Effect of Matrix Elasticity on Affinity Binding and Release of

Biological Cell Detachment from Poly(N-isopropyl acrylamide) and Its Applications. Marta A. Cooperstein and Heather E. Canavan. Langmuir 2010 26 (11),...
0 downloads 0 Views 377KB Size
Langmuir 2007, 23, 35-40

35

Effect of Matrix Elasticity on Affinity Binding and Release of Bioparticles. Elution of Bound Cells by Temperature-Induced Shrinkage of the Smart Macroporous Hydrogel† Igor Yu. Galaev,‡ Maria B. Dainiak,‡,§ Fatima Plieva,‡,§ and Bo Mattiasson*,‡ Department of Biotechnology, Lund UniVersity, P-O. Box 124, SE-22100 Lund, Sweden, and Protista Biotechnology AB, IDEON, SE-22370 Lund, Sweden ReceiVed May 24, 2006. In Final Form: August 9, 2006 The first step of bacterial or viral invasion is affinity and presumably multisite binding of bioparticles to an elastic matrix like a living tissue. We have demonstrated that model bioparticles such as inclusion bodies (spheres of about 1 µm in size) Escherichia coli cells (rods 1 × 3 µm), yeast cells (8 µm spheres), and synthetic microgel particles (0.4 µm spheres) are binding via different affinity interactions (IgG antibody-protein A, sugar-lectin, and metal ion-chelate) to a macroporous hydrogel (MH) matrix bearing appropriate ligands. The elastic deformation of the MH results in the detachment of affinity bound bioparticles. The particle detachment on elastic deformation is believed to be due to multipoint attachment of the particles to affinity matrix and the disturbance of the distance between affinity ligands when the matrix is deformed. No release of affinity bound protein occurred on elastic deformation. The efficiency of the particle release by the elastic deformation depends on the density of the ligands at the particle surface as well as on the elasticity of the matrix for relatively large particles. The release of the particles occurred irrespectively of whether the deformation was caused by external forces (mechanical deformation) or internal forces (the shrinkage of thermosensitive macroporous poly-N-isopropylacrylamide hydrogel on increase in temperature).

Introduction The first step of bacterial or any viral invasion is an adhesion to the solid-liquid interface at the surface of a living tissue. The interactions of mammalian cells with the surface of cell scaffolds are vital for the successful tissue engineering. The result of cellsurface contact is defined by three types of signals: chemical, topological, and mechanical.1 While cellular response to surface chemistry2-4 and topography5,6 and microbial adherence to stiff supports such as polystyrene, Teflon, and glass7-10 have been studied extensively, similar investigations regarding matrix mechanics have largely been overlooked. A pioneering study11 demonstrated that, compared with epithelial cells and fibroblasts growing on rigid substrates, those cultivated on flexible substrates showed reduced spreading and increased rates of motility or lamellipodial activity. These trends were independent of the adhesive ligand used (e.g., polylysine, fibronectin, collagen, and RGD).1 However, these studies have predominantly been carried out under static conditions. The dynamic studies were mainly addressing the changes in cell metabolism occurring on the matrix † Part of the Stimuli-Responsive Materials: Polymer, Colloids, and Multicomponent Systems special issue. * Corresponding author. Tel.: +46-46-222-8264. Fax: +46-46-222-4713. E-mail address: [email protected]. ‡ Lund University. § Protista Biotechnology AB.

(1) Wong, J. Y.; Leach, J. B.; Brown, X. Q. Surf. Sci. 2004, 570, 119-133. (2) Dainiak, M.; Galaev, I. Yu.; Mattiasson, B. J. Chromatogr. A 2002, 942, 123-131. (3) Castner, D. G.; Ratner, B. D. Surf. Sci. 2002, 500, 28-60. (4) Nath, N.; Hyun, J.; Ma, H.; Chilkoti, A. Surf. Sci. 2004, 570, 98-110. (5) Curtis, A.; Wilkinson, C. Biomaterials 1997, 18, 1573-1583. (6) Li, X.; Liu, T.; Chen, Y. Biochem. Eng. J. 2004, 22, 11-17. (7) Ming, F.; Whish, W. J. D.; Hubble, J. Enzyme Microbial Technol. 1998, 22, 94-99. (8) Rijnaarts, H. H. M.; Norde, W.; Bouwer, E. J.; Lyklema, J.; Zehnder, A. J. B. Colloids Surf. B.: Biointerfaces 1995, 4, 5-22. (9) Jucker, B. A.; Harms, H.; Zehnder, A. J. B. J. Bacteriol. 1996, 178, 54725479. (10) Parent, M. E.; Velegol, D. Colloids Surf. B. 2004, 39, 45-51. (11) Pelham, R. J.; Wang, Y. L. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 1366113665.

deformation12-16 rather than the efficiency of cell adhesion/ detachment. Alternatively, the effect of shear stress developed in the liquid flow, on the dynamics of cell adhesion/detachment to nondeforming matrixes has been studied.17,18 In the present work, the results of the studies of bioparticles interaction with and detachment from macroporous hydrogels (MHs) under elastic deformation are reported. Monolithic MHs from polyacrylamide (known also as cryogels) have recently been developed for the biotechnological applications19-24 and are characterized by high porosity and elasticity.25,26 Polyacrylamide-based MHs are elastic spongelike materials that can withstand large deformations and can be easily compressed 4-6fold without getting mechanically damaged, whereas traditional polyacrylamide gels are brittle and easily destroyed when deformed. Due to capillary forces, the MH monoliths retain the (12) Lehoux, S.; Tedgui, A. J. Biomechanics 2003, 36, 631-643. (13) Chiquet, M.; Renedo, A. S.; Huber, F.; Fluck, M. Matrix Biol. 2003, 22, 73-80. (14) Puk, C. K.; Miller, D. J.; Gamradt, S.; Wu, B. M.; McAllister, D. R. J. Biomed. Mater. Res. Part A 2006, 76A, 665-673. (15) Berry, C. C.; Cacou, C.; Lee, D. A.; Bader, D. L.; Shelton, J. C. Biorheology 2003, 40, 337-345. (16) Arnoczky, S. P.; Tian, T.; Lavagnino, M.; Gardner, K.; Schuler, P.; Morse, P. J. Orthop. Res. 2002, 20, 947-952. (17) Cao, X.; Eisenthal, R.; Hubble, J. Enzyme Microbial Technol. 2002, 31, 153-160. (18) Madrusov, E.; Houng, A.; Klein, E.; Leonard, E. F. Biotechnol. Prog. 1995, 11, 208-213. (19) Dainiak, M. B.; Kumar, A.; Plieva, F. M.; Galaev, I. Yu; Mattiasson, B. J. Chromatogr. A 2004, 1045, 93-98. (20) Kumar, A.; Plieva, F. M.; Galaev, I. Yu; Mattiasson, B. J. Immunol. Methods 2003, 283, 185-194. (21) Dainiak, M. B.; Plieva, F. M.; Galaev, I. Yu; Hatti-Kaul, R.; Mattiasson, B. Biotechnol. Prog. 2005, 21, 644-649. (22) Noppe, W.; Plieva, F. M.; Galaev, I. Yu.; Vanhoorelbeke, K.; Mattiasson, B.; Deckmyn, H. J. Chromatogr. A 2005, 1101, 79-85. (23) Teilum, M.; Hansson, M. J.; Dainiak, M. B.; Månsson, R.; Surve, S.; Elmer, E.; O ¨ nnerfjord, P.; Mattiasson, G. Anal. Biochem. 2006, 348, 209-221. (24) Kumar, A.; Bansal, V.; Nandakumar, K. S.; Galaev, I. Yu; Roychoudhury, P. K.; Holmdahl, R.; Mattiasson, B. Biotechnol. Bioeng. 2006, 93, 636-646. (25) Plieva, F.; Andersson, J.; Galaev, I. Yu.; Mattiasson, B. J. Sep. Sci. 2004, 27, 828-836. (26) Plieva, F. M.; Karlsson, M.; Aguilar, M.-R.; Gomez, D.; Mikhalovsky, S.; Galaev, I. Yu. Soft Matter 2005, 1, 303-309.

10.1021/la061462e CCC: $37.00 © 2007 American Chemical Society Published on Web 09/15/2006

36 Langmuir, Vol. 23, No. 1, 2007

liquid inside them and are drainage-protected. However, 7075% of the total liquid in the monoliths is easily pressed out mechanically and the compressed monoliths re-swell and adopt their initial shape upon contacting with a new portion of liquid.25-27 These properties of MHs are explained by their structure, namely large (10-100 µm) pores surrounded by thin walls composed of a concentrated polymer phase. The gels of poly-N-isopropylacrylamide (pNIPA) undergo reversible swelling and shrinkage in response to changes in temperature.28 When MHs are produced from pNIPA (pNIPAMH), the increase in temperature results in shrinkage of pNIPAMH hence providing a convenient means of MH compression by internal molecular forces rather than by applying an external mechanical force. To the best of our knowledge, the effect of the deformation of the soft matrix on the retention of specifically bound cells has been addressed only in our previous publication.29 The present manuscript presents a more detailed study of the phenomenon. Materials and Methods Materials. Copper sulfate, EDTA-tetrasodium salt, imidazole, sodium pyruvic acid (pyruvate), β-NADH, lysozyme, Concanavalin A (Type III) (ConA) from CanaValia ensiformis, and ethanolamine were bought from Sigma (St. Louis, USA). Iminodiacetic acid (IDA) was from Fluka (Buchs, Switzerland). HEPES was obtained from BDH Laboratory Supplies (Poole, England). High salt LB-Broth, chloramphenicol, sodium ampicillin, isopropyl-D-thiogalactopyranoside (IPTG), and micro agar were obtained from Duchefa (Haarlem, The Netherlands). Monolithic epoxy-activated polyacrylamide MHs (0.5 mL) (MHs) produced using 5 and 6% (w/v) solutions of comonomers in the reaction mixture were provided by Protista Biotechnology AB (IDEON, Lund, Sweden). Poly(N-isopropylpolyacrylamide)-based MHs (pNIPA-MHs) have been produced as follows: for the preparation of pNIPA-MHs with high cross-linking degree (N,N′methylenebis(acrylamide) (MBAAm)/pNIPA 1/10 mol/mol), the monomers (0.45 g pNIPA, 0.071 g MBAAm and 80 mL AGE) were dissolved in water (final concentration 6 or 10% (w/v)) and degassed afterward for 10 min. Free radical polymerization was initiated by N,N,N,′N′-tetramethylethylenediamine (0.9% (w/v)) and ammonium persulfate (0.9% (w/v)). The reaction mixture (cooled in ice bath till 6-8 °C), 0.5 mL, was poured into glass columns (i.d. 7 mm) and was frozen at -20 °C for 16 h. After washing with water the pNIPA cryogels were stored at 4 °C. For the preparation of pNIPA-MHs with low cross-linking degree (MBAAm/ pNIPA 1/100 mol/mol), the monomers (0.52 g NIPAAm, 0.0071 g MBAAm and 80 mL AGE) were dissolved in water till a final concentration of 6 or 10% (w/v). Further treatment was performed as described above. Recombinant strain of E. coli TG1 cells expressing a thermostable lactate dehydrogenase (from thermophilic B. stearothermophilus) carrying a tag of six histidine residues (His6-LDH) was a gift from Professor Leif Bu¨low, Department of Pure and Applied Biochemistry, Lund University. E. coli K12 strain pop 6510 (thr, leu, tonB, thi, lacY1, recA, dex5, metA, supE, and dex5) with plasmid pLH2 encoding the hybrid LamB-His (two 6 x His) monomers (His6-E. coli)30 was generously provided by Professor Victor de Lorenzo, Centro Nacional de Biotecnologia-CSIC, Campus de Cantoblanco, Madrid. Microgels prepared from cross-linked copolymer of N-isopropylacrylamide and vinylimidazole were prepared according to Dainiak et al.29 A 33 kD model protein expressed as inclusion bodies in recombinant E. coli and IgG anti A15 and anti B17 against (27) Galaev, I. Yu; Dainiak, M. B.; Plieva, F. M.; Hatti-Kaul, R.; Mattiasson, B. J. Chromatogr. A 2005, 1065, 169-175. (28) Schild, H. G. Prog. Polym. Sci. 1992, 17, 163-249. (29) Dainiak, M. B.; Kumar, A.; Galaev, I. Yu; Mattiasson, B. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 849-854. (30) Sousa, C.; Cebolla, A.; deLorenzo, V. Nat. Biotechnol. 1996, 14, 10171020.

GalaeV et al. the 15 amino acids on the N-terminus and 17 amino acids on the C-terminus of the 33 kD target protein, respectively 31 were generously provided by Dr. Helene Sundstro¨m, Department of Biotechnology, Royal Institute of Technology, Stockholm, Sweden and Dr. Gunnar Ho¨rnsten, SIK, IDEON, Lund, Sweden. Bakers yeast in the form of pressed blocks was from a local supplier. Estimation of uniaxial compression of MHs was conducted as described in ref 29. Briefly, a glass tube (20 × 7 mm i.d.) containing a MH (12.5 × 7.1 mm) was placed on a digital balance (1264 MP; Sartorius). A load was applied by placing a glass rod vertically on top of the cryogel monolith. The compression modulus of cryogel monoliths was estimated by using the equation E ) F/A x (∆l/l)-1 where E is the elastic modulus, F is the applied force, A is the cross-sectional area of the test sample, l is the initial length of the test sample, and ∆l is the change in length under the compressive force. The ∆l was proportional to F. Preparation of Immobilized Metal Affinity Chromatography (IMAC) MHs. Epoxy-activated MHs (16 plugs) were washed with 0.5 M Na2CO3, equilibrated with 0.5 M IDA in 1.0 M Na2CO3, pH 10.0, and incubated in 30 mL of this solution for 24 h at room temperature with gentle shaking. IDA-MHs were placed into the wells of a 96-well (each well has inner dimensions 3.2 × 0.7 cm i.d.) plate with drilled holes (0.3 cm diameter) at the bottom of each well and were washed extensively with an excess of water until pH became neutral. Cu(II) and Ni(II) ions were bound to the IDA-MH monoliths by passing 2 mL of 0.25 M CuSO4 or NiCl2 respectively, through each well. Finally, each well was washed with water and IDA-MH monoliths were equilibrated with 20 mM HEPES containing 0.2 M NaCl pH 7.0. Preparation of ConA-MHs. Epoxy-activated MHs were equilibrated with the solution of ConA (2 mg/mL) in 0.05 M carbonate buffer pH 9.0 containing 1 M NaCl, 1 mM CaCl2, and 1 mM MgCl2 and incubated with a fresh solution of ConA for 24 h at room temperature with gentle shaking. Nonreacted epoxy groups were blocked by incubating the MH with 0.1 M ethanolamine in 0.05 M carbonate buffer pH 9.0 containing 1 M NaCl, 1 mM CaCl2, and 1 mM MgCl2 for 2 h at room temperature with gentle shaking. ConA-MH were placed into the wells of a 96-well plate with holes of 0.3 cm diameter drilled at the bottom of each well and were washed with 0.1 M acetate buffer pH 6.5 containing 0.5 M NaCl, 1 mM CaCl2, and 1 mM MgCl2. Preparation of Protein A-MHs. Epoxy-activated MHs were equilibrated with 0.5 M ethylenediamine in 0.2 M Na2CO3 and incubated with a fresh portion of this solution overnight at room temperature with gentle shaking. After washing with water and 0.1 M sodium phosphate buffer, pH 7.2, MHs were equilibrated with glutaraldehyde solution (5% v/v) in the same buffer and incubated with a fresh portion of this solution for 5 h at room temperature with gentle shaking. The derivatized MHs with functional aldehyde groups were equilibrated with protein A solution (2 mg/mL in 0.1 M sodium phosphate buffer, pH 7.2) and incubated with a fresh portion of this solution for 48 h at 4 °C with gentle shaking. Finally, MHs were incubated with 30 mL of freshly prepared NaBH4 solution (0.1 M in sodium carbonate buffer, pH 9.2) for 3 h with gentle shaking to reduce Schiff base formed between the protein and the aldehydecontaining matrix. Determination of Ligand Density on Affinity MHs. The amount of immobilized IDA was determined by assaying the amount of bound Ni(II) ions as follows: 2 mL of 25 mM NiCl2 and 2 mL of deionized water were passed through IDA-MH monoliths. The amounts of Ni(II) ions in the applied NiCl2 solution and in the effluent were determined using Dr. Lange kit (LCK 337, Dr. Bruno Lange, GmbH, Du¨sseldorf, Germany) The samples were diluted 1:1000 with deionized water prior to the analysis. The amount of bound Ni(II) was calculated as a difference between the amounts of applied and nonbound Ni(II). The amounts of ConA and protein A immobilized on MH monoliths were determined by modified bicinchoninic acid assay according to Kumar et al.20

Elution of Bound Cells Cultivation of Cells and Protein Expression. E coli K12 strain with plasmid pLHb2 encoding the hybrid LamB-His6 monomers (His6-E. coli) and E. coli TG1 strain producing His6-LDH (E. coli TG1) were grown in Luria-Bertani (LB) medium (tryptone 10 g/l, yeast extract 5 g/l, NaCl 10 g/l) supplemented with 30 µg/mL chloramphenicol or 100 µg/mL ampicillin, respectively, at 37 °C in a shaking incubator at 175 rpm. His6-E. coli cells were harvested at middle log phase by centrifugation at 5800g for 5 min. The cell pellet was kept on ice and suspended in 20 mM HEPES and 200 mM NaCl pH 7.0 prior to adsorption tests. The cells were used within 1-2 days after cultivation. Expression of His6-LDH was carried out as follows: when the optical density at 600 nm of E. coli TG1 cell culture (200 mL inoculated with 10 mL of overnight culture) reached 0.7, IPTG and another portion of ampicillin were added to a final concentration of 48 and 100 mg/L, respectively. After 3.5 h the cells were harvested (5800 g for 5 min), re-suspended in 50 mL of 50 mM Tris-HCl pH 7.0, and sonicated. The obtained cell homogenate was divided into small fractions and stored at -20 °C. Enzyme Activity Assay. The assay of activity of lactate dehydrogenase was performed in 0.2 M Tris-HCl buffer pH 7.3 containing 1.0 mM pyruvate and 0.225 mM NADH, by monitoring the absorbance decrease of NADH at 340 nm. One enzyme unit represents the reduction of 1 µmol of pyruvate per minute. Binding and Recovery of Cells and Microgel Particles using Affinity MHs. The following running buffers were used in the adsorption tests: 0.1 M Tris-HCl, 150 mM NaCl, 5 mM CaCl2, and 5 mM MgCl2 pH 7.4 (buffer A) for yeast cells and ConA-MHs and 20 mM HEPES, 0.2 M NaCl pH 7.0 (buffer B) in the IMAC tests. Yeast cells, His6-E. coli cells, and microgel particles were suspended in the appropriate running buffer to an OD600 of 0.855 and OD450 of 0.727 and 0.674, respectively. Aliquotes (200 µL) of suspensions of yeast cells, His6-E. coli cells, and microgel particles were applied to ConA-, Ni(II)-IDA-, and Cu(II)-IDA-MHs, respectively, equilibrated with the corresponding running buffer. After 10-15 min of incubation, MHs were washed with 3.5 mL of the appropriate buffer to remove nonbound particles. The effluents were analyzed by measuring the absorbance at 600 nm in the case of yeast cells and at 450 nm in the case of His6-E. coli cells and microgel particles. The amount of bound nanoparticles was calculated as a difference between the amounts of applied and nonbound nanoparticles. Recovery of captured nanoparticles was carried out in two different ways. As a conventional way, the desorption of nanoparticles from affinity MHs was carried out by passing 1.5 mL of the appropriate eluent: R-D-manno-pyranoside (or glucose) in the case of yeast cells captured on ConA-MH and EDTA (or imidazole) in IMAC tests. The particles were also recovered by compressing the MHs. Compression was done by pressing MHs equilibrated with 0.5 mL of the appropriate eluent using a glass rod. Another portion of the eluent (0.5 mL) was added, and compression was repeated. The eluate was analyzed by measuring the absorbance at 600 or 450 nm. When using pNIPA-based MHs, yeast cells suspension (0.5 mL; 0.750 units OD600) in 0.1 M Tris-HCl, pH 7.4, containing 150 mM NaCl, 5 mM CaCl2 and 5 mM MgCl2, was passed through the columns. The columns were washed thoroughly with the binding buffer and then elution was carried out by passing the buffer at a temperature of 35 °C through the MHs. When the buffer at a temperature of 35 °C was passed through the columns, shrinking of the MHs occurred. Binding and Recovery of Inclusion Bodies using Protein A-MHs. The inclusion body slurry was diluted 8 or 30 times in 50 mM PBS, pH 7.2, and 0.5 mL of diluted slurry was incubated with 40 µL of antibody solution (1.0 mg ml-1) on ice for 15 min. The mixture was centrifuged at 10 000g for 2 min, and the pellet was re-suspended in 0.5 mL of 50 mM PBS and centrifuged once more for 2 min. The pellet was re-suspended in 1 or 4 mL of 50 mM PBS pH 7.2 and the suspension (0.150-0.220 mL) was applied to protein A-MHs equilibrated with 50 mM PBS pH 7.2. The MHs were incubated with the treated inclusion body suspension for 15 min and then washed with 2 mL of 50 mM PBS pH 7.2 to remove nonbound

Langmuir, Vol. 23, No. 1, 2007 37 material. The recovery of bound inclusion bodies was carried out by compressing protein A-MHs. Binding and Recovery of His6-LDH using Cu(II)-IDA-MHs. E. coli cell homogenate (0.2 mL) containing His6-LDH was applied to Cu(II)-IDA MHs equilibrated with buffer B (see above). After incubation for 10 min, Cu(II)-IDA-MHs were washed with 3.5 mL of buffer B to remove nonbound protein and cell debris. Bound His6-LDH was eluted with EDTA or imidazole in buffer B. Compression of Cu(II)-IDA-MHs was carried out as described above. His6-LDH was detected in the effluents and eluates by measuring enzyme specific activity. Recovery was determined as the percent of bound activity units that eluted in the purified pool. Swelling Measurements. Swelling dependence on temperature was measured at temperatures between 4 and 40 °C. The MHs were incubated in PBS, pH 7.4, for at least 10 h to achieve the swelling equilibrium at the given temperature. Swollen MHs were quickly removed from the media, blotted with filter paper to remove any excess surface moisture and weighted. This procedure was performed at least in triplicate. Swelling ratio was defined as % S ) (Ws - W0)/W0, where Ws is the weight of the swollen sample and W0 is the weight of the dried sample. The pulsed behavior in swelling of this MH was measured by alternating immersion of MHs into 0.1 M phosphate buffer pH 7.4 at two different temperatures, 19 or 37 °C (below and above Tc) respectively, and weighting them at different time intervals. The swelling ratio was determined as described above. Preparation of MH Samples L for Microscopy. Disks of 5 mm thickness were cut from the MHs. The MH discs were fixed in 2.5% glutaraldehyde in 0.12 M sodium phosphate buffer, pH 7.2, overnight, and postfixed in 1% osmium tetra-oxide for 1 h. Then the samples were dehydrated in ethanol (0, 50, 75, and 99.5%) and critical point dried. The dried samples were coated with gold/palladium (40/60) and examined using a JEOL JSM-5600LV scanning electron microscope.

Results and Discussion Contrary to rigid and relatively hydrophobic materials traditionally studied with respect to cell adhesion, biological tissues are soft, elastic and hydrophilic. Moreover, thanks to their elasticity biological tissues are subjected to pronounced deformation (e.g. muscle contraction-relaxation). MHs present a hydrophilic, highly porous and a highly elastic matrix (Figure 1). High porosity provides an ample solid-liquid interface for specific attachment of bioparticles while the size of the pores allows for unhindered convectional transport of bioparticles within the MH. Even such big particles as mammalian cells (about 20 µm in size) are transported with the flow through MH without any noticeable retention.20 However, when affinity ligands are coupled to MHs, then cell capture can take place. Hence, affinity MHs have been chosen as a model system to study the effect of mechanical deformation of the matrix on the retention of specifically bound bioparticles (bacterial and yeast cells as well as antibody-labeled inclusion bodies) and synthetic particles (microgels of cross-linked poly(N-isopropyl acrylamide-co-Nvinylimidazole). Intuitively, it was obvious that the matrix deformation would have no effect on the release of affinity bound macromolecules, even when these macromolecules have more than one binding site like recombinant His6-tagged tetrameric lactate dehydrogenase from Bacillus stearothermophilus.32 Indeed, the MH deformation had no effect on the release of lactate dehydrogenase bound to Cu(II)-loaded MH bearing iminodiacetate ligands (Cu(II)-IDA-MH) either in the absence or in the presence of a specific eluent (Figure 2). Most probably, lactate dehydrogenase despite (31) Ahlqvist, J.; Dainiak, M. B.; Kumar, A.; Ho¨rnsten, E. G.; Galaev, I. Yu; Mattiasson, B. Anal. Biochem. 2006, 354, 229-237. (32) Bernaudat, F.; Bu¨low, L. J. Chromatogr. A 2005, 1066, 219-224.

38 Langmuir, Vol. 23, No. 1, 2007

GalaeV et al.

Figure 3. Scanning electron micrograph of cross-linked poly(Nisopropyl acrylamide-co-N-vinylimidazole) microgel particles.

Figure 1. Scanning electron micrographs of “soft” (A) and “dense” (B) MHs prepared using 5 and 6% solutions of comonomers, respectively.

Figure 2. Release of captured recombinant His6-lactate dehydrogenase from Bacillus stearothermophilus by conventional elution and by squeezing Cu(II)-IDA-MHs equilibrated with different concentrations of imidazole. The amount of bound protein was taken as 100%. Experimental conditions: 0.2 mL of E. coli cell homogenate containing His6-lactate dehydrogenase was applied to Cu(II)-IDAMHs equilibrated with 20 mM HEPES, 0.2 M NaCl pH 7.0. (buffer B). After incubation for 10 min, Cu(II)-IDA-MHs were washed with 3.5 mL of buffer B to remove nonbound protein and cell debris. Captured His6-lactate dehydrogenase (30-35 µg protein per MH) was recovered either by passing 1.5 mL of the eluent in buffer B through Cu(II)-IDA-MHs or by squeezing Cu(II)-IDA-MHs equilibrated with 0.5 mL of the eluent in buffer B.

having a few binding sites binds to a single Cu(II)-IDA ligand as the probability of location of two or more Cu(II)-IDA ligands in the position which fits exactly the location of two binding sites on the enzyme molecule is very small. The situation is changed completely when transiting to larger particles with binding sites at the surface. Here the probability of multi-point attachment increases dramatically. As a model of such particle, we have used a microgel of cross-linked poly(N-isopropyl acrylamide-co-N-vinylimidazole) (Figure 3). The size of spherical microgel particles is 350-400 nm, which is somewhat larger, than the size of most virus particles, but smaller than the size of bacterial cells. Due to the imidazole groups of vinyl imidazole comonomer, the microgel particles are able of interacting with Cu(II)-IDA ligands. The suspension of microgel particles was applied on top of the MH monolith, allowed to penetrate in the monolith. No bound microgel particles were eluted with running buffer and elution

Figure 4. Release of captured microgel particles by conventional elution and by squeezing Cu(II)-IDA-MHs equilibrated with different concentrations of EDTA (A) and imidazole (B). The amount of bound microgel particles was taken as 100%. Experimental conditions: 0.2 mL of suspension of microgel particles (OD450 0.67) in 20 mM HEPES, 0.2 M NaCl, pH 7.0 (buffer B) was applied to Cu(II)-IDA-MHs equilibrated with the same buffer. After incubation for 10 min, Cu(II)-IDA-MHs were washed with 3.5 mL of buffer B to remove nonbound or weakly bound particles. Captured particles (0.13-0.14 units OD450 per MH) were recovered either by passing 1.5 mL of the eluent in buffer B through Cu(II)-IDA-MHs or by compressing Cu(II)-IDA-MHs equilibrated with 0.5 mL of the eluent in buffer B.

with a specific eluents, imidazole (up to 0.3 M) or EDTA (up to 50 mM), resulted in releasing of not more than a few percent of the bound microgel particles. However, when MH with bound microgel particles was compressed in the presence of either eluent, up to 60% of microgel particles was released from the MH (Figure 4). As the pulse of the eluent at 20-fold higher flow rate did not succeed in releasing more than 12% of captured microgel particles, one could assume that the primary reason for the release of bound particles is the elastic deformation of MH, rather than a flow generated in the pores during the MH compression. With a size of microgel particle in the micrometer range, one could assume that multisite interactions take place between the microgel particle and the MH. According to theoretical studies, it is unlikely that reasonable concentrations of a soluble monovalent competitor (specific eluent) can displace the binding equilibrium when the number of interactions is >10.33 For

Elution of Bound Cells

Figure 5. 5 Schematic presentation of the mechanism of detachment of captured particle induced by MH deformation.

bioparticles under typical chromatographic conditions (10101012 of ligands and receptors per cm2 and 10-10-10-8 cm2 of contact area) the number of specific binding interactions can be between 1 and 10 000.17 Thus, to detach the bioparticle from the matrix, an external force affecting the entire bioparticle is required to simultaneously disrupt multiple bonds.34 Alternatively, the matrix could be affected by an external force to promote the detachment of bioparticle. An elastic deformation of the matrix could be such a force resulting in changing the distances between the ligands at the interface and hence the development of stresses on the bound particle. As a result of these stresses, the efficiency of the interaction is reduced and some of the bonds are broken. The presence of the specific eluent prevents the re-formation of the broken bonds. As the result of these events, breakage of the existing bonds and impossibility to form new ones, the particle is released from the matrix. It was previously demonstrated by scanning electron microscopy studies that cells captured by affinity MHs are bound to the plain “flat” parts of the pore walls and are not entrapped in “dead flow” zones.35 The possible reasons for the disruption of affinity bonds can be the deformation of the plain surface (Figure 5). The compression of MH hardly affects soluble protein molecules bound to the gel but could have a dramatic effect on the bound bioparticle, which has a distinct solid-liquid interface of its own. The possibility of ligand damage as the result of MH deformation was excluded by showing that the compression did not have a pronounced effect on the binding properties of ConAand Cu(II)-IDA-MHs and on detachment efficiency in the repetitive cycle.29 Thus, the main driving force for the compression-induced detachment of bound particles from the surface is probably the physical dislodging of cells by microscopic deformation of the surface carrying affinity ligands and the removal of dislodged particles by the flow of squeezed out liquid. The presence of specific eluent promotes the detachment by decreasing the equilibrium number of bonds and preventing reattachment of free particles on their way down the column. To demonstrate the generic character of this phenomenon, the release of larger in size particles from affinity MHs was studied using two systems: (i) IgG labeled inclusion bodies captured by protein A-MH (particle size about 1 µm) and (ii) yeast cells captured by ConA-MH (particle size about 8 µm). Inclusion bodies are aggregates of denatured protein produced during the expression in recombinant E. coli. After labeling inclusion bodies with either of two types of antibodies, against 15 or against 17 amino acid residues at the N and C termini ends, respectively (anti-A15-IgG and anti-B17-IgG), it was possible to capture antibody-labeled inclusion bodies on protein A-MH.31 Captured inclusion bodies were rather efficiently released after elastic deformation of protein A-MH under very mild conditions (33) Hubble, J. Immunol. Today 1997, 18, 305. (34) Bell, G. I. Science 1978, 200, 618-627. (35) Arvidsson, P.; Plieva, F. M.; Savina, I. N.; Lozinsky, V. I.; Fexby, S.; Bu¨low, L.; Galaev, I. Yu; Mattiasson, B. J. Chromatogr. A 2002, 977, 27-38.

Langmuir, Vol. 23, No. 1, 2007 39

Figure 6. Release of captured inclusion bodies labeled with IgG anti A15 and anti B17 by squeezing protein A-MHs. The inclusion body slurry was diluted 8 (A, B) or 30 (C) times in 50 mM PBS, pH 7.2 and 0.5 mL of diluted slurry was incubated with 40 µL of antibody solution (1.0 mg mL-1). Different dilutions were used in order to ensure different density of antibodies on the particles surface. Experimental conditions: 0.150-0.220 mL of suspension of labeled inclusion bodies (OD470 3.0-2.1) in 50 mM PBS pH 7.2 was applied to protein A-MHs equilibrated with the same buffer. After incubation for 15 min, protein A-MHs were washed with 3.5 mL of the same buffer to remove nonbound or weakly bound particles. Captured inclusion bodies (0.38-0.39 units OD470 per monolith) were recovered by compressing A-MHs. The amount of bound inclusion bodies was taken as 100%.

at pH 7.0 (Figure 6). It is worthwhile to note that harsh conditions such as pH values around 2, are usually used to dissociate the complex formed between IgG and protein A.29 Hence, the mechanical stress appeared on elastic deformation, destabilized the affinity interactions at the particle-matrix interface. AntiB17-IgG have higher affinity to the antigens on inclusion bodies surface than anti-A15-IgG,31 thus one could expect that the preparations of anti-B17-IgG-labeled inclusion bodies have higher densities of respective IgG at the surface of inclusion bodies. Not surprisingly, anti-B17-IgG-labeled inclusion bodies were released with less efficiency by elastic deformation, than anti-A15-IgG-labeled inclusion bodies. Moreover, the increase in the density of labeling by using higher anti-B17-IgG concentration decreased the efficiency of deformation-induced release of particles (Figure 6). Two types of MHs, “soft” and “dense” MHs prepared using 5 and 6% solutions of comonomers, respectively, have been studied. “Dense” (6%) MHs have larger pores with thicker walls as compared with “soft” (5%) MHs (Figure 1). The difference in the structure results in different elastic modules of 0.016 and 0.065 MPa for “soft” and “dense” MHs, respectively. The 2.5fold difference in elasticity had no effect on the efficiency of the release of bound microgels (0.4 µm) and inclusion bodies (∼1 µm). However, the release of larger particles such as Escherichia coli cells (1 × 3 µm) and especially yeast cells (8 µm) was highly dependent on the MH elasticity (Figure 7). The more elastic the MH, the more efficient is the detachment of microbial cells upon the deformation. One should also keep in mind that, contrary to microgels and inclusion bodies, microbial cells are living species which are continuously changing their colloidal properties in response to environmental changes. For example, an incubation of yeast cells within ConA-MH was required for efficient capture of cells, whereas the amount of captured microgel particles bound to Cu(II)-IDA-MH was independent of the time of contact between the applied particles and the adsorbent. In both cases, the increase in the amount of yeast cells or microgel particles applied on ConA- or Cu(II)-IDA-MHs did not lead to an increased binding, and the excess of applied particles was washed with the flow (data not shown). In the above cases, the mechanical stress was caused by external force. However, it is possible to induce the mechanical deformation of elastic MH provided the latter is prepared from a smart

40 Langmuir, Vol. 23, No. 1, 2007

GalaeV et al.

Figure 9. Conventional elution with 0.3 M R-D-manno-pyranoside in the running buffer (0.1 M Tris-HCl with 150 mM NaCl, 5 mM CaCl2, and 5 mM MgCl2, pH 7.4) and temperature induced detachment of yeast cells bound to ConA-pNIPA-macroporous hydrogels, produced from polymerization feed with total monomer concentrations 6% and MBAAm(cross-linker)/NIPA ratios 1/10 and 1/100 mol/mol, respectively.

Figure 7. Release of captured yeast cells by conventional elution and by squeezing 5% (A) and 6% (B) ConA-MHs (MHs were prepared using 5 and 6% solution of comonomers, respectively). The amount of bound cells was taken as 100%. Experimental conditions: 0.2 mL of suspension of yeast cells (OD600 0.855) in 0.1 M TRIS, 150 mM NaCl, 5 mM CaCl2, and 5 mM MgCl2 pH 7.4 (buffer A) was applied to ConA-MHs equilibrated with the same buffer. After incubation for 15 min, ConA-MHs were washed with 3.5 mL of buffer A to remove nonbound cells. Captured cells (0.070.10 and 0.12-0.16 units OD600 on 5% and 6% MHs, respectively) were recovered either by passing 1.5 mL of the eluent in buffer A through ConA-MHs or by compressing ConA-MHs equilibrated with 0.5 mL of the eluent in buffer A.

with a relatively sharp transition around 30 °C (Figure 8a). The lower the total monomer concentration and the cross-linking degree, the higher was the change in swelling; hence, the larger amplitude of mechanical deformation occurred when the temperature was changed from 19 to 37 °C and backward (Figure 8b). The mechanical deformation of the pNIPA-MH originated due to temperature induced shrinkage of MHs happened to be sufficient to facilitate the release of bound bioparticles, namely yeast cells captured via concanavalin A (ConA) covalently coupled to the pNIPA-MH (Figure 9). When conventional elution with 0.3 M R-D-manno-pyranoside in the running buffer (0.1 M Tris-HCl with 150 mM NaCl, 5 mM CaCl2, and 5 mM MgCl2, pH 7.4) released only a minor amount of bound cells, the temperature induced detachment improved cell release to 16 and 37% for ConA-pNIPA-MHs produced from polymerization feed with total monomer concentrations 6% and MBAAm(crosslinker)/NIPA ratios 1/10 and 1/100 mol/mol, respectively. Not surprisingly, the higher release of captured cells was observed for the NIPA-MH with lower cross-linking density as the amplitude of changes was higher for this MH and hence one could expect higher elastic deformation of pore walls inside the MH monolith and as a result a stronger effect on the bound yeast cells. It is important to note that elasticity of MHs ensures retained viability even in the case of detachment of fragile mammalian cells.29

Conclusion

Figure 8. Temperature (A) and time (B) dependence of swelling of pNIPA-MHs produced under semi-frozen conditions from polymerization feeds with total monomer concentrations 6 and 10% and MBAAm(cross-linker)/NIPA ratios 1/10 and 1/100 mol/mol, respectively.

polymer like poly-N-isopropylacrylamide (pNIPA). In aqueous solutions pNIPA undergoes hydrophobic aggregation followed by the transition from soluble to insoluble state when the temperature is increased above a critical temperature of 32 °C.28 The swelling of MHs produced by cross-linking polymerization of NIPAM under semi-frozen conditions depends on temperature

The results presented above illustrate interactions of cells with an elastic affinity surface. Studies of affinity-mediated adhesion and recovery of bioparticles under various conditions using affinity MH-monoliths can serve as a model mimicking interactions of bacteria, viruses, macrophages etc. in biological systems. The release of specifically captured cells by mechanical compression of elastic MH monoliths is a novel efficient detachment strategy. Elasticity of MH monoliths ensures efficient recovery of captured cells with retained viability. This detachment strategy along with continuous porous structure makes these adsorbents very attractive for application in affinity cell separation. Acknowledgment. This work was financially supported by Protista International AB (Bjuv, Sweden), The Swedish Foundation for Strategic Research, The Swedish Foundation for International Cooperation in Research and Higher Education (STINT, IG2003-2089), The Swedish Institute (Visby Program, Projects 2886/2002 and 01211/2004), and Royal Swedish Academy of Sciences. LA061462E