seen that (hLrL)1/2 decreased more with the concentration of HA. This is because the hydrolysis rates increase with the concentration of HA and more obvious changes in viscosity and density arise during the degradation of HA by HAse.

3.5. Measurement of kinetic parameters of HAse A series of HA solutions with different concentration were hydrolyzed by HAse. During the

D. He et al. / Talanta 53 (2001) 1021–1029

1027

initial hydrolysis rates by the direct linear method of Lineweaver–Burk plot [34,39]: 1 1 1 Km × + = V Vmax [S] Vmax

Fig. 6. The dependence of the initial degradation rates on HA concentrations (37°C, pH 6.4).

(8)

where V and Vmax are the initial hydrolysis rate and maximum hydrolysis rate in kg·m − 2·s − 1/2· min − 1, Km is the Michaelis constant in mg·ml − 1, [S] is the HA concentration in mg·ml − 1. Thus, the curve of the reciprocal of initial rate and the reciprocal of HA concentration for experiments at a fixed enzyme concentration should be a straight line. Furthermore, the intercept with the ordinate will be 1/Vmax and the slope is Km/Vmax, from which Vmax and Km can be calculated. In this work, the concentration range of HA was from 0.1191 to 3.8096 mg·ml − 1, the temperature was controlled at 37°C. The dependence of 1/V and 1/[S] is shown in Fig. 7, it can be seen that 1/V is linear to 1/[S]. By fitting Eq. (8), the kinetic parameters are estimated, Km is 0.44 90.03 mg·ml − 1 and Vmax is −(5.299 0.36)× 10 − 3 kg·m − 2·s − 1/2·min − 1 (n=9, r= 0.998).

Fig. 7. The relationship between 1/[S] and 1/V (37°C, pH 6.4).

degradation process D[(hLrL)1/2] was measured by QCIT. The initial degradation rate (V) of enzymatic hydrolysis reaction of HA solutions was obtained by linear fitting from data of D[(hLrL)1/2] within 3 min. Fig. 6 shows the dependence of the initial hydrolysis rates on different concentrations of HA solutions. It can be seen that the curves of V and the concentration of HA [S] are basically coincidental to that of typical enzymatic reaction. Kinetic parameters of the enzymatic reaction can be estimated by using QCIT from the data of

4. Conclusion A new quartz crystal impedance sensing technique for the assay of HAse activity, which is based on the changes in viscosity and density during the enzymatic degradation of HA by HAse, is established here. The advantages of this method compared with measurement of reduction of viscosity and change in turbidity and production of N-acetylamino groups are shown in Table 1. The variations of equivalent circuit parameters of PQC were obtained during the enzymatic

Table 1 Comparison of QCIT with other methods applied in HAse assay Methods

Other reagents

Measuring procedures

Measuring objects

References

Reduction of viscosity Change in turbidity Production of NAG QCIT

Not necessary Requisite Requisite Not necessary

Troublesome and intermittent Troublesome and intermittent Troublesome and intermittent Easy and continuous

Viscosity of solution Absorbance of solution Production of NAG Motional resistance of PQC

[12–15] [16–18] [19–21] This paper

1028

D. He et al. / Talanta 53 (2001) 1021–1029

Table 2 Comparison of Km of HAse obtained by QCIT with other values reported in literatures Michaelis constant (Km, mg·ml−1)

Measuring conditions

Methods

References

0.18 9 0.03

Buffer, 0.1 M phosphate, 0.15 M NaCl, pH 5.3 (at 25°C)

[18]

0.46 90.05 0.449 0.04

Buffer, 0.1 M phosphate, 0.15 M NaCl, pH 5.3 (at 25°C) Buffer, 140 mM NaCl, 16 mM NaH2PO4, 7 mM Na2HPO4, pH 6.4 (at 37°C) Buffer, 15 M NaCl, 0.05 M potassium phosphate, 0.25 M acetate buffer, pH 6.0 (at 37°C) Buffer, 0.1 acetate, 0.15 M NaCl, pH 5.0 (at 27°C) Buffer, 20 mM MES, 0.1 M NaCl, pH 5.0 (at 37°C) Buffer, 0.15 M NaCl, 0.016 M phosphate, pH 6.4 (at 37°C)

The time-dependence of the scattered light Reductimetric assay A gel permeation chromatography Change in viscosity

[31]

Change in turbidity Production of NAG QCIT

[32] [33] This paper

0.09 0.3 0.6 0.44 90.03

degradation of HA. The curves of Rm and time indicated that the viscosity of HA solution decreased during the hydrolysis process. The initial hydrolysis rate of HA by HAse is obtained from (D[(hLrL)1/2]) within a short reaction period. Kinetic parameters (the Michaelis constant Km and the maximum rate Vmax) of the process were estimated by using a linear Lineweaver–Burk plot. The experimental results agreed well with that in the literature. Comparison of Km obtained by QCIT with that reported by other researchers is shown in Table 2. It can be anticipated that QCIT can be used increasingly to study the enzymatic degradation process by HAse, as well as to a wide variety of other depolymerization reaction.

References [1] R. Ruyssen, A. Lauwers, Pharmaceutical Enzymes, E Story Scientia, Ghent, 1978, p. 203. [2] S.S. Chen, D.S. Hsu, P. Hoffman, Clin. Chim. Acta 95 (1979) 277. [3] E. Belsky, B.P. Toole, Cell Differ. 12 (1983) 61. [4] W.M. Kulyk, R.A. Kosher, Dev. Biol. 120 (1987) 535. [5] S.N. Northup, R.O. Stasiw, H.D. Brown, Clin. Biochem. 6 (1973) 220. [6] M. Stern, M.T. Longaker, N.S. Adzick, et al., J. Natl. Cancer Inst. 83 (1991) 1569. [7] K. Hobarth, U. Maier, M. Marberger, Eur. Urol. 21 (1992) 206. [8] I. Repa, J.D. Garnic, N.K. Hollenberg, J. Cardiovasc. Pharmacol. 16 (1990) 286.

[18] [30]

[9] T.C. Laurent, J.R.E. Fraser, FASEB J. 6 (1992) 2397. [10] M.F. Meyer, G. Kreil, Proc. Natl. Acad. Sci. USA 93 (1996) 4543. [11] J. Demeester, O.M. Award, M. Bracke, A. Lauwers, Fresenius Z. Anal. Chem. 330 (1988) 366. [12] G.I. Wyer, C.W. Emmens, Biochem. J. 41 (1947) 29. [13] L.B. Cobbin, S.E. Dicker, J. Physiol. 163 (1962) 168. [14] M.D. Salegui, W.M. Pigman, Arch. Biochem. Biophys. 120 (1967) 60. [15] M.D. Salegui, H. Plonska, W. Pigman, Arch. Biochem. Biophys. 121 (1967) 548. [16] N. Ferrante, J. Biol. Chem. 220 (1956) 303. [17] R.H. Pryce-Jones, N.A. Lannigan, J. Pharm. Pharmacol. 31 (1979) 92P. [18] W.F. Reed, C.E. Reed, L.D. Byers, Biopolymers 30 (1990) 1073. [19] P. Gacesa, M.J. Savitsky, K.S. Dodgson, A.H. Olavesen, Anal. Biochem. 118 (1981) 76. [20] W.M. Bonner, E.Y. Cantey, Clin. Chim. Acta 13 (1966) 746. [21] C.R. Wilkinson, L.M. Bower, C. Warren, J. Pharm. Biomed. Anal. 14 (1996) 707. [22] J. Tung, G.E. Mark, G.F. Hollis, Anal. Biochem. 223 (1994) 149. [23] G.I. Frost, R. Stern, Anal. Biochem. 251 (1997) 263. [24] A.L. Kipling, M. Thompson, Anal. Chem. 62 (1990) 1514. [25] S.J. Martin, V.E. Granstaff, G.C. Frye, Anal. Chem. 63 (1991) 2272. [26] T.A. Zhou, L.H. Nie, S.Z. Yao, J. Electroanal. Chem. 293 (1990) 1. [27] J.C. Ernesto, C. Danilowicz, R. Etchenique, J. Chem. Soc. Faraday Trans. 91 (22) (1995) 4083. [28] Q.J. Xie, J. Wang, Z.N. Xu, H.W. Liu, Y.Y. Zhang, S.Z. Yao, Chem. J. Chin. Univ. 20 (1999) 549. [29] G. Keleti, W.H. Lederer, Handbook of Micromethods for the Biological Sciences, Van Nostrand Reinhold, New York, 1974, pp. 45 – 46.

D. He et al. / Talanta 53 (2001) 1021–1029 [30] K.P. Vercruysse, A.R. Lauwers, J.M. Demeester, J. Chromatogr. B 656 (1994) 179. [31] B. Friszer, Bull. Soc. Chim. Biol. 46 (1964) 1211. [32] J.C. Houck, R.H. Pearce, Biochim. Biophys. Acta 25 (1957) 555. [33] C.P. Jones, R.T. Sawyer, Thromb. Rev. 55 (1989) 791. [34] H. Lineweaver, D.J. Burk, J. Am. Chem. Soc. 56 (1934) 658. [35] Q.J. Xie, J. Wang, et al., Anal. Chem. 71 (1999) 4649.

.

1029

[36] M. Thompson, A.L. Kipling, W.C. Duncan-Hewitt, L.V. Rajakovic, B.A. Cavic-Vlasak, Analyst 116 (1991) 881. [37] H. Muramatsu, E. Tamiya, I. Karube, Anal. Chem. 60 (1988) 2142. [38] R.C. Weast (Ed.), CRC Handbook of Chemistry and Physics, 66th ed., CRC Press, Boca Raton, FL, 1985, pp. PD232 – PD262. [39] D.L. He, L.L. Bao, Y.M. Long, W.Z. Wei, S.Z. Yao, Talanta 50 (2000) 1267.