B cells and the antibodies they produce provide an effective immune response to pathogens and vaccines. Through antigendependent differentiation, naive B cells differentiate to antibodysecreting cells (ASCs) and memory cells [1]. In addition to providing a humoral immune response, B cells can function as antigen-presenting cells (APCs) [2]. Upon binding of antigen to B cell receptor (BCR), the complex is internalized by B cells and the processed peptides are then presented on their surface in a complex with MHCII for subsequent interaction with CD4⁺ T helper (Th) cells [3]. Key signals to B cells from cellular cooperation with Th cells include  CD40-CD40L interaction as well as secreted cytokines IL-21 and IL-4 [4]. Upon acquisition of T cell help B cells undergo proliferation and differentiate either into short-lived ASCs or germinal center (GC) B cells (a fate driven by master transcription factor BCL6) [5, 6].

The unique properties of B cells make the in vitro cell cultivation critical for various applications such as obtaining APCs for analyzing antigen-specific T-cell responses as well as ASCs for antibody production [7, 8], and creating cell models for studying lymphomas [9]. Numerous studies have investigated how CD40L, T-cell cytokines and microenvironment factors such as BAFF influence B-cell cultures to promote proliferation, class switching, and antibody-secreting cell (ASC) formation [10–12]. However, studies on in vitro BCL6⁺ GC B cell formation from human B cells are still limited. The aim of the work was to find optimal conditions to expand B cells and to efficiently turn them into antibody-secreting cells (ASCs) or germinal center (GC) cells in vitro.

In our study, we analyzed human peripheral blood naive B cells and CD27⁺ memory B cells cultured with T-cell signals (CD40L, IL-21, IL-4), BAFF, and additional medium components. We also assessed B-cell expansion, and accumulation of ASC and BCL6⁺ B cells.

METHODS

pcDNA-hCD40LG Plasmid Design

Human CD40LG gene sequence (NM_000074.3) was obtained using polymerase chain reaction (PCR) from total cDNA of human peripheral blood mononuclear cells (MC) using forward 5'-ATATGGATCCGCCACCATGATCGAAACATACAACCA-3' and reverse 5'-ATATGAATTCACACTGTTCAGAGTTTGAGTAAGCC-3' primers. The primer sequences included BamHI and EcoRI restriction sites at their 5' and 3' ends of CD40LG gene, respectively, as well as Kozak sequence (GCCACCATG) at the 5' end of the gene. After digestion of PCR product and pcDNA3.1+ vector (Thermo Fisher Scientific, USA, Cat. No. V79020) using restriction endonucleases BamHI-HF and EcoRI-HF (New England Biolabs, USA, Cat. Nos. R3136 and R3101), they were ligated using the Quick Ligation™ Kit (New England Biolabs, USA, Cat. No. M2200) and transformed into E. coli (strain NEB Stable, New England Biolabs, USA, Cat. No. C3040H). The plasmid was purified and isolated using the Plasmid Miniprep 2.0 kit (Evrogen, Russia, Cat. No. BC221) for further transfection.

Generation of a Mouse Fibroblast Feeder Line Expressing Human CD40L (3T3-hCD40L)

3T3-hCD40L feeder line was generated from NIH 3T3 mouse fibroblast cell line from the cell culture collection of the D. I. Ivanovsky Institute of Virology. NIH 3T3 cells were cultured in DMEM (PanEco, Russia, Cat. No. C415p) supplemented with 10% fetal bovine serum (FBS, STEMCELL Technologies, Canada, Cat. No. 06472) and 1× penicillin/streptomycin antibiotic mixture (PanEco, Russia, Cat. No. A065p). Cells were passaged every 3-4 days. To generate NIH 3T3 clones stably expressing hCD40L, cells were transfected with pcDNA-hCD40LG plasmid using Lipofectamine® 3000 reagent (Thermo Fisher Scientific, USA, Cat. No. L3000015) according to manufacturer’s instructions, followed by clone selection in DMEM medium supplemented with geneticin G418 (0.5 mg/ml; Thermo Fisher Scientific, USA, Cat. No. 10131027). Monoclonal cell lines were generated by limiting dilutions. CD40L expression was assessed by flow cytometry using anti-CD154-FITC antibodies (1:20; clone TRAP1, BD Bioscience, USA, Cat. No. 561721). The monoclonal cell line with the highest CD40L expression level was used in the study (fig. 1A).

Isolation of Mononuclear Cells (MC) from the Peripheral Blood of Healthy Donors

The study was conducted in accordance with the 2013 Declaration of Helsinki. Patient exclusion criteria were (a) infectious diseases in the acute or recovering stage; (b) autoimmune or chronic diseases. Peripheral blood from healthy donors was collected in vacuum tubes containing K3EDTA as anticoagulant. A total of four independent, unrelated donors, aged 21 to 65 years (median 39 years), participated in the study. The male to female ratio was 1:1. MCs were isolated on a Ficoll gradient (PanEco, Russia, Cat. No. P050E) in accordance with the protocol [13]. The obtained cells were stored in cooled cell sorting buffer (0.5% FBS in Dulbecco's phosphate buffered saline (DPBS)). Cells were counted using a CytoSMART Cell Counter (Corning, USA) using trypan blue. Cell viability was > 95%.

B cell Staining for Flow Cytometry Analysis (FACS)

B cell analysis was performed by flow cytometry using antibodies to the cell surface markers CD19, CD27, CD38, CD95 (Fas) as well as the intracellular transcription factor BCL6. Cells were stained and stored at 4 °C. For sorting of all B cells (CD3^–CD19⁺CD20⁺), 20 × 10⁶ MCs were preincubated in blocking solution with addition of 10 μg/ml Human Fc-block (clone K112-91, BD Biosciences, USA, cat. no. 564220) for 10 min on ice, after which they were stained with a mixture of antibodies anti-CD3-R718 (1 : 100, clone SK7; BD Biosciences, USA, cat. no. 751978), anti-CD19-BV510 (1 : 100, clone SJ25C1; BD Biosciences, USA, cat. no. 562947), anti-CD20-FITC (1:20, clone L27; BD Biosciences, USA, cat. 347673) MCs were incubated for 30 min on ice in the dark. To sort CD3^–CD19⁺CD20⁺CD27^– naive B cells and CD3^– CD19⁺CD20⁺CD27⁺ memory B cells, MCs were additionally stained with anti-CD27-PerCP-Cy5.5 (1:50, clone M-T271; BioLegend, USA, cat. no. 356408). After the staining, the cells were resuspended in sorting buffer. To assess the phenotype of B cell cultures on day 7, the supernatant with B cells was collected from the wells without subsequent trypsinization. The selected cells were resuspended in a blocking buffer and then stained with the following antibody mixture: anti-CD19-BV510 (1:100), anti-CD27-PerCP-Cy5.5 (1:50), anti-CD38-APC-R700 (1:50, clone HIT2; BD Biosciences, USA, cat. no. 564979), anti-CD95-PE (1:100, clone DX2; BioLegend, USA, cat. no. 305608).

Prior to flow cytometry analysis, the cells were stained with Helix NP Blue at a concentration of 25 nM (Biolegend, USA, cat. no. 425305) to exclude dead cells. Intracellular staining of cells was performed using the True-Nuclear™ Transcription Factor Buffer Set (BioLegend, USA, cat. No. 424401) and anti-BCL-6-AlexaFluor488 antibodies (1:20, clone: K112-91; BD Biosciences, USA, cat. No. 561524) for 12 h. The following subpopulations were identified using cytometric analysis of cultures: CD19⁺CD27^highCD38^high (defined as ASC), CD19⁺CD38^–CD27^–, CD19⁺CD38^–CD27⁺, CD19⁺CD38⁺CD27^– B cells, CD19⁺CD95^high B cells (corresponding to activated B cells), CD19⁺CD95^highBCL6⁺ B cells (defined as GC B cells). B-cell sorting and flow cytometry (FACS) analysis of B-cell cultures were performed on a BD FACSAria™ III cell sorter using FACSDiva™ software. FlowJo version 10.8.1 was used for data processing.

Co-cultivation of B cells with Feeder Cells

3T3-hCD40L fibroblasts and control non-transfected cells were inactivated in DMEM with mitomycin C (5 μg/ml; Sigma Aldrich, USA, cat. no. 50-07-7) for 2 h, washed with DPBS (PanEco, Russia, cat. no. P060E) at least 3 times, and plated onto a sixwell plate (Wuxi NEST, cat. no. 703002) at a density of 3 × 10⁵ cells/well. The following day, 1.5 × 10⁴ sorted CD19^–CD20^– B cells were added. Cultivation was performed in the following media: RPMI-1 based on RPMI-1640 (PanEco, Russia, cat. no. C330p) with the inclusion of 10% FBS, 1× sodium pyruvate (PanEco, Russia, cat. no. F023), 1× GlutaMAX™ (Thermo Fisher Scientific, USA, cat. no. 35050061), 1× antibiotic mixture and RPMI-2 based on the richRPMI (BioinnLabs, cat. no. bn-3A3R), containing human transferrin, insulin, albumin, as well as glutathione and additional microelements and vitamins, with the inclusion of 10% FBS, 1× GlutaMAX™, 1× antibiotics. Recombinant human IL-21 (rhIL-21, SCI-store, cat. #PSG260) was added to the media at a concentration of 50 ng/ml. CD20^–CD27^– and CD20^–CD27^– B cells were cultured in RPMI2 supplemented with IL-21 ± IL-4 (SCI-store, cat. #PSG040, 10 ng/ml) and ± BAFF (BioLegend, cat. #559604, 100 ng/ml). Cultivation was carried out for 7 days with medium replacement on day 3 and day 5. The experimental schemes for culturing all CD19^–CD20^– B cells, as well as CD19^–CD20^–CD27^– naive and CD19^–CD20^–CD27⁺ memory B cells under the test and control conditions are presented in Tables tab. 1 and tab. 2.

Statistical Analysis

For the independent experiments described in the study, we used MCs from the blood of four genetically unrelated adult donors. Statistical analysis was performed using GraphPad Prism version 9.5.1 (GraphPad Software Inc, USA). The statistical methods are specified in the figure legends. Statistically insignificant differences between groups are not indicated on the graphs. Outliers are not excluded from the analysis.

RESULTS

In the first stage of the study, we determined how the presence of CD40L, the cytokine IL-21, or specialized additives in the cell medium impacts B-cell survival and expansion as well as the formation of ASCs and germinal center B-cells (fig. 1, fig. 2). А transgenic NIH 3T3 murine fibroblast cell line stably expressing hCD40L (3T3-hCD40L) was generated for long-term co-culture of B cells with human CD40L (hCD40L) (fig. 1A).

B cells (CD19⁺CD20⁺) were obtained from human peripheral blood MCs by fluorescence sorting (with a population purity of > 99%). 1.5 × 10⁴ B cells were cultured for 7 days in a sixwell plate in the presence of 3T3-hCD40L feeders or control untransfected 3T3 cells. Two media were used for cultivation: standard RPMI-1640 supplemented with 10% fetal bovine serum, sodium pyruvate, and glutamine (designated as RPMI-1) or the same medium enriched with human serum supplements including recombinant insulin, transferrin, lipid-rich albumin, glutathione as well as additional trace elements and vitamins (RPMI-2). IL-21 was added at a concentration of 50 ng/ml (tab. 1).

Flow cytometry analysis of the resulting B-cell cultures (fig. 1B) showed that the combined action of hCD40L and IL-21 resulted in the highest level of B-cell expansion (fig. 1C) and increased B-cell survival, especially when supplemented with recombinant proteins from human serum (fig. 1D).

Cultures with hCD40L and IL-21 also showed the highest accumulation of CD27^highCD38^high ASC (fig. 1E, F), an increase in the number (but not percentage) of CD27^–CD38^– B cells (fig. 1G, H), and accumulation of CD27⁺ and CD38⁺ B cells (fig. 1 I-L), with the percentage of CD27^–CD38⁺ B cells being significantly higher when cultured in RPMI-1 (fig. 1K).

GC B cells are characterized by increased surface expression of CD95 (Fas) and the presence of the intracellular transcription factor BCL6 [5]. CD95 expression also increases on activated B cells [15]. Based on this, we assessed CD95 expression on the surface of B cells in culture as a marker of cellular activation (fig. 2A).

Based on the data analysis, up to 97% of B cells in the culture increased their surface expression of CD95 in the presence of 3T3-hCD40L-cells (fig. 2A-C) while the addition of IL-21 did not significantly affect the percentage of CD95^high B cells (fig. 2B). Moreover, up to 80% of all CD95^high B cells had the CD27^–CD38^– phenotype (fig. 2D, E). The number of CD27⁺CD38^–CD95^high and CD27^–CD38⁺CD95^high B cells in cultures increased with the combined addition of hCD40L and IL-21 (fig. 2G, I) while their proportion did not change (fig. 2F, H).

To assess the accumulation of CD95^high BCL6⁺ (GC-like) B cells, we fixed surface-stained B cells with further intracellular staining with antibodies to BCL6 (fig. 2J). Quantitative analysis revealed a low representation of this B cell subpopulation in the cultures (< 2%) (fig. 2K). The most noticeable accumulation of GC-like B cells was observed in the RPMI-2 culture supplemented with 3T3-hCD40L feeders and IL21 (fig. 2L). In the RPMI-2 medium, more than 50% of GC-like B cells had the CD27^–CD38^– phenotype which is significantly higher than in cultures with RPMI-1 (fig. 2M, N). CD27⁺CD38^–CD95^high BCL6⁺ B cells were equally abundant in cultures with RPMI-1 and RPMI-2 (fig. 2O, P) while the proportion of CD27^–CD38⁺ GC-like B cells was significantly increased in cultures with RPMI-1 (fig. 2Q, R).

Thus, co-cultivation of B cells with 3T3-hCD40L feeders and IL-21 allows for the greatest expansion and survival of B cells. Under these conditions, on day 7 of culture in RPMI-1, approximately 15% of living cells were ASCs and 0.2% were GC-like B cells, with 10% and 0.4% for RPMI-2, respectively. RPMI-2 was chosen for further work as it was compositionally richer for B cells.

In the next series of experiments, we investigated the effect of different cytokine compositions (IL-4, BAFF) on cell cultures of CD20⁺CD27^– B cells which were predominantly naive B cells, and CD20⁺CD27⁺ memory B cells. Individual B cell subpopulations were obtained from peripheral blood MCs by fluorescence sorting (with a population purity of > 95%). 1.5 × 10⁴ B cells were cultured in RPMI-2 medium in the presence of 3T3 hCD40L cells supplemented with IL-21 (50 ng/ml) and the cytokines IL-4 (10 ng/ml) and/or BAFF (100 ng/ml) (tab. 2).

Flow cytometric analysis of cell cultures on day 7 (fig. 3A) showed that various cytokine compositions had no significant effect on B-cell expansion or survival (fig. 3B, C). For all cytokine combinations, a trend toward increased expansion and a decreased proportion of dead B cells was observed in cultures of CD27^– naive B cells compared with CD27⁺ memory B cells.

Importantly, cultures of CD27⁺ memory B cells showed a significant accumulation of CD27^highCD38^high ASC (20–35% of all B cells) (fig. 3D, E). In the cultures of predominantly naive CD27^– B cells, the proportion of ASC was less than 2% (fig. 3D, E) while the proportion of CD27^–CD38^– B cells (fig. 3F, G) was significantly higher compared to the cultures of CD27⁺ B cells. CD27⁺CD38^– B cells were abundant in memory B cell cultures, and CD27^–CD38⁺ B cells were present in all cases without significant differences (fig. 3H–K). Regardless of the starting population, adding IL-4 decreased the proportion of ASC, CD27⁺CD38^–, and CD27^–CD38⁺ cells (fig. 3D, E, H–K).

Regardless of the conditions and starting cells, the majority of cells had CD95^high phenotype. Moreover, the proportion and number of CD95^high B cells were slightly increased in CD27^– naive B cell cultures (fig. 4A-C). Notably, a trend toward a decrease in the CD27^–CD38⁺CD95^high B cell subpopulation was observed in the presence of IL4 under all other tested conditions (fig. 4D, E).

Analysis of the number and proportion of CD95^highBCL6⁺ B cells (fig. 4F) in cultures from CD27^– and CD27⁺ B cells revealed no statistically significant differences (fig. 4G, H). However, the addition of IL-4 was shown to increase the proportion of GC-like B cells (especially in the case of CD27^– B cells). Notably, a trend toward increased BCL6 levels in CD27^– B-cell cultures compared to CD27⁺ B-cells was observed under all tested conditions except the addition of BAFF without IL-4 (fig. 4I, J).

When cultured in RPMI-2 medium with 3T3-hCD40L feeders and IL-21, a significantly greater expansion of the naive CD27^– B-cell cultures with minimal ASC accumulation was observed as compared to the CD27⁺ memory B-cell cultures where ASC reached 30%. No statistically significant differences in the accumulation of GC-like B-cells were detected. The addition of BAFF had no significant effect on B-cell cultures while IL-4 caused a slight decrease in the proportion of ASC and an increase in GC-like B-cells, especially in CD27^– B-cell cultures.

DISCUSSION

In vitro cultivation of human B cells is of fundamental importance for solving a wide range of biotechnological problems, in particular, obtaining ASCs for antibody production, using B cells as antigen-presenting cells for T cell activation, and testing transgenic immunotherapy products [16]. Although there are numerous studies on B cell cultivation conditions [11–13], developing effective protocols for specific applications still remains important.

The aim of this study was to find an optimal combinations of conditions for (a) efficient B cell expansion in cell culture; (b) maximal accumulation of ASC population; (c) BCL6-expressing B cells similar to GC B cells. We comprehensively explored the impact of baseline B-cell status based on CD27 marker (CD27^– and CD27⁺ subsets), a combination of proteins and supplements replacing human serum, the cytokine BAFF that is critical for B-cell survival, and also the stimuli that mimic T-cell support (CD40L, IL-21, IL-4), on the efficiency of cell culture protocols.

Expansion and Maintenance of Activated B-Cell Viability

Cytometric analysis of B-cell cultures on day 7 of co-culture demonstrated that the highest expansion and survival rates were achieved using a transgenic feeder line expressing human hCD40L in combination with the cytokine IL-21. These data are consistent with previously described results [17, 18]. At the same time, the addition of IL-4 and BAFF cytokines to the culture at concentrations of 10 and 100 ng/ml, respectively, did not have a statistically significant effect on cell proliferation or viability.

Optimization of the cell culture medium composition showed that adding recombinant human insulin, transferrin, and albumin into a base medium containing 10% FBS resulted in the greatest reduction in percentage of dead cells.

Based on the comparative analysis, cultures initiated from CD27^– B cells were, on average, 1.5 times more abundant than cultures derived from CD27⁺ cells and also exhibited a lower percentage of cell death as has been previously described [19].

Thus, it is advisable to use CD27^– cells as the initial population for optimal expansion and maintenance of activated B cells viability in vitro. The most effective protocol involves cocultivation with feeder cells expressing hCD40L in a medium enriched with IL-21, supplemented with a combination of serum proteins (including insulin, transferrin, and albumin).

Accumulation of ASC in B-cell Cultures

The next task was to determine the cell culture conditions that promote accumulation of CD27^highCD38^high ASC in B-cell cultures. It was shown that in the presence of both hCD40L and IL-21, CD27^highCD38^high cells accumulated in the cultures reaching 10-15% of the total cell pool. This is consistent with the previously published results [20, 21, 22]. The addition of recombinant serum proteins to the culture medium resulted in a decrease in the proportion of ASC in the cultures. Significant accumulation of CD27^highCD38^high B cells was observed in the cultures of memory CD27⁺ B cells (20–35% of all living cells), consistently with previous studies [21, 23] and the rapid recruitment of memory B cells to the ASC response upon repeated exposure to antigen and T-cell help [24]. In contrast to the memory B cell cultures, in the cultures from predominantly naive CD27^– B cells the proportion and number of ASC were significantly reduced (1–2% of the total cell number).

It was also demonstrated that in the presence of the cytokine BAFF, the number of CD27^highCD38^high and CD27⁺CD38^– B cells in the cultures of CD27⁺ B cells was maximal, consistent with other data available [25, 26]. Conversely, upon the addition of IL-4, abundance and proportion of ASC and CD27⁺CD38^– cells tended to decrease while the proportion of CD27^–CD38^– B cells tended to increase.

Our findings suggest that for ASC abundance in B cell culture, it is preferable to use CD27⁺ memory B cells in combination with hCD40L and IL-21, along with the addition of BAFF. It should be noted that the analysis was performed without considering the levels of ASC antibody secretion which can vary significantly depending on the level of ASC maturation and multiple additional factors, and requires separate studies to optimize antibody production in cell cultures.

Accumulation of GC B Cells in Cell Cultures

As mentioned earlier, the signal through the BCR as well as the Th cell "help" in the form of CD40L and IL-21 secretion induces the appearance of GC B cells. A key marker of GC B cells is the expression of BCL6 transcription factor [5]. Phenotypic analysis of lymphocytes from human secondary lymphoid organs identified CD95^highBCL6⁺CD38⁺CD27⁺/– B cells as GC B cells [27]. In our study, we were able to identify GC-like B cells in cultures by a combination of the surface marker CD95 and the intracellular transcription factor BCL6.

Accumulation of CD95^highBCL6⁺ B cells occurred in cultures with hCD40L and IL-21, consistently with in vivo studies [28]. The initial cell populations of CD27^– and CD27⁺ B cells did not significantly affect the proportion and number of CD95^highBCL6⁺ in the final B cell cultures. However, in cultures from CD27^– B cells, BCL6 level tended to increase. The addition of recombinant serum proteins and IL4 contributed to an increase in the proportion of GC-like B cells and a decrease in ASC. This is consistent with the results of studies suggesting that the combined action of IL-21 and IL-4 secreted by Th increases and stabilizes BCL6 expression in activated B cells [29], promoting their differentiation into GC B cells [5]. Moreover, BCL6 is an antagonist of the transcription factor Blimp-1 and inhibits further differentiation of B cells into ASC [24].

Thus, we were able to demonstrate that culturing B cells in the presence of hCD40L and IL-21 also allows for the formation of GC-like B cells to be observed through increased BCL6 expression. Moreover, a slightly higher accumulation of GC-like B cells was observed in a culture of CD27^‒ naive B cells when IL-4 and human serum substitutes were included in the medium.

However, it should be noted that none of the components used in this study resulted in the accumulation of more than 1% GC-like B cells. GC B cells are highly sensitive to apoptosis in the absence of support from Th and specialized cells — follicular dendritic cells (FDCs) [30]. It was shown that FDCbased cell lines is likely to be preferable for culturing GC B cells [10]. Based on our data and the results of previous studies, we suggest that maintaining GC B-cell cultures in vitro requires using more complex co-cultivation systems with the inclusion of FDC-like B-cell lines.

CONCLUSIONS

The study helped to identify optimal conditions for the effective expansion of B cells from human peripheral blood in vitro as well as for obtaining ASC-enriched cell culture. The presence of hCD40L-expressing feeder cells and IL-21 in the cultures was shown to be necessary for both proliferation and survival as well as for B cell differentiation in vitro. For optimal B cell culture expansion, it is preferable to use CD27^– naive B cells and a medium supplemented with a combination of human serum proteins (including insulin, transferrin, and albumin). Meanwhile, using CD27⁺ memory B cells and adding BAFF to medium are preferable for making ASC more abundant. In this study, we were able to assess the accumulation of CD95^highBCL6⁺ GC-like B cells. However, the combination of conditions used was insufficient to significantly increase the abundance of GC-like B cells in cultures. Presumably, the generation of GC-like B cell culture from peripheral blood B cells may require using FDClike feeder cells and additional soluble factors.