Development and optimization of a process for automated recovery of single cells identified by microengraving

Microfabricated devices are useful tools for manipulating and interrogating large numbers of single cells in a rapid and cost‐effective manner, but connecting these systems to the existing platforms used in routine high‐throughput screening of libraries of cells remains challenging. Methods to sort individual cells of interest from custom microscale devices to standardized culture dishes in an efficient and automated manner without affecting the viability of the cells are critical. Combining a commercially available instrument for colony picking (CellCelector, AVISO GmbH) and a customized software module, we have established an optimized process for the automated retrieval of individual antibody‐producing cells, secreting desirable antibodies, from dense arrays of subnanoliter containers. The selection of cells for retrieval is guided by data obtained from a high‐throughput, single‐cell screening method called microengraving. Using this system, 100 clones from a mixed population of two cell lines secreting different antibodies (12CA5 and HYB099‐01) were sorted with 100% accuracy (50 clones of each) in ～2 h, and the cells retained viability. © 2010 American Institute of Chemical Engineers Biotechnol. Prog., 2010     

Production of two hemopoietic growth factors is differentially regulated in single T lymphocytes activated with an anti-T cell receptor antibody

A method has been developed to measure the production by single activated T lymphocytes of two hemopoietic growth factors, granulocyte-macrophage CSF (GM-CSF) and multipotential CSF (multi-CSF or IL-3). When individual cells of the L3T4 (CD4)+ F23.1+ T cell clone E9.D4 were transferred by micromanipulation into wells coated with the monoclonal anti-T cell receptor antibody F23.1, up to 90% of cells produced CSF as detected by CSF-dependent hemopoietic cell lines. Production occurred in the absence of proliferation and did not require the addition of accessory cells or IL-2. Both the frequency of CSF-producing cells and the average production per positive cell depended on the density of the immobilized stimulating ligand, indicating that the response of each cell is not an all-or-none phenomenon but varies with the strength of stimulation. Individual cells of the clone varied over a 100-fold range in their total CSF titer with a mean value of about 0.05 U/cell. They also varied in their relative production of GM-CSF and multi-CSF. Thus, low producing cells secreted only GM-CSF whereas high producing cells also secreted multi-CSF. The failure of low producing cells to secrete multi-CSF was not genetically based since such cells could give rise to progeny that synthesized multi-CSF. These results suggest that the synthesis of these two lymphokines can be differentially regulated at the level of the single cell.     

Quantitative Imaging of Protein Secretions from Single Cells in Real Time

Protein secretions from individual cells create spatially and temporally varying concentration profiles in the extracellular environment, which guide a wide range of biological processes such as wound healing and angiogenesis. Fluorescent and colorimetric probes for the detection of single cell secretions have time resolutions that range from hours to days, and as a result, little is known about how individual cells may alter their protein secretion rates on the timescale of minutes or seconds. Here, we present a label-free technique based upon nanoplasmonic imaging, which enabled the measurement of individual cell secretions in real time. When applied to the detection of antibody secretions from single hybridoma cells, the enhanced time resolution revealed two modes of secretion: one in which the cell secreted continuously and another in which antibodies were released in concentrated bursts that coincided with minute-long morphological contractions of the cell. From the continuous secretion measurements we determined the local concentration of antibodies at the sensing array closest to the cell and from the bursts we estimated the diffusion constant of the secreted antibodies through the extracellular media. The design also incorporates transmitted light and fluorescence microscopy capabilities for monitoring cellular morphological changes and intracellular fluorescent labels. We anticipate that this technique can be adapted as a general tool for the quantitative study of paracrine signaling in both adherent and nonadherent cell lines.     

Quantitative Imaging of Protein Secretions from Single Cells in Real Time

Protein secretions from individual cells create spatially and temporally varying concentration profiles in the extracellular environment, which guide a wide range of biological processes such as wound healing and angiogenesis. Fluorescent and colorimetric probes for the detection of single cell secretions have time resolutions that range from hours to days, and as a result, little is known about how individual cells may alter their protein secretion rates on the timescale of minutes or seconds. Here, we present a label-free technique based upon nanoplasmonic imaging, which enabled the measurement of individual cell secretions in real time. When applied to the detection of antibody secretions from single hybridoma cells, the enhanced time resolution revealed two modes of secretion: one in which the cell secreted continuously and another in which antibodies were released in concentrated bursts that coincided with minute-long morphological contractions of the cell. From the continuous secretion measurements we determined the local concentration of antibodies at the sensing array closest to the cell and from the bursts we estimated the diffusion constant of the secreted antibodies through the extracellular media. The design also incorporates transmitted light and fluorescence microscopy capabilities for monitoring cellular morphological changes and intracellular fluorescent labels. We anticipate that this technique can be adapted as a general tool for the quantitative study of paracrine signaling in both adherent and nonadherent cell lines.     

Transfected Chinese hamster ovary cells as a model system for cytokine immunocytochemistry and in situ hybridisation.

The presence of cytokine producing cells is most easily revealed by techniques measuring the secreted cytokines in culture supernatants or body fluids. However, these techniques only measure the bulk cytokine release by a given, often mixed cell population. To demonstrate cytokine production at the single cell level, immunocytochemistry (ICC) and in situ hybridisation (ISH) are now widely used techniques. To establish these techniques, an easily accessible model system is needed which permits the evaluation of different ICC and ISH protocols. It can be used to demonstrate the specificity of the antibodies and may serve as a positive control for samples of unknown cytokine content. Here we propose the use of Chinese hamster ovary (CHO) cells transfected to express one specific cytokine as such a model system. Its usefulness is demonstrated by the characterisation of six monoclonal antibodies to human interleukin-4 and the establishment of two in situ hybridisation protocols.     

Rapid isolation of antigen-specific antibody-secreting cells using a chip-based immunospot array

Here we report a new method for isolating antigen-specific antibody-secreting cells (ASCs) using a microwell array chip, which offers a rapid, efficient and high-throughput (up to 234,000 individual cells) system for the detection and retrieval of cells that secrete antibodies of interest on a single-cell basis. We arrayed a large population of lymphoid cells containing ASCs from human peripheral blood on microwell array chips and detected spots with secreted antibodies. This protocol can be completed in less than 7 h, including 3 h of cell culture. The method presented here not only has high sensitivity and specificity comparable with enzyme-linked immunospot (ELISPOT) but it also overcomes the limitations of ELISPOT in recovering ASCs that can be used to produce antigen-specific human monoclonal antibodies. This method can also be used to detect cells secreting molecules other than antibodies, such as cytokines, and it provides a tool for cell analysis and clinical diagnosis.     

Transfected Chinese hamster ovary cells as model system for cytokine immunocytochemistry and in situ hybridisation.

The presence of cytokine producing cells is most easily revealed by techniques measuring the secreted cytokines in culture supernatants or body fluids. However, these techniques only measure the bulk cytokine release by a given, often mixed cell population. To demonstrate cytokine production at the single cell level, immunocytochemistry (ICC) and in situ hybridisation (ISH) are now widely used techniques. To establish these techniques, an easily accessible model system is needed which permits the evaluation of different ICC and ISH protocols. It can be used to demonstrate the specificity of the antibodies and may serve as a positive control for samples of unknown cytokine content. Here we propose the use of Chinese hamster ovary (CHO) cells transfected to express one specific cytokine as such a model system. Its usefulness is demonstrated by the characterisation of six monoclonal antibodies to human interleukin-4 and the establishment of two in situ hybridisation protocols.     

Basal and thyroliberin-stimulated prolactin synthesis in single-cell cultures and in populations of rat pituitary cells.

1. Newly synthesized prolactin was obtained from cultures of rat pituitary tumour cells (GH4C1 cells) after incubation with [35S]methionine. 2. Radioactive synthesized and secreted prolactin was quantified by an immunoprecipitation method by using disc-gel electrophoresis of the dissolved immunoprecipitate in the presence of sodium dodecyl sulphate. By using a microanalytical modification, hormone synthesis and secretion could also be studied in single-cell cultures. This technique was combined with a cytoimmunofluorescence method in which rhodamine-conjugated antibodies were used for studying intracellular prolactin. 3. The presence of radioactive synthesized and secreted prolactin was demonstrated in nine out of 13 single-cell cultures. Cell cultures containing 10 cells or more and clonal populations originating from one cell always secreted radioactive prolactin. 4. Thyroliberin treatment (2 muM) for 24h increased the extracellular accumulation of radioactive prolactin in five out of seven single-cell cultures and always in populations of cells. 5. The number of cells showing prolactin specific fluorescence increased from 20 to 50% and the intensity of this fluorescence became greater after thyroliberin treatment. 6. Studies of [35S]prolactin secretion from single cells and immunochemical detection of intracellular prolactin showed that some cells in an unsynchronized population did not secret radioactive prolactin or show prolactin specific fluorescence. 7. The quantitative effect of thyroliberin as studied in single-cell cultures suggested that the main if not the only effect was to increase prolactin synthesis in cells already producing hormone.     

Anti-CD3-stimulated T cells induce the production of multiple Ig H chain isotypes by individual human peripheral B lymphocytes.

Polyclonal activation of human B cells is achieved by coculture with T cells stimulated by mAb to the CD3 molecular complex. By formal limiting dilution analysis, approximately 60% of human peripheral blood B cells were found to produce Ig in this system. When individual B cells were cultured in microtiter wells with anti-CD3-activated T cells, more than one-third of cultures producing Ig contained multiple Ig H chain isotypes. Similar results were observed when individual IgM-expressing B cells, selected and dispersed by FACS were cultured with anti-CD3-activated T cells. The clonality of the B cells producing multiple Ig isotypes was supported by L chain analysis of the secreted Ig. Of the wells containing more than one H chain isotype, nearly 85% contained only a single L chain type. Clonality was further examined by polymerase chain reaction amplification of cDNA harvested from cultures originally seeded with individual B cells. In general, only a single VH gene family could be amplified from cultures producing more than one Ig isotype. Three separate VH regions were cloned and sequenced. One, a VHIV-mu was nearly identical to a previously described VH gene VH71.4; as second, a VHIV-gamma was very similar to a previously described VH gene segment V-79, whereas a third, a VHIII-gamma differed by 14% in nucleotide sequence from its closest germline counterpart VH3005. These results indicate that anti-CD3-activated T cells not only stimulate the majority of B cells to secrete Ig, but also induce individual B cells to produce multiple Ig H chain isotypes. Additionally, the procedure described provides a reliable method to sample a large proportion of the human peripheral B cell repertoire.     

Measurement of cytokine release at the single cell level using the ELISPOT assay

The enzyme-linked immunospot (ELISPOT) assay took its design concept from traditional ELISA techniques and evolved over the years from a method for detecting antibodies secreted from B cells to a method for detecting cytokines or other soluble mediators secreted from a variety of different cell types. The ELISPOT assay allows the quantitative measurement of frequency of cytokine secreting cells at the single cell level directly ex vivo without elaborate in vitro expansion or manipulation of cell populations. The function of cells can be inferred from the pattern of cytokines secreted by cells in response to diverse antigenic stimuli and thus the ELISPOT assay has become a powerful method for monitoring immune responses in health and disease. The ELISPOT assay like the ELISA assay is relatively easy to perform and it has the promise of robustness, reliability, and reproducibility of performance for use as a diagnostic tool. The history, applications, validation process, and future challenges of the ELISPOT assay are discussed in this chapter.     

Rapid selection of single cells with high antibody production rates by microwell array

Selection of single cells capable of producing target proteins at high rates is crucial for the development of protein manufacturing processes. Here, we present the rapid selection of single cells producing immunoglobulin antibodies at high specific rates by microwell array and microengraving. Chinese hamster ovary (CHO) cells secreting chimeric antibodies were deposited in a microwell array in a manner such that each microwell contained a single cell. Secreted antibodies in the microwells were transferred onto a glass slide by microengraving, followed by interrogation using fluorescence-based immunoassay. Single cells displaying high signal intensities were selected, retrieved, and clonally expanded to assess their specific antibody production rates. Three successive rounds of the process resulted in the selection of single cells showing significantly increased antibody production rates. The present approach can be applied to the selection of single cells for producing other therapeutic proteins in a high-throughput manner.     

Droplet-based microfluidic platforms for single T cell secretion analysis of IL-10 cytokine

Here we present a microfluidic method for the analysis of single cell secretions. The method co-encapsulates cells with microspheres conjugated with capture antibodies and detection fluorescence-labeled antibodies. The secreted substance captured on the microsphere surface and detected via detection antibodies generating a localized fluorescent signal on a microsphere surface. Using this method, CD4+CD25+ regulatory T cells were encapsulated and assayed to detect IL-10 secreting cell in population.     

Detection and isolation of antibody-forming clones by means of local cytolysis in gel

A method is suggested combining the screening and cloning of hybridomas producing cytotoxic antibodies. The method is based on the Erne local hemolysis principles. The cells from the preformed hybridoma line NATF 9.9 secreted monoclonal antibodies (McAb) against murine T lymphocyte differentiation antigen Lyt-3.2. A mixture of hybridoma cells and target cells was attached to the plastic Petri dish surface pretreated with Poly-L-lysine and covered with agarose. McAb production by hybridoma cells was elicited by complement-dependent lysis of target cells. The lysis was detected by incorporation in dead cells of the fluorescent dye ethidium bromide. Subsequent studies made it possible to evaluate the clonogenic capacity of McAb production and isolate active clones.     

Identification of rat IL-1beta, IL-2, IFN-gamma and TNF-alpha in activated splenocytes by intracellular immunostaining.

We have developed a sensitive three-step indirect immunofluorescence method to identify individual rat cells that produce cytokines including IL-1beta, IL-2, IFN-gamma and TNF-alpha. Cultured rat splenocytes were polyclonally activated to cytokine synthesis by mitogens such as lipopolysaccharide or a combination of a protein kinase C activator (phorbol 12-myristate 13-acetate) and a calcium ionophore (ionomycin). Careful selection of either antigen affinity-purified polyclonal or monoclonal cytokine-detecting antibodies combined with gentle fixation and permeabilization of the cells enabled discrimination of cytokine-producing cells based on distinct morphological staining criteria. Cells making IL-2, IFN-gamma and TNF-alpha could be identified by a characteristic, intracellular, rounded, juxtanuclear immunofluorescence signal. This staining pattern reflected the accumulation of the intracellularly processed cytokines in the Golgi organelle of producer cells. The staining of cells that synthesized IL-1beta, which is not transported intracellularly via the endoplasmatic reticulum-Golgi pathway, generated a different, but distinct and reproducible staining pattern, IL-1beta producing macrophages expressed intense nuclear immunofluorescence with additional reticular, cytoplasmic signals. Furthermore, the use of biologically neutralizing detecting antibodies against the cytokines under study prevented recognition of surface-stained target cells that had bound secreted cytokines by cytokine-specific receptors. This modified staining technology enabled analysis of the kinetic pattern and the frequency of cytokine-producing cells in cultures of rat splenocytes after various modes of polyclonal activation.     

The in vitro antibody response to cell surface antigens. II. Monoclonal antibodies to human leukemia cells.

A method has been developed for the production of monoclonal mouse antibody responses in vitro against human cell surface antigens. Limiting numbers of immune spleen cells were transferred to syngeneic, irradiated recipients whose spleen fragments were then cultured in vitro and stimulated to produce antibody. The majority of the antibody from any one fragment culture was likely to be the product of a single donor B cell and thus monoclonal. Evidence for this included a linear relationship between donor cell transferred and spleen fragments producing antibody, extremely restricted isoelectric focusing patterns of the individual antibody products, and unique reactivity patterns of these antibodies against a panel of human lymphoid cells. Different human B leukemia cells were seen as immunogenically distinct by the mouse. By using the monoclonal mouse antibodies as probes, a fine analysis of cell surface antigens is jow possible.     

Hybridomas producing monoclonal antibodies to Crimean hemorrhagic fever virus

A series of hybridomas producing monoclonal antibodies to Crimean hemorrhagic fever virus was obtained and their cultural properties were characterized. HEMA-12 and HEMA-24 secreted monoclonal IgG2b antibodies, HEMA-101 secreted monoclonal IgG1 antibodies, HEMA-31, HEMA-9 and HEMA-11 secreted monoclonal IgG2 antibodies. According to the results of the indirect immunofluorescence test, the titer of specific immunoglobulins in the culture fluid was 1:16-1:32, but sometimes reached 1:64-1:128. The titer of antibodies in ascitic fluid amounted to dilutions of 10(4)-10(5). Hybridomas were cloned by the method of ultimate dilutions. The injection of 5-15 million HEMA cells into the abdominal cavity of BALB/c mice induced the formation of ascitic tumors in the animals within 7-11 days and the accumulation of ascitic fluid in a volume of 1-4 ml. Hybridomas, found to be capable of passage from mouse to mouse, underwent 5-16 passages during the term of observation.     

Hybrid hybridoma cell lines which express anti-zeatin riboside, anti-abscisic Acid, and hybrid antibodies

Hybrid hybridoma cell lines that secreted antibodies which reacted with two distinct plant hormones, abscisic acid (ABA) and zeatin riboside (ZR) were prepared and cloned. These cell lines were derived by polyethylene glycol-mediated fusion of HAT-sensitive anti-ZR hybridoma cells with splenocytes harvested from a BALB/c mouse previously immunized with an ABA-bovine serum albumin conjugate. Chromatographic analyses indicated that these lines expressed two different isotypes, each associated with a specific immunologic reactivity, and that the populations of immunoglobulins secreted by these hybridomas included antibodies directed against each individual hapten as well as hybrid molecules which reacted simultaneously with both. Hybrid hybridomas such as these should provide antibody populations useful for simultaneous isolation of multiple plant hormones from individual plant samples.     

Acquisition and maturation of expressed B cell repertoires in normal and autoimmune mice

The expressed B cell repertoires of mice from 1 day to 5 mo of age were examined. The ELISA-spot assay was used to enumerate individual Ig-secreting splenic B cells and determine the proportion of these cells producing antibodies reactive with each of six autoantigens and two conventional Ag. Results indicate that i) neonatal B cells producing antibodies against specific members of the Ag panel arose in a temporally defined sequence, ii) antibodies produced by 1- to 5-wk-old mice appeared to be more cross-reactive than those produced by adult mice, iii) no bias toward autoantibody production was found in the neonatal repertoires of autoimmune-prone mice, and iv) as a whole, the pattern of repertoire development among diverse strains of mice was highly conserved, although individual mice varied considerably in the absolute number of B cells committed to the production of antibodies of a given specificity.     

Quantification of antibody production of individual hybridoma cells by surface plasmon resonance imaging

Surface plasmon resonance imaging (SPRi) is most frequently used for the label-free measurement of biomolecular interactions. Here we explore the potential of SPRi to measure antibody production of individual hybridoma cells. As a model system, cells from a hybridoma, producing monoclonal antibodies recognizing epithelial cell adhesion molecule (EpCAM), were used. Recombinant human EpCAM protein was immobilized on an SPR sensor and hybridoma cells were introduced into an IBIS MX96 SPR imager and the SPRi response was followed for 10 h. SPRi responses were detected on the spots of the sensor only where ligands of the produced antibody were present. By measuring the SPRi signals on individual cells the antibody production of the individual cells was measured and production rates were calculated. For 53 single EpCAM hybridoma cells the production ranged from 0.16 to 11.95 pg (mean 2.96 pg per cell, SD 2.51) over a period of 10 h. Antibody excretion per cell per hour ranged from 0.02 to 1.19 pg (mean 0.30, SD 0.25). Here we demonstrate for the first time that antibody production of individual cells can be measured and quantified by SPRi, opening a new avenue for measuring excretion products of individual cells.