- Email: [email protected]
Isolation of single cells for protein therapeutics using microwell selection and Surface Plasmon Resonance imaging
Analytical Biochemistry 531 (2017) 45e47
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Isolation of single cells for protein therapeutics using microwell selection and Surface Plasmon Resonance imaging F. Abali a, M. Stevens b, A.G.J. Tibbe b, L.W.M.M. Terstappen a, *, P.N. van der Velde c, R.B.M. Schasfoort a, d a
Medical Cell Biophysics Group, MIRA Institute, Faculty of Science and Technology, University of Twente, PO Box 217, 7500AE Enschede, The Netherlands VyCAP, Abraham Rademakerstraat 41, 7425PG Deventer, The Netherlands IBIS Technologies, Pantheon 5, 7521 PR Enschede, The Netherlands d Interfluidics BV, Duizendblad 28, 7483 AL Haaksbergen, The Netherlands b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 March 2017 Received in revised form 17 May 2017 Accepted 19 May 2017 Available online 22 May 2017
Here the feasibility is demonstrated that by combining Surface Plasmon Resonance Imaging (SPRi) and self-sorting microwell technology product secretion of individual cells can be monitored. Additionally isolation of the selected cells can be performed by punching the cells from the microwells using coordinates of the positions of microwells obtained with SPRi. Cells of interest can be retrieved sterile from the microwell array for further cultivation. © 2017 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Keywords: SPRi EpCAM Hybridoma Antibody/protein excretion Single cells SPR cytometry Microwell McSPRInter
A range of biopharmaceuticals, therapeutic recombinant monoclonal antibodies (mAbs) [1], are currently produced in high amounts for therapeutic applications. It requires cell lines capable of producing the desired product at high productivities. However, various cell lines used for mAb synthesis are highly heterogeneous in yield, which consequently results in the need to screen a large library of cells to identify the best stable high producing clones [2]. There is an interest of biopharmaceutical industries for a technology that enables rapid and reliable screening a large number of candidate cells for cell line development [3,4]. The most widely used technologies for isolation of single cells are limiting dilution and ClonePIX FL™ (Molecular Devices, LLC.) [5]. Traditionally, limiting dilution is the method most used for single cell cloning. However, although simple and low in costs the process is laborious, time-consuming and involves long-term cell culture which results in low probability of monoclonality and the
* Corresponding author. E-mail address: [email protected] (L.W.M.M. Terstappen).
efficiency is highly dependent on the total number of cell than can be screened [3,4]. ClonePix FL™ uses random cell seeding in high viscosity semisolid medium to immobilize cells and retain secreted proteins/antibodies in the vicinity of the growing cells in combination with automatic picking of colonies. High producing colonies are detected via immune precipitation or fluorescence techniques [4]. However, a relative long incubation time (6e12 days) is needed in order to let the cells grow to sufficient numbers enabling detection of a significant amount of product. In addition, although the method is automated, the selection of single cells is not possible and therefore monoclonality is not always ensured. Therefore it is expected that the combination of a single cell based isolation technology and a sensitive detection technique could be an effective method for cellline development. Surface Plasmon Resonance (SPR) enables the label free detection of biomolecular interactions in real time using relatively small quantities of materials [6]. The conventional SPR analysis involves the binding of one of the two interacting molecules to the sensor surface while the other is flowed as an analyte over the sensor
http://dx.doi.org/10.1016/j.ab.2017.05.021 0003-2697/© 2017 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
46
F. Abali et al. / Analytical Biochemistry 531 (2017) 45e47
surface. Although SPR has been considered as a label free method for mainly protein quantification [7e9] only recently, a study reported the quantification of proteins derived from whole intact single cells using SPR. Stojanovic et al. demonstrated the feasibility to monitor and quantify the amount of product excreted by whole intact single cells on top of an SPR sensor surface using real-time label free SPR imaging (SPRi) [10]. Furthermore, in a different study it was shown that 99.1% of excreted cell product binds directly to the sensor surface and only 0.9% was lost due to diffusion into the bulk [11]. In that study however, after monitoring and determining individual cells in the SPR image, high producer cells could not be retrieved from the sensor surface. In this report, we show how self-sorting microwell technology [12] and SPR can be applied to isolate the cells of interest. Combining the two technologies enables monitoring, tracking and quantifying the secretion of antibodies from individual cells label free and in real time using SPRi in combination with microwell technology. A pool of thousands of cells can be screened after an overnight incubation, in hours instead of weeks. Fig. 1 demonstrates the working principle of the Microwell cell selection printer (McSPRinter). An array of isolated single cells in a microwell chip is used to spot the secreted product (e.g antibodies) onto a functionalized SPR sensor chip. In this manner a proper selection of cells can be carried out based on the specific interaction of the molecules produced by single cells. Fig. 2 illustrates the procedure for selecting single cells based on criteria as the quality and quantity of the secreted product. First indicated with I in the figure, a SPRi sensor is functionalized with a specific capture/selection surface. The microwell chip depicted in II contains an array of 6400 microwells in an effective area of 8 8 mm2 with a single pore of 5 mm at the bottom which enables the fast sorting of single cells in individual wells. Each microwell has a diameter of 70 ± 2 mm, a depth of 360 ± 10 mm and a volume of 1.4 nL. The sorting principle enabling the seeding of a single cell per well is described elsewhere [12]. Subsequently, the microwell chip is connected to the SPRi sensor depicted in III and incubated for a short period of time (e.g overnight) allowing the cells to secrete specific molecules which will be captured by a ligand functionalized sensor surface. Next in IV & V, the microwell is disconnected from the SPRi sensor and incubated, while the captured molecules on the SPRi sensor are analyzed on a SPRi instrument to determine for each cell the amount of specific product, but also the on- and off rate. In VI the criteria for selecting interesting cells are set for each position of the array with respect to corner reference points which are set when the microwells and sensor are connected and
coordinates of the cells of interest in the microwells are determined. Rare high-producers cells identified with SPRi measurement can be retrieved from the microwell array into a microtiter plate. In practice, the microwell chip is scanned using an inverted fluorescence microscope that acquires images of the entire chips in an automated fashion. Once the coordinates of the array are known individual cells are punched out from the microwell into a microtiter plate under sterile and optimal cultivation conditions for further cultivation of the specific cell as illustrated in VII. To demonstrate the feasibility of the McSPRinter a proof of principle study was performed. The EpCAM antibody producing cells line (IgG1, VU1dD9), were seeded (Vycap, Enschede, The Netherlands) and attached onto a SPRi imaging chip Easy2Spot® pre-activated G-type Senseye® sensors (Ssens bv, Enschede, the Netherlands) with immobilized rhEpCAM Active human EpCAM protein fragment (ab155712, Abcam Cambridge, UK). The recombinant human EpCAM protein was immobilized on a SPRi sensor to specifically interact and bind antibodies produced by the hybridoma cells. Next, microwells and SensEye sensor were incubated under sterile culture conditions at 37 C and 5% CO2. After an overnight incubation the SensEye sensor and microwell chip were detached and the SPRi sensor was inserted in the IBIS MX96 instrument (IBIS technologies B.V., Enschede, the Netherlands). First an SPR image at fixed height and contrast angle was made at the initiation of the experiment. In order to detect small quantities of produced antibody from the single hybridoma cells an amplification cascade was used consisting of a biotinylated Goat antimouse IgG injection, followed by injection of neutravidine and finally biotinylated gold nanoparticles. A second image is acquired and normalized by subtracting the first image. Small regions of interest (ROIs) were placed to observe the products secreted by single hybridoma cells. Step VI in Fig. 2 shows a typical example of regions of secreted product. Here, EpCAM antibody dots produced by single VU1D9 hybridoma cell line were captured onto the SPRi imaging chip. Each spot represents antibodies derived from a single cell in a single microwell. Practically, the rare high producing cells can be punched out and be cultivated for monoclonal expansion. Experiments with different cell lines showed that punching resulted in a survival rate above 70%. Overall these results demonstrated the feasibility that the production of antibodies by individual hybridoma cells embedded in microwells can be measured in real time using SPRi. The quantification method in the McSPRinter is label free and enables to set selection criteria for the molecules based not only on
Fig. 1. The working principle of McSPRinter. A microwell chip containing an array of microwells filled with single cells is attached to a ligand functionalized SPR sensor surface. This combination is incubated. Product secreted by a cell diffuses via the 5 mm pore at the bottom of the microwell towards the sensor surface where it is captured.
F. Abali et al. / Analytical Biochemistry 531 (2017) 45e47
47
Fig. 2. Flow chart of the McSPRinter. The process is separated in SPRi (left) and Punching (right). The two technologies meet at the centre where the SPRi sensor surface is combined with the microwell chip (I, II &III). Protein products derived from viable single cells are spotted onto a SPRi capture surface (IV & VI) and the array coordinates are passed to the Puncher for identification of the wells of interest and isolation of the cells in these wells (V &VI).
a specific quantity but also on affinity criteria (e.g. on-and off rates and equilibrium dissociation constants of the product of a cell). Additionally, the process cycle period in the McSPRinter is extremely short (overnight), ensures monoclonality while screening of unstable cells (e.g. matured B-lymphocytes) can in principle also be applied. Conclusion The McSPRinter uses a self-sorting microwell chip for printing of cell secreted products onto a SPRi sensor surface. The microwell is used to sort single cells in individual microwells. Sorted cells secrete protein products and the molecules diffuse from the cell through the 5 mm pore at the bottom of the microwell to the sensor surface. The printed array is analyzed label free using SPRi and obtained coordinates of the ROI are used to select and punch cells of interest from the microwell chip. These cells can be isolated in a relative short time, which is important for large-scale cell screening purposes for the product of interest.
[2]
[3] [4]
[5]
[6] [7]
[8]
[9]
[10]
Acknowledgements As of April 1, 2017 work described in this manuscript is supported by NWO Applied and Engineering Sciences project number: 15327. References [1] A. Rita Costa, M. Elisa Rodrigues, M. Henriques, J. Azeredo, R. Oliveira,
[11]
[12]
Guidelines to cell engineering for monoclonal antibody production, Eur. J. Pharm. Biopharm. 74 (2010) 127e138. T. Lai, Y. Yang, S.K. Ng, Advances in mammalian cell line development technologies for recombinant protein production, Pharmaceuticals 6 (2013) 579e603. S.M. Browne, M. Al-Rubeai, Selection methods for high-producing mammalian cell lines, Trends Biotechnol. 25 (2007) 425e432. T. Nakamura, T. Omasa, Optimization of cell line development in the GS-CHO expression system using a high-throughput, single cell-based clone selection system, J. Biosci. Bioeng. 120 (2015) 323e329. J.J.C. Hou, B.S. Hughes, M. Smede, K.M. Leung, K. Levine, S. Rigby, P.P. Gray, T.P. Munro, High-throughput ClonePix FL analysis of mAb-expressing clones using the UCOE expression system, New Biotechnol. 31 (2014) 214e220. R.B.M. Schasfoort, A.J. Tudos, Handbook of Surface Plasmon Resonance, RSC Pub., Cambridge, UK, 2008. N. Chavane, R. Jacquemart, C.D. Hoemann, M. Jolicoeur, G. De Crescenzo, Atline quantification of bioactive antibody in bioreactor by surface plasmon resonance using epitope detection, Anal. Biochem. 378 (2008) 158e165. Y. Kikuchi, S. Uno, M. Nanami, Y. Yoshimura, S.-i. Iida, N. Fukushima, M. Tsuchiya, Determination of concentration and binding affinity of antibody fragments by use of surface plasmon resonance, J. Biosci. Bioeng. 100 (2005) 311e317. M.S. Mehand, B. Srinivasan, G. De Crescenzo, Estimation of analyte concentration by surface plasmon resonance-based biosensing using parameter identification techniques, Anal. Biochem. 419 (2011) 140e144. I. Stojanovi c, T.J.G. van der Velden, H.W. Mulder, R.B.M. Schasfoort, L.W.M.M. Terstappen, Quantification of antibody production of individual hybridoma cells by surface plasmon resonance imaging, Anal. Biochem. 485 (2015) 112e118. I. Stojanovi c, W. Baumgartner, T.J.G. van der Velden, L.W.M.M. Terstappen, R.B.M. Schasfoort, Modeling single cell antibody excretion on a biosensor, Anal. Biochem. 504 (2016) 1e3. J.F. Swennenhuis, A.G. Tibbe, M. Stevens, M.R. Katika, J. van Dalum, H.D. Tong, C.J. van Rijn, L.W. Terstappen, Self-seeding microwell chip for the isolation and characterization of single cells, Lab. Chip 15 (2015) 3039e3046.
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Isolation of single cells for protein therapeutics using microwell selection and Surface Plasmon Resonance imaging F. Abali a, M. Stevens b, A.G.J. Tibbe b, L.W.M.M. Terstappen a, *, P.N. van der Velde c, R.B.M. Schasfoort a, d a
Medical Cell Biophysics Group, MIRA Institute, Faculty of Science and Technology, University of Twente, PO Box 217, 7500AE Enschede, The Netherlands VyCAP, Abraham Rademakerstraat 41, 7425PG Deventer, The Netherlands IBIS Technologies, Pantheon 5, 7521 PR Enschede, The Netherlands d Interfluidics BV, Duizendblad 28, 7483 AL Haaksbergen, The Netherlands b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 March 2017 Received in revised form 17 May 2017 Accepted 19 May 2017 Available online 22 May 2017
Here the feasibility is demonstrated that by combining Surface Plasmon Resonance Imaging (SPRi) and self-sorting microwell technology product secretion of individual cells can be monitored. Additionally isolation of the selected cells can be performed by punching the cells from the microwells using coordinates of the positions of microwells obtained with SPRi. Cells of interest can be retrieved sterile from the microwell array for further cultivation. © 2017 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Keywords: SPRi EpCAM Hybridoma Antibody/protein excretion Single cells SPR cytometry Microwell McSPRInter
A range of biopharmaceuticals, therapeutic recombinant monoclonal antibodies (mAbs) [1], are currently produced in high amounts for therapeutic applications. It requires cell lines capable of producing the desired product at high productivities. However, various cell lines used for mAb synthesis are highly heterogeneous in yield, which consequently results in the need to screen a large library of cells to identify the best stable high producing clones [2]. There is an interest of biopharmaceutical industries for a technology that enables rapid and reliable screening a large number of candidate cells for cell line development [3,4]. The most widely used technologies for isolation of single cells are limiting dilution and ClonePIX FL™ (Molecular Devices, LLC.) [5]. Traditionally, limiting dilution is the method most used for single cell cloning. However, although simple and low in costs the process is laborious, time-consuming and involves long-term cell culture which results in low probability of monoclonality and the
* Corresponding author. E-mail address: [email protected] (L.W.M.M. Terstappen).
efficiency is highly dependent on the total number of cell than can be screened [3,4]. ClonePix FL™ uses random cell seeding in high viscosity semisolid medium to immobilize cells and retain secreted proteins/antibodies in the vicinity of the growing cells in combination with automatic picking of colonies. High producing colonies are detected via immune precipitation or fluorescence techniques [4]. However, a relative long incubation time (6e12 days) is needed in order to let the cells grow to sufficient numbers enabling detection of a significant amount of product. In addition, although the method is automated, the selection of single cells is not possible and therefore monoclonality is not always ensured. Therefore it is expected that the combination of a single cell based isolation technology and a sensitive detection technique could be an effective method for cellline development. Surface Plasmon Resonance (SPR) enables the label free detection of biomolecular interactions in real time using relatively small quantities of materials [6]. The conventional SPR analysis involves the binding of one of the two interacting molecules to the sensor surface while the other is flowed as an analyte over the sensor
http://dx.doi.org/10.1016/j.ab.2017.05.021 0003-2697/© 2017 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
46
F. Abali et al. / Analytical Biochemistry 531 (2017) 45e47
surface. Although SPR has been considered as a label free method for mainly protein quantification [7e9] only recently, a study reported the quantification of proteins derived from whole intact single cells using SPR. Stojanovic et al. demonstrated the feasibility to monitor and quantify the amount of product excreted by whole intact single cells on top of an SPR sensor surface using real-time label free SPR imaging (SPRi) [10]. Furthermore, in a different study it was shown that 99.1% of excreted cell product binds directly to the sensor surface and only 0.9% was lost due to diffusion into the bulk [11]. In that study however, after monitoring and determining individual cells in the SPR image, high producer cells could not be retrieved from the sensor surface. In this report, we show how self-sorting microwell technology [12] and SPR can be applied to isolate the cells of interest. Combining the two technologies enables monitoring, tracking and quantifying the secretion of antibodies from individual cells label free and in real time using SPRi in combination with microwell technology. A pool of thousands of cells can be screened after an overnight incubation, in hours instead of weeks. Fig. 1 demonstrates the working principle of the Microwell cell selection printer (McSPRinter). An array of isolated single cells in a microwell chip is used to spot the secreted product (e.g antibodies) onto a functionalized SPR sensor chip. In this manner a proper selection of cells can be carried out based on the specific interaction of the molecules produced by single cells. Fig. 2 illustrates the procedure for selecting single cells based on criteria as the quality and quantity of the secreted product. First indicated with I in the figure, a SPRi sensor is functionalized with a specific capture/selection surface. The microwell chip depicted in II contains an array of 6400 microwells in an effective area of 8 8 mm2 with a single pore of 5 mm at the bottom which enables the fast sorting of single cells in individual wells. Each microwell has a diameter of 70 ± 2 mm, a depth of 360 ± 10 mm and a volume of 1.4 nL. The sorting principle enabling the seeding of a single cell per well is described elsewhere [12]. Subsequently, the microwell chip is connected to the SPRi sensor depicted in III and incubated for a short period of time (e.g overnight) allowing the cells to secrete specific molecules which will be captured by a ligand functionalized sensor surface. Next in IV & V, the microwell is disconnected from the SPRi sensor and incubated, while the captured molecules on the SPRi sensor are analyzed on a SPRi instrument to determine for each cell the amount of specific product, but also the on- and off rate. In VI the criteria for selecting interesting cells are set for each position of the array with respect to corner reference points which are set when the microwells and sensor are connected and
coordinates of the cells of interest in the microwells are determined. Rare high-producers cells identified with SPRi measurement can be retrieved from the microwell array into a microtiter plate. In practice, the microwell chip is scanned using an inverted fluorescence microscope that acquires images of the entire chips in an automated fashion. Once the coordinates of the array are known individual cells are punched out from the microwell into a microtiter plate under sterile and optimal cultivation conditions for further cultivation of the specific cell as illustrated in VII. To demonstrate the feasibility of the McSPRinter a proof of principle study was performed. The EpCAM antibody producing cells line (IgG1, VU1dD9), were seeded (Vycap, Enschede, The Netherlands) and attached onto a SPRi imaging chip Easy2Spot® pre-activated G-type Senseye® sensors (Ssens bv, Enschede, the Netherlands) with immobilized rhEpCAM Active human EpCAM protein fragment (ab155712, Abcam Cambridge, UK). The recombinant human EpCAM protein was immobilized on a SPRi sensor to specifically interact and bind antibodies produced by the hybridoma cells. Next, microwells and SensEye sensor were incubated under sterile culture conditions at 37 C and 5% CO2. After an overnight incubation the SensEye sensor and microwell chip were detached and the SPRi sensor was inserted in the IBIS MX96 instrument (IBIS technologies B.V., Enschede, the Netherlands). First an SPR image at fixed height and contrast angle was made at the initiation of the experiment. In order to detect small quantities of produced antibody from the single hybridoma cells an amplification cascade was used consisting of a biotinylated Goat antimouse IgG injection, followed by injection of neutravidine and finally biotinylated gold nanoparticles. A second image is acquired and normalized by subtracting the first image. Small regions of interest (ROIs) were placed to observe the products secreted by single hybridoma cells. Step VI in Fig. 2 shows a typical example of regions of secreted product. Here, EpCAM antibody dots produced by single VU1D9 hybridoma cell line were captured onto the SPRi imaging chip. Each spot represents antibodies derived from a single cell in a single microwell. Practically, the rare high producing cells can be punched out and be cultivated for monoclonal expansion. Experiments with different cell lines showed that punching resulted in a survival rate above 70%. Overall these results demonstrated the feasibility that the production of antibodies by individual hybridoma cells embedded in microwells can be measured in real time using SPRi. The quantification method in the McSPRinter is label free and enables to set selection criteria for the molecules based not only on
Fig. 1. The working principle of McSPRinter. A microwell chip containing an array of microwells filled with single cells is attached to a ligand functionalized SPR sensor surface. This combination is incubated. Product secreted by a cell diffuses via the 5 mm pore at the bottom of the microwell towards the sensor surface where it is captured.
F. Abali et al. / Analytical Biochemistry 531 (2017) 45e47
47
Fig. 2. Flow chart of the McSPRinter. The process is separated in SPRi (left) and Punching (right). The two technologies meet at the centre where the SPRi sensor surface is combined with the microwell chip (I, II &III). Protein products derived from viable single cells are spotted onto a SPRi capture surface (IV & VI) and the array coordinates are passed to the Puncher for identification of the wells of interest and isolation of the cells in these wells (V &VI).
a specific quantity but also on affinity criteria (e.g. on-and off rates and equilibrium dissociation constants of the product of a cell). Additionally, the process cycle period in the McSPRinter is extremely short (overnight), ensures monoclonality while screening of unstable cells (e.g. matured B-lymphocytes) can in principle also be applied. Conclusion The McSPRinter uses a self-sorting microwell chip for printing of cell secreted products onto a SPRi sensor surface. The microwell is used to sort single cells in individual microwells. Sorted cells secrete protein products and the molecules diffuse from the cell through the 5 mm pore at the bottom of the microwell to the sensor surface. The printed array is analyzed label free using SPRi and obtained coordinates of the ROI are used to select and punch cells of interest from the microwell chip. These cells can be isolated in a relative short time, which is important for large-scale cell screening purposes for the product of interest.
[2]
[3] [4]
[5]
[6] [7]
[8]
[9]
[10]
Acknowledgements As of April 1, 2017 work described in this manuscript is supported by NWO Applied and Engineering Sciences project number: 15327. References [1] A. Rita Costa, M. Elisa Rodrigues, M. Henriques, J. Azeredo, R. Oliveira,
[11]
[12]
Guidelines to cell engineering for monoclonal antibody production, Eur. J. Pharm. Biopharm. 74 (2010) 127e138. T. Lai, Y. Yang, S.K. Ng, Advances in mammalian cell line development technologies for recombinant protein production, Pharmaceuticals 6 (2013) 579e603. S.M. Browne, M. Al-Rubeai, Selection methods for high-producing mammalian cell lines, Trends Biotechnol. 25 (2007) 425e432. T. Nakamura, T. Omasa, Optimization of cell line development in the GS-CHO expression system using a high-throughput, single cell-based clone selection system, J. Biosci. Bioeng. 120 (2015) 323e329. J.J.C. Hou, B.S. Hughes, M. Smede, K.M. Leung, K. Levine, S. Rigby, P.P. Gray, T.P. Munro, High-throughput ClonePix FL analysis of mAb-expressing clones using the UCOE expression system, New Biotechnol. 31 (2014) 214e220. R.B.M. Schasfoort, A.J. Tudos, Handbook of Surface Plasmon Resonance, RSC Pub., Cambridge, UK, 2008. N. Chavane, R. Jacquemart, C.D. Hoemann, M. Jolicoeur, G. De Crescenzo, Atline quantification of bioactive antibody in bioreactor by surface plasmon resonance using epitope detection, Anal. Biochem. 378 (2008) 158e165. Y. Kikuchi, S. Uno, M. Nanami, Y. Yoshimura, S.-i. Iida, N. Fukushima, M. Tsuchiya, Determination of concentration and binding affinity of antibody fragments by use of surface plasmon resonance, J. Biosci. Bioeng. 100 (2005) 311e317. M.S. Mehand, B. Srinivasan, G. De Crescenzo, Estimation of analyte concentration by surface plasmon resonance-based biosensing using parameter identification techniques, Anal. Biochem. 419 (2011) 140e144. I. Stojanovi c, T.J.G. van der Velden, H.W. Mulder, R.B.M. Schasfoort, L.W.M.M. Terstappen, Quantification of antibody production of individual hybridoma cells by surface plasmon resonance imaging, Anal. Biochem. 485 (2015) 112e118. I. Stojanovi c, W. Baumgartner, T.J.G. van der Velden, L.W.M.M. Terstappen, R.B.M. Schasfoort, Modeling single cell antibody excretion on a biosensor, Anal. Biochem. 504 (2016) 1e3. J.F. Swennenhuis, A.G. Tibbe, M. Stevens, M.R. Katika, J. van Dalum, H.D. Tong, C.J. van Rijn, L.W. Terstappen, Self-seeding microwell chip for the isolation and characterization of single cells, Lab. Chip 15 (2015) 3039e3046.