Biosensors& Bioeiectronics6 (1991) 353-358
A quartz crystal viscosity sensor for monitoring coagulation reaction and its application to a multichannel coagulation detector H. Muramatsu, Seiko Instruments
K. Kimura, T. Ataka
Inc.. Applied Research Department,
563 Takatsukashinden,
Matsudo-shi,
Chiba 271, Japan
R. Homma, Y. Miura National
Institute of Health, Department
of Blood Products, 2-10-35 Kamiosaki, Japan
Shinagawa-ku.
Tokyo 141,
& I. Karube University
of Tokyo, Research Center for Advanced Science and Technology, Tokyo 153, Japan
(Received 4 May 1990: revised version received 10 September
4-6-l Komaba. Meguro-ku,
1990; accepted 24 September
1990)
Abstract: A quartz crystal viscosity sensor was applied to a coagulation
reaction monitoring system. The system consists of 16 oscillating circuits. a channel selector, a frequency counter, a temperature controller and a microcomputer. The system is named the Quartz Chemical Analyzer (QCA). AT-cut quartz crystals (9 MHz) were used as viscosity detectors and were attached to a cell in order to expose only one side of the quartz plate. The system was applied to the detection of the blood coagulation factors VIII (F VIII) and IX (F IX). The activity of these factors was assayed by a single-stage method. A linear relationship was obtained in a double-logarithmic diagram of concentration versus coagulation time with respect to F VIII and F IX in the range @OS-O*4 unit cm-) and @025-O-2 unit cmm3. respectively.
Keywords: quartz crystal, viscosity sensor, blood coagulation
INTRODUCTION In
chemical
analysis
a quartz
crystal
micro-
balance (Sauerbrey, 1959) has been applied to the sensing of gases (King, 1964, Hlavay & Guilbault.
factors.
1977). ions (Nomura & Tijima, 1981). biopolymers (Shons et al., 1972; Thompson ec al., 1986; Muramatsu efal., 1987) and microbes (Muramatsu et al., 1986; Muramatsu et al., 1989a). The property of the quartz crystal in contact with liquid has
353 Biosensors CcBioekmonics 0956~5663/91/$03.50@1991 Elsevier Science Publishers Ltd. England. Printed in Great Britain
354
H. Muramatsu et al.
also been studied (Kanazawa & Gordon, 1985: Muramatsu et al., 1988a. b; Muramatsu et al., 19893). It has been shown that the resonant frequency of the quartz crystal changes depending upon the viscosity and density of the liquid. Kanazawa and Gordon derived the following equation (Kanazawa & Gordon, 1985): AF = -Fs”*(tl~~_lw~q)“*
(1)
where F, is the resonant frequency of the quartz crystal, AF is the change in resonant frequency, pi is the density of the liquid, n is the viscosity of the liquid, pQ is the density of the quartz and p is the elastic modulus of the quartz. The authors have also derived an equation for the resonant resistance, which is one of the electrical equivalent circuit parameters of the quartz crystal, as follows (Muramatsu et al., 1988a): R = (2nFspLq)“*A/k2
(2)
where k is the electro-mechanical coupling factor andA is the electrode area of the quartz crystal. A quartz crystal has been used as a viscosity monitoring sensor. The authors have applied this sensor to the determination of (a) endotoxin concentrations using the gelation reaction of Limulus amebocyte lysate (Muramatsu et al.. 1988b) and (b) fibrinogen concentrations using the coagulation reaction of thrombin (Muramatsu et al., 19896). The mechanism of the coagulation process of blood has been elucidated (Davie & Fujikawa, 1975). Blood coagulation factors VIII and IX are known to be factors that are deficient in the blood of patients with hemophilia A and hemophilia B, respectively. The analysis of the blood coagulation factors is widely performed at clinical inspection. The estimation of the activity coagulation factor preparations is also important (Miura, 1984; Miura et al., 1989). The single-stage method has been used for the determination of F VIII and F IX activity (Hardisty et al., 1962). In this method, coagulation time was measured for the activity determination by mixing the deficient plasma of F VIII (or F IX), the activation factors, Ca*+ ions and the sample solution. In the present work, the quartz crystal viscosity sensor was applied to the monitoring of and the 16channel coagulation reaction, coagulation detector was made up using the quartz crystals. The system was applied to the detection of F VIII and F IX.
Fig. 1. Schematic diagram of the measuring system: A, quartz crystal: B. oscillating circuit: C. heater: D. signal selector: E. frequency counter: F, microcomputer.
EXPERIMENTAL Apparatus and materials
The measuring system (Quartz Chemical Analyzer: QCA) consists of 16 oscillating circuits, a channel selector, a frequency counter and a temperature controller (Fig. 1). The oscillating circuits and the channel selector were fabricated using TI’L-IC. The temperature controller was constructed from an aluminum block, a rubber heater and a temperature controlling unit. The frequency counter was fabricated using a counting IC (Intersil, ICM7226B). The system was controlled by a microcomputer (SeikoEpson, PC-286). The measuring cells consisted of 9 MHz AT-cut quartz crystals. The quartz crystals were positioned at the bottom of the cells in order to expose only one side of them. The 16 quartz crystals in QCA are used for multiple measurement at the same time. QCA is easily composed as a multichannel system. An activated partial thromboplastin time reagent, deficient plasma and Owren’s Verona1 buffer solution (pH 7.35) were obtained from Kokusai-shiyaku. CaC12 (0.025 M) was obtained from Fuji-Rebio. Standard blood coagulation factors from the National Institute of Health were used for a sample solution of F VIII and F IX. The sample solution was diluted with Owren’s veronal buffer solution for preparing the concentration. A light scattering detector (Kokusai-shiyaku, Coagurostat) was used for light scattering measurement. Procedure
F VIII (or F IX) activity was assayed by the singlestage method as follows (Hardisty et al., 1962). The sample solution of F VIII (or F IX) and an
355
Quartz crystal viscosity sensor for monitoring coagulation reaction
activated partial thromboplastin time reagent were added to pre-incubated plasma deficient in F VIII (or F IX) for 1 min at 37°C. After incubating the mixture for 2 min, O-025 M CaClz was added. Immediately, the mixture was poured onto the quartz crystals. The resonant frequency was monitored at intervals of a few seconds. The gate time for resonant frequency measurement was 1 s.
RESULTS AND DISCUSSION Resonant frequency response to the coagulation reaction
Figure 2 shows the resonant frequency response to the coagulation reaction with F VIII sample solution of 0.0543 unit cmV3. In Fig. 2, the resonant frequency changed as a result of viscosity change caused by the coagulation reaction. It took 5-10 s to start changing the frequency. Because the blood resonant coagulation system consists of cascade enzyme reaction, it appeared to take 5-10 s to begin the last stage of the coagulation reaction. The frequency change decreased to zero when coagulation was completed. Two methods of calculating the coagulation time were considered in this work. In the first method, the time to reach a certain ratio of total frequency change (ratio coagulation time) was used. In the second method, the time to attain a
Fig. 2. Resonant frequency response to the coagulation reaction with F VIIIsamplesolution of 0.0543 unit cm-! A is the ratio for calculation of coagulation time, and B show-s the ratio coagulation time. C i.sthe thresholdjkquency for calculation of coagulation time. and D sh0w.s the threshold coagulation time.
threshold certain frequency coagulation time) was used. This shown in Fig. 2.
(threshold scheme is
Study of the coagulation ratio and the threshold frequency
Table 1 shows the ratio coagulation time for a series of F VIII concentrations in duplicate measurements of the resonant frequency response. The regression slope and coefftcient between the coagulation time and concentration of F VIII are also shown in Table 1. The results were dealt with by a double-logarithmic scale, because it is usually used for a wide-range measurement and it is known that the relation for the light scattering method shows a linear relation in a double-logarithmic diagram. The calculation was based on the equation of log T = a log C + b, where T is the coagulation time, C is the concentration of coagulation factor, and a and b are regression constants. In Table 1 the regression slope shows close agreement with the coagulation ratio in the range of 20-70%. The correlation curves of 60% and 70% showed a good regression coefticient. The slope of the correlation curve of 60% is steeper than that of 70%. This result corresponds to results from other experiments which were not shown in this paper. A value of 60% was adopted for the coagulation ratio. Table 2 shows the threshold coagulation time for a series of F VIII concentrations in duplicate measurement ofthe resonant frequency response. The regression slope and coefficient between the coagulation time and concentration of F VIII are also shown in Table 2, as for Table 1. In Table 2. the regression slope is seen to be in close agreement with the threshold frequency in the range of 200-800 Hz. The correlation curves of 200 Hz and 300 Hz showed good regression coefficients. The slope of the correlation curve for 300 Hz is seen to be steeper than that for 200 Hz. This result corresponds with results from other experiments. A threshold frequency of 300 Hz was adopted. Correlation between coagulation time and the concentration of coagulation factors
Figure 3(a) shows the correlation between the ratio of coagulation time and the concentration of F VIII. A linear relationship in the double
356
H. Muramatsu et
Ratio coagulation
TABLE 1 time (min) for a series of F VIII concentrations in duplicate measurement of resonant frequency response. The regression constants and coefficients are also shown
Concentration (unit cm-3
Ratio in totalfrequency 90%
80%
70%
60%
50%
40%
30%
20%
0.76 0.83 0.93 1.30 1.18 I.11 1.37 1.24
0.69 0.72 0.87 11.1 1.11 I.04 1.25 1.17
064 a68 0.83 a85 I.06 a99 1.20 1.11
0.61 0.65 0.80 0.76 1.01 0.95 1.14 1.05
0.58 a61 0.76 a73 0.97 0.91 1.08 1xIo
0.56 0.58 0.73 0.70 0.92 0.88 1.02 O.%
0.53 0.55 0.70 0.67 0.88 a84 0.96 0.91
0.51 0.52 0.67 0.63 0.83 0.81 0.91 0.86
-0.153 -0.223
-0.215 -0.248
-0.263 -a268
-0.291 -0.272
-0.312 -0.273
-0.351 -0270
-0.371 -0.267
Regression coetlicient
0945
a972
a980
log(time) = a log(concentration)
+ b.
0.434 a434 0.217 a217 0.109 a109 0.054 a054 :
Threshold
al.
0.980
0.978
-0.27 1 -0.33 0.979
0.975
TABLE 2 coagulation time (min) for a series of F VIII concentrations in duplicate measurement frequency response. The regression constants and coefficients are also shown
Concentration (unit cm --),
0.434 a434 0.217 a217 0.109 al09 0.054 a054 a b
Regression coefficient log(time) = a log(concentration)
0.973
of resonant
Threshold frequency 100 Hz
200 Hz
300 Hz
400 Hz
500 Hz
046
0.47 0.61 0.59 0.75 0.73 0.59 0.78
0.50 0.51 0.68 0.63 0.82 0.78 090 0.88
0.52 a54 0.72 a66 0.87 0.80 0.96 a93
0.54 0.57 0.75 0.69 0.92 0.82 1.01 O-98
0.56 a60 0.79 a72 0.96 a85 1.06 1.04
-0.194 -0.37 1 0.861
-0.267 -0.375 0.976
-0.276 -a359 0.976
-0.279 -0.342 0.974
-0.283 -0.325 0.972
&WHz
O-58 0.64 0.83 0.75 1.01 0.87 1.12 1.10 -0.286 -0.308 0.%9
700 Hz
800 Hz
0.61 067 0.88 082 1.05 089 1.23 1.17
0.63 0.70 0.95 I.05 1.10 0.91 1.23 1.27
-0.284 -0.284 0.963
-0.274 -0.243 0.966
+ b.
logarithmic diagram was obtained in Fig. 3(a). The regression slope was -0.272 min unit-’ cm3. The linear relation is useful for the application of quantitative analysis. The regression slope of the logarithmic plot between the threshold coagulation time and the concentration of F VIII was -0.276 min unit-’ cm3. Figure 3(b) shows the logarithmic correlation between the ratio coagulation time and the concentration of F IX. A linear relation was obtained in Fig. 3(b). The regression slope was
-0.314 min unit-’ cm3. The regression slope of the logarithmic plot between the threshold coagulation time and the concentration of F IX was -0.279 min unit-’ cm3. Comparison with the light scattering method
The coagulation time of the light scattering method was obtained from a maximum changing point of the light scattering intensity. Figure 4(a) shows the correlation in a double-logarithmic
Quartz crystal viscosity sensor for monitoring
coagulation
357
reaction
(a)
,o-o.
5 I ,o-l.
5
I ,o-o.
10-l
CONCENTRATION
5
1 CONCENTRATION
(unlt.cm-3)
(unlt.cm-3)
(b)
(b)
to-2
10
-1.
5
CONCENTRATION
Fig. 3. Correlation concentration
,o-o.
10-I (U”lt.cm-3)
(regression slope, -0.272
unit-‘cm+
regression coejicient,
(regression
slope.
-0.325
min
coeflcient,
0.980) unit-’
,0-l.
5
and cm.‘:
min
Fig. 4. (a) Correlation
between F VIII
lIMC+A
concentration
and
coagulation
time’ obtained
with light scattermg method
(b) F IX
(rqyession
slope,
min
regression
coefjicient. 0.999). (h) Correlation
0.981).
-0.152
unit-’
cm-‘:
regression
between F IX concen-
tration and coagulation time obtained with light scattering
diagram of F VIII concentration versus the coagulation time obtained with a light scattering method. Figure 4(b) also shows the logarithmic correlation of the ratio coagulation time versus the concentration of F IX. The slopes obtained from Fig. 4(a) and Fig. 4(b) were about -0.15 min unit-’ cm3. The slope obtained from the QCA was about double that of the slope obtained with the light scattering method. This difference in slopes shows the response curves of the two methods given by the different processes. The intensity R of the light scattering is given by =
5
(UnIt.cm-3)
method (rqression slope, -0209
K/R
,o-o.
10-l
CONCENTRATION
between the ratio coagulation time and
of (a) F VIII
10-2
5
(3)
where Mis the molecular weight of the particle, C is the concentration of the particle, and K and A
coeflcient.
min unit-’
cm-‘: regression
0.999).
are constants. The amplitude of light scattering intensity change depends on the molecular weight and the concentration. The reciprocal expression of light scattering intensity correlates to the reciprocal expression of the molecular weight. The viscosity q of the polymer solution is given by (tl - tlo)/rlo = KM’”
(4)
where no is the viscosity of water. M is an average molecular weight of polymer molecule, C is a concentration of polymer, and K is a constant. The viscosity of the liquid depends on the molecular weight and the concentration. The
358
H. Muramatsu
resonant frequency varies with the change in viscosity as shown in eqn (1). The resonant frequency correlates to the square root of the viscosity, and the viscosity correlates to the square root of the molecular weight. This relationship shows that the change in resonant frequency is more sensitive than the viscosity change for the lower molecular weight suspension. This relation indicates that in the early stage of the coagulation reaction the resonant frequency changes more markedly than the light scattering intensity. Consequently. the sigmoidal type of response curve of the resonant frequency change should rise in the early stages of the reaction compared with the curve of the light scattering intensity. The difference in the two types of response curve is also apparent for the various concentrations of the coagulation factors in the early stages of the coagulation reaction. When the coagulation time is calculated at the earlier stage of the reaction. the slope of the coagulation time versus the concentration for the QCA becomes steeper than that for the light scattering method. On the whole. we believe both the QCA and the light scattering method to be usable for the determination of blood coagulation factors. REFERENCES Davie, E. W. & Fujikawa. K. (1975). Basic mechanism in blood coagulation. Annu. Rev. Biochem.. 44, 799-829.
Hardisty, R. M.. Macpherson. J. C. & Ingram. G. 1. C. (1962). A one-stage factor VIII (anti-haemophilic globulin) assay and its use on venous and capillary plasma. Thromhos. Diathes Haemorrh.. 7, 215-29. Hlavay, J. & Guilbault, G. G. (1977). Application of the piezoelectric crystal detector in analytical chemistry. Anal. Chem.. 49, 1890-g. Kanazawa, K. K. & Gordon, J. G. (1985). The oscillation frequency of a quartz resonator in contact with a liquid. Anal. Chim. Acta, 175, 99-105.
King, W. H. (1964). Piezoelectric Anal. Chem.. 36, 1735-8.
sorption
detector.
et al.
Miura. Y. (1984). Behaviour of activated clotting factors in the potency estimation of coagulation factor preparations. J Jpn. Sot. Blood Transfusion, 30, 114-20.
Miura. Y., Zhu. W., Hara. S. & Homma, R. (1989). Effects of concentrated factors in concentrated human antihaemophilic globulin on the potency test. 12th Congr. Int. Sot. Thrombosis and Haemostasis. Vol. 62. Schattener. New York, p. 78. Muramatsu, H.. Kajiwara. K., Tamiya, E. & Karube, I. (1986). Piezoelectric immuno sensor for the detection of Candida albicans microbes. Anal. Chim. Acta, 188, 257-61.
Muramatsu. H., Dicks, J. M.. Tamiya, E. & Kaube, I. (1987). Piezoelectric crystal biosensor modified with protein A for determination of immunoglobulins. Anal. Chem., 59, 2760-3. Muramatsu, H., Tamiya. E. & Karube. 1. (1988a). Computation of equivalent circuit parameters of quartz crystal in contact with liquids and study of liquid properties. Anal. Chem.. 60, 2142-6. Muramatsu. H., Tamiya. E., Suzuki, M. & Karube, I. (1988b). Viscosity monitoring with a piezoelectric quartz crystal and its application to determination of endotoxin by gelation of Limulus amebocyte lysate. Anal. Chim. Acta 215, 91-8. Muramatsu, H., Watanabe, Y.. Hikuma, M.. Ataka, T.. Kubo, I., Tamiya. E. & Karube. I. (1989a). Piezoelectric crystal biosensor system for detection of Escherichia coli. Anal. Len., 22, 2155-66. Muramatsu, H.. Tamiya. E., Suzuki, M. & Karube, I. (19896). Quartz-crystal gelation detector for the determination of fibrinogen concentration. Anal. Chim. Acta. 217, 321-6. Nomura, T. & Iijima. M. (1981). Electrolytic determination of nanomolar concentrations of silver in solution with a piezoelectric quartz crystal. Anal. Chim. Acta. 131,97-102.
Sauerbrey. G. (1959). Verwendung von Schwingquarzen zur Wagung dunner Schichten und zur Mikrowagung. Z. Phys.. 155, 206-22. Shons. A., Dorman. F. & Najarian. J. (1972). An immunospecitic microbalance. J. Biomed. Mater. Rex, 6, 565-70.
Thompson. M.. Arthur. C. L. & Dhaliwal, G. K (1986). Liquid-phase piezoelectric and acoustic transmission studies of interfacial immunochemistry. Anal. Chem.. 58, 1206-9.
A quartz crystal viscosity sensor for monitoring coagulation reaction and its application to a multichannel coagulation detector H. Muramatsu, Seiko Instruments
K. Kimura, T. Ataka
Inc.. Applied Research Department,
563 Takatsukashinden,
Matsudo-shi,
Chiba 271, Japan
R. Homma, Y. Miura National
Institute of Health, Department
of Blood Products, 2-10-35 Kamiosaki, Japan
Shinagawa-ku.
Tokyo 141,
& I. Karube University
of Tokyo, Research Center for Advanced Science and Technology, Tokyo 153, Japan
(Received 4 May 1990: revised version received 10 September
4-6-l Komaba. Meguro-ku,
1990; accepted 24 September
1990)
Abstract: A quartz crystal viscosity sensor was applied to a coagulation
reaction monitoring system. The system consists of 16 oscillating circuits. a channel selector, a frequency counter, a temperature controller and a microcomputer. The system is named the Quartz Chemical Analyzer (QCA). AT-cut quartz crystals (9 MHz) were used as viscosity detectors and were attached to a cell in order to expose only one side of the quartz plate. The system was applied to the detection of the blood coagulation factors VIII (F VIII) and IX (F IX). The activity of these factors was assayed by a single-stage method. A linear relationship was obtained in a double-logarithmic diagram of concentration versus coagulation time with respect to F VIII and F IX in the range @OS-O*4 unit cm-) and @025-O-2 unit cmm3. respectively.
Keywords: quartz crystal, viscosity sensor, blood coagulation
INTRODUCTION In
chemical
analysis
a quartz
crystal
micro-
balance (Sauerbrey, 1959) has been applied to the sensing of gases (King, 1964, Hlavay & Guilbault.
factors.
1977). ions (Nomura & Tijima, 1981). biopolymers (Shons et al., 1972; Thompson ec al., 1986; Muramatsu efal., 1987) and microbes (Muramatsu et al., 1986; Muramatsu et al., 1989a). The property of the quartz crystal in contact with liquid has
353 Biosensors CcBioekmonics 0956~5663/91/$03.50@1991 Elsevier Science Publishers Ltd. England. Printed in Great Britain
354
H. Muramatsu et al.
also been studied (Kanazawa & Gordon, 1985: Muramatsu et al., 1988a. b; Muramatsu et al., 19893). It has been shown that the resonant frequency of the quartz crystal changes depending upon the viscosity and density of the liquid. Kanazawa and Gordon derived the following equation (Kanazawa & Gordon, 1985): AF = -Fs”*(tl~~_lw~q)“*
(1)
where F, is the resonant frequency of the quartz crystal, AF is the change in resonant frequency, pi is the density of the liquid, n is the viscosity of the liquid, pQ is the density of the quartz and p is the elastic modulus of the quartz. The authors have also derived an equation for the resonant resistance, which is one of the electrical equivalent circuit parameters of the quartz crystal, as follows (Muramatsu et al., 1988a): R = (2nFspLq)“*A/k2
(2)
where k is the electro-mechanical coupling factor andA is the electrode area of the quartz crystal. A quartz crystal has been used as a viscosity monitoring sensor. The authors have applied this sensor to the determination of (a) endotoxin concentrations using the gelation reaction of Limulus amebocyte lysate (Muramatsu et al.. 1988b) and (b) fibrinogen concentrations using the coagulation reaction of thrombin (Muramatsu et al., 19896). The mechanism of the coagulation process of blood has been elucidated (Davie & Fujikawa, 1975). Blood coagulation factors VIII and IX are known to be factors that are deficient in the blood of patients with hemophilia A and hemophilia B, respectively. The analysis of the blood coagulation factors is widely performed at clinical inspection. The estimation of the activity coagulation factor preparations is also important (Miura, 1984; Miura et al., 1989). The single-stage method has been used for the determination of F VIII and F IX activity (Hardisty et al., 1962). In this method, coagulation time was measured for the activity determination by mixing the deficient plasma of F VIII (or F IX), the activation factors, Ca*+ ions and the sample solution. In the present work, the quartz crystal viscosity sensor was applied to the monitoring of and the 16channel coagulation reaction, coagulation detector was made up using the quartz crystals. The system was applied to the detection of F VIII and F IX.
Fig. 1. Schematic diagram of the measuring system: A, quartz crystal: B. oscillating circuit: C. heater: D. signal selector: E. frequency counter: F, microcomputer.
EXPERIMENTAL Apparatus and materials
The measuring system (Quartz Chemical Analyzer: QCA) consists of 16 oscillating circuits, a channel selector, a frequency counter and a temperature controller (Fig. 1). The oscillating circuits and the channel selector were fabricated using TI’L-IC. The temperature controller was constructed from an aluminum block, a rubber heater and a temperature controlling unit. The frequency counter was fabricated using a counting IC (Intersil, ICM7226B). The system was controlled by a microcomputer (SeikoEpson, PC-286). The measuring cells consisted of 9 MHz AT-cut quartz crystals. The quartz crystals were positioned at the bottom of the cells in order to expose only one side of them. The 16 quartz crystals in QCA are used for multiple measurement at the same time. QCA is easily composed as a multichannel system. An activated partial thromboplastin time reagent, deficient plasma and Owren’s Verona1 buffer solution (pH 7.35) were obtained from Kokusai-shiyaku. CaC12 (0.025 M) was obtained from Fuji-Rebio. Standard blood coagulation factors from the National Institute of Health were used for a sample solution of F VIII and F IX. The sample solution was diluted with Owren’s veronal buffer solution for preparing the concentration. A light scattering detector (Kokusai-shiyaku, Coagurostat) was used for light scattering measurement. Procedure
F VIII (or F IX) activity was assayed by the singlestage method as follows (Hardisty et al., 1962). The sample solution of F VIII (or F IX) and an
355
Quartz crystal viscosity sensor for monitoring coagulation reaction
activated partial thromboplastin time reagent were added to pre-incubated plasma deficient in F VIII (or F IX) for 1 min at 37°C. After incubating the mixture for 2 min, O-025 M CaClz was added. Immediately, the mixture was poured onto the quartz crystals. The resonant frequency was monitored at intervals of a few seconds. The gate time for resonant frequency measurement was 1 s.
RESULTS AND DISCUSSION Resonant frequency response to the coagulation reaction
Figure 2 shows the resonant frequency response to the coagulation reaction with F VIII sample solution of 0.0543 unit cmV3. In Fig. 2, the resonant frequency changed as a result of viscosity change caused by the coagulation reaction. It took 5-10 s to start changing the frequency. Because the blood resonant coagulation system consists of cascade enzyme reaction, it appeared to take 5-10 s to begin the last stage of the coagulation reaction. The frequency change decreased to zero when coagulation was completed. Two methods of calculating the coagulation time were considered in this work. In the first method, the time to reach a certain ratio of total frequency change (ratio coagulation time) was used. In the second method, the time to attain a
Fig. 2. Resonant frequency response to the coagulation reaction with F VIIIsamplesolution of 0.0543 unit cm-! A is the ratio for calculation of coagulation time, and B show-s the ratio coagulation time. C i.sthe thresholdjkquency for calculation of coagulation time. and D sh0w.s the threshold coagulation time.
threshold certain frequency coagulation time) was used. This shown in Fig. 2.
(threshold scheme is
Study of the coagulation ratio and the threshold frequency
Table 1 shows the ratio coagulation time for a series of F VIII concentrations in duplicate measurements of the resonant frequency response. The regression slope and coefftcient between the coagulation time and concentration of F VIII are also shown in Table 1. The results were dealt with by a double-logarithmic scale, because it is usually used for a wide-range measurement and it is known that the relation for the light scattering method shows a linear relation in a double-logarithmic diagram. The calculation was based on the equation of log T = a log C + b, where T is the coagulation time, C is the concentration of coagulation factor, and a and b are regression constants. In Table 1 the regression slope shows close agreement with the coagulation ratio in the range of 20-70%. The correlation curves of 60% and 70% showed a good regression coefticient. The slope of the correlation curve of 60% is steeper than that of 70%. This result corresponds to results from other experiments which were not shown in this paper. A value of 60% was adopted for the coagulation ratio. Table 2 shows the threshold coagulation time for a series of F VIII concentrations in duplicate measurement ofthe resonant frequency response. The regression slope and coefficient between the coagulation time and concentration of F VIII are also shown in Table 2, as for Table 1. In Table 2. the regression slope is seen to be in close agreement with the threshold frequency in the range of 200-800 Hz. The correlation curves of 200 Hz and 300 Hz showed good regression coefficients. The slope of the correlation curve for 300 Hz is seen to be steeper than that for 200 Hz. This result corresponds with results from other experiments. A threshold frequency of 300 Hz was adopted. Correlation between coagulation time and the concentration of coagulation factors
Figure 3(a) shows the correlation between the ratio of coagulation time and the concentration of F VIII. A linear relationship in the double
356
H. Muramatsu et
Ratio coagulation
TABLE 1 time (min) for a series of F VIII concentrations in duplicate measurement of resonant frequency response. The regression constants and coefficients are also shown
Concentration (unit cm-3
Ratio in totalfrequency 90%
80%
70%
60%
50%
40%
30%
20%
0.76 0.83 0.93 1.30 1.18 I.11 1.37 1.24
0.69 0.72 0.87 11.1 1.11 I.04 1.25 1.17
064 a68 0.83 a85 I.06 a99 1.20 1.11
0.61 0.65 0.80 0.76 1.01 0.95 1.14 1.05
0.58 a61 0.76 a73 0.97 0.91 1.08 1xIo
0.56 0.58 0.73 0.70 0.92 0.88 1.02 O.%
0.53 0.55 0.70 0.67 0.88 a84 0.96 0.91
0.51 0.52 0.67 0.63 0.83 0.81 0.91 0.86
-0.153 -0.223
-0.215 -0.248
-0.263 -a268
-0.291 -0.272
-0.312 -0.273
-0.351 -0270
-0.371 -0.267
Regression coetlicient
0945
a972
a980
log(time) = a log(concentration)
+ b.
0.434 a434 0.217 a217 0.109 a109 0.054 a054 :
Threshold
al.
0.980
0.978
-0.27 1 -0.33 0.979
0.975
TABLE 2 coagulation time (min) for a series of F VIII concentrations in duplicate measurement frequency response. The regression constants and coefficients are also shown
Concentration (unit cm --),
0.434 a434 0.217 a217 0.109 al09 0.054 a054 a b
Regression coefficient log(time) = a log(concentration)
0.973
of resonant
Threshold frequency 100 Hz
200 Hz
300 Hz
400 Hz
500 Hz
046
0.47 0.61 0.59 0.75 0.73 0.59 0.78
0.50 0.51 0.68 0.63 0.82 0.78 090 0.88
0.52 a54 0.72 a66 0.87 0.80 0.96 a93
0.54 0.57 0.75 0.69 0.92 0.82 1.01 O-98
0.56 a60 0.79 a72 0.96 a85 1.06 1.04
-0.194 -0.37 1 0.861
-0.267 -0.375 0.976
-0.276 -a359 0.976
-0.279 -0.342 0.974
-0.283 -0.325 0.972
&WHz
O-58 0.64 0.83 0.75 1.01 0.87 1.12 1.10 -0.286 -0.308 0.%9
700 Hz
800 Hz
0.61 067 0.88 082 1.05 089 1.23 1.17
0.63 0.70 0.95 I.05 1.10 0.91 1.23 1.27
-0.284 -0.284 0.963
-0.274 -0.243 0.966
+ b.
logarithmic diagram was obtained in Fig. 3(a). The regression slope was -0.272 min unit-’ cm3. The linear relation is useful for the application of quantitative analysis. The regression slope of the logarithmic plot between the threshold coagulation time and the concentration of F VIII was -0.276 min unit-’ cm3. Figure 3(b) shows the logarithmic correlation between the ratio coagulation time and the concentration of F IX. A linear relation was obtained in Fig. 3(b). The regression slope was
-0.314 min unit-’ cm3. The regression slope of the logarithmic plot between the threshold coagulation time and the concentration of F IX was -0.279 min unit-’ cm3. Comparison with the light scattering method
The coagulation time of the light scattering method was obtained from a maximum changing point of the light scattering intensity. Figure 4(a) shows the correlation in a double-logarithmic
Quartz crystal viscosity sensor for monitoring
coagulation
357
reaction
(a)
,o-o.
5 I ,o-l.
5
I ,o-o.
10-l
CONCENTRATION
5
1 CONCENTRATION
(unlt.cm-3)
(unlt.cm-3)
(b)
(b)
to-2
10
-1.
5
CONCENTRATION
Fig. 3. Correlation concentration
,o-o.
10-I (U”lt.cm-3)
(regression slope, -0.272
unit-‘cm+
regression coejicient,
(regression
slope.
-0.325
min
coeflcient,
0.980) unit-’
,0-l.
5
and cm.‘:
min
Fig. 4. (a) Correlation
between F VIII
lIMC+A
concentration
and
coagulation
time’ obtained
with light scattermg method
(b) F IX
(rqyession
slope,
min
regression
coefjicient. 0.999). (h) Correlation
0.981).
-0.152
unit-’
cm-‘:
regression
between F IX concen-
tration and coagulation time obtained with light scattering
diagram of F VIII concentration versus the coagulation time obtained with a light scattering method. Figure 4(b) also shows the logarithmic correlation of the ratio coagulation time versus the concentration of F IX. The slopes obtained from Fig. 4(a) and Fig. 4(b) were about -0.15 min unit-’ cm3. The slope obtained from the QCA was about double that of the slope obtained with the light scattering method. This difference in slopes shows the response curves of the two methods given by the different processes. The intensity R of the light scattering is given by =
5
(UnIt.cm-3)
method (rqression slope, -0209
K/R
,o-o.
10-l
CONCENTRATION
between the ratio coagulation time and
of (a) F VIII
10-2
5
(3)
where Mis the molecular weight of the particle, C is the concentration of the particle, and K and A
coeflcient.
min unit-’
cm-‘: regression
0.999).
are constants. The amplitude of light scattering intensity change depends on the molecular weight and the concentration. The reciprocal expression of light scattering intensity correlates to the reciprocal expression of the molecular weight. The viscosity q of the polymer solution is given by (tl - tlo)/rlo = KM’”
(4)
where no is the viscosity of water. M is an average molecular weight of polymer molecule, C is a concentration of polymer, and K is a constant. The viscosity of the liquid depends on the molecular weight and the concentration. The
358
H. Muramatsu
resonant frequency varies with the change in viscosity as shown in eqn (1). The resonant frequency correlates to the square root of the viscosity, and the viscosity correlates to the square root of the molecular weight. This relationship shows that the change in resonant frequency is more sensitive than the viscosity change for the lower molecular weight suspension. This relation indicates that in the early stage of the coagulation reaction the resonant frequency changes more markedly than the light scattering intensity. Consequently. the sigmoidal type of response curve of the resonant frequency change should rise in the early stages of the reaction compared with the curve of the light scattering intensity. The difference in the two types of response curve is also apparent for the various concentrations of the coagulation factors in the early stages of the coagulation reaction. When the coagulation time is calculated at the earlier stage of the reaction. the slope of the coagulation time versus the concentration for the QCA becomes steeper than that for the light scattering method. On the whole. we believe both the QCA and the light scattering method to be usable for the determination of blood coagulation factors. REFERENCES Davie, E. W. & Fujikawa. K. (1975). Basic mechanism in blood coagulation. Annu. Rev. Biochem.. 44, 799-829.
Hardisty, R. M.. Macpherson. J. C. & Ingram. G. 1. C. (1962). A one-stage factor VIII (anti-haemophilic globulin) assay and its use on venous and capillary plasma. Thromhos. Diathes Haemorrh.. 7, 215-29. Hlavay, J. & Guilbault, G. G. (1977). Application of the piezoelectric crystal detector in analytical chemistry. Anal. Chem.. 49, 1890-g. Kanazawa, K. K. & Gordon, J. G. (1985). The oscillation frequency of a quartz resonator in contact with a liquid. Anal. Chim. Acta, 175, 99-105.
King, W. H. (1964). Piezoelectric Anal. Chem.. 36, 1735-8.
sorption
detector.
et al.
Miura. Y. (1984). Behaviour of activated clotting factors in the potency estimation of coagulation factor preparations. J Jpn. Sot. Blood Transfusion, 30, 114-20.
Miura. Y., Zhu. W., Hara. S. & Homma, R. (1989). Effects of concentrated factors in concentrated human antihaemophilic globulin on the potency test. 12th Congr. Int. Sot. Thrombosis and Haemostasis. Vol. 62. Schattener. New York, p. 78. Muramatsu, H.. Kajiwara. K., Tamiya, E. & Karube, I. (1986). Piezoelectric immuno sensor for the detection of Candida albicans microbes. Anal. Chim. Acta, 188, 257-61.
Muramatsu. H., Dicks, J. M.. Tamiya, E. & Kaube, I. (1987). Piezoelectric crystal biosensor modified with protein A for determination of immunoglobulins. Anal. Chem., 59, 2760-3. Muramatsu, H., Tamiya. E. & Karube. 1. (1988a). Computation of equivalent circuit parameters of quartz crystal in contact with liquids and study of liquid properties. Anal. Chem.. 60, 2142-6. Muramatsu. H., Tamiya. E., Suzuki, M. & Karube, I. (1988b). Viscosity monitoring with a piezoelectric quartz crystal and its application to determination of endotoxin by gelation of Limulus amebocyte lysate. Anal. Chim. Acta 215, 91-8. Muramatsu, H., Watanabe, Y.. Hikuma, M.. Ataka, T.. Kubo, I., Tamiya. E. & Karube. I. (1989a). Piezoelectric crystal biosensor system for detection of Escherichia coli. Anal. Len., 22, 2155-66. Muramatsu, H.. Tamiya. E., Suzuki, M. & Karube, I. (19896). Quartz-crystal gelation detector for the determination of fibrinogen concentration. Anal. Chim. Acta. 217, 321-6. Nomura, T. & Iijima. M. (1981). Electrolytic determination of nanomolar concentrations of silver in solution with a piezoelectric quartz crystal. Anal. Chim. Acta. 131,97-102.
Sauerbrey. G. (1959). Verwendung von Schwingquarzen zur Wagung dunner Schichten und zur Mikrowagung. Z. Phys.. 155, 206-22. Shons. A., Dorman. F. & Najarian. J. (1972). An immunospecitic microbalance. J. Biomed. Mater. Rex, 6, 565-70.
Thompson. M.. Arthur. C. L. & Dhaliwal, G. K (1986). Liquid-phase piezoelectric and acoustic transmission studies of interfacial immunochemistry. Anal. Chem.. 58, 1206-9.
