- Email: [email protected]
Toxin–antigen conjugates as selection tools for antibody producing cells
Journal of Immunological Methods 387 (2013) 167–172
Contents lists available at SciVerse ScienceDirect
Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim
Research paper
Toxin–antigen conjugates as selection tools for antibody producing cells Katrin Messerschmidt, Katja Heilmann ⁎ Junior Research Group “Antibody Technologies”, Department of Biotechnology, Institute of Biochemistry and Biology, Potsdam University, Karl-Liebknecht-Str. 24-25, D-14476 Potsdam-Golm, Germany
a r t i c l e
i n f o
Article history: Received 7 April 2012 Received in revised form 16 October 2012 Accepted 16 October 2012 Available online 22 October 2012 Keywords: Monoclonal antibody Hybridoma technology Selection of antibody producing cells
a b s t r a c t The generation of antibodies with designated specificity requires cost-intensive and time-consuming screening procedures. Here we present a new method by which hybridoma cells can be selected based on the specificity of the produced antibody by the use of antigen–toxin-conjugates thus eliminating the need of a screening procedure. Initial experiments were done with methotrexate as low molecular weight toxin and fluorescein as model antigen. Methotrexate and a methotrexate– fluorescein conjugate were characterized regarding their toxicity. Afterwards the effect of the fluorescein-specific antibody B13-DE1 on the toxicity of the methotrexate–fluorescein conjugate was determined. Finally, first results showed that hybridoma cells that produce fluorescein specific antibodies are able to grow in the presence of fluorescein–toxin-conjugates. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Monoclonal antibodies (mAbs) are the most important binding molecules in biomedical research as well as in clinical diagnostics and therapy. Although mAbs are such invaluable tools the conventional technology used for the generation and screening of antibody-producing hybridoma cells has not much changed since its invention in the 1970s (Köhler and Milstein, 1975). For the production of monoclonal antibodies by hybridoma technology B lymphocytes of an immunized laboratory animal – in most cases of a mouse (Mus musculus) – are fused with myeloma cells. B cells provide the information for antibody production, and myeloma cells the information for unlimited cell division. After HAT selection of hybrids all generated hybridoma cells have to be screened for the production of the antibody of interest. Furthermore, the hybridoma cells have to be cloned using limiting dilution techniques before mass production can be
Abbreviations: ATCC, American Type Culture Collection; mAbs, monoclonal antibodies; OVA, ovalbumin; BSA, bovine serum albumine; FITC, fluoresceinisothiocyanate; OVA-FITC, fluoresceinisothiocyanate linked to ovalbumin; BSA-FITC, fluoresceinisothiocyanate linked to bovine serum albumin; PBS, phosphate-buffered saline; HAT, hypoxanthine-azaserinethymidine; FCS, fetal calf serum; NCS, neonatal calf serum; MTT, methylthiazolyldiphenyltetrazolium bromide; OD, optical density. ⁎ Corresponding author. Tel.: +49 331 977 5348; fax: +49 331 977 5061. E-mail address: [email protected] (K. Heilmann). 0022-1759/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jim.2012.10.010
established. This process is up to now a very expensive, timeconsuming and low-throughput process. Although several selection strategies for high producing cell lines are described the conventional screening procedures are still state of the art. This might be due to complex sorting processes with specific equipment as described in several publications (Lee et al., 2008; Caron et al., 2008; Hanania et al., 2005). To find a general, simple and well-working selection principle is difficult because the combination of multiple chromosome sets often leads to instability which could affect the antibody related genes of the cell resulting in a loss of antibody production. The present study was aimed to find a procedure by which the generation of antigen-specific monoclonal antibodies by the hybridoma technique will turn out to become easier, faster and more cost effective. The new technique used a selection process which allows the survival of only those hybridoma cells that produce the antibody of interest. This process is based on toxin–antigen conjugates which induce cell death when hybridoma cells secrete unspecific antibodies. When an antigen-specific antibody is bound to the antigen the toxin will be neutralized i.e. it is no longer toxic for the cell. That means the addition of the toxin–antigen conjugate to a mixture of antigen-specific and non-specific hybridoma cells should enable only those cells to survive that produce the antigen-specific antibody (Fig. 1). Therefore screening and cloning procedures could be avoided to find and isolate those hybridoma cells that
168
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
1.
Ab producing cell
2.
Ab producing cell
3.
Ab producing cell
T
Ag
Ag
T
T
Fig. 1. Principles of the described selection process. 1. A toxin that kills hybridoma cells is identified. 2. Toxin is covalently linked to an antigen. Toxin–antigen conjugate is still toxic to hybridoma cells. 3. Toxin–antigen conjugate is no longer toxic if an antigen-specific antibody is bound to the conjugate.
produce antigen-specific antibodies in the vast multitude of antibody-producing cells after a hybridoma fusion. To show the feasibility of this new selection process a methotrexate– fluorescein conjugate was used as model conjugate. Using this conjugate a fluorescein-specific antibody-producing cell line could be selected and established after hybridomas were generated from a mouse immunized with a fluorescein carrier protein conjugate.
exclusively. Therefore the quantity of formazan produced correlates directly to the number of viable cells. At first, cell culture medium was removed completely and a fresh medium (100 μl/well) was added to the cells, followed by the addition of MTT (5 mg/ml in PBS, 10 μl/well). After incubation for 3 h (under 7% CO2 and 100% humidity) the medium was removed completely and cells were lysed using dimethylsulfoxide (100 μl/well). The amount of produced formazan was determined by measuring the optical density (OD) at 550 nm.
2. Methods 2.3. Conjugation of fluoresceinisothiocyanate to proteins 2.1. Toxicity assay All assays to identify the concentration of optimum toxicity of methotrexate (Hexal, Holzkirchen, Germany) and the methotrexate–fluorescein conjugate (Invitrogen, Darmstadt, Germany) were performed in 96 well tissue culture plates. Hybridoma cells (clone E17-FH10, producing antibodies specifically binding horseradish peroxidase) were seeded (5000 cells/ ml, 100 μl/well) in cell culture medium (RPMI 1640, PAN Biotech, Aidenbach, Germany) containing 10% fetal calf serum (FCS, Gibco Invitrogen, Darmstadt, Germany) and 2 mM glutamine. Methotrexate or methotrexate–fluorescein was added at different concentrations (100 μl/well). Cells without toxin or toxin–antigen-conjugate were cultivated in parallel as positive control. After incubation for seven days (under 7% CO2 and 100% humidity) the amount of viable cells was determined using a cell proliferation assay. To evaluate the effect of the fluorescein-specific antibody B13-DE1 (Micheel et al., 1988) on the toxicity of methotrexate and methotrexate–fluorescein (both at a concentration of 6.25 μM) identical toxicity assays were performed with and without the addition of the purified fluorescein-specific antibody B13-DE1 (600 μg/ml=4 μM).
Fluoresceinisothiocyanate (FITC) was linked to bovine serum albumin (BSA) or ovalbumin (OVA) as described in Gani et al. (1980). Briefly, a mixture of BSA or OVA (5 mg/ml), FITC (400 μg/ml) and NaH3PO4 (10 mM) in phosphate buffered saline (PBS) was incubated for 4 h at room temperature. The reaction products (designated BSA-FITC and OVA-FITC) were dialyzed against PBS at 4 °C overnight. 2.4. Purification of murine monoclonal antibody Culture supernatants of antibody producing hybridoma cells were centrifuged (16,000 ×g, 5 min, 4 °C), filtered (0.45 μm) and mixed 4/1 with binding buffer (4 M NaCl, 2 M Glycin NaOH pH 8.9) before applying to a protein A column (HiTrap, Amersham GE Healthcare, Buckinghamshire, UK). The column was washed with diluted binding buffer (1/4) and antibodies were eluted using 0.1 M citrate pH 5.5 or 3.5. The eluted antibodies were neutralized with 500 μl 1 M Tris–HCl, pH 9.0 immediately. Purified antibodies were characterized using indirect and competitive enzyme-linked immunosorbent assays (ELISA).
2.2. Cell proliferation assay
2.5. Indirect enzyme-linked immunosorbent assay (ELISA)
Methylthiazolyldiphenyltetrazolium bromide (MTT) was used to determine the amount of viable cells. MTT (characterized by a yellow color) is cleaved to formazan (characterized by a purple color) by mitochondrial enzymes of living cells
Production of fluorescein-specific antibodies by hybridoma cells was shown using indirect ELISA as follows. Microtiter plates were coated with BSA-FITC (incubation overnight with 5 μg/ml protein in PBS, 50 μl per well), washed with tap water
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
and blocked with 50 μl PBS containing 5% NCS per well for 1 h at room temperature. The wells were then incubated for 1 h at room temperature with 50 μl/well of hybridoma cell culture supernatant. Plates were washed with tap water and incubated for 1 h with an HRP-conjugated goat anti-mouse IgG antibody (Sigma, Steinheim, Germany), diluted 1/3001 in PBS/NCS. Fifty μl/well TMB solution (0.12 mg/ml tetra-methylbenzidine in 50 mM NaH2PO4 with 0.04% CH4N2O·H2O2) was used as substrate to detect solid phase bound HRP. The reaction was stopped after 5 min with 50 μl/well 1 M H2SO4 and OD of the reaction product was measured at 450 nm in an ELISA reader and a reference wavelength of 630 nm (Beckman Coulter GmbH, Krefeld, Germany). Recognition of soluble unconjugated fluorescein and methotrexate–fluorescein by purified antibodies was shown using competitive ELISA. The procedure was identical to the indirect ELISA except that a mixture of fluorescein (12.8 pM to 5 μM) or methotrexate–fluorescein (12.8 pM to 5 μM) and the purified antibody (1 μg/ml) was added to the BSAFITC-coated microtiter plates. 2.6. Selection of antigen-specific hybridomas Balb/c mice (minimum 12 weeks old) were immunized three times intraperitoneally with fluoresceinisothiocyanateconjugated ovalbumin (OVA-FITC). For the first immunization 100 μg OVA-FITC in 100 μl PBS with 100 μl complete Freund's adjuvant was injected. Three weeks later a second immunization followed using 50 μg OVA-FITC in 200 μl PBS. Seven days later, sera were tested for fluorescein-specific antibodies in an enzyme-linked immunosorbent assay (ELISA). Responding mice were boosted with 50 μg OVA-FITC in 200 μl PBS. Four days after booster injection spleen cells of the mice were fused with X63-Ag8.653 myeloma cells (ATCC: CRL-1580) using a modified electrofusion technique (Stoicheva and Hui, 1994). All cell cultivation steps were done using RPMI 1640 without phenol red. Following the fusion (day 0), cells were plated into 96-well plates (TPP, Trasadingen, Germany) and cultured in HAT-medium (RPMI 1640 with 10% FCS, 2 mM glutamine, 100 μM hypoxanthine, 5.8 μM azaserine and 16 μM thymidine, 200 μl/well). At day 3 the medium was removed completely and fresh HAT-medium was added (200 μl/well). At day 7 the selection process was started by removing the HAT medium completely and adding fresh medium (RPMI 1640 with 10% FCS and 2 mM glutamine, 200 μl/well) containing methotrexate–fluorescein (2.5 μM). At day 14 the selection process was terminated by changing medium to RPMI 1640 with 10% FCS and 2 mM glutamine (200 μl/well). One of the surviving clones was chosen, cloned using limited dilution and cultivated for production of the monoclonal antibody.
169
unmodified methotrexate this effect was achieved at a concentration of about 50 nM. The concentration dependent differences were caused by coupling methotrexate to fluorescein which prevents the active transport of the substance into the cells.
3.2. Binding of fluorescein-specific monoclonal antibody B13DE1 to the methotrexate–fluorescein conjugate In order to show whether the model antibody B13-DE1 (monoclonal mouse-anti-FITC-antibody, produced in our own facility) can detoxify the methotrexate–fluorescein conjugate we checked the binding in competitive ELISA. The results shown in Fig. 3 revealed that the methotrexate–fluorescein conjugate is recognized by the fluorescein-specific antibody B13-DE1 with the same affinity as non-modified fluorescein. As expected pure methotrexate was not bound by the antibody B13-DE1.
3.3. Neutralization of methotrexate–fluorescein toxicity by antibody B13-DE1 Hybridoma cells (producing a biotin-specific monoclonal antibody) were incubated with methotrexate–fluorescein conjugate with and without fluorescein-binding antibody B13-DE1 in order to investigate the detoxifying effect of the fluoresceinbinding antibody towards the toxin–fluorescein conjugate. Controls were performed with methotrexate or medium only, each with and without antibody (Fig. 4). When the cells were incubated with methotrexate–fluorescein in the absence of the antibody only 51.5% of the cells survived compared to the control without toxin (100%). In contrast 85.1% of the cells survived in the presence of the fluorescein-specific antibody (n=6, p= 0.000028). This indicated that the purified fluorescein-specific antibody B13DE1 is able to detoxify the antigen–toxin conjugate based on the principle shown in Fig. 1 and so improve cell survival of antigen-specific hybridomas.
3. Results 3.1. Toxicity of methotrexate and methotrexate–fluorescein to hybridoma cells In our model system both methotrexate and the methotrexate–fluorescein conjugate were toxic to hybridoma cells (Fig. 2). Compared to unmodified methotrexate the methotrexate–fluorescein conjugate is less toxic but is still able to kill hybridoma cells at a molar concentration of 10 μM. With
Fig. 2. Toxicity of methotrexate and methotrexate–fluorescein conjugate to hybridoma cells. Different concentrations of methotrexate (open circles) and methotrexate–fluorescein conjugate (closed circles) were added to hybridoma cells. After incubation for seven days the percentage of viable cells was determined using a cell proliferation assay and compared to cells cultivated without toxin or toxin–antigen-conjugate. Measurements were done eightfold; data shown here are mean values with standard deviation.
170
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
Fig. 3. Recognition of methotrexate–fluorescein conjugate by antibody B13-DE1. Different concentrations of fluorescein (open squares), methotrexate–fluorescein conjugate (closed circles) and methotrexate (open circles) were mixed with a constant concentration of B13-DE1 and added to wells coated with BSA-FITC. Antibody binding was detected using a HRP-conjugated goat anti-mouse IgG antibody and TMB as substrate. Measurements were done sixfold; data shown here are mean values with standard deviation.
Fig. 5. Recognition of BSA-FITC by antibody E109-FG3. Different concentrations of antibody E109-FG3 were added to wells coated with BSA-FITC. Antibody binding was detected using a HRP-conjugated goat anti-mouse IgG antibody and TMB as substrate. Measurements were done fourfold; data shown here are mean values with standard deviation.
4. Discussion
To test the selection benefit for freshly fused cells we applied the methotrexate–fluorescein conjugate seven days after fusing myeloma cells with spleen cells from a mouse immunized with OVA-FITC. One of the surviving cell clones was chosen for more detailed investigations. A solid-phase ELISA showed that this clone (designated E109-FG3) produced fluorescein-specific antibodies and a competitive ELISA proved that the produced antibodies were able to bind the fluorescein hapten in solution (Figs. 5 and 6). As compared to B13-DE1 the selected FITCspecific antibody E109-FG3 is less affine but showed no cross reactivity with phenol red.
Although monoclonal antibodies are such important reagents the same basic method is still applied worldwide to generate them as was published in 1975 (Köhler and Milstein, 1975). There have been made several attempts to improve this method and to increase the yield of hybridomas that produce the antibody of interest (Browne and Al-Rubeai, 2007; Shirahata et al., 1998). But most of these techniques were not applicable as routine method or were just quantitative. Electrofusion generally increases the number of hybridoma clones after fusion up to more than 10 times (Karsten et al., 1988; Stoicheva and Hui, 1994). It is therefore applied in several laboratories but does increase the fusion rates not only of antigen-specific B lymphocytes but also of all cells. First attempts to increase only the yield of antigen-specific hybridomas were made by an antigendirected fusion (Kranz et al., 1980; Micheel et al., 1988). But this method is obviously not applicable for all antigens because
Fig. 4. Effect of antibody B13-DE1 on toxicity of methotrexate–fluorescein conjugate. Hybridoma cells (E17-FH10) were incubated with medium (control), methotrexate (6.25 μM) and methotrexate–fluorescein conjugate (6.25 μM) either with (+) or without (−) the addition of antibody B13-DE1 (4 μM). After incubation for seven days the amount of viable cells was determined using a cell proliferation assay. Measurements were done sixfold; data shown here are mean values with standard deviation.
Fig. 6. Recognition of soluble fluorescein by antibody E109-FG3. Different concentrations of haptens fluorescein (open squares) and phenol red (closed squares) were mixed with a constant concentration of antibody E109-FG3 and added to wells coated with BSA-FITC. Antibody binding was detected using a HRP-conjugated goat anti-mouse IgG antibody and TMB as substrate. Measurements were done twice; data shown here are mean values with standard deviation.
3.4. Selection of fluorescein-specific antibody-producing hybridoma cells by methotrexate–fluorescein conjugate directly after fusion
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
the conjugation of the antigen to the cell surface of viable cells requires considerable efforts. Some other procedures to directly isolate antigen-specific hybridomas were invented which are based e.g. on the cultivation of the hybridomas on semi-solid medium to obtain and then easily identify and isolate single-cell clones (Davis et al., 1982), the labeling of the cell surface of a fusion partner and the isolation of the antigen-specific hybridomas after fusion by fluorescence-activated cell sorting (Manz et al., 1995), or the encapsulation of the fused cells in agarose microdroplets and isolation of the fluorescence-labeled antigenspecific hybridomas by flow cytometry (Gray et al., 1995). But all these methods use either difficult experimental procedures or expensive equipment. Therefore, easier methods which are applicable in each laboratory are still urgently needed. In this study we presented a method for the isolation of antigen-specific hybridomas which can be applied in each cell culture laboratory and does not need additional expensive equipment. The method is based on the fact that a substance which is toxic to cultivated cells does retain its toxicity also after conjugation to another substance. The toxic effect is blocked if hybridoma cells producing antigen-specific antibodies which are able to bind the conjugate. This binding prevents the cellular uptake of methotrexate and avoids any toxic effects to the hybridoma cell. In order to reduce bystander effects hybridoma cells were plated in 96 well plates as monocultures and cultivated in the presence of toxin–antigen-conjugates. Directly after fusion the cells are slow-growing cells with low production rates. The first assays for specific antibody production are routinely performed after 7 till 10 days. At this stage bystander effects are rare and should not influence the selection process. Even with established hybridomas we could not observe a significant impact of bystander effects at different cultivation temperatures (unpublished results). We could show that antibodies added to the toxin conjugate protect the cells in culture against the toxic effects, and that cells producing the corresponding antibody itself are protected against the toxic effects, too. In contrast, cells producing no antibody or a non-related one are killed by the toxin conjugate. That this is principally possible was already shown in previous experiments using fluorescein-specific scFvproducing Escherichia coli as antibody-producing cells and ampicillin–FITC conjugate as toxin (Sellrie and Micheel, 2008). The transfer of the selection strategy to mammalian cells was more difficult than we anticipated. First of all, the number of suitable toxins was significantly reduced by the required properties such as functional groups for conjugation, availability in high quality sufficient for cell culture and toxicity at low concentrations in hybridoma cell culture. Further, the fusion process is a random process and the outcome of specific hybridomas could not be standardized in contrast to bacterial systems. Further the use of haptens as antigens might be difficult to evaluate selection benefits because the output of positive clones is low in general. Till now there are only two monoclonal anti-FITC antibodies commercially available (4-4-20 produced by Kranz et al., 1982 and B13-DE1 produced by Micheel et al., 1988. Nevertheless, we think, that the usage of toxin–antigen conjugates leads to a selection benefit directly after fusion. The monoclonal antibody E109-FG3 selected by a toxin–antigen conjugate in this study was able to recognize fluorescein in a competitive principle and showed no cross-reactivity to phenol red in contrast to the monoclonal
171
anti-fluorescein antibody B13-DE1 generated by conventional hybridoma technology (Micheel et al., 1988). Our experiments showed that the neutralization of the methotrexate conjugates required actually a comparatively high concentration of antibodies. We do not exactly know whether this is due to the fact that antibody B13-DE1 also binds phenol red since our cell cultures contained phenol red. Phenol red is used as pH indicator but could act also as growth factor (Devleeschouwer et al., 1992). Further studies are necessary to investigate the biological function of toxin action on hybridomas. Ongoing experiments are performed to evaluate the correlation between membrane bound receptors and selection success by toxin– antigen conjugates. For this also other toxins e.g. ricin were currently tested for suitability as conjugate in this selection principle. The application of further toxins is necessary to increase the specificity of selection and to eliminate toxinspecific hybridomas. To select antibodies against proteins or other antigens of higher molecular weights toxins have to be used which react directly with receptors on the cell surface or which activate their entry into the cells themselves. We are checking at present several different toxins to find the one which fulfills these needs. 5. Conclusions This new method seems to be applicable for the selection of antibody-producing hybridomas by an easy and fast procedure which does not need time-consuming manual work and does not require additional expensive equipment. It should, therefore, be possible to be performed in each normal cell culture laboratory. Acknowledgments This work was supported by the Federal Ministry of Education and Research (InnoProfile 703). We thank Professor Burkhard Micheel for providing the fluorescein-specific antibody B13-DE1, for proofreading the manuscript and for the constructive comments. References Browne, S.M., Al-Rubeai, M., 2007. Selection methods for high-producing mammalian cell lines. Trends Biotechnol. 25, 425. Caron, A.W., Nicolas, C., Gaillet, B., Ba, I., Pinard, M., Garnier, A., Massie, B., Gilbert, R., 2008. Fluorescent labelling in semi-solid medium for selection of mammalian cells secreting high-levels of recombinant proteins. BMC Biotechnol. 9, 42. Davis, J.M., Pennington, J.E., Kubler, A.M., Conscience, J.F., 1982. A simple, single-step technique for selecting and cloning hybridomas for the production of monoclonal antibodies. J. Immunol. Methods 50, 161. Devleeschouwer, N., Body, J.J., Legros, N., Muquardt, C., Donnay, I., Wouters, P., Leclercq, G., 1992. Growth factor-like activity of phenol red preparations in the MCF-7 breast cancer cell line. Anticancer. Res. 12, 789. Gani, M.M., Hunt, T., Summerell, J.M., 1980. A simple method of labelling mouse Thy-1 antibodies with FITC. J. Immunol. Methods 34, 133. Gray, F., Kenney, J.S., Dunne, J.F., 1995. Secretion capture and report web: use of affinity derivatized agarose microdroplets for the selection of hybridoma cells. J. Immunol. Methods 182, 155. Hanania, E.G., Fieck, A., Stevens, J., Bodzin, L.J., Palsson, B.O., Koller, M.R., 2005. Automated in situ measurements of cell-specific antibody secretion and laser-mediated purification for rapid cloning of highly-secreting producers. Biotechnol. Bioeng. 91, 872. Karsten, U., Stolley, P., Walther, I., Papsdorf, G., Weber, S., Conrad, K., Pasternak, L., Kopp, J., 1988. Direct comparison of electric field-mediated
172
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
and PEG-mediated cell fusion for the generation of antibody producing hybridomas. Hybridoma 7, 627. Köhler, G., Milstein, C., 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 495. Kranz, D.M., Billing, P.A., Herron, J.N., Voss Jr., E.W., 1980. Modified hybridoma methodology: antigen-directed chemically mediated cell fusion. Immunol. Commun. 9, 639. Kranz, D.M., Herron, J.N., Voss, E.W., 1982. Mechanisms of ligand-binding by monoclonal anti-fluorescyl antibodies. J. Biol. Chem. 257 (12), 6987. Lee, R., Tran, M., Nocerini, M., Liang, M., 2008. A high-throughput hybridoma selection method using fluorometric microvolume assay technology. J. Biomol. Screen. 13, 210. Manz, R., Assenmacher, M., Pflüger, E., Miltenyi, S., Radbruch, A., 1995. Analysis and sorting of live cells according to secreted molecules, relocated to a cellsurface affinity matrix. Proc. Natl. Acad. Sci. 92, 1921.
Micheel, B., Jantscheff, P., Böttger, V., Scharte, G., Kaiser, G., Stolley, P., Karawajew, L., 1988. The production and radioimmunoassay application of monoclonal antibodies to fluorescein isothiocyanate (FITC). J. Immunol. Methods 111, 89. Sellrie, F., Micheel, B., 2008. Selection of recombinant antibody-producing E. coli cells by means of toxin conjugates. Biochem. Eng. J. 38, 309. Shirahata, S., Katakura, Y., Teruya, K., 1998. Cell hybridization, hybridomas and human hybridomas. Methods Cell Biol. 57, 111. Stoicheva, N.G., Hui, S.W., 1994. Electrically induced fusion of mammalian cells in the presence of polyethylene glycol. J. Membr. Biol. 141, 177.
Contents lists available at SciVerse ScienceDirect
Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim
Research paper
Toxin–antigen conjugates as selection tools for antibody producing cells Katrin Messerschmidt, Katja Heilmann ⁎ Junior Research Group “Antibody Technologies”, Department of Biotechnology, Institute of Biochemistry and Biology, Potsdam University, Karl-Liebknecht-Str. 24-25, D-14476 Potsdam-Golm, Germany
a r t i c l e
i n f o
Article history: Received 7 April 2012 Received in revised form 16 October 2012 Accepted 16 October 2012 Available online 22 October 2012 Keywords: Monoclonal antibody Hybridoma technology Selection of antibody producing cells
a b s t r a c t The generation of antibodies with designated specificity requires cost-intensive and time-consuming screening procedures. Here we present a new method by which hybridoma cells can be selected based on the specificity of the produced antibody by the use of antigen–toxin-conjugates thus eliminating the need of a screening procedure. Initial experiments were done with methotrexate as low molecular weight toxin and fluorescein as model antigen. Methotrexate and a methotrexate– fluorescein conjugate were characterized regarding their toxicity. Afterwards the effect of the fluorescein-specific antibody B13-DE1 on the toxicity of the methotrexate–fluorescein conjugate was determined. Finally, first results showed that hybridoma cells that produce fluorescein specific antibodies are able to grow in the presence of fluorescein–toxin-conjugates. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Monoclonal antibodies (mAbs) are the most important binding molecules in biomedical research as well as in clinical diagnostics and therapy. Although mAbs are such invaluable tools the conventional technology used for the generation and screening of antibody-producing hybridoma cells has not much changed since its invention in the 1970s (Köhler and Milstein, 1975). For the production of monoclonal antibodies by hybridoma technology B lymphocytes of an immunized laboratory animal – in most cases of a mouse (Mus musculus) – are fused with myeloma cells. B cells provide the information for antibody production, and myeloma cells the information for unlimited cell division. After HAT selection of hybrids all generated hybridoma cells have to be screened for the production of the antibody of interest. Furthermore, the hybridoma cells have to be cloned using limiting dilution techniques before mass production can be
Abbreviations: ATCC, American Type Culture Collection; mAbs, monoclonal antibodies; OVA, ovalbumin; BSA, bovine serum albumine; FITC, fluoresceinisothiocyanate; OVA-FITC, fluoresceinisothiocyanate linked to ovalbumin; BSA-FITC, fluoresceinisothiocyanate linked to bovine serum albumin; PBS, phosphate-buffered saline; HAT, hypoxanthine-azaserinethymidine; FCS, fetal calf serum; NCS, neonatal calf serum; MTT, methylthiazolyldiphenyltetrazolium bromide; OD, optical density. ⁎ Corresponding author. Tel.: +49 331 977 5348; fax: +49 331 977 5061. E-mail address: [email protected] (K. Heilmann). 0022-1759/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jim.2012.10.010
established. This process is up to now a very expensive, timeconsuming and low-throughput process. Although several selection strategies for high producing cell lines are described the conventional screening procedures are still state of the art. This might be due to complex sorting processes with specific equipment as described in several publications (Lee et al., 2008; Caron et al., 2008; Hanania et al., 2005). To find a general, simple and well-working selection principle is difficult because the combination of multiple chromosome sets often leads to instability which could affect the antibody related genes of the cell resulting in a loss of antibody production. The present study was aimed to find a procedure by which the generation of antigen-specific monoclonal antibodies by the hybridoma technique will turn out to become easier, faster and more cost effective. The new technique used a selection process which allows the survival of only those hybridoma cells that produce the antibody of interest. This process is based on toxin–antigen conjugates which induce cell death when hybridoma cells secrete unspecific antibodies. When an antigen-specific antibody is bound to the antigen the toxin will be neutralized i.e. it is no longer toxic for the cell. That means the addition of the toxin–antigen conjugate to a mixture of antigen-specific and non-specific hybridoma cells should enable only those cells to survive that produce the antigen-specific antibody (Fig. 1). Therefore screening and cloning procedures could be avoided to find and isolate those hybridoma cells that
168
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
1.
Ab producing cell
2.
Ab producing cell
3.
Ab producing cell
T
Ag
Ag
T
T
Fig. 1. Principles of the described selection process. 1. A toxin that kills hybridoma cells is identified. 2. Toxin is covalently linked to an antigen. Toxin–antigen conjugate is still toxic to hybridoma cells. 3. Toxin–antigen conjugate is no longer toxic if an antigen-specific antibody is bound to the conjugate.
produce antigen-specific antibodies in the vast multitude of antibody-producing cells after a hybridoma fusion. To show the feasibility of this new selection process a methotrexate– fluorescein conjugate was used as model conjugate. Using this conjugate a fluorescein-specific antibody-producing cell line could be selected and established after hybridomas were generated from a mouse immunized with a fluorescein carrier protein conjugate.
exclusively. Therefore the quantity of formazan produced correlates directly to the number of viable cells. At first, cell culture medium was removed completely and a fresh medium (100 μl/well) was added to the cells, followed by the addition of MTT (5 mg/ml in PBS, 10 μl/well). After incubation for 3 h (under 7% CO2 and 100% humidity) the medium was removed completely and cells were lysed using dimethylsulfoxide (100 μl/well). The amount of produced formazan was determined by measuring the optical density (OD) at 550 nm.
2. Methods 2.3. Conjugation of fluoresceinisothiocyanate to proteins 2.1. Toxicity assay All assays to identify the concentration of optimum toxicity of methotrexate (Hexal, Holzkirchen, Germany) and the methotrexate–fluorescein conjugate (Invitrogen, Darmstadt, Germany) were performed in 96 well tissue culture plates. Hybridoma cells (clone E17-FH10, producing antibodies specifically binding horseradish peroxidase) were seeded (5000 cells/ ml, 100 μl/well) in cell culture medium (RPMI 1640, PAN Biotech, Aidenbach, Germany) containing 10% fetal calf serum (FCS, Gibco Invitrogen, Darmstadt, Germany) and 2 mM glutamine. Methotrexate or methotrexate–fluorescein was added at different concentrations (100 μl/well). Cells without toxin or toxin–antigen-conjugate were cultivated in parallel as positive control. After incubation for seven days (under 7% CO2 and 100% humidity) the amount of viable cells was determined using a cell proliferation assay. To evaluate the effect of the fluorescein-specific antibody B13-DE1 (Micheel et al., 1988) on the toxicity of methotrexate and methotrexate–fluorescein (both at a concentration of 6.25 μM) identical toxicity assays were performed with and without the addition of the purified fluorescein-specific antibody B13-DE1 (600 μg/ml=4 μM).
Fluoresceinisothiocyanate (FITC) was linked to bovine serum albumin (BSA) or ovalbumin (OVA) as described in Gani et al. (1980). Briefly, a mixture of BSA or OVA (5 mg/ml), FITC (400 μg/ml) and NaH3PO4 (10 mM) in phosphate buffered saline (PBS) was incubated for 4 h at room temperature. The reaction products (designated BSA-FITC and OVA-FITC) were dialyzed against PBS at 4 °C overnight. 2.4. Purification of murine monoclonal antibody Culture supernatants of antibody producing hybridoma cells were centrifuged (16,000 ×g, 5 min, 4 °C), filtered (0.45 μm) and mixed 4/1 with binding buffer (4 M NaCl, 2 M Glycin NaOH pH 8.9) before applying to a protein A column (HiTrap, Amersham GE Healthcare, Buckinghamshire, UK). The column was washed with diluted binding buffer (1/4) and antibodies were eluted using 0.1 M citrate pH 5.5 or 3.5. The eluted antibodies were neutralized with 500 μl 1 M Tris–HCl, pH 9.0 immediately. Purified antibodies were characterized using indirect and competitive enzyme-linked immunosorbent assays (ELISA).
2.2. Cell proliferation assay
2.5. Indirect enzyme-linked immunosorbent assay (ELISA)
Methylthiazolyldiphenyltetrazolium bromide (MTT) was used to determine the amount of viable cells. MTT (characterized by a yellow color) is cleaved to formazan (characterized by a purple color) by mitochondrial enzymes of living cells
Production of fluorescein-specific antibodies by hybridoma cells was shown using indirect ELISA as follows. Microtiter plates were coated with BSA-FITC (incubation overnight with 5 μg/ml protein in PBS, 50 μl per well), washed with tap water
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
and blocked with 50 μl PBS containing 5% NCS per well for 1 h at room temperature. The wells were then incubated for 1 h at room temperature with 50 μl/well of hybridoma cell culture supernatant. Plates were washed with tap water and incubated for 1 h with an HRP-conjugated goat anti-mouse IgG antibody (Sigma, Steinheim, Germany), diluted 1/3001 in PBS/NCS. Fifty μl/well TMB solution (0.12 mg/ml tetra-methylbenzidine in 50 mM NaH2PO4 with 0.04% CH4N2O·H2O2) was used as substrate to detect solid phase bound HRP. The reaction was stopped after 5 min with 50 μl/well 1 M H2SO4 and OD of the reaction product was measured at 450 nm in an ELISA reader and a reference wavelength of 630 nm (Beckman Coulter GmbH, Krefeld, Germany). Recognition of soluble unconjugated fluorescein and methotrexate–fluorescein by purified antibodies was shown using competitive ELISA. The procedure was identical to the indirect ELISA except that a mixture of fluorescein (12.8 pM to 5 μM) or methotrexate–fluorescein (12.8 pM to 5 μM) and the purified antibody (1 μg/ml) was added to the BSAFITC-coated microtiter plates. 2.6. Selection of antigen-specific hybridomas Balb/c mice (minimum 12 weeks old) were immunized three times intraperitoneally with fluoresceinisothiocyanateconjugated ovalbumin (OVA-FITC). For the first immunization 100 μg OVA-FITC in 100 μl PBS with 100 μl complete Freund's adjuvant was injected. Three weeks later a second immunization followed using 50 μg OVA-FITC in 200 μl PBS. Seven days later, sera were tested for fluorescein-specific antibodies in an enzyme-linked immunosorbent assay (ELISA). Responding mice were boosted with 50 μg OVA-FITC in 200 μl PBS. Four days after booster injection spleen cells of the mice were fused with X63-Ag8.653 myeloma cells (ATCC: CRL-1580) using a modified electrofusion technique (Stoicheva and Hui, 1994). All cell cultivation steps were done using RPMI 1640 without phenol red. Following the fusion (day 0), cells were plated into 96-well plates (TPP, Trasadingen, Germany) and cultured in HAT-medium (RPMI 1640 with 10% FCS, 2 mM glutamine, 100 μM hypoxanthine, 5.8 μM azaserine and 16 μM thymidine, 200 μl/well). At day 3 the medium was removed completely and fresh HAT-medium was added (200 μl/well). At day 7 the selection process was started by removing the HAT medium completely and adding fresh medium (RPMI 1640 with 10% FCS and 2 mM glutamine, 200 μl/well) containing methotrexate–fluorescein (2.5 μM). At day 14 the selection process was terminated by changing medium to RPMI 1640 with 10% FCS and 2 mM glutamine (200 μl/well). One of the surviving clones was chosen, cloned using limited dilution and cultivated for production of the monoclonal antibody.
169
unmodified methotrexate this effect was achieved at a concentration of about 50 nM. The concentration dependent differences were caused by coupling methotrexate to fluorescein which prevents the active transport of the substance into the cells.
3.2. Binding of fluorescein-specific monoclonal antibody B13DE1 to the methotrexate–fluorescein conjugate In order to show whether the model antibody B13-DE1 (monoclonal mouse-anti-FITC-antibody, produced in our own facility) can detoxify the methotrexate–fluorescein conjugate we checked the binding in competitive ELISA. The results shown in Fig. 3 revealed that the methotrexate–fluorescein conjugate is recognized by the fluorescein-specific antibody B13-DE1 with the same affinity as non-modified fluorescein. As expected pure methotrexate was not bound by the antibody B13-DE1.
3.3. Neutralization of methotrexate–fluorescein toxicity by antibody B13-DE1 Hybridoma cells (producing a biotin-specific monoclonal antibody) were incubated with methotrexate–fluorescein conjugate with and without fluorescein-binding antibody B13-DE1 in order to investigate the detoxifying effect of the fluoresceinbinding antibody towards the toxin–fluorescein conjugate. Controls were performed with methotrexate or medium only, each with and without antibody (Fig. 4). When the cells were incubated with methotrexate–fluorescein in the absence of the antibody only 51.5% of the cells survived compared to the control without toxin (100%). In contrast 85.1% of the cells survived in the presence of the fluorescein-specific antibody (n=6, p= 0.000028). This indicated that the purified fluorescein-specific antibody B13DE1 is able to detoxify the antigen–toxin conjugate based on the principle shown in Fig. 1 and so improve cell survival of antigen-specific hybridomas.
3. Results 3.1. Toxicity of methotrexate and methotrexate–fluorescein to hybridoma cells In our model system both methotrexate and the methotrexate–fluorescein conjugate were toxic to hybridoma cells (Fig. 2). Compared to unmodified methotrexate the methotrexate–fluorescein conjugate is less toxic but is still able to kill hybridoma cells at a molar concentration of 10 μM. With
Fig. 2. Toxicity of methotrexate and methotrexate–fluorescein conjugate to hybridoma cells. Different concentrations of methotrexate (open circles) and methotrexate–fluorescein conjugate (closed circles) were added to hybridoma cells. After incubation for seven days the percentage of viable cells was determined using a cell proliferation assay and compared to cells cultivated without toxin or toxin–antigen-conjugate. Measurements were done eightfold; data shown here are mean values with standard deviation.
170
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
Fig. 3. Recognition of methotrexate–fluorescein conjugate by antibody B13-DE1. Different concentrations of fluorescein (open squares), methotrexate–fluorescein conjugate (closed circles) and methotrexate (open circles) were mixed with a constant concentration of B13-DE1 and added to wells coated with BSA-FITC. Antibody binding was detected using a HRP-conjugated goat anti-mouse IgG antibody and TMB as substrate. Measurements were done sixfold; data shown here are mean values with standard deviation.
Fig. 5. Recognition of BSA-FITC by antibody E109-FG3. Different concentrations of antibody E109-FG3 were added to wells coated with BSA-FITC. Antibody binding was detected using a HRP-conjugated goat anti-mouse IgG antibody and TMB as substrate. Measurements were done fourfold; data shown here are mean values with standard deviation.
4. Discussion
To test the selection benefit for freshly fused cells we applied the methotrexate–fluorescein conjugate seven days after fusing myeloma cells with spleen cells from a mouse immunized with OVA-FITC. One of the surviving cell clones was chosen for more detailed investigations. A solid-phase ELISA showed that this clone (designated E109-FG3) produced fluorescein-specific antibodies and a competitive ELISA proved that the produced antibodies were able to bind the fluorescein hapten in solution (Figs. 5 and 6). As compared to B13-DE1 the selected FITCspecific antibody E109-FG3 is less affine but showed no cross reactivity with phenol red.
Although monoclonal antibodies are such important reagents the same basic method is still applied worldwide to generate them as was published in 1975 (Köhler and Milstein, 1975). There have been made several attempts to improve this method and to increase the yield of hybridomas that produce the antibody of interest (Browne and Al-Rubeai, 2007; Shirahata et al., 1998). But most of these techniques were not applicable as routine method or were just quantitative. Electrofusion generally increases the number of hybridoma clones after fusion up to more than 10 times (Karsten et al., 1988; Stoicheva and Hui, 1994). It is therefore applied in several laboratories but does increase the fusion rates not only of antigen-specific B lymphocytes but also of all cells. First attempts to increase only the yield of antigen-specific hybridomas were made by an antigendirected fusion (Kranz et al., 1980; Micheel et al., 1988). But this method is obviously not applicable for all antigens because
Fig. 4. Effect of antibody B13-DE1 on toxicity of methotrexate–fluorescein conjugate. Hybridoma cells (E17-FH10) were incubated with medium (control), methotrexate (6.25 μM) and methotrexate–fluorescein conjugate (6.25 μM) either with (+) or without (−) the addition of antibody B13-DE1 (4 μM). After incubation for seven days the amount of viable cells was determined using a cell proliferation assay. Measurements were done sixfold; data shown here are mean values with standard deviation.
Fig. 6. Recognition of soluble fluorescein by antibody E109-FG3. Different concentrations of haptens fluorescein (open squares) and phenol red (closed squares) were mixed with a constant concentration of antibody E109-FG3 and added to wells coated with BSA-FITC. Antibody binding was detected using a HRP-conjugated goat anti-mouse IgG antibody and TMB as substrate. Measurements were done twice; data shown here are mean values with standard deviation.
3.4. Selection of fluorescein-specific antibody-producing hybridoma cells by methotrexate–fluorescein conjugate directly after fusion
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
the conjugation of the antigen to the cell surface of viable cells requires considerable efforts. Some other procedures to directly isolate antigen-specific hybridomas were invented which are based e.g. on the cultivation of the hybridomas on semi-solid medium to obtain and then easily identify and isolate single-cell clones (Davis et al., 1982), the labeling of the cell surface of a fusion partner and the isolation of the antigen-specific hybridomas after fusion by fluorescence-activated cell sorting (Manz et al., 1995), or the encapsulation of the fused cells in agarose microdroplets and isolation of the fluorescence-labeled antigenspecific hybridomas by flow cytometry (Gray et al., 1995). But all these methods use either difficult experimental procedures or expensive equipment. Therefore, easier methods which are applicable in each laboratory are still urgently needed. In this study we presented a method for the isolation of antigen-specific hybridomas which can be applied in each cell culture laboratory and does not need additional expensive equipment. The method is based on the fact that a substance which is toxic to cultivated cells does retain its toxicity also after conjugation to another substance. The toxic effect is blocked if hybridoma cells producing antigen-specific antibodies which are able to bind the conjugate. This binding prevents the cellular uptake of methotrexate and avoids any toxic effects to the hybridoma cell. In order to reduce bystander effects hybridoma cells were plated in 96 well plates as monocultures and cultivated in the presence of toxin–antigen-conjugates. Directly after fusion the cells are slow-growing cells with low production rates. The first assays for specific antibody production are routinely performed after 7 till 10 days. At this stage bystander effects are rare and should not influence the selection process. Even with established hybridomas we could not observe a significant impact of bystander effects at different cultivation temperatures (unpublished results). We could show that antibodies added to the toxin conjugate protect the cells in culture against the toxic effects, and that cells producing the corresponding antibody itself are protected against the toxic effects, too. In contrast, cells producing no antibody or a non-related one are killed by the toxin conjugate. That this is principally possible was already shown in previous experiments using fluorescein-specific scFvproducing Escherichia coli as antibody-producing cells and ampicillin–FITC conjugate as toxin (Sellrie and Micheel, 2008). The transfer of the selection strategy to mammalian cells was more difficult than we anticipated. First of all, the number of suitable toxins was significantly reduced by the required properties such as functional groups for conjugation, availability in high quality sufficient for cell culture and toxicity at low concentrations in hybridoma cell culture. Further, the fusion process is a random process and the outcome of specific hybridomas could not be standardized in contrast to bacterial systems. Further the use of haptens as antigens might be difficult to evaluate selection benefits because the output of positive clones is low in general. Till now there are only two monoclonal anti-FITC antibodies commercially available (4-4-20 produced by Kranz et al., 1982 and B13-DE1 produced by Micheel et al., 1988. Nevertheless, we think, that the usage of toxin–antigen conjugates leads to a selection benefit directly after fusion. The monoclonal antibody E109-FG3 selected by a toxin–antigen conjugate in this study was able to recognize fluorescein in a competitive principle and showed no cross-reactivity to phenol red in contrast to the monoclonal
171
anti-fluorescein antibody B13-DE1 generated by conventional hybridoma technology (Micheel et al., 1988). Our experiments showed that the neutralization of the methotrexate conjugates required actually a comparatively high concentration of antibodies. We do not exactly know whether this is due to the fact that antibody B13-DE1 also binds phenol red since our cell cultures contained phenol red. Phenol red is used as pH indicator but could act also as growth factor (Devleeschouwer et al., 1992). Further studies are necessary to investigate the biological function of toxin action on hybridomas. Ongoing experiments are performed to evaluate the correlation between membrane bound receptors and selection success by toxin– antigen conjugates. For this also other toxins e.g. ricin were currently tested for suitability as conjugate in this selection principle. The application of further toxins is necessary to increase the specificity of selection and to eliminate toxinspecific hybridomas. To select antibodies against proteins or other antigens of higher molecular weights toxins have to be used which react directly with receptors on the cell surface or which activate their entry into the cells themselves. We are checking at present several different toxins to find the one which fulfills these needs. 5. Conclusions This new method seems to be applicable for the selection of antibody-producing hybridomas by an easy and fast procedure which does not need time-consuming manual work and does not require additional expensive equipment. It should, therefore, be possible to be performed in each normal cell culture laboratory. Acknowledgments This work was supported by the Federal Ministry of Education and Research (InnoProfile 703). We thank Professor Burkhard Micheel for providing the fluorescein-specific antibody B13-DE1, for proofreading the manuscript and for the constructive comments. References Browne, S.M., Al-Rubeai, M., 2007. Selection methods for high-producing mammalian cell lines. Trends Biotechnol. 25, 425. Caron, A.W., Nicolas, C., Gaillet, B., Ba, I., Pinard, M., Garnier, A., Massie, B., Gilbert, R., 2008. Fluorescent labelling in semi-solid medium for selection of mammalian cells secreting high-levels of recombinant proteins. BMC Biotechnol. 9, 42. Davis, J.M., Pennington, J.E., Kubler, A.M., Conscience, J.F., 1982. A simple, single-step technique for selecting and cloning hybridomas for the production of monoclonal antibodies. J. Immunol. Methods 50, 161. Devleeschouwer, N., Body, J.J., Legros, N., Muquardt, C., Donnay, I., Wouters, P., Leclercq, G., 1992. Growth factor-like activity of phenol red preparations in the MCF-7 breast cancer cell line. Anticancer. Res. 12, 789. Gani, M.M., Hunt, T., Summerell, J.M., 1980. A simple method of labelling mouse Thy-1 antibodies with FITC. J. Immunol. Methods 34, 133. Gray, F., Kenney, J.S., Dunne, J.F., 1995. Secretion capture and report web: use of affinity derivatized agarose microdroplets for the selection of hybridoma cells. J. Immunol. Methods 182, 155. Hanania, E.G., Fieck, A., Stevens, J., Bodzin, L.J., Palsson, B.O., Koller, M.R., 2005. Automated in situ measurements of cell-specific antibody secretion and laser-mediated purification for rapid cloning of highly-secreting producers. Biotechnol. Bioeng. 91, 872. Karsten, U., Stolley, P., Walther, I., Papsdorf, G., Weber, S., Conrad, K., Pasternak, L., Kopp, J., 1988. Direct comparison of electric field-mediated
172
K. Messerschmidt, K. Heilmann / Journal of Immunological Methods 387 (2013) 167–172
and PEG-mediated cell fusion for the generation of antibody producing hybridomas. Hybridoma 7, 627. Köhler, G., Milstein, C., 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 495. Kranz, D.M., Billing, P.A., Herron, J.N., Voss Jr., E.W., 1980. Modified hybridoma methodology: antigen-directed chemically mediated cell fusion. Immunol. Commun. 9, 639. Kranz, D.M., Herron, J.N., Voss, E.W., 1982. Mechanisms of ligand-binding by monoclonal anti-fluorescyl antibodies. J. Biol. Chem. 257 (12), 6987. Lee, R., Tran, M., Nocerini, M., Liang, M., 2008. A high-throughput hybridoma selection method using fluorometric microvolume assay technology. J. Biomol. Screen. 13, 210. Manz, R., Assenmacher, M., Pflüger, E., Miltenyi, S., Radbruch, A., 1995. Analysis and sorting of live cells according to secreted molecules, relocated to a cellsurface affinity matrix. Proc. Natl. Acad. Sci. 92, 1921.
Micheel, B., Jantscheff, P., Böttger, V., Scharte, G., Kaiser, G., Stolley, P., Karawajew, L., 1988. The production and radioimmunoassay application of monoclonal antibodies to fluorescein isothiocyanate (FITC). J. Immunol. Methods 111, 89. Sellrie, F., Micheel, B., 2008. Selection of recombinant antibody-producing E. coli cells by means of toxin conjugates. Biochem. Eng. J. 38, 309. Shirahata, S., Katakura, Y., Teruya, K., 1998. Cell hybridization, hybridomas and human hybridomas. Methods Cell Biol. 57, 111. Stoicheva, N.G., Hui, S.W., 1994. Electrically induced fusion of mammalian cells in the presence of polyethylene glycol. J. Membr. Biol. 141, 177.