- Email: [email protected]
Separation method based on affinity reaction between magnetic and nonmagnetic particles for the analysis of particles and biomolecules
Journal of Chromatography A, 1130 (2006) 227–231
Separation method based on affinity reaction between magnetic and nonmagnetic particles for the analysis of particles and biomolecules H.Y. Tsai a , F.H. Hsu b , Y.P. Lin b , C. Bor Fuh b,∗ a
b
School of Applied Chemistry, Chung Shan Medical University, Taichung, Taiwan Department of Applied Chemistry, National Chi Nan University, Puli, Nantou 545, Taiwan Available online 12 June 2006
Abstract A separation method is reported for particle and biochemical analysis based on affinity interactions between particle surfaces under magnetic field. In this method, magnetic particles with immunoglobulin G (IgG) or streptavidin on the surface are flowed through a separation channel to form a deposition matrix for selectively capturing nonmagnetic analytes with protein A or biotin on the surface due to specific antigen (Ag)–antibody (Ab) interactions. This separation method was demonstrated using model reactions of IgG–protein A and streptavidin–biotin on particle surface. The features of this new separation method are (1) the deposited Ag–Ab complex can be examined and further analyzed under the microscope, (2) a kinetic study of complex binding is possible, and (3) the predeposited matrix can be formed selectively and changed easily. The detection limits were about 10−11 g. The running time was less than 10 min. The selectivities of studied particles were 94% higher than those of label-controlled particles. This method extends the applications of analytical magnetapheresis to nonmagnetic particles. Preliminary study shows that this separation method has a great potential to provide a simple, fast, and selective analysis for particles, blood cells, and immunoassay related applications. © 2006 Elsevier B.V. All rights reserved. Keywords: Affinity
1. Introduction Magnetic separation has been used in many biochemical related applications [1–12]. Analytical magnetapheresis is a relatively new technique of magnetic separation for separating magnetically susceptible particles [1–4]. Analytical magnetapheresis uses a thin (<0.03 cm) channel and the magnetic force is applied perpendicular to the channel flow axis for separating magnetically susceptible particles. Particles with high fieldinduced velocities are selectively attracted by the magnetic force and deposited on the channel bottom plate to separate from those particles with low field-induced velocities as both particles pass along the channel. The field-induced velocity is a function of particle magnetic susceptibility, particle diameter, carrier viscosity, and magnetic field strength [1,2,4]. Previously, magnetic particles were selectively deposited on the unpacked channel plate for examination under the microscope in analytical magnetapheresis. It would be difficult for analytical magnetapheresis to analyze nonmagnetic particles
∗
Corresponding author. Tel.: +886 4 9291 9779; fax: +886 4 9291 7956. E-mail address: [email protected] (C.B. Fuh).
0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2006.05.045
like cells without magnetic derivation. Various magnetic ion labels were used for magnetic derivation but these derivations were not very selective. It is important to have selective magnetic derivation or have alternative ways to analyze nonmagnetic particles. The purpose of this new approach is to extend the applications of analytical magnetapheresis to nonmagnetic particles like cells via affinity reactions on the particle surface. A separation method is based on affinity interactions between particle surfaces. In this method, magnetic particles with antibody (Ab) [or antigen (Ag)] on the surface are flowed through a separation channel to deposit as a matrix above the interpolar gap under magnetic fields, as shown in Fig. 1. Nonmagnetic particles with specific Ag (or Ab) on the surface are selectively captured to the predeposited zone and separated from those particles without specific Ag (or Ab) on the surface due to Ag–Ab interactions as they flow through the channel. The deposited Ag–Ab complex can be further analyzed under microscopic examinations which cannot be easily done in a capillary column. The predeposited zone can be formed easily and changed conveniently. Therefore, this method is very useful for analyzing particles of micron size like cells in simple, fast, and selective ways. This method can provide simple and economical ways for cell analysis as compared with flow cytometer. Magnetic particles of
228
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
Fig. 1. Top view of a separation method. (A) Glass plate above magnet. (B) Deposition zone formed. (C) Affinity complex formed. ( ) Magnetic particles with antibody on the surface; ( ) particles with antigen on the surface; ( ) antibody; ( ) antigen.
submicron sizes were chosen to form the matrix for easier observation of captured nonmagnetic particle under the microscope in this study. A separation method was investigated to extend its applications to nonmagnetic particles using model reactions of immunoglobulin G (IgG)–protein A and streptavidin–biotin on particle surfaces. 2. Experimental The channel assembly was the same as that used in previous works in the literature [1,2,4]. Briefly, the channel components contained a cut-out Mylar spacer sandwiched between acrylic sheet and glass plate using silicone sealant and all these layers were pressed evenly with clamps. The thin glass plate (150 m) was used as the channel bottom plate for particle depositions. The channel length, breadth, and thickness used were 1.0, 0.05, and 0.012 cm, respectively. The calculated void volume was 0.0006 ml. A permanent magnet assembly from rare earth magnets (Nd–Fe–B, neodymium–iron–boron, Super Electronic, Taipei, Taiwan), connected with soft-iron pole pieces, was used to generate magnetic fields for experiments. The maximum energy product of Nd–Fe–B magnets was 3.50 × 107 G Oe. An interpolar gapwidth of 1.5 mm was used for all experiments, and the saturation field Bo was 2.1 × 104 G. Magnetic field measurements were made using a Gaussmeter and a Hall-effect probe (Model Gauss 5080, F.W. Bell, Orlando, FL, USA) with adjustable microstages. The probe measured magnetic flux perpendicular to a sensing area with a diameter of 0.4 mm. A multichannel syringe pump (Model 200, KD Scientific, Boston, MA, USA) was used for sample and carrier delivery. The carriers used were phosphate-buffered saline (PBS) solutions and Hank balanced salt (HBS) solutions with pHs of 7.02. A light microscopy (Olympus BX-50, Tokyo, Japan) was used for cell and particle counting and verification. A Hemacytometer was used to count cells and particles and to calculate their concentrations in experiments. A refrigerated centrifuge (Model EBA-12, Hettich Zentrifugen, Tuttlingen, Germany) was used for cell and particle centrifugations. Streptavidin- and IgGconjugated magnetic particles with sizes of 0.86 m were purchased from Merck Eurolab (Paris, France). Protein A, d-biotin, Vitamin b1, and bovine serum albumin (BSA) were purchased from Sigma Chemical (St. Louis, MO, USA). Silica particles (7.8 ± 0.4 m) were purchased from Hypersil (London, Eng-
land). Yeasts were from a bakery store in a nearby market of Puli, Taiwan. 2.1. Preparation of labeled particles In order to compare specific and nonspecific binding, protein A-labeled particles were compared to BSA-labeled particles, biotin-labeled particles were compared to Vitamin b1labeled particles and BSA-labeled particles. Protein A-labeled particles were prepared by mixing 1.0 × 104 particles with 2.84 × 10−6 M protein A in 3 ml of HBS solutions under stirring for 10 min. Label-controlled particles were prepared by mixing 2.84 × 10−6 M BSA with 1.0 × 104 particles in 3 ml of HBS solutions under stirring for 10 min. In order to confirm the specific binding of IgG–protein A on particle surface, protein A-labeled particles were further blocked by BSA prepared as the same as label-controlled particles. Biotin-labeled particles were prepared by mixing 1.0 × 104 particles with 2.74 × 10−6 M d-biotin in 3 ml of HBS solutions under stirring for 10 min. Biotin label-controlled particles were prepared by mixing 1.0 × 104 particles with 2.74 × 10−6 M BSA and Vitamin b1 in 3 ml of HBS solutions under stirring for 5 min, respectively. The labeled particles were centrifuged and washed with 5 ml of HBS solutions three times to remove excess labels and stored at 4 ◦ C before use. The sample injection volume was 0.1 ml. 2.2. A separation method Ab-conjugated magnetic particles of submicron sizes were flowed through the separation channel and deposited on the channel bottom plate above the interpolar gap at a flow-rate of 0.01 ml min−1 . Ag-labeled particles were then flowed through the separation channel for selective interactions with predeposited magnetic particles at controlled flow-rates. Binding kinetic data were collected for both streptavidin– biotin and streptavidin–BSA interactions by loading different amounts of labeled yeasts. The rate of complex formation was calculated from dividing the number of deposited complex by the time of flowing biotin-labeled yeasts over the predeposited zone containing streptavidin-conjugated magnetic particles. The equation for method of initial rate is dp/dt = ka M0 Lbulk , where P is the number of deposited complex, Ka the rate constant of complex formation, M0 the amount of predeposited streptavidin-
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
229
conjugated magnetic particles and Lbulk is the amount of biotinlabeled yeasts. 3. Results and discussion The amounts of predeposited magnetic particles, the selectivities of Ag–Ab reactions, the flow-rates of labeled particles, and the amounts of labeled particles were studied for their effects on Ab–Ag reactions in a separation method. 3.1. The amounts of predeposited magnetic particles The effect of different amounts of IgG-conjugated magnetic particles (1.0 × 103 to 1.0 × 107 ) with 2.0 × 103 protein A-labeled particles on a separation method was studied at a flow-rate of 0.02 ml min−1 . The deposition zone can be clearly seen when the number of IgG-conjugated magnetic particles for predeposition was equaled or greater than 1.0 × 104 . The depositions of protein A-labeled particles increased as IgGconjugated magnetic particles increased. The optimal number of IgG-conjugated magnetic particles for Ag–Ab reactions was 1.0 × 106 , thus was used in rest of the experiments. The amounts of IgG-conjugated magnetic particles roughly filled the predeposition area.
Fig. 3. The deposited percentages of labeled particles at different flow-rates in a separation method. The number of injected particles was 200.
about 94% higher than those of label-controlled particles for IgG- and streptavidin-conjugated magnetic particles, respectively. The amounts of particles captured on the deposition zone for negative control of protein A-labeled particles and Vitamin b1-labeled particles averaged 7% or less than protein A- and biotin-labeled particles, respectively. 3.3. The flow-rates of labeled particles
3.2. The selectivities of Ag–Ab reactions The selectivity of protein A–IgG and streptavidin–biotin between particle surfaces was studied in a separation method. Fig. 2 shows the selectivity of protein A- and biotin-labeled yeasts for IgG- and streptavidin-conjugated magnetic particles using some controlled-labels for comparison. Protein A- and biotin-labeled silica particles also showed similar selectivities. The controlled labels of protein A were bare particles and particles labeled with BSA. Protein A-labeled particles blocking with BSA were also used to confirm the specific binding of IgG–protein A complex. The controlled labels of biotin were bare particles, and particles labeled with Vitamin b1 and BSA. The selectivities of protein A- and biotin-labeled particles were
The effect of flow-rates (0.005–0.135 ml min−1 ) of labeled particles on Ag–Ab formations in a separation method was studied. Protein A-labeled particles could be captured by IgGconjugated magnetic particles and be seen on the deposition zone as their flow-rates equaled or less than 0.045 ml min−1 . The amount of deposition from protein A-labeled particles increased about 60% as their flow-rates decreased from 0.045 to 0.005 ml min−1 . The deposited percentages of labeled particles at different flow-rates are shown in Fig. 3. The deposited percentages of model reactions from two complexes increased linearly as the flow-rate decreased from 0.015 to 0.002 ml min−1 with correlation coefficients greater than 0.996. The percentages of complex formation from protein A–IgG were about 5% higher
Fig. 2. Selectivity of Ab-conjugated magnetic particles and Ag-labeled particles in a separation method. (A) IgG-conjugated magnetic particles and protein A-labeled yeasts. Flow-rate was 0.01 ml min−1 . (B) Streptavidin-conjugated magnetic particles and biotin-labeled yeasts. Flow-rate was 0.005 ml min−1 .
230
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
Fig. 4. Microscopic deposition of different amounts of protein A-labeled silica particles with 1.0 × 106 IgG-conjugated magnetic particles under 200× magnification at a flow-rate of 0.01 ml min−1 . The number of protein A-labeled particles used for injection was: (1) 0, (2) 125, (3) 500, (4) 2000.
than those of streptavidin–biotin at the same flow-rates. The optimal flow-rates could be obtained from Fig. 3 for required deposition percentages of a given complex. 3.4. The amounts of labeled particles Fig. 4 shows some of the microscopic depositions from different amounts of protein A-labeled silica particles (0–2000) with 1.0 × 106 IgG-conjugated magnetic particles in a separation method. The micrographs of deposition zone were mainly from protein A-labeled silica particles due to their larger sizes (7.8 m) than magnetic particles (0.9 m). The number of protein A-labeled silica particles captured on the deposition zone increased with the number of injected particles. The amounts of protein A-labeled silica particles need to be larger than 125 at a flow-rate of 0.01 ml min−1 in order to see and analyze the deposition zone under the microscope. The running time was around 10 min at a flow-rate of 0.01 ml min−1 . 3.5. Estimation of labels on each particle The difference of protein A concentration before and after labeling silica particles was 1.82 × 10−7 M as determined by spectrometer. All protein A molecules were assumed to be labeled on the surface of silica particles. Therefore, the number of protein A labels was around 3.29 × 1010 on each silica particle, as calculated from dividing 3 ml of 1.82 × 10−7 M molecules by 1.0 × 104 silica particles. There was about 4.98 × 10−9 g of protein A on each silica particle. This separation method was also studied using streptavidinconjugated magnetic particles with biotin-labeled particles. Microscopic depositions of biotin-labeled particles with streptavidin-conjugated magnetic particles increased as flowrates decreased and as the amount of magnetic and labeled particles increased in a manner analogous to the IgG–protein A reaction described above. The difference of biotin concentration before and after labeling particles was 4.7 × 10−7 M as determined by spectrometer. Therefore, the number of labeled biotin was around 6.61 × 1010 on each particle. There were about 2.67 × 10−13 g of biotin on each silica particle. The detection limit was about 2.67 × 10−11 g since the depositions can be
surely confirmed for injection of 100 biotin-labeled silica particles. 3.6. The possible applications for kinetic study of complex binding The applications of this separation method for kinetic study of complex binding were also investigated using method of initial rate. The rate of complex formation for different numbers of biotin-labeled yeasts at a flow-rate of 0.05 ml min−1 was plotted as a straight line. The reaction rate of complex formation was linearly proportional to the number of biotin-labeled yeasts over a range of 100–1000 with the correlation coefficient equaled to 0.999. The slope of fitted line was proportional to the rate constant of complex formation between particle surface. Therefore, the relative rate constant of complex formation can be obtained by dividing two equations. The relative rate constant of streptavidin–biotin between particle surface was 39 times greater than those of controlled streptavidin–BSA. This study indicated the possible applications of analytical magnetapheresis for cell related study like cell adhesion or cell interactions. 4. Conclusions A separation method has been demonstrated using model reactions of protein A–IgG and biotin–streptavidin on particle surface. The running time was less than 10 min. The predeposited matrix can be formed selectively and changed easily as needed. The deposited Ag–Ab complex can be examined and further analyzed under the microscope. Possible applications of analytical magnetapheresis on kinetic study for cell related analysis was indicated. The detection of Ag–Ab complex can be automated using density scanning or optical spectroscopic method. Therefore, this method would be very useful for analyzing particles of micron size like cells. The selectivities of biotin- and protein A-labeled particles were about 94% higher than those of label-controlled particles for streptavidin- and IgGconjugated magnetic particles, respectively. The detection limits were about 10−11 g. A lower limit could be achieved by further modification and optimization. This method extends the applications of analytical magnetapheresis to nonmagnetic particles. It has a great potential to provide a simple, fast, and selective
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
analysis for particle, blood cells, and immunoassay related applications. Acknowledgment This work was supported by the National Science Council of Taiwan. References [1] C.B. Fuh, M.H. Lai, L.Y. Lin, S.Y. Yeh, Anal. Chem. 72 (2000) 3590. [2] C.B. Fuh, L.Y. Lin, M.H. Lai, J. Chromatogr. A 874 (2000) 131. [3] M. Zborowski, C.B. Fuh, R. Green, L. Sun, J.J. Chalmers, Anal. Chem. 67 (1995) 3702.
231
[4] C.B. Fuh, Y.S. Su, H.Y. Tsai, J. Chromatogr. A 1027 (2004) 289. [5] J.P. Hancock, J.T. Kemshed, J. Immunol. Methods 164 (1993) 51. [6] S. Funderud, K. Nustad, T. Lea, F. Vartdal, G. Guadernack, P. Stensted, L. Ugelstad, in: G.B. Klaus (Ed.), Lymphocytes: A Practical Approach, Oxford University Press, New York, 1987, p. 55. [7] J. Ugelstad, P. Stensted, L. Kilaas, W.S. Prestvik, R. Herje, A. Berge, E. Hornes, Blood Purif. 11 (1993) 347. [8] R. Tyagi, M.N. Gupta, Biotechnol. Appl. Biochem. 21 (1995) 217. [9] Y. Morimoto, H. Nattsume, Jpn. J. Clin. Med. 56 (1998) 649. [10] C.M. Schweitzer, C.E. ven der Schoot, A.M. Drager, P. van der Valk, A. Zevenbergen, B. Hooibrink, A.H. Westra, M.M. Langenhuijsen, Exp. Hematol. 23 (1995) 41. [11] N. Maher, H.K. Dillon, S.H. Vermund, T.R. Unnasch, Appl. Environ. Microbiol. 67 (2001) 449. [12] T. Matsunaga, H. Nakayama, M. Okochi, H. Takeyama, Biotechnol. Bioeng. 73 (2001) 400.
Separation method based on affinity reaction between magnetic and nonmagnetic particles for the analysis of particles and biomolecules H.Y. Tsai a , F.H. Hsu b , Y.P. Lin b , C. Bor Fuh b,∗ a
b
School of Applied Chemistry, Chung Shan Medical University, Taichung, Taiwan Department of Applied Chemistry, National Chi Nan University, Puli, Nantou 545, Taiwan Available online 12 June 2006
Abstract A separation method is reported for particle and biochemical analysis based on affinity interactions between particle surfaces under magnetic field. In this method, magnetic particles with immunoglobulin G (IgG) or streptavidin on the surface are flowed through a separation channel to form a deposition matrix for selectively capturing nonmagnetic analytes with protein A or biotin on the surface due to specific antigen (Ag)–antibody (Ab) interactions. This separation method was demonstrated using model reactions of IgG–protein A and streptavidin–biotin on particle surface. The features of this new separation method are (1) the deposited Ag–Ab complex can be examined and further analyzed under the microscope, (2) a kinetic study of complex binding is possible, and (3) the predeposited matrix can be formed selectively and changed easily. The detection limits were about 10−11 g. The running time was less than 10 min. The selectivities of studied particles were 94% higher than those of label-controlled particles. This method extends the applications of analytical magnetapheresis to nonmagnetic particles. Preliminary study shows that this separation method has a great potential to provide a simple, fast, and selective analysis for particles, blood cells, and immunoassay related applications. © 2006 Elsevier B.V. All rights reserved. Keywords: Affinity
1. Introduction Magnetic separation has been used in many biochemical related applications [1–12]. Analytical magnetapheresis is a relatively new technique of magnetic separation for separating magnetically susceptible particles [1–4]. Analytical magnetapheresis uses a thin (<0.03 cm) channel and the magnetic force is applied perpendicular to the channel flow axis for separating magnetically susceptible particles. Particles with high fieldinduced velocities are selectively attracted by the magnetic force and deposited on the channel bottom plate to separate from those particles with low field-induced velocities as both particles pass along the channel. The field-induced velocity is a function of particle magnetic susceptibility, particle diameter, carrier viscosity, and magnetic field strength [1,2,4]. Previously, magnetic particles were selectively deposited on the unpacked channel plate for examination under the microscope in analytical magnetapheresis. It would be difficult for analytical magnetapheresis to analyze nonmagnetic particles
∗
Corresponding author. Tel.: +886 4 9291 9779; fax: +886 4 9291 7956. E-mail address: [email protected] (C.B. Fuh).
0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2006.05.045
like cells without magnetic derivation. Various magnetic ion labels were used for magnetic derivation but these derivations were not very selective. It is important to have selective magnetic derivation or have alternative ways to analyze nonmagnetic particles. The purpose of this new approach is to extend the applications of analytical magnetapheresis to nonmagnetic particles like cells via affinity reactions on the particle surface. A separation method is based on affinity interactions between particle surfaces. In this method, magnetic particles with antibody (Ab) [or antigen (Ag)] on the surface are flowed through a separation channel to deposit as a matrix above the interpolar gap under magnetic fields, as shown in Fig. 1. Nonmagnetic particles with specific Ag (or Ab) on the surface are selectively captured to the predeposited zone and separated from those particles without specific Ag (or Ab) on the surface due to Ag–Ab interactions as they flow through the channel. The deposited Ag–Ab complex can be further analyzed under microscopic examinations which cannot be easily done in a capillary column. The predeposited zone can be formed easily and changed conveniently. Therefore, this method is very useful for analyzing particles of micron size like cells in simple, fast, and selective ways. This method can provide simple and economical ways for cell analysis as compared with flow cytometer. Magnetic particles of
228
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
Fig. 1. Top view of a separation method. (A) Glass plate above magnet. (B) Deposition zone formed. (C) Affinity complex formed. ( ) Magnetic particles with antibody on the surface; ( ) particles with antigen on the surface; ( ) antibody; ( ) antigen.
submicron sizes were chosen to form the matrix for easier observation of captured nonmagnetic particle under the microscope in this study. A separation method was investigated to extend its applications to nonmagnetic particles using model reactions of immunoglobulin G (IgG)–protein A and streptavidin–biotin on particle surfaces. 2. Experimental The channel assembly was the same as that used in previous works in the literature [1,2,4]. Briefly, the channel components contained a cut-out Mylar spacer sandwiched between acrylic sheet and glass plate using silicone sealant and all these layers were pressed evenly with clamps. The thin glass plate (150 m) was used as the channel bottom plate for particle depositions. The channel length, breadth, and thickness used were 1.0, 0.05, and 0.012 cm, respectively. The calculated void volume was 0.0006 ml. A permanent magnet assembly from rare earth magnets (Nd–Fe–B, neodymium–iron–boron, Super Electronic, Taipei, Taiwan), connected with soft-iron pole pieces, was used to generate magnetic fields for experiments. The maximum energy product of Nd–Fe–B magnets was 3.50 × 107 G Oe. An interpolar gapwidth of 1.5 mm was used for all experiments, and the saturation field Bo was 2.1 × 104 G. Magnetic field measurements were made using a Gaussmeter and a Hall-effect probe (Model Gauss 5080, F.W. Bell, Orlando, FL, USA) with adjustable microstages. The probe measured magnetic flux perpendicular to a sensing area with a diameter of 0.4 mm. A multichannel syringe pump (Model 200, KD Scientific, Boston, MA, USA) was used for sample and carrier delivery. The carriers used were phosphate-buffered saline (PBS) solutions and Hank balanced salt (HBS) solutions with pHs of 7.02. A light microscopy (Olympus BX-50, Tokyo, Japan) was used for cell and particle counting and verification. A Hemacytometer was used to count cells and particles and to calculate their concentrations in experiments. A refrigerated centrifuge (Model EBA-12, Hettich Zentrifugen, Tuttlingen, Germany) was used for cell and particle centrifugations. Streptavidin- and IgGconjugated magnetic particles with sizes of 0.86 m were purchased from Merck Eurolab (Paris, France). Protein A, d-biotin, Vitamin b1, and bovine serum albumin (BSA) were purchased from Sigma Chemical (St. Louis, MO, USA). Silica particles (7.8 ± 0.4 m) were purchased from Hypersil (London, Eng-
land). Yeasts were from a bakery store in a nearby market of Puli, Taiwan. 2.1. Preparation of labeled particles In order to compare specific and nonspecific binding, protein A-labeled particles were compared to BSA-labeled particles, biotin-labeled particles were compared to Vitamin b1labeled particles and BSA-labeled particles. Protein A-labeled particles were prepared by mixing 1.0 × 104 particles with 2.84 × 10−6 M protein A in 3 ml of HBS solutions under stirring for 10 min. Label-controlled particles were prepared by mixing 2.84 × 10−6 M BSA with 1.0 × 104 particles in 3 ml of HBS solutions under stirring for 10 min. In order to confirm the specific binding of IgG–protein A on particle surface, protein A-labeled particles were further blocked by BSA prepared as the same as label-controlled particles. Biotin-labeled particles were prepared by mixing 1.0 × 104 particles with 2.74 × 10−6 M d-biotin in 3 ml of HBS solutions under stirring for 10 min. Biotin label-controlled particles were prepared by mixing 1.0 × 104 particles with 2.74 × 10−6 M BSA and Vitamin b1 in 3 ml of HBS solutions under stirring for 5 min, respectively. The labeled particles were centrifuged and washed with 5 ml of HBS solutions three times to remove excess labels and stored at 4 ◦ C before use. The sample injection volume was 0.1 ml. 2.2. A separation method Ab-conjugated magnetic particles of submicron sizes were flowed through the separation channel and deposited on the channel bottom plate above the interpolar gap at a flow-rate of 0.01 ml min−1 . Ag-labeled particles were then flowed through the separation channel for selective interactions with predeposited magnetic particles at controlled flow-rates. Binding kinetic data were collected for both streptavidin– biotin and streptavidin–BSA interactions by loading different amounts of labeled yeasts. The rate of complex formation was calculated from dividing the number of deposited complex by the time of flowing biotin-labeled yeasts over the predeposited zone containing streptavidin-conjugated magnetic particles. The equation for method of initial rate is dp/dt = ka M0 Lbulk , where P is the number of deposited complex, Ka the rate constant of complex formation, M0 the amount of predeposited streptavidin-
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
229
conjugated magnetic particles and Lbulk is the amount of biotinlabeled yeasts. 3. Results and discussion The amounts of predeposited magnetic particles, the selectivities of Ag–Ab reactions, the flow-rates of labeled particles, and the amounts of labeled particles were studied for their effects on Ab–Ag reactions in a separation method. 3.1. The amounts of predeposited magnetic particles The effect of different amounts of IgG-conjugated magnetic particles (1.0 × 103 to 1.0 × 107 ) with 2.0 × 103 protein A-labeled particles on a separation method was studied at a flow-rate of 0.02 ml min−1 . The deposition zone can be clearly seen when the number of IgG-conjugated magnetic particles for predeposition was equaled or greater than 1.0 × 104 . The depositions of protein A-labeled particles increased as IgGconjugated magnetic particles increased. The optimal number of IgG-conjugated magnetic particles for Ag–Ab reactions was 1.0 × 106 , thus was used in rest of the experiments. The amounts of IgG-conjugated magnetic particles roughly filled the predeposition area.
Fig. 3. The deposited percentages of labeled particles at different flow-rates in a separation method. The number of injected particles was 200.
about 94% higher than those of label-controlled particles for IgG- and streptavidin-conjugated magnetic particles, respectively. The amounts of particles captured on the deposition zone for negative control of protein A-labeled particles and Vitamin b1-labeled particles averaged 7% or less than protein A- and biotin-labeled particles, respectively. 3.3. The flow-rates of labeled particles
3.2. The selectivities of Ag–Ab reactions The selectivity of protein A–IgG and streptavidin–biotin between particle surfaces was studied in a separation method. Fig. 2 shows the selectivity of protein A- and biotin-labeled yeasts for IgG- and streptavidin-conjugated magnetic particles using some controlled-labels for comparison. Protein A- and biotin-labeled silica particles also showed similar selectivities. The controlled labels of protein A were bare particles and particles labeled with BSA. Protein A-labeled particles blocking with BSA were also used to confirm the specific binding of IgG–protein A complex. The controlled labels of biotin were bare particles, and particles labeled with Vitamin b1 and BSA. The selectivities of protein A- and biotin-labeled particles were
The effect of flow-rates (0.005–0.135 ml min−1 ) of labeled particles on Ag–Ab formations in a separation method was studied. Protein A-labeled particles could be captured by IgGconjugated magnetic particles and be seen on the deposition zone as their flow-rates equaled or less than 0.045 ml min−1 . The amount of deposition from protein A-labeled particles increased about 60% as their flow-rates decreased from 0.045 to 0.005 ml min−1 . The deposited percentages of labeled particles at different flow-rates are shown in Fig. 3. The deposited percentages of model reactions from two complexes increased linearly as the flow-rate decreased from 0.015 to 0.002 ml min−1 with correlation coefficients greater than 0.996. The percentages of complex formation from protein A–IgG were about 5% higher
Fig. 2. Selectivity of Ab-conjugated magnetic particles and Ag-labeled particles in a separation method. (A) IgG-conjugated magnetic particles and protein A-labeled yeasts. Flow-rate was 0.01 ml min−1 . (B) Streptavidin-conjugated magnetic particles and biotin-labeled yeasts. Flow-rate was 0.005 ml min−1 .
230
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
Fig. 4. Microscopic deposition of different amounts of protein A-labeled silica particles with 1.0 × 106 IgG-conjugated magnetic particles under 200× magnification at a flow-rate of 0.01 ml min−1 . The number of protein A-labeled particles used for injection was: (1) 0, (2) 125, (3) 500, (4) 2000.
than those of streptavidin–biotin at the same flow-rates. The optimal flow-rates could be obtained from Fig. 3 for required deposition percentages of a given complex. 3.4. The amounts of labeled particles Fig. 4 shows some of the microscopic depositions from different amounts of protein A-labeled silica particles (0–2000) with 1.0 × 106 IgG-conjugated magnetic particles in a separation method. The micrographs of deposition zone were mainly from protein A-labeled silica particles due to their larger sizes (7.8 m) than magnetic particles (0.9 m). The number of protein A-labeled silica particles captured on the deposition zone increased with the number of injected particles. The amounts of protein A-labeled silica particles need to be larger than 125 at a flow-rate of 0.01 ml min−1 in order to see and analyze the deposition zone under the microscope. The running time was around 10 min at a flow-rate of 0.01 ml min−1 . 3.5. Estimation of labels on each particle The difference of protein A concentration before and after labeling silica particles was 1.82 × 10−7 M as determined by spectrometer. All protein A molecules were assumed to be labeled on the surface of silica particles. Therefore, the number of protein A labels was around 3.29 × 1010 on each silica particle, as calculated from dividing 3 ml of 1.82 × 10−7 M molecules by 1.0 × 104 silica particles. There was about 4.98 × 10−9 g of protein A on each silica particle. This separation method was also studied using streptavidinconjugated magnetic particles with biotin-labeled particles. Microscopic depositions of biotin-labeled particles with streptavidin-conjugated magnetic particles increased as flowrates decreased and as the amount of magnetic and labeled particles increased in a manner analogous to the IgG–protein A reaction described above. The difference of biotin concentration before and after labeling particles was 4.7 × 10−7 M as determined by spectrometer. Therefore, the number of labeled biotin was around 6.61 × 1010 on each particle. There were about 2.67 × 10−13 g of biotin on each silica particle. The detection limit was about 2.67 × 10−11 g since the depositions can be
surely confirmed for injection of 100 biotin-labeled silica particles. 3.6. The possible applications for kinetic study of complex binding The applications of this separation method for kinetic study of complex binding were also investigated using method of initial rate. The rate of complex formation for different numbers of biotin-labeled yeasts at a flow-rate of 0.05 ml min−1 was plotted as a straight line. The reaction rate of complex formation was linearly proportional to the number of biotin-labeled yeasts over a range of 100–1000 with the correlation coefficient equaled to 0.999. The slope of fitted line was proportional to the rate constant of complex formation between particle surface. Therefore, the relative rate constant of complex formation can be obtained by dividing two equations. The relative rate constant of streptavidin–biotin between particle surface was 39 times greater than those of controlled streptavidin–BSA. This study indicated the possible applications of analytical magnetapheresis for cell related study like cell adhesion or cell interactions. 4. Conclusions A separation method has been demonstrated using model reactions of protein A–IgG and biotin–streptavidin on particle surface. The running time was less than 10 min. The predeposited matrix can be formed selectively and changed easily as needed. The deposited Ag–Ab complex can be examined and further analyzed under the microscope. Possible applications of analytical magnetapheresis on kinetic study for cell related analysis was indicated. The detection of Ag–Ab complex can be automated using density scanning or optical spectroscopic method. Therefore, this method would be very useful for analyzing particles of micron size like cells. The selectivities of biotin- and protein A-labeled particles were about 94% higher than those of label-controlled particles for streptavidin- and IgGconjugated magnetic particles, respectively. The detection limits were about 10−11 g. A lower limit could be achieved by further modification and optimization. This method extends the applications of analytical magnetapheresis to nonmagnetic particles. It has a great potential to provide a simple, fast, and selective
H.Y. Tsai et al. / J. Chromatogr. A 1130 (2006) 227–231
analysis for particle, blood cells, and immunoassay related applications. Acknowledgment This work was supported by the National Science Council of Taiwan. References [1] C.B. Fuh, M.H. Lai, L.Y. Lin, S.Y. Yeh, Anal. Chem. 72 (2000) 3590. [2] C.B. Fuh, L.Y. Lin, M.H. Lai, J. Chromatogr. A 874 (2000) 131. [3] M. Zborowski, C.B. Fuh, R. Green, L. Sun, J.J. Chalmers, Anal. Chem. 67 (1995) 3702.
231
[4] C.B. Fuh, Y.S. Su, H.Y. Tsai, J. Chromatogr. A 1027 (2004) 289. [5] J.P. Hancock, J.T. Kemshed, J. Immunol. Methods 164 (1993) 51. [6] S. Funderud, K. Nustad, T. Lea, F. Vartdal, G. Guadernack, P. Stensted, L. Ugelstad, in: G.B. Klaus (Ed.), Lymphocytes: A Practical Approach, Oxford University Press, New York, 1987, p. 55. [7] J. Ugelstad, P. Stensted, L. Kilaas, W.S. Prestvik, R. Herje, A. Berge, E. Hornes, Blood Purif. 11 (1993) 347. [8] R. Tyagi, M.N. Gupta, Biotechnol. Appl. Biochem. 21 (1995) 217. [9] Y. Morimoto, H. Nattsume, Jpn. J. Clin. Med. 56 (1998) 649. [10] C.M. Schweitzer, C.E. ven der Schoot, A.M. Drager, P. van der Valk, A. Zevenbergen, B. Hooibrink, A.H. Westra, M.M. Langenhuijsen, Exp. Hematol. 23 (1995) 41. [11] N. Maher, H.K. Dillon, S.H. Vermund, T.R. Unnasch, Appl. Environ. Microbiol. 67 (2001) 449. [12] T. Matsunaga, H. Nakayama, M. Okochi, H. Takeyama, Biotechnol. Bioeng. 73 (2001) 400.