Rapid Proliferation and Differentiation of a Subset of Circulating IgM Memory B Cells to a CpG/Cytokine Stimulus In Vitro

Circulating human IgM expressing memory B cells have been incompletely characterized. Here, we compared the phenotype and in vitro functional response (capacity to proliferate and differentiate to antibody secreting cells) in response to CpG and a cytokine cocktail (IL-2, IL-6, and IL-10) of sorted naïve B cells, IgM memory B cells and isotype-switched circulating memory B cells. Compared to naïve B cells, IgM memory B cells had lower integrated mean fluorescence intensity (iMFI) of BAFF-R, CD38, CD73, and IL-21R, but higher iMFI of CD95, CD11c, TLR9, PD-1, and CD122. Compared to switched memory B cells, IgM memory B cells had higher iMFI of BAFF-R, PD-1, IL-21R, TLR9, and CD122, but lower iMFI of CD38, CD95, and CD73. Four days after receiving the CpG/cytokine cocktail, higher frequencies of IgM than switched memory B cells—and these in turn greater than naïve cells—proliferated and differentiated to antibody secreting cells. At this time point, a small percentage (median of 7.6%) of stimulated IgM memory B cells changed isotype to IgG. Thus, among the heterogeneous population of human circulating IgM memory B cells a subset is capable of a rapid functional response to a CpG/cytokine stimulus in vitro.


Introduction
B cells (Bc) and the antibodies they produce after becoming antibody secreting cells (ASC) are critical for protecting the individual from pathogens and maintaining tissue homeostasis [1,2]. Instrumental to the function of Bc is their capacity to differentiate from naïve (nonantigen experienced) to antigen experienced memory Bc (mBc) [1]. This process renders mBc capable of promptly responding to a second encounter with a pathogen and is reflected by their propensity to rapidly proliferate and differentiate in vitro, in an antigen independent fashion, to T cell dependent (CD40L) and independent stimuli [3,4]. As an example of the latter, it has been shown that naïve cells depend on Bc receptor signaling, while mBc are activated by CpG and cytokines, without need for Bc receptor stimulation [5]. In mice, mBc can be derived independently of a germinal center (GC -, mostly IgM with low frequency of . Samples were acquired on a LSR Fortessa (BD Biosciense) and analyzed using FlowJo v.9.3.2. The data are represented as an integrated mean fluorescent intensity (iMFI = percentage of positive cells × MFI), a metric measure that combines magnitude (percentage of cells expressing a marker) and quality (MFI of this same marker) and reflects the total potential functional response of each cell subset [26,27].
Stimulation and analyses of sort purified Bc B cells subpopulations were labeled with carboxyfluorescein succinimidyl ester (CFSE), as previously described [28], with minor modifications. Bc were washed twice with sterile PBS and stained (0.4-1 x 10 6 cells/ml) with CFDA-SE (Cell-TraceTM CFSE Cell Proliferation Kit, Invitrogen Molecular Probes) for 5 min at room temp protected from light. After being washed three times with 10ml of PBS 5%, FBS 5%, naïve Bc and IgM mBc and switched mBc were stimulated, as previously described [18], adapting the limiting dilution assays conditions to stimulation in a 48 well format: Bc were stimulated with 2.5 μg/ml of CpG (ODN 2006; InvivoGen, San Diego, CA), 10 ng/ml human recombinant IL-2, 10 ng/ml human recombinant IL-6, 15 ng/ml IL-10 (all cytokines from PreproTech, Rocky Hill, New Jersey), and NIH 3T3 murine fibroblasts (ATCC Manassas, VA, USA). The NIH 3T3 feeder cells were irradiated with 3,000 rads (Radiotherapy unit, Centro Javeriano de Oncología, Bogotá) and then used at a concentration of 5,000 cells/well. B cells (20,000 in 200 μl of complete medium with 10% FBS) were cultured at 37°C with 5% CO 2 in flat bottom 48 well plates for different periods of time. At the end of the cultures, cells were washed twice with sterile PBS and stained on ice with AQUA (Invitrogen Molecular probes) and subsequently with antibodies (all from BD Bioscience) against (labeled): CD19 (APC-H7), CD38 (PerCp-Cy5.5), CD27 (PE-CF594). After washing, cells were resuspended in 100 μl PBS, 0.5% BSA (Sigma-Aldrich) plus 250 μl of cytofix/cytoperm (BD Pharmingen) and incubated for 20 min at 4°C. After washing twice with perm/wash (BD Pharmingen) cells were stained with antibodies against IgM APC (Jackson ImmunoResearch) or IgG APC (BD Bioscience) for 30 min on ice. Samples were acquired on a LSR Fortessa (BD Bioscience) and analyzed using FlowJo v.9.3.2 (FlowJo, LLC, U.S. OR).

Statistical analyses
Since data was not normally distributed, non parametric tests were used for comparisons. Differences among groups were determined using Kruskal-Wallis tests followed by a Wilcoxon signed-rank test using GraphPad Prism software v.5.0a for Mac OS X, GraphPad Software (La Jolla, CA). Significance was established if p<0.05. Data are shown as median and interquartile range.

Results
Phenotypic differences of naïve Bc, IgM mBc, and switched mBc Bc (CD19+) were separated into naïve (CD27 -IgD + : the transitional Bc subset was not excluded from the analysis), IgM mBc (CD27 + IgD + IgM + : which included the IgM only but not the IgD only subpopulations) and switched mBc (CD27 + IgD -IgM -), as described in Fig  1A and 1B  The iMFI of CD40 and CD21, a part of the Bc co-receptor [29] and considered a molecule highly expressed on marginal zone Bc [30], were comparable among Bc subsets (Fig 2, first row). BAFF-R iMFI was high among Bc and significantly higher in naïve Bc (Fig 2, first row).
The iMFI of CD38, CD73, and CD95 were intermediate to low in the three Bc subsets ( Fig  2, second row). CD38 is an ADP-ribosyl cyclase and its iMFI was higher in naïve Bc than both switched mBc and IgM mBc. In agreement with previous results, the iMFI of CD73 (an ectoenzyme that degrades extracellular nucleoside monophosphates to adenosine) of naïve Bc was comparable to that of switched mBc, but higher than that of IgM mBc [31]. The expression of CD95 is associated with caspase dependent apoptosis [32] and, as previously reported, CD95 iMFI of naïve Bc was comparatively very low [4,32]. Also, CD95 iMFI of IgM mBc was significantly lower than that of switched mBc.
The iMFI of CD11c a marker shown to be expressed by IgM mBc specific for bacteria in mice [33], was intermediate and similar for IgM and switched mBc, but lower for naïve Bc (Fig  2, third row).
Similar low iMFI of CD43, a marker that has been proposed to identify B1 Bc in humans [16] were identified in the three Bc subset studied (Fig 2, third row).
The iMFI of CD122, the beta chain of the IL-2 and IL-15 receptors, TLR9, the receptor for CpG, and PD-1, which regulates activation and survival of Bc [34], were low for the three Bc subsets, but significantly higher in IgM mBc than naïve Bc or switched mBc (Fig 2, third row and fourth row, respectively). However, both IgM and switched mBc express more intracellular TLR9 than naïve cells, but no differences were seen between IgM and switched mBc (S3 Fig).
Finally, in agreement with previous reports [35], higher iMFI of IL-21R, a key molecule critically implicated in Bc activation [36], was detected for naïve cells than switched mBc (Fig 2, fourth row). We also compared the expression of CD45RO (data not shown), which resembled the expression of CD43 (Fig 2).
A subset of IgM mBc rapidly proliferate and differentiate to ASCs in vitro in response to CpG and cytokines To compare the capacity to proliferate and differentiate to ASC (CD27 hi CD38 hi ) of naïve Bc and IgM and switched mBc, Bc were enriched from buffy coats of blood bank volunteers by rosette formation and subsequently each subset was purified by the sorting strategy described in Fig 3. Bc were then labeled with CFSE and stimulated with CpG, a cocktail of cytokines (IL-2, IL-6, and IL-10) and murine fibroblasts (as feeder cells) for two, three, four, and five days. The peak response for proliferation and differentiation to ASC (CD27 hi CD38 hi ) was between days three and four, for both types of mBc (S4 Fig). Four days after stimulation a higher  frequency of IgM mBc proliferated (Fig 4A and 4B) and acquired the phenotype of ASC, compared to switched mBc and naïve Bc (Fig 4D). It has been previously shown that, in general, naïve Bc have undergone a lower number of divisions (two divisions) than IgM mBc (seven divisions) and switched mBc (10 divisions) [15], and that the number of divisions a Bc has performed limits the maximum additional divisions it can undertake [37]. In agreement with this, we found that the MFI CFSE-which is inversely proportional to the number of divisions a Bc has performed in vitro after stimulation-was lower in the IgM mBc vs. switched mBc ( Fig  4C).

A subset of stimulated IgM mBc rapidly change isotype in vitro after stimulation
Our group has previously shown that a median of 62% IgM mBc stimulated for seven days in a limiting dilution assay, with the same stimulus as we have used here, switch to IgG [18]. To determine if isotype switch occurred in IgM mBc under the present stimulation conditions (four days), we stained Bc for intracellular IgM and IgG (Fig 5A). Coherent with our initial results, we observed an isotype switch to IgG in a median of 7.6% (range 2-20%) of IgM mBc four days after stimulation (Fig 5A and 5C). As expected, very low or no naïve Bc expressed IgG and no, or very low, numbers of switched mBc expressed IgM. Of note the median IgG expression on switched mBc was 34.4% because the subset was sorted as IgD-, IgM-cells that include IgA an IgE cells. Intracellular IgG expression was not found on purified naïve Bc and IgM mBc prior in vitro stimulation (S5 Fig).

Discussion
Comparison of the functional capabilities of switched and IgM mBc may help in better discriminating human Bc subsets. Here, we have shown that four days after stimulation with a CpG/cytokine cocktail (IL-2, IL-6, and IL-10) (Fig 4), a higher frequency of circulating IgM mBc proliferate and differentiate in vitro to ASC (CD38 hi CD27 hi ) compared to switched mBc and naïve Bc, and a subset of IgM mBc are already capable of switching to IgG (Fig 5). We have also identified/corroborated phenotypic differences between the three Bc subpopulations, which may help explain this functional behavior (Fig 2).
Antigen independent functional response to T cell dependent (CD40L) or independent (CpG) stimuli (in the absence or presence of multiple cytokines) of human spleen naïve Bc and IgM and switched mBc have been extensively compared [3,4,38,39]. In general, these studies have reported important differences between the naïve and mBc, but only minor or no dissimilarities between the two mBc subsets: higher frequencies of IgM and switched spleen mBc differentiate to ASC in response to CD40L, compared to naïve Bc [40]. A greater proportion of human spleen switched mBc than IgM mBc was induced to proliferate after stimulation with CD40L, IL-2, IL-10, and IL-21 [3,4,38]. In contrast, and in agreement with our results with blood Bc (Fig 4), when spleen IgM and switched mBc were stimulated with CpG a higher percentage of IgM mBc than switched mBc proliferated [4]. Also, the differentiation rates to ASC of the two spleen mBc have been comparable or slightly different, depending on the cytokine used for co-stimulation [40]. Extrapolation of these results to Bc from other organs has been put in doubt by experiments showing that Bc from human tonsils differ functionally from Bc subsets with the same phenotype isolated from blood [22]. Thus, a comprehensive analysis of B cell subsets must include functional studies of B cells from different organs including blood, like in the present study.
Studies that have compared blood naïve Bc and IgM and switched mBc are scarce: while (in agreement with our results, Fig 4) higher frequencies of blood IgM mBc seem to proliferate in response to CpG, switched mBc appear to better respond to an alloreactive T cell clone [5]. After stimulation with Staphylococcus aureus Cowan, IL-2, IL-10, and anti-CD40 for eight days, a similar differentiation of CD27 + IgD + mBc (most probably IgM mBc), and CD27 + IgD -mBc (most probably switched mBc) to CD38 + ASC was observed [23]. More recently, it has been shown that blood IgM mBc (excluding IgM only mBc) share a similar gene expression pattern with switched mBc [25]. Notwithstanding, in response to a BCR stimulus, in the presence or absence of CD40L, IgM mBc up regulate markers that permit them to migrate to B-cell follicles, whereas activated IgG+ mBc preferentially showed a plasma cell differentiation. In this study it was also shown that T-independent stimulus, like terbutaline and CEACAM8, preferentially stimulated IgM mBc, while CD40L (a T-dependent stimulus) preferentially stimulates switched mBc [25]. This last study highlights the plasticity of IgM mBc that respond differently depending on the stimulation protocol. For this reason it is important to note that our results maybe be restricted to our stimulation procedure (based on CpG and cytokines), which was previously standardized for the study of antigen specific mBc [17,18,21]. In our previous experiments stimulating purified IgM and switched mBc, but in a limiting dilution assay for seven days, we found a higher cloning efficiency in the latter, which was accompanied, at this time point, by induction of higher frequencies of ASC identified by flow cytometry and ELI-SPOT [18]. Altogether, these results and our new findings (Fig 4) suggest that IgM mBc proliferate and differentiate more rapidly than switched mBc in response to CpG/cytokines, but that with time a greater proportion of switched mBc can become functional [18]. Finally, in support for a selective effect of CpG on IgM mBc, it has been shown that in blood CD27 + mBc CpG induced increased IgM secretion, compared to CD40L [41].
Our results showing that IgM mBc, but not naïve Bc, can change isotype to IgG by day four (Fig 5A and 5B) are coherent with previous studies in which cells were stimulated with CD40L, IL-2, and IL-10 for 10 days [24]. However, in the study that compared CD27 + IgD + and CD27 + IgD -mBc stimulated with CD40L and IL-2 or SAC plus IL-2 for eight days no isotype switch was seen in the IgD + CD27 + (mainly IgM mBc) [23]. Nonetheless, when IL-10 was added to the Bc cultures, IgM mBc were shown to secrete IgG two to three days after stimulation [23]. Altogether, these reports are in agreement with our findings (Fig 5), indicating that IgM mBc can rapidly change isotype to a non BCR stimulus, but that this is dependent on the presence of cytokines like IL-10, similar to what has been reported for spleen IgM mBc [40].
Naïve Bc and IgM and switched mBc were shown to have several phenotypic differences (Fig 2). Compared to switched mBc, IgM mBc had higher iMFI of molecules involved in the activation of Bc, such as TLR9 [5,41,42], IL-21-R [43], and CD122 [44] (Fig 2). Of these markers the expression of TLR9 and CD122 may partially explain the higher proliferative and capacity to differentiate to ASC of IgM mBc compared to switch mBc in our assay (Fig 4). Activation through TLR9 has been shown to selectively provoke proliferation of mBc and in particular of IgM mBc [5,41,42]. In agreement with our results for surface protein (Fig 2), a tendency to higher expression of TLR9 mRNA has been shown in IgM mBc than in naïve Bc and switch mBc [45]. Although this difference may explain in part the functional differences of IgM and switched mBc (Figs 3, 4 and 5), this interpretation must be taken with precaution because most of this receptor is expressed intracellularly [4] and when we measured intracellular TLR9 no differences were observed between IgM and switched mBc (S3 Fig). Since we included IL-2 in the Bc cultures, higher expression of CD122 (the IL-2 receptor) on IgM mBc could also partially explain the increased proliferation of these cells (Figs 3, 4 and 5). Because we also included IL-10 and IL-6 in our cultures, higher expression of IL-10R but lower expression of IL-6R mRNA by IgM mBc than in switched mBc [25] may also modulate the increased proliferation/differentiation of the former. Compared to switched mBc lower frequencies of IgM mBc express CD95 (Fig 2), which regulates Bc survival [32]. The expression of this marker on mBc may also give them an in vitro survival advantage that could be reflected on their increased proliferation frequencies (Fig 4).
The capacity of a cell to switch has been related to the presence of CD40 [46], BAFF-R [47], CD73 [31], IL21R [43] and TLR9 [48]. Of these markers, CD73 seems like a good marker to identify the subset of IgM mBc that switch in vitro in response to our stimulus (Fig 5). In favor of this hypothesis, it was recently reported that the presence of CD73 correlated with the capacity to switch isotype of IgM mBc [31]. Moreover, in mice, CD73 + mBc are enriched in those with somatic mutations [9] and GC -IgM mBc were mainly CD73 + [7]. However, the frequencies of CD73 expressing IgM mBc [31] (Fig 2) are higher than those of cells undergoing isotype switch (Fig 5). Using a combination of CD73 with one or more of the above mentioned differentially expressed markers involved in activation, survival, or isotype switch may be useful to identify IgM mBc that switch isotype after this stimulus.
The nature of human blood IgM + CD27 + Bc is a subject of controversy [11,30]. At least a part of the subset that expresses IgD seem to be innate Bc with a pre-diversified repertoire of Ig [49,50], which share many markers with spleen marginal zone-and similar to those present in mice-but capable of circulating in blood [10]. Besides, some studies [51], but not other [52], have shown that the CD27 + IgM + IgD + Bc are clonally related to switched mBc, suggesting that at least a fraction of these Bc may be true IgM mBc. In support for this hypothesis, CD27 + IgM + IgD + cells transferred to immunodeficient mice immunized with Streptococcus pneumoniae [53] can develop specific IgG mBc. The existence of human blood IgM + IgD + mBc specific for rotavirus [17,18], tetanus toxoid [17,54], and the D antigen [54] also suggest that at least a fraction of CD27 + IgM + IgD + Bc are true mBc. Finally, as previously mentioned, IgM mBc (excluding IgM only mBc) share a similar gene expression pattern with switched mBc [25]. Thus, the CD27 + IgM + IgD + Bc seem to be heterogenous, a part being innate (related to marginal zone Bc) while other being antigen dependent mBc (probably GC -) [11].
In conclusion, we have shown that a subset of IgM mBc rapidly proliferate and differentiate to ASC early after stimulation with CpG/cytokines. The functional assay we have described (Figs 4 and 5), applied to subsets of purified Bc selected based on combinations of the markers that differentiate naïve Bc and IgM and switched mBc (Fig 2), may be useful to identify various subpopulations among the IgM mBc. Kinetic experiment of proliferation and differentiation to ASC (CD38 hi CD27 hi ) of IgM (brown) and switched mBc (blue) four days after stimulation with CpG, a cocktail of cytokines (IL-2, IL-6, and IL-10), and murine fibroblasts (as feeder cells) for two, three, four, and five days. Kinetic experiments for early (one, two or three days) and late responses (five, seven and ten days) were also performed using ELISPOT as readout (data not shown). Since the early time point experiments showed comparable results to the flow cytometry presented in this figure and cell mortality was above 80% in the late time point experiments (data not shown), we chose day four for the experiments reported elsewhere in this paper.