Hemolysis tests necessary for the development of circulatory artificial organs have been carried out as relative evaluations using animal blood. 1–5 To evaluate the hemolytic properties of devices under comparable conditions, a new standard testing tool is needed as a substitute for animal blood. We have been developing microcapsule suspensions for use as standard hemolysis test solutions. Microcapsule suspensions, as described here, do not simulate the biological properties of animal blood, but can simulate static mechanical and fatigue properties of erythrocytes. Each microcapsule has a leuco dye inside, which changes to a red color in acid solutions. The microcapsule suspension, or Model Blood, was prepared by adding the microcapsules to aqueous solutions, where the viscosity was controlled with ethylene glycol. It has been reported that the hemolytic properties of a prototype centrifugal blood pump, using polyurethane microcapsule suspensions, showed results similar to those obtained with bovine blood. However, similar properties for a commercial pump could not be detected due to low hemolysis, that is, the microcapsules did not break in the commercial pump. 6 In this study, we synthesized microcapsules with membranes made of polyurethane and melamine resin. A melamine resin membrane is more fragile than one of polyurethane. The diameter of each capsule was controlled to be approximately 100 μm to enhance the sensitivity of the capsule destruction, as well as 10 μm, similar to that of erythrocytes, to evaluate hemolysis caused by the wall surface roughness in centrifugal pumps. Therefore, we prepared the following four types of microcapsule suspensions: 100 μm polyurethane, 10 μm polyurethane, 100 μm melamine resin, and 10 μm melamine resin. This study focuses on the diameters and membrane materials using polyurethane capsules and melamine resin capsules.
The fractural characteristics of these microcapsule suspensions secondary to fluid dynamic shear stress, one of the most important factors in hemolytic evaluation in centrifugal blood pumps, 7,8 were examined using a rotational shear stressor. Although the fractural characteristics of animal blood using viscometers have been reported, 9–14 the shear stressor described here is a double-cone, double-cylinder type, designed to shear homogenously with an 8 ml gap volume. The outer cylinder rotated and the inner cylinder fixed to stabilize the flow in the gap as shearing occurred. 15 The fractural properties of the four suspensions were compared with bovine blood using the rotating shear stressor.
Materials and Methods
Production of Microcapsule Slurry
The following four types of microcapsule slurry (Three Bond Co., Ltd., Tokyo, Japan) were prepared with the maximum diameter of the microcapsule as the target: (1) 100 μm maximum diameter capsules with polyurethane membranes; (2) 10 μm maximum diameter capsules with polyurethane membranes; (3) 100 μm maximum diameter capsules with melamine resin membranes; (4) 10 μm maximum diameter capsules with melamine resin membranes.
Leuco dye (Yamada Chemicals Co., Ltd., Kyoto, Japan) was dissolved i-propylnaphthalene (Kureha Chemical Industry Co., Ltd., Tokyo, Japan) to obtain 109.7 mg/ml. The slurries in (1) and (2), were synthesized together with the leuco dye solution, aliphatic polyisocyanate (Sumitomo Bayer Urethane Co., Ltd., Tokyo, Japan), and denatured aliphatic polyamine (Japan Epoxy Resins Co., Ltd., Tokyo, Japan) by interfacial polycondensation. The melamine resin capsule slurries in (3) and (4) were synthesized together in situ with leuco dye solution, melamine (Showa Chemicals Co., Ltd., Tokyo, Japan), and formaldehyde (Showa Chemicals Co., Ltd.). The diameters of these microcapsules were changed by controlling the emulsion particle size versus reaction time, irrespective of the amount of membrane polymer.
Preparation of Microcapsule Suspensions.
An aqueous solution, used as artificial plasma, with 2.49 mPa·s viscosity, was added to the slurries, contained 38.8 w/v% ethylene glycol (Wako Co., Ltd., Osaka, Japan) and 1.8 w/v% sodium chloride (Wako Co., Ltd.). Equal volumes of slurry and aqueous solution were mixed to prepare the four microcapsule suspensions as follows: MS-100U, microcapsule suspensions prepared from (1); MS-10U, microcapsule suspensions prepared from (2); MS-100 M, microcapsule suspensions prepared from (3); MS-10 M, microcapsule suspensions prepared from (4).
The chemical components of these suspensions are shown in Table 1. Amounts of each component were similar in the final concentrations. The basic properties of these suspensions and bovine blood were examined. To prevent thrombus formation, 15 v/v% of acid citrate dextrose (ACD-A, Terumo Co., Ltd., Tokyo, Japan) and 6 units/ml of heparin sodium (Wako Co., Ltd.) were initially added to bovine blood, which was then diluted with physiologic saline to a 30% hematocrit.
Observation of Microcapsules.
The macroscopic appearance of microcapsules were evaluated with capsule shapes observed using an optical microscope (CK-2, OLYMPUS Co., Ltd., Tokyo, Japan).
A needle was pressed against the microcapsules, and deformation of the capsules was qualitatively observed under the microscope.
Granular Variations in Microcapsule Suspensions.
Granular variations in the microcapsules in suspension were measured with a particle size analyzer (LS-100, Beckman Coulter, CA).
The membrane thickness of the microcapsules was measured using transmission electron microscopy (JEM-200EX, JEOL LTD, Tokyo Japan).
The density was measured using a pycnometer at 25°C.
Microcapsule suspensions and bovine blood were placed into a capillary tube (Propper Manufacturing Co., Ltd., NY) with an inside diameter of 1.2 mm and a length of 75 mm. The tubes were centrifuged at 12,000 rpm (9,338 × g) for 10 min using a centrifuge with a TH-1 rotor (MC-150, TOMY SEIKO Co., Ltd., Tokyo, Japan). The volume fraction of microcapsules in the suspensions were measured to produce hematocrit values.
Capsule concentration in the suspensions was detected with a Fuchs Rosenthal hemocytometer (Nippon Rinsho Kikaikogyo Co., Ltd., Tokyo, Japan).
Viscosities of the microcapsule suspensions and bovine blood were measured using a rotational viscometer (DVM-II, TOKIMEC, Inc., Tokyo, Japan) with a standard rotor (cone angle of 1 degree and 34 min) at rotational speeds of 20, 50, and 100 rpm at 25°C.
Fractural Properties of Microcapsule Suspensions Evaluated with the Rotational Shear Stressor.
A photograph of the shear stressor apparatus (Rheometer NRM-100BR, Nihon Rheology Kiki Co., Ltd., Chiba, Japan) is shown in Figure 1, and the test unit and a cross-section of the shear stressed portion are represented in Figure 2A and B. The shear stressor contained 8 ml of test solution. The outer cylinder was rotated at a maximum speed of 4,000 rpm, which corresponded to 15,000 s−1 of shear velocity. Tests were carried out for 15 min in the microcapsule suspensions, and for 5 min in bovine blood against the increasing shear velocity. Moreover, torque (Tq) generated at the inner cylinder was measured, and the suspension viscosity (η) was calculated from Tq to observe the flow states in the shear stressor as given below: MATHwhere τmean = average shear stress at the wall between the outer cylinder and the inner cylinder (Pa), γmean = average shear velocity at the wall between the outer cylinder and the inner cylinder (s−1), R2 = radius of the outer cylinder (m), R1 = radius of the inner cylinder (m), h = average height of the test solution (m), n = rotational speed of the outer cylinder (rad/s), a = 0.159 (a coefficient for shear stress), and b = 0.209 (a coefficient for shear velocity).
Concentrations of leuco dye in the test suspensions and free hemoglobin in the plasma were detected as follows: For microcapsule suspensions, 2 ml of n-hexane (Wako Co., Ltd.) were added to 2 ml of each test suspension and mixed with a vortex mixer (NS-80, Iuchi Seieido Co., Ltd., Osaka, Japan) for 6 s to extract the free leuco dye into the n-hexane phase. These mixtures were centrifuged at 12,000 rpm (8,855 × g) for 20 s using a centrifuge with a TMP-21 rotor (MX-150, TOMY SEIKO Co., Ltd.). The supernatants were then removed from the mixed solution. To obtain color solutions, 1 ml of each supernatant was mixed with 1 ml of ethanol containing 0.1 M hydrochloric acid (Wako Co., Ltd.). Optical densities (UV-1600, Shimadzu Co., Ltd., Kyoto, Japan) of these colored solutions were measured at 520 nm, and evaluated to obtain the degree of hemolysis. The bovine blood was centrifuged at 12,000 rpm (8,855 × g) for 5 min at 4°C, before removing 200 μl of plasma. This was colored using a commercial kit based on the 3,3′,5,5′-tetramethylbenzidine (TMB) method (527-A, Hemoglobin, Plasma, Sigma Diagnostics, Inc., MO). The optical density was measured with a spectrophotometer (UV-1600, Shimadzu Co., Ltd.) to obtain the free hemoglobin concentration. Mean values of the leuco dye and free hemoglobin concentrations were then calculated from three repeated experiments.
Table 2 compares the macroscopic appearance, capsule shapes, membrane hardness, obtained diameters, membrane thickness, densities, hematocrits, capsule concentrations, and viscosities of the four types of microcapsules and bovine blood. Macroscopic appearance of MS-100U, MS-10U, and MS-100 M were seen as white suspensions, whereas MS-10 M was a pink suspension. Optical photographs of the four types of suspensions are shown in Figure 3. The diameters of the microcapsules had a wide distribution, as a result of granular variations. As all the suspensions were evident as spherical capsules, the capsule reactions were successful. Membrane hardness showed that the melamine resin capsule membranes were hard, whereas the polyurethane capsule membranes were elastic. The average diameters of MS-100U and MS-10U were 64 μm and 5 μm, respectively. The maximum diameters of MS-100U and MS-10U were 122 μm and 13 μm, which were close to the target diameters; these values were obtained from granular variation (Figure 4). On the other hand, the average diameter of MS-100 M and MS-10 M were 62 μm and 12 μm, respectively, which have a wider distribution than the polyurethane capsules. The maximum diameters of MS-100 M and MS-10 M were 177 μm and 30 μm, which were larger in diameter than the target values. Although the diameter distributions of the melamine resin capsules did not agree completely with those of polyurethane, and the tensile strength, elasticity, and molecular weight of the capsule membranes were not determined, a fractural property comparison of different diameters and membrane materials was the focus of this study. Membrane thickness was 200, 150, 35, and 35 nm for MS-100U, MS-10U, MS-100 M, and MS-10 M, respectively (Figure 5A–D). Membrane thickness/average diameter, as a kind of capsule fragility index, was calculated, and the decline followed the sequence MS-10U (0.03), MS-10 M (0.003) = MS-100U (0.003), and MS-100 M (0.0006). Densities of the four suspensions were almost the same at 1.02 g/cm3. Hematocrits for MS-100U and MS-10U were both 25%, similar to that in bovine blood (30%), and lower than those in MS-100 M (42%) and MS-10 M (46%). The capsule concentration in MS-100U was 1.83 × 107 capsules/ml, which had a concentration similar to MS-100 M (1.98 × 107 capsules/ml). However, capsule concentrations differed between the 1.80 × 109 capsules/ml of MS-10U and 1.28 × 109 capsules/ml of MS-10 M. From the viscosity characteristics, MS-100U, MS-100 M, and MS-10 M were non-Newtonian, because their viscosities decreased with the viscometer’s cone rotational speed, whereas MS-10U was Newtonian. The viscosities of the melamine-resin capsules obtained from a viscometer were smaller than those of the polyurethane capsules and close to that of bovine blood.
The fractural characteristics of microcapsule suspensions using a rotational shear stressor will be addressed next and are shown in Figure 6A–E). We have reported a relationship between a decrease in capsule numbers and the concentration of leuco dye due to the rupture of microcapsules. 6 In MS-100U, the leuco dye concentrations slightly increased for shearing velocities between 0 s−1 and 7,500 s−1, remained constant between 7,500 s−1 and 11,250 s−1, and markedly increased between 11,250 s−1 and 15,000 s−1, reaching a maximum value of 0.344 mg/dl. In control experiments without shearing, the concentration was constant at 0.110 mg/dl (Figure 6A). The leuco dye concentration in MS-100U increased at higher shear velocity, whereas MS-10U concentrations showed results similar to those of the control experiment (Figure 6B). For MS-100 M, the concentration of leuco dye increased between 0 s−1 and 7,500 s−1, slightly decreased between 7,500 s−1 and 9,375 s−1, and then sharply decreased between 9,375 s−1 and 11,250 s−1 (Figure 6C). The leuco dye concentration of MS-100 M at 7,500 s−1 was 2.560 mg/dl, which was 7.4 times higher than that of MS-100U at 15,000 s−1. Thus, the fractural properties of MS-100 M showed that capsule destruction was inhibited at high shear velocities, although the microcapsules can be broken with low shear velocity of approximately 7,500s−1. For MS-10 M, the mean values of leuco dye concentration were very low, and had tendencies similar to MS-10U (Figure 6D). For bovine blood, hemolysis increased at higher shear velocity, and the concentration of free hemoglobin at 15,000 s−1 was 25.4 mg/dl (Figure 6E). Because the concentration curve of bovine blood showed a tendency similar to that of MS-100U, we may conclude that the fractural characteristics of MS-100U using the rotational shear stressor had properties similar to that of bovine blood. Although the sensitivity to damage of MS-100 M was the highest of the four types of microcapsule suspensions, it exhibited different shearing characteristics when compared with that of bovine blood.
Viscosities calculated from the torque of microcapsule suspensions against shear velocity measured with the shear stressor are shown in Figure 7A–E. For all test suspensions and bovine blood, the figures clearly show that viscosity, which is related to torque, increased at higher shear velocity. Therefore, it is evident that shear stress, as generated in the shear stressor, will certainly increase at higher shear velocity. Despite this, destruction of the microcapsules in the MS-100 M decreased in the high shear stress region. Moreover, the viscosities of all test suspensions and bovine blood in the shear stressor increased 1.4 to 1.9 times over the initial value. For the standard solution for torque calibration, which is a Newtonian fluid, there was an increase of approximately 1.7 times that of the initial value, in the range of 3,500 s−1 to 15,000 s−1.
The maximum diameters of the melamine capsules differed from the target values, because the melamine capsule reactions were more complicated than the polyurethane capsule reactions. As such, controlling the emulsion particle sizes during the capsule reactions is difficult, causing diameter distributions to become wider. Hence, the maximum diameters of MS-100 M and MS-10 M were larger than the target values. Hematocrits in the polyurethane capsules were lower than those with a melamine resin membrane. This finding seems to depend on the fragility of the membrane material. Polyurethane is easier to compress by centrifugation than melamine because polyurethane is more elastic than melamine resin. In fact, the capsule concentrations of MS-100U (1.83 × 107 capsules/ml), and MS-100 M (1.98 × 107 capsules/ml) were almost the same. On the other hand, the capsule concentrations of MS-10U and MS-10 M were slightly different. This finding may be because the average diameter of MS-10 M was larger than that of MS-10U (Figure 4).
It was difficult to unify the granular variations, viscosities, and hematocrits of the four suspensions. However, we focused on the fact that diameter size and membrane material affected capsule destruction, and the behavior was compared with that of erythrocytes. Therefore, additional conditions were not controlled for in this study.
The fractural properties of MS-100U showed that destruction of the microcapsules increased with higher shear stress, a result similar to that seen when using bovine blood. We think that the combination of the polyurethane membrane elasticity and membrane thickness/average diameter ratio (0.003, Table 2) produced capsule destruction with increasing shear stress generated in the shear stressor. However, the leuco dye concentration of MS-100U at 15,000 s−1 was lower than that of MS-100 M in all regions. The concentration baseline of MS-100 M was higher than those of the other suspensions. We think that, because the membrane synthesis of the microcapsules was slightly incomplete and free leuco dye remained on the outside membrane surface (Figure 5), even the concentration of the control experiment reached approximately 1.0 mg/dl. MS-100 M experienced maximum destruction and was ruptured in the velocity range 0 s−1 and 11,250 s−1; this tendency was different from that in many reports e.g., hemolysis increases with higher shear stress. 10,11,13,16 This is the region where both turbulence and laminar flows coexisted, as reported by Aouidef et al.17 Moreover, flow conditions in the shear stressor changed from laminar to turbulent with the increase in shear velocity, because the viscosities of all suspensions, even the standard test solutions for torque calibration, became higher than the initial value (Figure 7). We think that these flow conditions affected the destruction of MS-100 M, because the microcapsules would only roll at the wall without being sheared at high speed, as its melamine resin membrane hardness is different from the polyurethane membrane. However, further studies using computational fluid dynamic analysis and flow visualization will be needed for verification. Broken capsule numbers, considering the total leuco dye concentration of the suspensions (1,831 mg/dl for MS-100U, and 1,784 mg/dl for MS-100 M), and extraction efficiency with n-hexane (22.6%), were 0.08% (0.344 mg/dl) for MS-100U at 15,000 s−1 and 0.64% (2.560 mg/dl) for MS-100 M at 7,500 s−1 after a 15 min rotation. A similar test for bovine blood gave 0.32% at 15,000 s−1 after a 5 min rotation. Microcapsule destruction for MS-100U was only a quarter of that of bovine blood, whereas destruction of MS-100 M was double that of bovine blood. However, coloring sensitivity of the leuco dye in this study is 100 times higher than that of hemoglobin. 6 In fact, even 0.03% capsule destruction can be detected, which makes this method applicable to hemolytic evaluation; therefore, destruction of MS-100U can be easily detected.
MS-10U and MS-10 M both failed to rupture during our experiments. The membrane thickness/average diameter ratio of MS-10 M, however, was the same as that of MS-100U, which showed fractural characteristics similar to bovine blood. Thus, it is clear that the membrane material is more important than the membrane thickness/average diameter ratio in the case of polyurethane or melamine resin membranes. A small amount of capsule destruction was observed as shown in Figure 6D, and it is thought that only microcapsules with larger diameters actually ruptured. Therefore, biological materials should be considered next, such as lipids or proteins with 10 μm diameter capsules.
In our study, we did not change the exposure time. Giersiepen et al.18 have reported that there were relationships of shear stress and exposure time to hemolysis amount using human blood. It is important to determine these relationships. However, the purpose of this experiment was to conduct screening tests to find out the interaction of capsule size and degree of hemolysis. In the near future, if we can obtain microcapsules suitable for hemolysis testing, it will be necessary to survey additional parameters.
It is known that the size of erythrocytes in bovine blood changes with the magnitude of shear stress. 19,20 Moreover, it has also been reported that hemolysis changes with aging of erythrocytes. 21 We also intend, in future work, to study the destruction differences between microcapsule suspension and bovine blood by selecting the membrane materials or controlling the strength of the membrane, while considering the factors of size, age, and deformation.
Four types of microcapsule slurries with different diameters and membrane materials were synthesized to prepare microcapsule suspensions for testing. For suspensions with maximum 100 μm diameter melamine resin membranes, the microcapsules broke at a relatively low shear stress and exhibited high fragility. Suspensions with maximum 100 μm diameter polyurethane membranes ruptured at higher shear stress regions, as seen in bovine erythrocytes. As such, these microcapsules can be used as a substitute for animal blood in hemolysis tests.
The authors are grateful to Ms. K. Fujisawa for her technical assistance, and to the Three Bond Co., Ltd. (Hachioji, Tokyo, Japan) for the production of the microcapsule slurries and measurement of capsule size analysis. The authors also thank Prof. T. Akamatsu of Setsunan University (Osaka, Japan) for his much appreciated suggestions.
1. Mizuguchi K, Damm G, Aber GA, et al: Does hematocrit affect in vitro hemolysis test results? Preliminary study with Baylor/NASA prototype axial flow pump. Artif Organs 18: 650–656, 1994.
2. Yeleswarapu KK, Antaki JF, Kameneva MV, Rajagopal KR: A mathematical model for shear-induced hemolysis. Artif Organs 19: 576–582, 1995.
3. Shimono T, Makinouchi K, Nosé Y: Total erythrocyte destruction time: The new index for the hemolytic performance of rotary blood pumps. Artif Organs 19: 571–575, 1995.
4. Tayama E, Nakazawa T, Takami Y, et al: The hemolysis test of Gyro C1E3 pump in pulsatile mode. Artif Organs 21: 675–679, 1997.
5. Kobayashi S, Nitta S, Yambe T, Sonobe T, Naganuma S, Hashimoto H: Hemolysis test of disposable type vibrating flow pump. Artif Organs 21: 691–693, 1997.
6. Maruyama O, Yamane T, Tsunemoto N, Nishida M, Tsutsui T, Jikuya T: A preliminary study of microcapsule suspension for hemolysis evaluation of artificial organs. Artif Organs 23: 274–279, 1999.
7. Yamane T, Asztalos B, Nishida M, et al: Flow visualization as a complementary tool to hemolysis testing in the development of centrifugal blood pumps. Artif Organs 22: 375–380, 1998.
8. Masuzawa T, Tsukiya T, Endo S, et al: Development of design methods for a centrifugal blood pump with a fluid dynamic approach: results in hemolysis tests. Artif Organs 23: 757–761, 1999.
9. Blackshear PL, Dorman FD, Steinbach JH: Some mechanical effects that influence hemolysis. Trans ASAIO 11: 112–117, 1965.
10. Shapiro SI, Williams MC: Hemolysis in simple shear flows. AIChE J 16: 575–580, 1970.
11. Champion JV, North PF, Coakley WT, Williams AR: Shear fragility of human erythrocytes. Biorheology 8: 23–29, 1971.
12. Williams AR: Shear-induced fragmentation of human erythrocytes. Biorheology 10: 303–311, 1972.
13. Leverett LB, Hellums JD, Alfrey CP, Lynch EC: Red blood cell damage by shear stress. Biophys J 12: 257–273, 1972.
14. Blackshear PL, Blackshear GL:Mechanical Hemolysis. Handbook of Bioengineering
. New York: McGraw-Hill, 1987, pp. 15.1–15.19.
15. Taylor GI: Fluid friction between rotating cylinders. I. Torque measurements. Proc R Soc Lond A 157: 546–564, 1936.
16. Nevaril CG, Lynch EC, Alflrey CP, Hellums JD: Erythrocyte damage and destruction induced by shearing stress. J Lab Clin Med 71: 784–790, 1968.
17. Aouidef A, Yamane T, Maruyama O, Nishida M: Fluid dynamic characteristics of a rotating shear stressor simulating rotary blood pumps. Life Support 13: 14–18, 2001.
18. Giersiepen M, Wurzinger LJ, Opitz R, Reul H: Estimation of shear stress-related blood damage in heart valve prostheses-in vitro comparison of 25 aortic valves. Int J Artif Organs 13: 300–306, 1990.
19. Sutera SP, Mehrjardi MH: Deformation and fragmentation of human red blood cells in turbulent shear flow. Biophys J 15: 1–10, 1975.
20. Bessis M, Mohandas N: A diffractometric method for the measurement of cellular deformability. Blood Cells 1: 307–313, 1975.
Copyright © 2002 by the American Society for Artificial Internal Organs
21. Jikuya T, Tsutsui T, Shigeta O, Sankai Y, Mitsui T: Species differences in erythrocyte mechanical fragility comparison of human, bovine, and ovine cells. ASAIO J 44; M452–M455, 1998.