Modulation of hybridoma cell growth and antibody production by coating cell culture material with extracellular matrix proteins

Biochemical Engineering Journal 35 (2007) 301–308

Modulation of hybridoma cell growth and antibody production by coating cell culture material with extracellular matrix proteins K. Heilmann a , T. Groth b,c , M. Schossig c , A. Lendlein c , B. Micheel a,∗ a

Department of Biotechnology, Institute of Biochemistry and Biology, University of Potsdam, Karl-Liebknecht-Strasse 24-25, D-14476 Golm, Germany b Biomedical Materials Group, Institute of Pharmaceutical Technology and Biopharmacy, University of Halle-Wittenberg, D-06099 Halle (Saale), Germany c GKSS Research Centre, Institute of Polymer Research, Kantstrasse 55, D-14513 Teltow, Germany Received 18 May 2006; received in revised form 21 December 2006; accepted 16 January 2007

Abstract The influence of coating polystyrene tissue culture plates with different proteins on murine hybridoma cell growth and antibody production was investigated. Fibronectin, collagen I, bovine serum albumin and laminin were used to coat NUNC® and COSTAR® cell culture plates. Cell number and antibody concentration in culture fluids were quantified as indicators for cell viability, proliferation and productivity. Adhesive behaviour, morphology, expression of surface receptors of hybridoma cells and the presence of tyrosine-phosphorylated proteins in cell lysates were characterized by cell adhesion experiments, microscopy, flow cytometry and Western Blot analysis. It was shown that coatings with fibronectin (0.2 ␮g/ml) lead to a substantial improvement of cell growth by 50–70% and an increase of monoclonal antibody production by 100–120%. Collagen I coatings showed an improvement in cell growth by 30–70% and by 60% for the production of monoclonal antibodies. Coatings with BSA and laminin had minor effects on these parameters. It was found that the hybridoma cell lines used in this study did not express the ␣2 -chain of the ␣2 ␤1 -integrin, which is responsible for binding to collagen and laminin. However, the presence of ␤1 -integrin on the cell surface was shown, which should enable hybridoma cells to bind fibronectin. We propose, therefore, that fibronectin adsorption to cell culture materials may be a promising approach to enhance the production of monoclonal antibodies by cultivated hybridoma cells. © 2007 Elsevier B.V. All rights reserved. Keywords: Hybridoma cells; Tissue culture plates; Cell growth; Antibody production; Extracellular matrix; Fibronectin adsorption

1. Introduction Morphology, growth and function of most cells are strongly dependent on the adhesive interactions with neighbouring cells and extracellular matrices [1]. Therefore adhesive interactions of cells with foreign substrata used in biomedical and biotechnological applications may have a significant impact on their growth and function [2]. Since, protein adsorption occurs rapidly if materials contact protein fluids, such as serum-containing culture media, the cellular behaviour may be controlled by the composition of the adsorption layer [3]. In particular, the adsorption of adhesive proteins, such as fibronectin as a component

∗

Corresponding author. Tel.: +49 331 977 5242; fax: +49 331 977 5061. E-mail address: [email protected] (B. Micheel).

1369-703X/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2007.01.035

of the extracellular matrix on biomaterials, has been shown to promote adhesion, growth and function of cells greatly [4]. Extravasation, homing and programming of B lymphocytes are processes that require adhesive interactions with ECM components and different cell types [5]. Integrins are involved in these processes controlling cellmatrix and cell–cell adhesion by specific ligand binding [6]. Integrins represent a family of ␣␤-heterodimeric adhesion receptors and are constitutively expressed on lymphocytes or can be activated by antigens, chemokines and phorbol esters [7]. The ligation of integrins by its specific ligand induces a signal transfer called outside-in-signalling, which triggers signal cascades involved in expression of genes important for survival, growth and differentiation of cells [8]. The major integrins on B cells are lymphocyte function-associated-antigen1 (LFA-1 or ␣L ␤2 -integrin), very late antigen-4 (VLA-4 or

302

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308

␣4 ␤1 -integrin) and ␣4 ␤7 -integrin. LFA-1 binds to intercellular adhesion molecule-1 (ICAM-1 or CD54). For example, ␣4 ␤7 -integrin binds to fibronectin which is expressed on the luminal surface of endothelial as one of the first steps during extravasation of lymphocytes [9]. The subsequent migration of lymphocytes to lymphoid organs is crucial for encountering antigens and differentiation of cells [10]. Differentiation of B lymphocytes results in the secretion of monoconal antibodies (mAb) to label invading pathogens and guide the activity of other leukocytes. Such antibodies can now be produced in large quantities by cultivating hybridoma cells in vitro. Since hybridoma cells are fusion products of B lymphocytes (plasma cells) and myeloma cells the knowledge about their adhesive phenotype and expression of surface receptors may be useful to control their growth and antibody production. Myeloma cells represent malignant plasma cells located in the bone marrow [11]. Adhesion molecules expressed by myeloma cells are CD44 (HCAM), CD49e (VLA-5 or ␣5 -integrin), CD50/54 (ICAM1) and CD56 [12]. While plasma cells may circulate in the blood and lymph vessels, myeloma cells remain located in the bone marrow. Therefore, a fusion of both cell types may combine different adhesive behaviour and functions. It should be noted that a fusion of a plasma cell with a myeloma cell may not simply result in a receptor repertoire of both cell types. Some chromosomes may get lost after fusion and the expression of surface integrins may also depend on specific extracellular signals [13]. Because of their growth in suspension, hybridoma cells do not require adhesion to substrata to survive and to produce antibodies [14]. This also fits with the observation that plasma cells, which represent differentiated B cells, do not require adhesion to extracellular matrices (ECM) or intercellular adhesion to survive and function [15]. However, we have recently shown that adhesion of hybridoma cells to certain polymer membranes may result in a stimulation of their growth and in an inhibition of antibody production [16]. Similar results were also obtained by Luo and Yang [17], who showed an inverse relation between adhesion and specific mAb production of hybridoma cells. On the other hand, our previous investigation has also demonstrated that a membrane provoking moderate adhesion and growth of hybridoma cells stimulated antibody secretion to an extend that greatly exceeded that of suspended cells [16]. Since polymer membranes may adsorb different quantities of proteins dependent on their chemical composition [18], we were interested to know whether proteins, which can be adsorbed from tissue culture media or represent components of the ECM have an impact on behaviour of hybridoma cells. Therefore, we studied the effect of different proteins adsorbed to two different types of polystyrene tissue culture material (TCPS) on hybridoma growth and antibody production. The proteins fibronectin, laminin and collagen I were selected, as components of the extracellular matrix, while bovine serum albumin (BSA) was chosen as a major component of serum-containing culture media. A major finding of this study was that coating TCPS with low quantities of fibronectin leads to a substantial improvement of hybridoma growth and monoclonal antibody production.

2. Materials and methods 2.1. Cell culture The hybridoma cell line H10 (mouse IgG2b, anti-rat-Fc␧RI), kindly provided by I.Pecht, Weizman Institut of Science, Rehovot, Israel [19], was grown in RPMI 1640 complete medium (Biochrom AG, Berlin, Germany) supplemented with 2 mM l-glutamine (Biochrom AG, Berlin, Germany), 50 ␮M 2-mercaptoethanol (Roth, Karlsruhe, Germany), 10% heatinactivated fetal calf serum (Biochrom AG, Berlin, Germany) at 37 ◦ C and 5% CO2 . All cell culture experiments were performed using this medium. 2.2. Cell growth experiments We compared NUNC® and COSTAR® cell culture plates, which were coated with protein solutions of fibronectin (0.2–20 ␮g/ml), collagen I (1–80 ␮g/ml), BSA (10–800 ␮g/ml) and laminin (0.4–20 ␮g/ml) prior to cell culture experiments. Uncoated TCPS from the same manufacturers were used as controls. Dilutions of protein solutions were prepared with phosphate-buffered saline without calcium and magnesium (PBS, 150 mM NaCl, 5.8 mM NaH2 PO4 , 5.8 mM Na2 HPO4 , pH = 7.4). After coating the plates at 37 ◦ C for 2 h, the wells were washed with PBS and the cells were placed in the wells (100 ␮l/well) with a starting density of 4 × 104 cells/ml and incubated at 37 ◦ C and 5% CO2 . The cells were completely harvested after 3 days. The cell numbers were quantified after complete cell lysis with 1% Triton X-100 using a cytotoxicity detection kit based on the release of lactate dehydrogenase (LDH, Roche Applied Science, Penzberg, Germany). The activity of the cytosolic enzyme LDH can be used to estimate the amount of cells [20]. Standard curves were established by counting cells and comparing the numbers with LDH activity. Each experiment was repeated eight times. Mean values and standard deviations were calculated. BSA, collagen I (from rat tail) and laminin (from human placenta) were purchased from Sigma. Fibronectin (from human plasma) was supplied from Roche Diagnostics, Mannheim, Germany. 2.3. Quantification of antibody production To determine the concentration of monoclonal antibodies (mAb) produced by the hybridoma cells, 96-well plates (Greiner bio-one GmbH, Frickenhausen, Germany) were coated with purified goat-anti-mouse-Ig antibodies (produced in our laboratory—50 ␮l/well at a concentration of 8 ␮g/ml in PBS) for 2 h at room temperature (RT) or overnight at 4 ◦ C. The wells were washed and blocked with PBS containing 5% newborn calf serum (NCS, Biochrom AG, Berlin, Germany) for 30 min at RT. After washing, 50 ␮l cell-free culture supernatant was added per well and incubated at RT for 1 h. Mouse antibody binding was revealed by incubation with a peroxidaseconjugated anti-mouse-Ig antibody (dilution 1:5000, Sigma) for 1 h at RT. After each incubation step the wells were washed

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308

three times under running water. After a final washing step 50 ␮l of substrate solution was added to the wells. One milligram per milliliter o-phenylendiamine and 0.01% H2 O2 in 50 mM sodium citrate/100 mM NaH2 PO4 at a pH of 5.0 was used. After 20 min the reaction was stopped by the addition of 100 ␮l/well stop solution (1 M H2 SO4 , 50 mM Na2 SO3 ). The absorbance was measured at 492 nm with a reference wavelength of 630 nm using an ELISA-Reader (Tecan easy-Win, Spectrafluor Plus, Austria). These experiments were also performed eight times and mean values and standard deviations were calculated. 2.4. Inhibition of cell adhesion by RGD peptide These experiments were performed with the RGD peptide (Gly-Arg-Gly-Asp-Thr-Pro, Calbiochem, Germany) to inhibit specifically the binding of RGD-recognizing integrins to adsorbed adhesive proteins, particularly to fibronectin. Therefore, TCPS (96-well plates, COSTAR® ) were coated with fibronectin (3, 5 and 10 ␮g/ml) diluted in PBS. The plates were incubated at 4 ◦ C for 24 h. The protein solutions were removed and the wells were saturated with 50 ␮l/well PBS containing 3% BSA at 37 ◦ C for 2 h to block non-specific adhesion of cells to the substratum. Then, the plates were washed with cold PBS (50 ␮l/well). Hybridoma cells were harvested and adjusted to a final concentration of 1 × 106 cells/ml. Two hundred microliters of cell suspension were combined with 300 ␮l of RGD solution in different concentrations. The mixture was agitated at 4 ◦ C for 2 h to allow the integrins on the cell surface to bind the RGD peptides [21]. The prepared cell suspensions were then added to the wells (50 ␮l/well) and incubated at 37 ◦ C for 2 h. After that the supernatant was removed and the plates were washed gently with 50 ␮l/well cold PBS. Adhering cells were fixed with 50 ␮l/well methanol for 15 min. Methanol was removed and the plates were dried for 30 min at RT. The fixed cells were stained with 0.07% crystal violet in PBS (100 ␮l/well) for 5 min. The plates were then washed twice with 100 ␮l/well PBS and agitated with 50 ␮l/well 1% sodium dodecyl sulfate (SDS) for 10 min. The absorbance was measured at 562 nm using an ELISA-Reader (Tecan easy-Win, Spectrafluor Plus, Austria). The experiments were performed in quadruplicates. Mean values and standard deviations were determined. 2.5. Detection of tyrosine-phosphorylated proteins To learn if adhesion of hybridoma cells to fibronectin coated TCPS generates signals that can be observed in adhesiondependent cells, cell lysates of H10 cells were investigated for tyrosine-phosphorylated proteins. Therefore, TCPS (6-well plates, COSTAR® ) were coated with different concentrations of fibronectin (1, 3, 5 and 10 ␮g/ml; 2 ml/well; diluted in PBS). The plates were incubated at 37 ◦ C for 2 h. After that the wells were washed with PBS. Then the cells were placed in the wells (2 ml/well) at a starting density of 7 × 105 cells/ml and incubated at 37 ◦ C and 5% CO2 for 24 h. After that the plates were centrifuged (200 × g, 4 ◦ C, 5 min) and the cells were washed twice with icecold PBS (500 ␮l/well).

303

Thereafter, the cells were lysed in 500 ␮l/well lysis buffer (50 mM tris HCl, pH 7.4, 20% glycerol, 0.1 mM dithiothreitol, 1 mM sodium orthovanadate, protease inhibitor cocktail) at 4 ◦ C for 20 min. During the incubation period, tightly adhering cells were additionally removed from the bottom of the wells using a cell scraper (TPP AG, Trasadingen, Switzerland). The plates were then centrifuged (200 × g, 4 ◦ C, 5 min) to remove cell debris and DNA. Hundred microliters of the cell lysates were used to determine the protein content with a DC Protein Assay (BioRad, Munich, Germany). Samples (30 ␮g protein per slot) were boiled for 5 min in Roti-Load-buffer (Roth GmbH & Co., Karlsruhe, Germany) and were loaded on a SDS PAGE. SDS PAGE was carried out under reducing conditions on 1 mm thick polyacrylamide gels (12.5%) for 30 min at 50 V and after that for 1 h at 20 mA. The samples were transferred to a nitrocellulose membrane for the detection of tyrosine-phosphorylated proteins. The primary antibody (murine anti-phosphotyrosine-antibody PY20, Biomol GmbH, Hamburg, Germany) was used at a concentration of 1 ␮g/ml. The detection of tyrosine-phosphorylated proteins were carried out with the “Western Breeze Chemiluminescent Western Blot Immunodetection Kit” according to the instructions of the producer (Invitrogen GmbH, Karlsruhe, Germany). 2.6. Scanning electron microscopy of hybridoma cells Hybridoma cells were grown on plain, fibronectin (0.2 ␮g/ml) and collagen I (60 ␮g/ml) coated TCPS discs (Thermanox, Nunc GmbH & Co.KG, Wiesbaden, Germany) by seeding 1 × 104 cells/well in 24-well culture plates. One milliliter of cell culture medium was added to each well. Cells were cultivated at 37 ◦ C and 5% CO2 for 3 days. Then, the medium was removed and 1 ml/well 4% glutaraldehyde diluted in PBS was added for 20 min at RT. The TCPS discs were washed for 30 min with distilled water. Cells were then treated with a 1% osmiumtetroxid solution diluted in PBS for 20 min. After fixation, the samples were washed several times with distilled water to remove traces of fixatives and buffer salts. The following dehydration procedure was carried out with increasing ethanol solutions starting with 30% up to 100% for 15 min per step. Then the samples were stored in 100% ethanol at 4 ◦ C and prepared by critical point drying as described elsewhere [22]. Sputtering of samples was performed with gold–palladium. The images were taken using a LEO Gemini field emission scanning electron microscope, type 1550 VP (Carl Zeiss, Jena, Germany). 2.7. Flow cytometric analysis Hybridoma cells (1 × 105 in 100 ␮l PBS) were incubated with CD11a-, CD18-, CD29- and CD72-specific rat-anti-mouse antibodies and a FITC-conjugated goat-anti-rat antibody as secondary antibody or FITC-conjugated antibodies (CD49b-specific hamster-anti-rat antibody, CD69-specific ratanti-mouse antibody) to label cell surface antigens. Analysis was performed twice by a Cytomics FC 500 CXP cytometer (Beckman Coulter GmbH, Germany).

304

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308

2.8. Statistical analysis Statistic evaluation of the experimental data was done with a Student’s t-test calculator software from GraphPad Inc. (San Diego, CA, USA). 3. Results 3.1. Effect of protein coating of TCPS on hybridoma growth and antibody production In these experiments different protein concentrations of fibronectin, collagen I, BSA and laminin were used for coating. The tests were performed with 96-well plates from both Nunclon (NUNC® ) and Corning (COSTAR® ). Cell number and antibody concentration was determined after 3 days of cultivation to avoid the change of media. In general, an enhanced growth of hybridoma cells was detected using plates from both manufacturers after protein coating. The strongest stimulation of cell growth and antibody secretion in plates from both manufacturers was observed using low fibronectin and collagen I coatings (Figs. 1 and 2). A concentration of 0.2 ␮g fibronectin per ml coating solution resulted e.g. in an improvement by 72.2% growth on NUNC® plates (Fig. 1B) but only by 48.5% growth on COSTAR® plates (Fig. 1A). That means cell growth was in general higher on NUNC® plates (Fig. 1B). A further observation was that an

Fig. 2. Production of monoclonal antibodies by H10 hybridoma cells on precoated COSTAR® (A) and NUNC® (B) cell culture plates. The plates were precoated with the same protein concentrations as described in Fig. 1 and the same selected concentrations of fibronectin, collagen I, BSA and laminin were presented. Hybridoma cells (4 × 104 ml−1 ) were cultivated in complete medium for 72 h. The concentration of monoclonal antibodies was estimated by a sandwich ELISA (mean ± S.D., n = 8, p ≤ 0.0001).

increase in the concentration of coating proteins caused in most cases a decrease in cell growth. An increase of monoclonal antibody production was however observed only after coating the plates with fibronectin and collagen I (Fig. 2). It was found that hybridoma cells cultured on coated NUNC® plates (Fig. 2B) produced lower antibody quantities if the cell growth was stimulated by more than 50%. A strong decrease of antibody secretion was also observed after increasing coating concentrations of fibronectin. In contrast, hybridoma cells cultured on fibronectin coated COSTAR® plates (Fig. 2A) expressed higher antibody concentrations in all cases. But BSA and laminin did not show such high promoting effects. 3.2. Inhibition of cell adhesion by RGD peptides

Fig. 1. Growth of H10 hybridoma cells on precoated COSTAR® (A) and NUNC® (B) cell culture plates. Plates were preadsorbed with fibronectin (0.2–20 ␮g/ml), collagen I (1–80 ␮g/ml), BSA (10–800 ␮g/ml) and laminin (0.4–20 ␮g/ml) at 37 ◦ C for 2 h. Selected concentrations for fibronectin (0.2, 1, 5, 20 ␮g/ml), collagen I (1, 5, 60 ␮g/ml), BSA (100, 800 ␮g/ml) and laminin (0.4, 1, 5, 20 ␮g/ml) were shown in figure. TCPS was used as uncoated control. Hybridoma cells (4 × 104 ml−1 ) were cultivated in complete medium for 72 h. The cell number was estimated with the LDH assay (mean ± S.D., n = 8, p ≤ 0.0001).

Since the strongest effects on cell growth and antibody production were exerted by fibronectin coatings, the role of integrin-mediated adhesion was investigated by using RGD peptides as inhibitor of cell adhesion. We found on uncoated and fibronectin coated COSTAR® plates a decrease of cell adhesion after incubation with RGD peptides (Fig. 3). A concentration of 70 ␮g/ml RGD peptides abolished adhesion on plain TCPS almost completely. 3.3. Detection of tyrosine-phosphorylated proteins Tyrosine-phosphorylated proteins are involved in signal transduction processes, which are triggered by clustering of integrins on extracellular matrix proteins, such as fibronectin. Therefore cell lysates of H10 cells were prepared and the presence of

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308

305

of cells was observed (Fig. 5A). On fibronectin coated TCPS discs (Fig. 5B) hybridoma cells showed an increased spreading with formation of pseudopodia in comparison to collagen I coated discs (Fig. 5C). Here we observed that collagen I formed fibres on the substratum and cells adhered in aggregates. 3.5. Flow cytometry

Fig. 3. Inhibition of cell adhesion of H10 hybridoma cells growing on fibronectin coated and uncoated COSTAR® plates. Hybridoma cells were incubated with an RGD peptide at 4 ◦ C for 2 h. After that the cells were seeded at a concentration of 2 × 104 ml−1 cells per well and incubated at 37 ◦ C for 2 h. Wells were washed, adhesional cells were fixed with 100% methanol and stained with 0.07% crystal violett. OD was measured at 560 nm (mean ± S.D., n = 4, p ≤ 0.05).

tyrosine-phosphorylated proteins was determined. As shown in Fig. 4 tyrosine-phosphorylated proteins between 10 and 30 kDa were detectable in the cell lysates (Fig. 4, black arrows). Furthermore the quantities of tyrosine-phosphorylated proteins in H10 cells showed a dependence on the fibronectin coating concentration, e.g. the intensity increased with increasing coating concentration of fibronectin. The heavy and light chain of the antibody secreted by H10 cells were visible at 50 and 30 kDa. 3.4. Scanning electron microscopy Scanning electron microscopy was used to visualize the degree of hybridoma cell spreading and the strength of cellsurface interaction (Fig. 5). On plain TCPS discs no spreading

Fig. 4. Detection of tyrosine-phosphorylated proteins in the cell lysates of H10 hybridoma cells and human fibroblasts (F) after cultivation on fibronectin coated and uncoated TCPS. (A) TCPS, B-F) TCPS coated with fibronectin: (B) 1 ␮g/ml, (C) 3 ␮g/ml, (D) 5 ␮g/ml, (E) 10 ␮g/ml, (F) 10 ␮g/ml. H10 cells and human fibroblasts (7 × 105 cells/ml) were cultivated for 24 h. The cells were lysed and the protein content in the lysates was determined. 30 ␮g protein/slot were analysed with a SDS PAGE and Western Blot. Tyrosine-phosphorylated proteins were detected with a PY20 antibody and verificated with the “Western Breeze Chemiluminescent Western Blot Immunodetection Kit” (Invitrogen GmbH, Karlsruhe, Germany).

To verify the presence of specific integrins that can interact with adsorbed matrix proteins flow cytometry was performed. Antibodies against fibronectin and collagen I receptors were used to demonstrate the presence of these receptors. As shown in Fig. 6 there were no detectable fluorescence signals for CD49b (␣2 -integrin) in comparison to CD11a (␣L -integrin), CD18 (␤2 integrin) and CD29 (␤1 -integrin). Binding to collagen I and laminin is mediated by the ␣2 ␤1 -integrin. The lack of fluorescence signals for CD49b indicates a lack of the ␣2 -chain and, therefore, the lack of the complete receptor. Antibodies against CD69 (VEA) and CD72 (Ly-b2), which are not present at the surface of plasma cells, were used as negative controls in these experiments. 4. Discussion It is generally assumed that hybridoma cells grow adhesionindependent, which is reflected by their growth in suspension culture. However, we have recently shown that growth and function of hybridoma cells were dependent on the composition of polymer membranes, which was related to the adhesion of these cells [16]. Here we extended these findings and showed that growth and antibody secretion of hybridoma cells is related to their adhesion on protein-coated substrata. It was shown that coating of TCPS with different proteins increased cell growth and antibody secretion. Fibronectin had the most pronounced impact on both parameters. It improved cell growth by nearly 50–70% and the production of monoclonal antibodies by 70–120%, but only if minor concentrations were used for coating. Concentrations of only 0.2–0.6 ␮g/ml were able to induce such an effect (Fig. 2). A surprising result was that an increase of protein coating resulted in a reduced cell proliferation and antibody secretion. This indicates that the conformation of the adsorbed proteins is important for their interaction with hybridoma cells. Proteins like fibronectin tend to change their conformation upon adsorption at low coating concentrations [23]. It might be speculated that such conformational changes favour the adhesion of hybridoma cells, while the conformation and orientation of proteins at higher coating concentrations may inhibit cell attachment. Indeed, a clear indication for a specific interaction between the hybridoma cells and adsorbed fibronectin came from the experiments with RGD peptides. Here it was visible that addition of RGD peptides inhibited adhesion of hybridoma cells on fibronectin coated TCPS in a concentration dependent manner. This indicates a specific interaction between the fibronectin receptor or related integrins (␤1 integrins) in hybridoma cells and adsorbed fibronectin. However, this effect was also observed on the control surface (TCPS

306

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308

Fig. 5. Scanning electron microscopy of H10 hybridoma cells on uncoated (D), fibronectin coated (A, B) and collagen I-coated (C) TCPS discs. Cells were seeded on uncoated and coated (0.2 ␮g/ml fibronectin, 60 ␮g/ml collagen I) TCPS discs with a starting density of 1 × 104 cells/well and were cultivated for 72 h in cell culture medium. Images were taken using a LEO Gemini field emission scanning electron microscope, type 1550 VP (Carl Zeiss, Jena, Germany) (bar 20 and 10 ␮m (B)).

without fibronectin pre-adsorption). This suggests also that coadsorption of proteins from serum can make a contribution to the adhesive process observed here. At low coating concentrations co-adsorption of other proteins from serum is possible. So-called attachment factors in serum are fibronectin (at minor concentrations) and particularly vitronectin [24]. Indeed, vitronectin might be a further candidate for the observed effects because it has a high affinity to surfaces and has a small size, which would still enable its binding between the preadsorbed proteins [25]. Moreover, binding of cells via integrins to vitronectin can be inhibited by RGD peptides, too [26]. It has been also reported that lower quantities of adsorbed vitronectin are stronger adhesive for lymphocytes than adsorbed fibronectin [27]. The assumption that binding of hybridoma cells to fibronectin is a regulator of their growth and function is also supported by the results of immunoblotting. Here an increasing concentration of tyrosine-phosphorylated proteins was detectable in cells grown on fibronectin coated surfaces. This indicates a fibronectin-dependent signal transduction in hybridoma cells. Signal transduction via integrins is known to be a regulator of cell spreading, cell growth, differentiation and survival [28].

While coating with fibronectin resulted in the strongest effect, only moderate improvements were achieved by collagen I coating, while pre-adsorption of albumin and laminin had only marginal effects. However, all proteins produced a more or less pronounced increase in cell growth and antibody secretion, which suggests that processes of co-adsorption of proteins from serum as described above play a role. Flow cytometry showed that the ␣2 -chain of the ␣2 ␤1 -integrin important for the binding of collagen I and laminin was not detectable on the surface of hybridoma cells. This result could possibly explain the lower effect of collagen I and laminin coatings on hybridoma cell growth and antibody production. At least, the presence of ␤1 integrin was an additional hint that the fibronectin receptor plays a crucial role for the adhesion of hybridoma cells to substrata. Additional evidence for a different adhesive behaviour of cells came from scanning electron microscopy. Here, the morphology of hybridoma cells on fibronectin coated surfaces was characterized by an increased spreading and the forming of pseudopodia. In contrast, on collagen I coated or plain TCPS no such spreading could be detected. These observations confirmed the results for a more pronounced adhesion if fibronectin coatings were used for cultivation.

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308

307

Fig. 6. Flow cytometry of H10 hybridoma cells stained with antibodies to CD11a, CD18, CD29, CD49b, CD69 and CD72. Analysis was performed twice by Cytomics FC 500 CXP cytometer (Beckman Coulter GmbH, Krefeld, Germany).

The cultivation in NUNC® plates compared to COSTAR® plates resulted generally in higher cell numbers but lower yields of monoclonal antibodies. The difference in both types of plates may be dependent on the chemical composition of the plate surface. Normally, TCPS is modified by physical plasma treatment to become more wettable. Depending on the type of plasma, such as oxygen- or nitrogen-plasma, a different chemical composition is obtained. In this context it was observed that TCPS having nitrogen-containing functional groups binds preferably fibronectin while oxygen-containing groups adsorb higher quantities of vitronectin [29]. Since, we do not know the chemical composition of the plates we can here only speculate that the observed differences are due to a different chemical composition of both plates. Summarizing our results, low quantities of fibronectin seem to be an efficient surface coating, to support and improve the growth and productivity of the hybridoma cells. With the highest improvement by 120% compared to uncoated polystyrene, this treatment could be a useful approach for the industrial largescale production of monoclonal antibodies. Acknowledgment This work was supported by GKSS Research Centre Ph.D. programme to support cooperation with the University of Potsdam (Grant no. 6.T9.00.G01-HS1).

References [1] A.Q. Rafi-Janajreh, P.S. Nagarkatti, M. Nagarkatti, Role of CD44 in CTL and NK cell activity, Front. Biosci. 3 (1998) 665–671. [2] M. Maeda, M. Kimura, S. Inoue, K. Kataoka, T. Okano, Y. Sakurai, Adhesion behaviour of rat lymphocyte subpopulations (B cell and T cell) on the surface of polystyrene/polypeptide graft copolymer, J. Biomater. Res. 20 (1) (1986) 25–30. [3] S.G. Olenych, M.D. Moussallem, D.S. Salloum, J.B. Schlenoff, T.C.S. Keller, Fibronectin and cell attachment to cell and protein resistant polyelectrolyte surfaces, Biomacromolecules 6 (2005) 3252–3258. [4] G. Altankov, V. Thom, T. Groth, K. Jankova, G. Jonsson, M. Ulbricht, Modulating the biocompatibility of polymer surfaces with poly(ethylene glycol): effect of fibronectin, J. Biomed. Mater. Res. 52 (2000) 219–230. [5] T.A. Springer, Adhesion receptors of the immune system, Nature 346 (1990) 425–433. [6] R.O. Hynes, Integrins: bidirectional, allosteric signaling machines, Cell 110 (2002) 673–687. [7] S.J. McLeod, A.J. Shum, R.L. Lee, F. Takei, M.R. Gold, The Rap GTPases regulate integrin-mediated adhesion, cell spreading, actin polymerization and Pyk2 tyrosine phosphorylation in B lymphocytes, J. Biol. Chem. 279 (2004) 12009–12019. [8] A.E. Aplin, A. Howe, S.K. Alahari, R.L. Juliano, Signal transduction and signal modulation by cell adhesion receptor: The role of integrins, cadherins, immunglobulin-adhesion molecules and selectins, Pharmacol. Rev. 50 (1998) 198–255. [9] T.M. Carlos, J.M. Harlan, Leukocyte-endothelial adhesion molecules, Blood 84 (7) (1994) 2068–2101. [10] C.E. Leuker, M. Labow, W. Muller, N. Wagner, Neonatally induced inactivation of the vascular cell adhesion molecule 1 gene impairs B cell

308

[11]

[12] [13]

[14] [15] [16]

[17]

[18]

[19]

K. Heilmann et al. / Biochemical Engineering Journal 35 (2007) 301–308 localization and T cell-dependent humoral immune response, J. Exp. Med. 193 (6) (2001) 755–768. C.A. Janeway Jr., P. Travers, M. Walport, M. Shlomchik, Immunbiology— The Immune System in Health and Disease, Garland Science and Taylor & Francis Books Inc., New York, 2001. M. Hallek, L.P. Bergsagel, K. Anderson, Multiple myeloma: increasing evidence for a multistep transformation process, Blood 91 (1) (1998) 3–21. D. Chauhan, H. Uchiyama, Y. Akbarali, M. Urashima, K. Yamamoto, T.A. Libermann, K.C. Anderson, Multiple myeloma cell adhesion-induced interleukin-6 expression in bone marrow stromal cells involves activation of NF-kappa B, Blood 87 (3) (1996) 1104–1112. J.P. Tharakan, P.C. Chau, A radial flow hollow fiber bioreactor for the largescale culture of mammalian cells, Biotechnol. Bioeng. 18 (1985) 329–342. M. Shapiro-Shelef, K. Calame, Plasma cell differentiation and multiple myeloma, Curr. Opin. Immunol. 16 (2) (2004) 226–234. K. Heilmann, T. Groth, O. Behrsing, W. Albrecht, M. Schossig, A. Lendlein, B. Micheel, The influence of the chemical composition of cell culture material on the growth and antibody production of hybridoma cells, J. Biotechnol. 115 (2005) 291–301. J. Luo, S.T. Yang, Effects of three-dimensional culturing in a fibrous matrix on cell cycle, apoptosis, and MAb production by hybridoma cells, Biotechnol. Prog. 20 (1) (2004) 306–315. J.A. Koehler, M. Ulbricht, G. Belfort, Intermolecular forces between polymer films and proteins with relevance to filtration, Langmuir 13 (1997) 4162–4171. E. Ortega, R. Schweitzer-Stenner, I. Pecht, Possible orientational constraints determine secretory signals induced by aggregation of IgE receptors on mast cells, EMBO J. 7 (1988) 4101–4109.

[20] Y. Tamada, E.A. Kulik, Y. Ikada, Simple method for platelet counting, Biomaterials 16 (3) (1995) 259–261. [21] S. Zubler, Diplomarbeit. Herstellung und Charakterisierung von verschiedenen Trx-polyRGD-Konstrukten, ETH Z¨urich (2002). [22] A.L. Cohen, Critical point drying-principles and procedures, Scanning Electron Microsc. 2 (1979) 303–324. [23] F. Grinnell, Fibronectin adsorption on material surfaces, Ann. N Y Acad. Sci. 516 (1987) 280–290. [24] T. Groth, G. Altankov, Studies on cell-biomaterial interaction: role of tyrosine phosphorylation during fibroblast spreading on surfaces varying in wettability, Biomaterials 17 (1996) 1227–1234. [25] M.D. Bale, L.A. Wohlfahrt, D.F. Mosher, B. Tomasini, R.C. Sutton, Identification of vitronectin as a major plasma protein adsorbed on polymer surfaces of different copolymer composition, Blood 74 (1989) 2698–2706. [26] K.R. Gehlsen, W.S. Argraves, M.D. Pierschbacher, E. Ruoslahti, Inhibition of in vitro tumor cell invasion by Arg-Gly-Asp-containing synthetic peptides, J. Cell Biol. 106 (1988) 925–930. [27] T. Groth, I. Zlatanov, G. Altankov, Adhesion of human peripheral lymphocytes on biomaterials preadsorbed with fibronectin and vitronectin, J. Biomater. Sci. Polym. Ed. 6 (1994) 729–739. [28] D.D. Schlaepfer, C.R. Hauck, D.J. Sieg, Signaling through focal adhesion kinase, Prog. Biophys. Mol. Biol. 71 (3–4) (1999) 435–478. [29] J.G. Steele, G. Johnson, C. McFarland, B.A. Dalton, T.R. Gengenbach, R.C. Chatelier, P.R. Underwood, H.J. Griesser, Roles of serum vitronectin and fibronectin in initial attachment of human vein endothelial cells and dermal fibroblasts on oxygen- and nitrogen-containing surfaces made by radiofrequency plasmas, J. Biomater. Sci. Polym. Ed. 6 (1994) 511–532.