Targeted fusion of antibody-secreting cells: Unlocking monoclonal antibody production with hybridoma technology

ABSTRACT

Hybridomas, the first method for creating monoclonal antibodies (mAbs), were reported 50 years ago. This approach, which transformed biomedical research and laid the foundation for many of the current therapeutic, diagnostic, and research reagent applications of mAbs, is still used today, despite reported low fusion yields between short-lived B cells and immortal myeloma cells. To improve hybridoma production yields and accelerate development of new mAbs, we addressed two key limitations: 1) random pairing between myeloma cells and antibody-producing cells, and 2) low efficiency of the polyethylene-glycol-mediated fusion process. We first characterized and isolated antibody-secreting cells (ASCs) from the spleen of immunized mice before cell fusion to increase the probability of successive pairing between the most suitable cell fusion partners and favor the generation of functional hybridomas. Specifically, we developed an optimized workflow combining fluorescence-activated cell sorting with antibody secretion assays, using a panel of five cell-surface markers (CD3, TACI, CD138, MHC-II, and B220) to identify a distinct ASC subset with key characteristics. Such ASCs exhibited a plasmablast phenotype with high MHC-II expression and secreted high levels of antigen (Ag)-specific antibodies in immunized mice. We then implemented a cell electrofusion procedure adapted to low cell numbers (<10⁶ cells), in order to perform the targeted electrofusion of TACI^highCD138^high sorted ASCs. This targeted approach yielded viable hybridomas in 100% of seeded culture wells compared to only 40% for the electrofusion of unsorted cells. In particular, over 60% of hybridomas generated from TACI^highCD138^high sorted ASCs secreted Ag-specific mAbs, including IgGs with high Ag binding affinity (<10⁻⁹ M). These results pave the way for a high-yield mAb production method via cell fusion, with the potential to streamline hybridoma generation and thereby expand access to mAbs.

GRAPHICAL ABSTRACT

Introduction

Antibodies, also known as immunoglobulins (Ig), are large glycoproteins that are produced, cell-surface expressed, and secreted by B lymphocytes, typically in response to an infection or immunization. Each mature B cell produces only one type of antibody with the same single antigenic specificity. These monoclonal antibodies (mAbs) are highly specific and have established themselves as one of the largest groups of biologics (biotherapeutic proteins) now used in various therapeutic applications.^1,2

Since the approval of the first therapeutic mAb, muromonab-CD3 (Orthoclone OKT3), in 1986, the Food and Drug Administration (FDA) has approved more than 100 therapeutic mAbs for various diseases, including autoimmune disorders and inflammatory diseases and cancer.³ In addition to their therapeutic potential, mAbs are widely used as in vitro and in vivo diagnostic tools due to their high specificity and targeted binding affinity. In medical imaging, mAbs or their derivatives are also tools to detect tumors in early stages of development.⁴ In research, antibodies are essential in many techniques, such as ELISA, flow cytometry, immunocytochemistry and immunohistochemistry, immunofluorescence, Western blotting, and immunoprecipitation.^5,6 The increasing importance of mAbs as therapeutics and diagnostics highlights the critical need to accelerate and optimize their production and development methods.

Fifty years ago, in 1975, Köhler and Milstein reported a method to immortalize B cells producing mAbs, since these cells cannot survive in culture for more than a few days.⁷ Using this method, hybridoma cells are formed by fusion between a short-lived, antibody-producing B cell and an immortal myeloma cell. This fusion typically uses spleen cells harvested after mouse immunization with the antigen (Ag) of interest, generally using polyethylene-glycol (PEG). After a few days of culture in a selective culture medium that allows only fused cells to survive, Ag-specific antibodies in hybridoma culture supernatants are evaluated. Cells secreting Ag-specific antibodies are subsequently cloned, usually by limiting dilution, to isolate single hybridomas secreting Ag-specific Abs and achieve monoclonality. Each hybridoma clone constitutively produces one specific mAb, and cell lines are cryopreserved to ensure long-lasting mAb production. Hybridoma technology has been used extensively in antibody discovery campaigns.⁸ As the resulting antibodies are typically of murine origin, chimerization and humanization strategies have been used to reduce their immunogenicity and enhance their therapeutic efficacy.

By using B cells from animals such as mouse, rat or rabbit, the hybridoma technology preserves the natural ability of their immune system to generate highly specific mAbs. Indeed, the B cells that enter the cell fusion process to produce Ab-secreting hybridomas have undergone all steps of the differentiation process, such as somatic hypermutations in the complementary-determining regions (CDRs) of the variable genes, thus leading to a strong increase in affinity for the antigen. This differentiation process occurs in the secondary lymphoid organs, in the germinal centers formed during an immune response, as a result of follicular helper T and activated B cells collaboration, leading to plasmablasts and plasma cells that can secrete substantial amounts (14 000 Abs/min) of Ag-specific antibodies.^9–12

One major limitation of hybridoma technology is the rarity of the fusion events between B cells and myeloma cells (5 × 10⁻⁶ efficiency with conventional PEG-based fusion),¹³ resulting from the random pairing between the two cell partners and the low efficiency of the PEG-based cell fusion mechanism. Consequently, many B cells that may produce high-affinity antibodies to the target antigen are lost. To address this challenge, strategies such as B cell enrichment prior to cell fusion have been used to facilitate successive pairing, thereby enhancing cell fusion efficiency.^14,15 In addition to PEG-based fusion methods, alternative approaches, including electrofusion, have also been investigated to enable the generation of stable antibody-producing hybridomas. This technique uses electrical pulses to induce membrane fusion, allowing for greater control and improved hybridoma yield.¹⁶

The overall yield of the hybridoma procedure can be evaluated in several ways, such as identifying the number of ELISA-positive wells indicative of the presence of Ag-specific Abs in cell supernatants,^17,18 enumerating the individual colonies of growing hybridomas, which can then be compared to the number of input myeloma or spleen cells,¹⁹ or use of membrane or nuclear fluorescence markers to confirm cell fusion events by microscopy or flow cytometry, when no information regarding antibody secretion is required.^19,20

During the 50 years since the hybridoma method was first reported, numerous other methods for producing antibodies directly from B cells have evolved as alternatives to cell fusion. As early as 1995, Lagerkvist et al. isolated single, Ag-specific human B cells using Ag-coated magnetic beads to generate Ag-specific recombinant antibody fragments.²¹ O Starkie et al. developed a multi-parameter flow cytometry single, IgG⁺ memory B cells sorting technique, followed by direct amplification of V_H and V_L region encoding genes and subsequent expression in cell culture systems, allowing the preservation of the original V_H and V_L pairing.²² It is undeniable that single-cell recombinant antibody technologies offer the essential advantage of long-term stability of transfected cell lines compared to hybridomas. Following the rearrangement of their genetic material, hybridomas can, in some cases, contain one or more additional productive heavy or light chains, which can negatively affect properties of the antibodies, including their specificity.^23,24 However, single B cell-based techniques remain time-consuming and technically challenging, requiring many micromanipulations and further work to produce recombinant antibodies derived from the B cells that are selected.

Despite its limitations, hybridoma technology remains robust and requires no specific instrumentation, making it easy to implement in research laboratories. For example, we have used the method to develop numerous mAbs for immunological tests for the diagnosis/detection of infectious diseases or biological warfare agents (including commercially available tests) and immunotherapies.^25,26

To improve hybridoma technology further, we developed a new approach based on the selection of antibody-secreting cells (ASCs), the terminally differentiated form of B cells, before the cell fusion step.

Specifically, by selecting ASCs, including plasmablasts and plasma cells, based on cell-surface markers expression, the probability of successive fusion between cell fusion partners of interest can be increased, leading to a high proportion of functional hybridomas among all fused cells (high yield).

First, we implemented an optimized labeling panel and workflow to characterize splenic cells using flow cytometry and cell sorting. To identify the cell populations of interest (i.e., ASCs), we measured their Ab secretion levels in culture supernatant several days after sorting, using dedicated ELISA assays. We compared the results from naive and immunized mice to highlight the differentiation process of B cells and identify the activated cell populations following immunization. We then applied the developed labeling panel to analyze a hybridoma cell mixture obtained by conventional fusion (before culture and cloning). We analyzed the expression of ASC cell-surface markers on the hybridomas generated and the potential of the different sorted populations to secrete Ag-specific antibodies. Finally, we performed spleen cell sorting by flow cytometry to enrich the sample in ASCs and targeted fusion of these cells to favor the production of hybridomas-secreting specific antibodies with high binding affinity for the target antigen.

Results

Characterization of ASCs phenotype using spleen cells from naive and immunized mice

To improve the efficiency of hybridoma generation from spleen cells, we aimed to select the most suitable ASCs, including plasmablasts and plasma cells, before cell fusion.

We selected a panel of five cell-surface markers specific to the B cell lineage (i.e., CD3, TACI, CD138, MHC class-II, and B220), and used this panel to identify ASCs among spleen cells from a naive and an immunized Balb-c mouse by flow cytometry (Figure 1). We chose these markers because recent studies have evaluated the B cell maturation process and have characterized mouse splenic plasma cells and plasmablasts using specific cell-surface proteins.^12,27–29 We implemented a gating strategy (used for all experiments) to select single viable B cells from the spleen and determined the best marker combination to discriminate distinct ASC subsets. We first analyzed the expression of TACI and CD138 markers to select TACI^highCD138^high plasmablasts (PB) and plasma cells (PC), as previously described.²⁷ We observed that the immunization process often resulted in an increase of this spleen cell subset compared to the naive mice (i.e., 0.74% in Figure 1a vs 2.6% in Figure 1b). Then, we examined the expression of MHC class II and B220, two well-known B cell markers whose expression decreases during the maturation of PB into PC.^28,30,31 This gating strategy allowed us to distinguish three TACI^highCD138^high cell populations according to B220 and MHC-II expression levels (Figure 1a,b), namely B220^highMHC-II^high (P1), B220^low/intMHC-II^high (P2), and B220^lowMHC-II^low/int (P3). In this experiment using NheB as the target antigen, P1 cells were more prominent in the naive mouse (about 66% in Figure 1a) than in the immunized mouse (about 24% in Figure 1b). P2 cells were underrepresented in the naive mouse (13% in Figure 1a) compared to the immunized mouse (57% in Figure 1b). P3 cells were present at similar levels, i.e., around 19%, for both naive and immunized mice.

Figure 1.

Analysis of ASCs from naive mouse and immunized mouse spleens. Flow cytometry analysis of ASCs from naive (a) vs. immunized (b) mice using the following gating strategy. A first gate was set on single lymphocytes through physical parameters (FSC-A vs. SSC-A) and on (SSC-H vs. SSC-A) to eliminate doublets, and a second gate on fluorescence parameters (viability dye) to exclude dead cells. Then, we focused on CD3⁻ cells to exclude the CD3⁺ T lymphocytes and targeted the CD138^highTACI^high population that we subdivided into P1, P2 and P3 subsets according to MHC-II and B220 expression. P1 cells are included in the red gate, P2 cells in the blue gate, and P3 cells in the orange gate. hi = high, int = intermediate, lo = low. Levels of total (c) and Ag-specific secreted antibodies (d) in cell culture supernatants 4 days after sorting of spleen cells from immunized and naive mice. For each cell population, the absorbance was normalized to the number of cells per well. Error bars = ± SD (technical duplicates). Data from a single experiment using two individual mice, one mouse having been immunized against NheB.

We evaluated the antibody secretion levels in culture supernatant from each cell subset using ELISA assays to characterize the three identified cell populations. Specifically, we sorted around 10,000 cells from P1, P2, and P3 populations using spleen cells from the mice described above, and cultured them for 4 days in complete media supplemented with IL-4 and CD40L to mimic T-cell interactions and promote cell survival. We then measured the absorbance levels resulting from total secreted antibodies in cell supernatants, including IgG and IgM, by spectrophotometry (Figure 1c). The absorbance levels resulting from Ag-specific secreted antibodies were only measured for the immunized mouse (Figure 1d). For the naive mouse, P1 and P2 cells did not secrete antibodies at measurable levels, contrary to P3 cells whose secretion level was considerably higher. For the NheB-immunized mouse, P1 cells did not secrete any measurable antibodies (total or Ag-specific Figure 1c,d). In contrast, P2 and P3 cells secreted antibodies at high levels similar to P3 in naive mouse, with P2 secreting about twice as many Ag-specific antibodies than P3 cells (Figure 1d). P2 and P3 were thus consistent with an ASC phenotype, while P1 cells appeared as non-secreting for all tested conditions.

Additional data obtained using two other mice immunized with VIM-I from independent experiments showed similar results for phenotyping and antibody secretion profiles (Supplementary data Figure S1 and Figure S2), indicating that P1, P2, and P3 cells could also be found in splenocytes obtained using different sets of immunization conditions and were not specific to a single antigen or mouse.

Evaluation of IgG expression and Ab secretion levels in ASCs

To characterize further the different P1, P2, and P3 cells, we used the membrane expression of IgG as an additional cell-surface marker and measured the secretion levels of the resulting cell subpopulations (Figure 2). We generated these data from the same immunized mouse as in Figure 1. Results showed that 75% of P1 cells highly express membrane IgG, while 47% of P2 cells express membrane IgG at a medium level, and 28% of P3 cells at a medium to low level (Figure 2a). Following subsequent cell sorting of P1, P2, and P3 based on IgG expression, we performed dedicated immunoassays to measure Ab secretion levels in cell culture supernatants (Figure 2b,c). Both IgG⁺ and IgG⁻ cells within the P2 and P3 cell subsets secreted antibodies at similarly high levels, although secretion levels from IgG⁻ cells tended to be slightly higher (Figure 2b). Regarding Ag-specific antibody secretion, once more IgG⁺ and IgG⁻ cells from the P2 and P3 cell subsets secreted antibodies (Figure 2c). Follow-up experiments conducted using spleen cells from mice immunized with VIM-I resulted in similar tendencies regarding IgG expression and secretion (Supplementary data, Figures S1 and S2). In these additional studies, we also evaluated IgM expression in P1, P2, and P3 cell subsets (Supplementary data, Figure S1a) and observed that P3, which only expressed membrane IgG at 24%, expressed IgM at 75%. Overall, these results showed that P2 and P3 can be considered as ASCs, and it is interesting to note that they can both secrete and express Abs on the cell surface.

Figure 2.

Analysis of membrane-expressed IgG and secreted Abs from P1, P2, and P3 ASCs from the spleen of a mouse immunized with NheB. ASCs expressing IgG at the cell surface were defined using IgG1, 2ab-PE-Vio770 antibody (a). The percentage of each ASC population expressing IgG (IgG⁺) or not (IgG⁻) is indicated in each histogram whose color is the same as the ASCs identified on the dot plots from Figure 1b. Around 5,000 cells of each ASCs subset were isolated. Detections of corresponding total antibodies (b) and Ag-specific antibody secretion (c) were performed in cell culture supernatants 4 days after cell sorting. For each cell population, the absorbance is normalized to the number of cells per well. Error bars = ± SD (technical duplicates). Representative data from a single experiment are shown. Additional experiments performed using two other mice immunized with VIM-I gave similar results (see Supplementary data, Figures S1 and S2).

Determination of the phenotype of hybridomas secreting Ag-specific Abs

We aimed at identifying more precisely which of the spleen cell populations has a higher probability of leading to hybridomas that can secrete Ag-specific Abs. Here, around 10⁸ spleen cells from another mouse immunized against VIM-I were fused with myeloma cells using a conventional PEG-mediated cell fusion procedure. After 10 days of culture in the selective medium, we analyzed, before cloning, the resulting crude mixture of cells including visible viable hybridomas and dead cells (e.g., non-desired fused cells and non-fused dead cells). Cells were stained with the five-marker panel described above by flow cytometry. We also evaluated the expression level and specificity of membrane IgGs using biotinylated Ag combined with Streptavidin-PE as described by Parks et al..³² The CD138 marker was not analyzed here since we considered all fused cells to be CD138^high, due to its high expression on myeloma cells (Supplementary data, Figure S3). Fluorescence-activated cell sorting (FACS) dot plots in Figure 3a show that we obtained 8% of viable cells corresponding to hybridomas. As expected, the majority of events on this plot were consistent with the presence of cell debris resulting from the death of non-fused cells due to the selective medium and the rarity of PEG-mediated cell fusion events. Among the selected living cells, 87% of hybridomas were TACI⁺, of which 85% were MHC-II⁺B220⁻, thus corresponding to the identified ASC phenotype.

Figure 3.

Identification of hybridoma secreting Ag-specific antibodies after fusion of spleen and myeloma cells. After ten days of culture in selective media, all hybridoma cells were collected and stained using a viability dye (viobility dye 405/520), TACI-APC, B220-VioBlue, MHC-II-PerCP-Vio700, IgG-PE-Vio770 antibodies and Ag-biotin/Streptavidin-PE. A first gate was set on viable hybridoma cell singlets through physical parameters (FSC-A vs. SSC-A) and on (FSC-H vs. FSC-A), and a second gate on fluorescent parameters (viability dye) to exclude dead cells. Then, we selected the TACI⁺ cells to analyze the expression of MHC-II and B220 together with cell surface IgG and their Ag specificity (a). Each of the 3 cell subsets identified (IgG⁺Ag⁺, IgG⁺Ag⁻, IgG⁻Ag⁻) was isolated in single cell mode in separate wells. After 7 days, wells containing growing cells were tested for Ag specificity with dedicated immunoassays (b). Absorbance values are represented in box and whiskers plots with minimum, first quartile, median, third quartile and maximum values. Representative data from a single experiment using one mouse having been immunized against VIM-I is shown. An additional experiment performed using a second mouse, gave similar results (not shown).

We then analyzed membrane expression of Ag-specific IgGs to divide TACI⁺ sorted cells into three new cell subsets: IgG⁺Ag⁺ (26%), IgG⁺Ag⁻ (42%), and IgG⁻Ag⁻ (31%). About 400 cells from each subset were isolated using single-cell sorting mode by flow cytometry in separate 96-well plates. After 7 days of culture, we evaluated Ag-specific antibody secretion in the supernatants of the wells containing growing cells originated from a single isolated clone (35–43% of wells). Figure 3b shows that 100% of IgG⁺Ag⁺ cells secreted Ag-specific antibodies with a median absorbance value of 1.4, while none of the IgG⁺Ag⁻ and IgG⁻Ag⁻ cells secreted them at measurable levels. These results suggest that the only hybridoma cells capable of secreting Ag-specific antibodies also expressed them on the cell surface.

Targeted fusion of TACI^highCD138^high sorted spleen cells

To verify that fusion between ASCs and myeloma cells favored the generation of hybridomas secreting Ag-specific antibodies, compared to random spleen cells, we sorted TACI^highCD138^high cells (i.e., expecting >60% of ASCs) from the spleen of a VIM-I-immunized mouse by flow cytometry (Figure 4). We isolated around 300 000 cells that we fused with myeloma cells using an electrofusion protocol that was better suited for cell fusion at low cell numbers (<10⁶ cells) than the conventional PEG-based procedure. As a control, 300 000 non-sorted spleen cells from the same immunized mouse were fused in parallel (see workflow Figure 4a). Overall fusion yields for both conditions were calculated using the number of wells containing growing hybridomas divided by the total number of plated wells.¹⁷ Due to the low number of isolated cells, only seven wells were plated for each condition, at 1 to 2 × 10⁵ cells/mL to preserve cell viability. As a first demonstration, we obtained a 100% fusion yield for TACI^highCD138^high cells (i.e., seven of seven wells containing growing hybridomas), while for non-sorted spleen cells, cell growth was only observed in three of seven wells (about 40% fusion yield).

Figure 4.

Identification of hybridoma secreting Ag-specific antibodies after targeted fusion of TACI^highCD138^high spleen cells. Workflow of the entire process (a): TACI^highCD138^high cells from a mouse immunized against VIM-I were sorted by FACS and electrofused with NS1 myeloma cells. As a control, non-sorted spleen cells from the same mouse were electrofused with NS1 myeloma cells. After 14 days of culture in selective media, cells were collected and stained using TACI-APC, B220-VioBlue, MHC-II-PerCP-Vio700, IgG-PE-Vio770, IgM-APC-Vio770 antibodies and Ag-biotin/Streptavidin-PE. For both TACI^highCD138^high fused cells (b) and non-sorted fused cells (c), a first gate was set on single viable hybridoma cells through physical parameters (FSC-A vs. SSC-A) and a second gate on fluorescent parameters (viability dye) to exclude dead cells. Then, we excluded the TACl⁻ cells to analyze the expression of MHC-II and B220 together with cell surface IgG and their Ag specificity. Each of the four cell subsets for TACI^highCD138^high fused cells and the one subset for non-sorted fused cells were isolated in single cell mode in separate wells. For IgG^lowAg^low cells, cell surface IgM expression was evaluated and IgM⁺ and IgM⁻ cells were also isolated in single cell mode in separate wells. After 7 days, wells containing growing cells were tested for Ag specificity (d) and total Ab (e) secretion with dedicated immunoassays. Absorbance values are represented in box and whiskers plots with minimum, first quartile, median, third quartile and maximum values. An additional experiment performed using another mouse, gave preliminary results (see Supplementary data, Figure S6). high = hi/intermediate = int/low = lo.

After 14 days of incubation in the selective medium, we analyzed the membrane expression of TACI, B220, MHC-II, IgG, and the Ag specificity of the resulting hybridomas from both tested conditions by flow cytometry (Figure 4b,c). After gating for TACI⁺, more than 90% of selected cells displayed an ASC phenotype, with MHC-II⁺B220⁻ expression, for both cell fusions. Subsequently, the analysis of cell-surface expressed Ag-specific IgGs revealed distinct profiles depending on the initial cells used for fusion. For the hybridomas produced by fusion with non-sorted cells, there was only one population with IgG^lowAg^low phenotype. For the hybridomas produced by fusion with TACI^highCD138^high sorted cells, we could distinguish four different cell subsets: the major one was IgG^intAg^int (60% of cells), a second only present at 1.5% was IgG^intAg^high, and the two others were IgG^highAg^low and IgG^lowAg^low found at 4% and 33%, respectively. IgM expression analysis on IgG^low cells from both non-sorted and TACI^highCD138^high fused cells showed that these respective subsets expressed IgM at 33% and 84%.

We then clonally isolated the following cell subsets: IgG^intAg^int, IgG^intAg^high, IgG^highAg^low, IgG^lowAg^lowIgM⁺, and IgG^lowAg^lowIgM⁻ from the cell fusions using flow cytometry in single-cell mode in separate 96-well plates (plated at one cell per well) and evaluated Ag-specific Ab secretion levels in supernatants after 7 days of culture (Figure 4d). We observed that only cells expressing Ag-specific IgGs at the cell surface could secrete them (median values of 1.59 and 1.78 for IgG^intAg^int and IgG^intAg^high, respectively). It is important to note that, for the non-sorted fused cells, no hybridomas were able to secrete Ag-specific Abs.

To confirm these findings, we performed a new targeted fusion of TACI^highCD138^high spleen cells using another mouse immunized with VIM-I from an independent experiment, including a control experiment using non-sorted spleen cells. We observed similar results to those in Figure 4: a cell subset with a IgG⁺Ag⁺ phenotype that secreted Ag-specific Abs was only present in hybridomas generated from the TACI^highCD138^high sorted cells (Supplementary data, Figure S6).

Since the other cell subsets displayed an ASCs phenotype without secretion of Ag-specific Abs (IgG^highAg^low, IgG^lowAg^lowIgM⁺, and IgG^lowAg^lowIgM⁻), we investigated the secretion of total Abs (IgG and IgM) in cell supernatants (Figure 4e). Interestingly, we found that all cell subsets secreted Abs at varying levels. While IgG^intAg^int, IgG^highAg^low, IgG^lowAg^lowIgM⁺, and IgG^lowAg^lowIgM⁻ from TACI^highCD138^high fused cells secreted high levels of total Abs (median value >0.5), IgG^intAg^high from sorted cells and IgG^lowAg^lowIgM⁺, IgG^lowAg^lowIgM⁻ from non-sorted cells secreted lower levels of Abs (median value ≤0.5). Isotyping was then conducted for each subset to ascertain the phenotype between IgG and IgM (see Supplementary data, Figure S4). We observed that IgM⁺ cells secreted IgM antibodies, while IgG^int/high cells secreted IgG1_kappa antibodies, and IgG^intAg^high and IgG^lowAg^lowIgM⁻ from non-sorted cells secreted IgA antibodies.

Finally, to characterize further the Ag-specific Abs produced with the targeted fusion of TACI^highCD138^high ASCs, we measured the binding affinities to VIM-I of four monoclonal IgGs produced by four individual IgG^intAg^int hybridoma clones by label-free kinetic analysis, using the Octet biolayer interferometry system. All four purified IgGs demonstrated specific high-affinity binding to VIM-I. Specifically, three of them showed binding affinities ranging from 0.21 to 0.35 nM and the forth an affinity of 3.37 nM due to a higher dissociation rate (K_off = 1.07 × 10⁻³ s⁻¹ vs. 1.5 to 2.0 × 10⁻⁴ s⁻¹, Figure 5). These results confirm that our approach enables the production of high-affinity Ag-specific mAbs.

Figure 5.

Binding affinities of four purified anti-VIM-I mAbs. The binding affinities of four different VIM-I specific IgGs secreted by four hybridomas selected among the IgG^intAg^int population of Figure 4 and purified from culture supernatants were measured by biolayer interferometry assay using an Octet® RED96 system. Graphics represent VIM-I:anti-VIM-I antibodies binding using 10 ng/ml immobilized anti-VIM-I antibodies against VIM-I titrated from 32 to 4 nM in a 2-fold serial dilution with corresponding kinetic constants of association, dissociation and affinity, of the four anti-VIM-I mAbs. K_D is the equilibrium dissociation constant calculated by k_off/k_on; k_on is the association rate constant; k_off is the dissociation rate constant.

Discussion

The study presented here provides a better understanding and an improvement of hybridoma technology by identifying and characterizing ASCs as the most suitable partners among spleen cells for the generation of hybridomas secreting Ag-specific antibodies.

ASCs are differentiated from activated B cells following immunization. We analyzed several combinations of markers specific to the B cell lineage to select this population over other spleen cells. The cell-surface markers were chosen due to their upregulation or downregulation during B cell maturation. The expression of CD19 and CXCR4³³ was initially tested for B cell assessment, but flow cytometry analyses using these markers did not allow for sufficiently precise discrimination of different cell subsets of interest (data not shown). As a result, a panel including CD3, TACI, CD138, MHC-II, and B220 markers, which allowed the definition of three distinct B cell subsets in naive and immunized mice spleen, named P1, P2, and P3, was selected for subsequent analyses. In this study, the P1 cell subset TACI^highCD138^highB220^highMHC-II^high appeared as non-secretory cells for both tested conditions but displayed a high cell surface expression of IgG for the immunized mouse. For a more detailed assessment of IgGs at the surface of P1, we performed an additional flow cytometry analysis using another sample of spleen cells from the same mouse as in Figure S2, replacing the TACI staining with a biotinylated Ag combined with Streptavidin-APC (limited number of fluorophores available). This approach revealed a P1* population analogous to P1 (CD138^highB220^highMHC-II^high, defined in supplementary data Figure S5), and confirmed that these surface IgGs were specific for the target antigen. Therefore, although P1 did not secrete Abs, this cell subset is similar to what is usually referred to as early plasmablasts.³⁴ Similarly, we demonstrated that P2* and P3* surface expressed Ag-specific IgGs.

gThe P2 cell subset TACI^highCD138^highB220^low/intMHC-II^high are ASCs along with the P3 cell subset TACI^highCD138^highB220^lowMHC-II^low/int since both expressed and secreted IgG, as previously described.^10,35–37 However, while P2 cells expressed membrane IgGs and secreted Ag-specific Abs, P3 cells expressed mainly IgMs with secretion of low-affinity Ag-specific Abs or non-specific Abs. Comparison of the representation of the three cell subsets in the immunized and naive mice revealed a significant decrease of P1 cells under immunization and a substantial increase of P2 cells in the spleen, suggesting a maturation process of P1 cells into P2 cells through Ag exposure. On the other hand, P3 cells were present in the same proportion in both naive and immunized mice. Taking into account the phenotype, IgM membrane expression, and Ab secretion levels of P3 cells, we suggest that these plasma-like cells may originate from an extrafollicular response in the spleen that does not involve T cell interaction, which could lead to the development of plasma cells that secrete low-affinity IgM.^11,38,39 P1 and P2 cells, however, seem to derive from a germinal center response where T cell interactions allow B cells class switch and affinity maturation by somatic hypermutations that result in the formation of highly Ag-specific cells. Nevertheless, P2 cells, which conserve PB characteristics with high MHC-II expression, surprisingly secreted high levels of Ag-specific antibodies. This observation could be explained by the fact that P2 cells underwent further maturation in plate wells after cell sorting due to IL-4 and CD40L enrichment in the culture media, leading to PC formation. However, our additional experiments revealed that neither IL-4 and CD40L nor the incubation time after cell sorting affected antibody secretion (Supplementary data, Figure S1). We also expected to identify a higher percentage of plasma cells secreting Ag-specific IgGs in the spleen, described as long-lived plasma cells (LLPCs)⁴⁰ and featuring P3 phenotype, but potentially some of these cells had already left the spleen and migrated to plasma cell niches in the bone marrow.⁴¹

P2 cells appeared as the most interesting ASCs to fuse and generate Ag-specific IgG-secreting hybridomas. To confirm this hypothesis, we used our five-marker panel to analyze the phenotype of a hybridoma mixture obtained after cell fusion, and revealed that these cells had an ASC-like phenotype with high expression of TACI and CD138, a low expression on B220 and an intermediate/high expression of MHC class II. Moreover, ELISA analysis demonstrated that the clones secreting Ag-specific Abs also expressed Ag-specific IgGs at the cell surface. We can thus hypothesize that this IgG expression in hybridomas may originate from spleen cells, probably from P2 and P3 ASCs.

Furthermore, the absence of phenotypes other than ASC on the hybridoma cells (>70% of hybridoma cells with TACI⁺/MHC-II⁺/B220⁻ phenotype) suggests that fusion events occur preferentially with ASCs. This alleged phenomenon could be explained by various factors influencing the cell fusion process, including the cell type, as well as cell size and the curvature of the cell membrane (either for PEG-based^42–45 or electrofusion^46–48). We speculate that there is a higher probability for NS1 myeloma cells, which are abnormal PC, to undergo fusion with cells of a similar type and a comparable size. Due to their extended Golgi apparatus, ASCs are larger in size and could be a more suitable fusion partner for NS1 cells than other B cells. Notably, the ASC phenotype of hybridomas cannot be attributed to NS1 cells that are TACI⁻, MHC-II⁻, and B220⁻ (Supplementary data, Figure S3). Additional investigations focusing on the fusion of myeloma cell lines different from NS1 and using a large panel of target antigens may provide further insight into this question.

Finally, as an early demonstration we sorted TACI^highCD138^high cells to increase the proportion of ASCs before fusion, successfully showing an enhancement in both the fusion yield and the number of generated Ag-specific-secreting hybridomas (Figure 4). In contrast, while the fusion of non-sorted spleen cells (i.e., when starting with the same number of 300,000 cells) allowed the generation of viable hybridomas (with a lower yield of 40% vs. 100%), it did not produce any Ag-specific-secreting clones. This demonstration was also supported by data from another independent experiment (Supplementary data, Figure S6). These results appear to support the hypothesis of a preferential fusion between myeloma cells and ASCs. Indeed, the lower fusion yield observed with non-sorted fused cells, along with the absence of antigen-specific hybridomas, could be due to the relatively low percentage of ASCs in the spleen (0.5–2%), which reduces the probability of successful pairings and fusion events with myeloma cells, especially when dealing with a limited number of cells. Enriching the spleen cell suspension with ASCs favored the generation of hybridoma, including two types of hybridoma cells: 1) IgG-expressing and IgG-secreting hybridomas, with some of the IgG being Ag-specific mAbs, and 2) IgM-expressing and IgM-secreting hybridomas, which are expected to display low affinity for the target antigen. By linking these results with spleen cell profiles, we can assume that IgG-producing cells may derive from P2 cells, while IgM-producing hybridomas may originate from P3 cells.

Preliminary attempts were made to fuse P2 cells exclusively (TACI^highCD138^highB220^low/intMHC-II^high) using the conventional PEG-based method. However, the very limited number of P2 cells recovered from the spleen, in conjunction with the low fusion efficiency of PEG, prevented successful hybridoma generation. Consequently, we opted to work with a larger population of cells (TACI^highCD138^high) and developed an electrofusion procedure adapted to the fusion of low numbers of cells. This procedure was found to be successful in the generation of hybridoma when fusing as few as 300,000 cells, but was not suited for fusion of a larger number of cells (such as a whole spleen). Therefore, no experiments of electrofusion on a whole spleen were carried out in this study. Although the results from PEG-based fusion using a high number of non-sorted spleen cells (Figure 3) or from electrofusion using a small number of ASC-enriched cells (Figure 4) cannot be directly compared, the latter method led to a five-fold increase in the number of generated Ag-specific secreting-hybridoma cells. This estimation was based on the IgG⁺Ag⁺ cells percentage reported to the total number of viable hybridomas counted before cell sorting. In addition, working with fewer cells simplifies the post-fusion process by reducing plate handling and facilitating downstream selection.

In conclusion, this study demonstrated that selecting ASC populations before cell fusion can enhance the efficiency of hybridoma production. Our findings suggest that fusion between spleen and myeloma cells primarily occurs with ASCs, which might explain the low frequency of fusion events in the conventional method and the improvements of our approach. In addition, by first isolating ASC populations, we were able to streamline the hybridoma method, and produce specific mAbs that bind Ag with nanomolar or sub-nanomolar affinities (Figure 5).

Overall, enhancing hybridoma technology may facilitate the broader use of this foundational technique and reduce reliance on animals in certain contexts (i.e., higher fusion yields would require fewer animals to be immunized).

Data limitations and perspectives

The FACS-based isolation of TACI^highCD138^high cells is labor-intensive and costly in terms of time and reagents. Efforts are underway to develop magnetic cell sorting as a more efficient alternative for ASC selection from spleen cells. Moreover, future studies will focus on the efficient isolation and exclusive fusion of Ag-specific ASCs to generate only Ag-specific secreting-hybridomas. A more detailed analysis of cell fusion efficiency at the cellular level will also be conducted, alongside a larger-scale study involving a larger number of mice for statistical validation, as well as the inclusion of a broader range of antigens to assess further the versatility of our approach.

Materials and methods

Mice immunization

gBalb-c mice (8–12 weeks), bred in the animal care unit at CEA, were immunized intraperitoneally with 50 μg per injection of purified antigen with alum hydroxide (1:1) after being anesthetized with isoflurane delivered through a vaporizer. The mice received four immunizations every 3 weeks and were left at rest for 2 months after the last injection. Mice were bled 2 weeks after each immunization to evaluate and monitor the polyclonal anti-antigen response in the sera using a specific enzyme immuno assay (see detailed procedure below). The mouse presenting the highest titer was selected for preparation of mAbs as previously described.⁴⁹ Before use, the animals were injected intravenously with 50 µg of antigen for three successive days and then were killed by cervical dislocation.

The NheB-immunized mice were already available in our facilities and used in the first experiments. NheB (produced and supplied by ANSES) is a subunit derived from Nhe, an enterotoxin secreted by Bacillus cereus. For the following experiments, mice were immunized against VIM-I, an enzyme responsible for antibiotic resistance.

Cell culture

The myeloma cell line NS1 was cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mM glutamine, 2 mM sodium pyruvate, 1% penicillin/streptomycin, 1% MEM with non-essential amino acids (all purchased from Sigma Aldrich) at 37°C and 7% CO₂.

PEG-mediated cell fusion

Splenocytes from mice immunized with the target antigen were recovered. Splenocytes and myeloma cells were washed and then suspended in RPM1 1640 serum-free medium (Sigma) at 3:1 ratio. Then, both cells were centrifuged together at 1,000 rpm for 10 min. After removing the supernatant, one milliliter of 50% w/v PEG 1500 (Sigma) was added dropwise to the cell pellet over 1 min with gentle stirring. After a 90 s incubation at 37°C, 50 mL of the serum-free culture medium were added dropwise. The cells were then centrifuged for 10 min at 1,000 rpm, washed twice, and resuspended in selective medium (RPM1 1640 supplemented with 15% heat-inactivated FCS, 1X hypoxanthine (ThermoFisher), 2 mM glutamine, 2 mM sodium pyruvate, 1% penicillin/streptomycin (all purchased from Sigma Aldrich). Cells were finally distributed in 6-well microtiter plates containing macrophages as feeder layer (5 × l0³ Balb/c peritoneal exsudate cells per well) and incubated at 37°C and 7% CO₂.

Cell electrofusion

Spleen cells were recovered from cell sorting and added to myeloma cells at 1:1 ratio. Cells were then washed with RPMI 1640, suspended in 80 µL of electrofusion buffer (100 mM Sorbitol, 0.5 mM MgCl₂, 0.1 mM Ca(CH₃COO)₂, 1 mg/mL BSA, pH = 7.5), transferred into 0.1 cm gap electroporation cuvette (BIORAD) and centrifuged for 5 min at 300 × g. The cuvette was placed in the Gene Pulser Xcell Electroporation System (BIORAD) and electrofusion was performed using a Square Wave Protocol of 1500 kV/cm electric field, 50 µs pulse duration, 3 repetitions with 0.1 s interval. After electrofusion, the cuvette was incubated for 10 min at 37°C. Cells were then slowly suspended, transferred into a 50 mL tube and diluted at about 1 × 10⁵ cells/mL with selective medium. Finally, after 20 min incubation at 37°C, cells were distributed in 96-well plates, 200 µL/wells leading to 7 wells plated containing around 43 000 cells, and incubated at 37°C and 7% CO₂. After several days of culture in selective media, fusion yield was calculated as the number of wells with growing hybridomas divided by the total number of wells plated just after fusion.

ASCs cell sorting

Mouse spleens were harvested and crushed through a 40 µm cell strainer (ThermoFisher) to extract splenocytes. Red bloods cells were lysed using Red Blood Cell Lysis Buffer (Sigma). Cells were then washed twice with phosphate-buffered saline (PBS) (-)CaCl₂ (-)MgCl₂ (Sigma) and stained for 15 min at 4°C according to manufacturer instructions using the following fluorescent antibodies (Miltenyi Biotec): Viobility 405/520 Fixable Dye (#130-130-404), CD3e-Vio®Bright FITC (#130-109-246), CD267(TACI)-APC (#130-124-101), CD138-PE (#130-120-810), CD45R(B220)-Vio®Blue (#130-110-851), MHC Class II-PerCP-Vio®700 (#130-112-391), IgG2ab-PE-Vio®770 (#130-099-064), IgG1-PE-Vio®770 (#130-117-103). Then, cells were washed twice with PBS and sorted using BD FACSAria™ III and BD FACSDiva software. Sorted cells were recovered in 15% FCS complete medium containing RPMI 1640 supplemented with 15% heat-inactivated FCS, 2 mM glutamine, 2 mM sodium pyruvate, 1% penicillin/streptomycin, 1% MEM with non-essential amino acids, 2% HFCS (Hybridoma Fusion and Cloning Supplement) (all from Sigma) and incubated at 37°C and 7% CO₂ for several days. FACS data analysis was performed using the FlowJo software V10.10.0 (BD Biosciences).

Hybridoma cell sorting

Cells were washed twice with PBS (-)CaCl₂, (-)MgCl₂ (Sigma) and incubated for 30 min at 4°C with biotinylated antigen at 5 µg/mL. Then, cells were washed and stained for 15 min at 4°C according manufacturer instructions using fluorescent antibodies (Miltenyi Biotec): Viobility 405/520 Fixable Dye (#130-130-404), CD3e-Vio®Bright FITC (#130-109-246), CD267(TACI)-APC (#130-124-101), CD138-PE (#130-120-810), CD45R(B220)-Vio®Blue (#130-110-851), MHC Class II-PerCP-Vio®700 (#130-112-391), IgG2ab-PE-Vio®770 (#130-099-064), IgG1-PE-Vio®770 (#130-117-103), IgM-APC-Vio®770 (#130-116-315) and Streptavidin-PE. After staining, cells were washed twice with PBS and sorted using BD FACSMelody™. Sorted cells were recovered in 15% FCS complete medium and incubated in 37°C and 7% CO₂ for several days. FACS data analysis was performed using FlowJo software V10.10.0.

Evaluation of polyclonal response and antibody detection in cell culture supernatants

Anti-antigen antibodies were detected in sera of immunized mice or in hybridoma and spleen cells culture supernatants using ELISA. 100 µL of serial dilutions of mouse serum diluted in an Enzyme Immuno Assay (EIA) buffer (0.1 M phosphate buffer pH=7.4 containing 0.15 M NaCl, 0.1% bovine serum albumin and 0.01% sodium azide) or cell culture supernatants were transferred into microtiter plates coated with goat anti-mouse IgG + IgM antibodies (Jackson ImmunoResearch Laboratories, #115-005-044) and incubated overnight at 4°C. For Ag-specific antibodies detection, plates were washed and incubated with 100 µL/well of biotinylated antigen at 100 ng/mL for 4 h at room temperature (RT). Acetylcholinesterase (AChE)-conjugated streptavidin was then added at 1 Ellman units [UE]/mL after washing and incubated for 1 h at RT. For total antibodies detection, plates were washed and incubated with 100 µL of goat anti-mouse-AChE antibodies at 2 Ellman units [UE]/mL for 1 h at RT. Finally, plates were washed and incubated with 200 µL of Ellman’s reagent (protected from light) and the absorbance was measured at λ = 414 nm after 30 minutes or 1 hour using Epoch spectrophotometer (BioTek instruments).

Antibody isotyping

Antibody isotyping was performed using a Pierce Rapid ELISA Mouse mAB Isotyping Kit (Invitrogen) according to the manufacturer’s instructions. Briefly, cell culture supernatants were diluted at 1:50 in 1 mL Tris Buffered Saline. Then, 50 µL of the diluted supernatant and 50 µL of Goat anti-Mouse IgG + IgA + IgM - HRP were added to each well of the 8-well pre-coated strip, and covered for 1 h at RT. After three washings using the wash buffer (potassium phosphate buffer 10 mM, 0.05% Tween 20), 75 µL of TMB substrate were added to each well and incubated for 15 min protected from light. Then, 75 µL of Stop Solution were added and absorbance values were measured at λ = 450 nm using Epoch spectrophotometer (BioTek instruments).

Affinity determination of anti-VIM-I antibodies

The affinities of antibodies for the VIM-I antigen were determined by bio-layer interferometry using the ForteBio system (Pall Laboratory). The evaluated antibodies were prepared at 10 µg/mL in EIA buffer + 0.02% Tween 20 (Sigma) and dispensed in a 96-well microplate at a volume of 200 µL per well, along with the recombinant VIM-I protein (used between 250 and 0 nM with a 2 in 2 dilution), a glycine regeneration solution (Sigma, pH=1.4) and EIA buffer + 0.02% Tween 20. Prior to binding measurements, the anti-mouse Fc (AMC) sensors tips were hydrated in EIA buffer + 0.02% Tween 20. The loading step was then performed by dipping the sensors in mAb-containing wells for 300 sec, followed by the baseline step in EIA buffer + 0.02% Tween 20 for 60 sec. The binding kinetics were then measured by dipping the mAb-coated sensors into the wells containing recombinant VIM-I at varying concentrations. The binding interactions were monitoring over 600 sec associated period and followed by a 600 sec dissociation period in the wells containing EIA buffer + 0.02% Tween 20. The AMC sensors tips were regenerated with glycine [pH=1.4] and neutralized in the EIA buffer + 0.02% tween 20 between each binding cycle. The equilibrium dissociation constant (K_D) was calculated using the ratio between the dissociation rate constant (k_off) and the association rate constant (k_on), obtained with global Langmuir 1:1 fit (Octet Data Analysis software, vHT.10).

Funding Statement

This work was supported by the Commissariat à l’Energie Atomique et aux Energies Alternatives (CEA). This information was already in the acknowledments but it can be moved to the funding section for more clarity.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Author contributions

Fanny Rousseau designed and performed experiments, analyzed the data, and wrote the paper. Catherine Menier contributed to FACS experiments and data analysis and revised the manuscript. Patricia Brochard contributed to fusion experiments and provided technical guidance. Stéphanie Simon supervised the complete study. Anne Wijkhuisen and Karla Perez-Toralla contributed to some experiments, conceived the manuscript, wrote the introduction, and revised the manuscript. All authors read and approved the final manuscript.

Data availability statement

This study includes no data deposited in external repositories.

Ethics statement

All experiments were performed in compliance with French and European regulations on the care of laboratory animals (European Community Directive 86/609, French Law 2001–486, 6 June 2001) and with the agreements of the Ethics Committee of the Commissariat à l’Energie Atomique (CEtEA ‘Comité d’Ethique en Expérimentation Animale’ n° 44) no. 12–026 and 15–055 delivered to S.S. by the French Veterinary Services and CEA agreement D-91-272-106 from the Veterinary Inspection Department of Essonne (France).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19420862.2025.2510336

Abbreviations

mAb

monoclonal antibody

Ab

Antibody

ASC

Antibody-Secreting Cell

FACS

Fluorescence-Activated Cell Sorting

Ig

Immunoglobulin

Ag

Antigen

PEG

Polyethylene glycol

PB

Plasmablast

PC

Plasma cell

ELISA

Enzyme-Linked ImmunoSorbent Assay

EIA

Enzyme Immuno Assay