Asymmetric Segregation of Polarized Antigen on B Cell Division Shapes Presentation Capacity

See allHide authors and affiliations

Science  27 Jan 2012:
Vol. 335, Issue 6067, pp. 475-479
DOI: 10.1126/science.1214100


During the activation of humoral immune responses, B cells acquire antigen for subsequent presentation to cognate T cells. Here we show that after mouse B cells accumulate antigen, it is maintained in a polarized distribution for extended periods in vivo. Using high-throughput imaging flow cytometry, we observed that this polarization is preserved during B cell division, promoting asymmetric antigen segregation among progeny. Antigen inheritance correlates with the ability of progeny to activate T cells: Daughter cells receiving larger antigen stores exhibit a prolonged capacity to present antigen, which renders them more effective in competing for T cell help. The generation of progeny with differential capacities for antigen presentation may have implications for somatic hypermutation and class switching during affinity maturation and as B cells commit to effector cell fates.

The recognition of antigen through the B cell receptor (BCR) triggers B cell proliferation and differentiation that results in the formation of plasma cells capable of antibody production and long-lasting memory cells (1). The majority of B cell activation is initiated by antigen on the surface of presenting cells, including follicular dendritic cells [FDCs (2)], dendritic cells (3), and macrophages (46). The binding of cognate antigen to the BCR initiates intracellular signaling [reviewed in (79)], resulting in antigen extraction from the cell surface (10). BCR-driven uptake targets antigen to intracellular endosomal compartments containing major histocompatibility complex [MHC (11)], which allows for presentation to cognate T cells that go on to promote maximal B cell activation (12, 13). During the course of a typical humoral immune response, B cells present antigen to T cells in two distinct situations within secondary lymphoid tissues: at the B-T cell border after initial antigen acquisition (14, 15) and within the germinal center (GC) to generate affinity-matured antibodies (1619). The importance of B cell antigen presentation to T cells is highlighted by the observation that disruptions of B-T conjugate stability result in a profound defect in GC formation and humoral immunity (20, 21) and has been associated with X-linked lymphoproliferative disease in humans (22).

After BCR-mediated acquisition in vivo, B cells initially concentrate internalized antigen in a single cargo at the uropod as they migrate to search for cognate T cells (4). This observation raises the question of whether this polarized distribution of antigen is maintained over time and during subsequent B cell division. To address this question, we used intensely labeled fluorescent particulates equivalent in diameter to a typical pathogen and coated with hen egg lysozyme (HEL; bHEL, beads with HEL conjugated HEL) to monitor antigen over extended periods of time. Transgenic MD4 B cells that express HEL-specific BCR were labeled with the vital dye SNARF-1 and adoptively transferred into C57BL/6 wild-type (WT) mice 1 day before immunization with bHEL. Using multiphoton microscopy (MPM) of explanted lymph nodes, we observed that MD4 B cells acquired antigen in a polarized fashion from CD169+ macrophages (Fig. 1A and movie S1) and maintained this polarization at the uropod as they detached from the subcapsular sinus (fig. S1A and movie S2). Similarly, even up to 3 days after acquisition, the polarized distribution of antigen was maintained in migratory B cells that were pre-loaded with bHEL before being transferred to a WT host (to minimize continued antigen exposure) and visualized in the spleen (Fig. 1A). Moreover, we observed that antigen contained in a small number of endosomal compartments was also distributed toward one pole in B cells pre-incubated with bHEL and cultured with CD40 stimulation and interleukin-4 (IL-4), signals that B cells normally receive by interacting with T cells (Fig. 1B; fig. S1, B to E; and movie S3). Thus, the establishment and preservation of antigen polarity do not appear to require interactions with other cells.

Fig. 1

Antigen polarization in B cells is maintained during B cell division. (A) MPM time-lapse images (minutes:seconds) from (top) a popliteal lymph node showing MD4 B cells (red) and CD169+ macrophages (cyan) 3 hours after injection of bHEL (green); (middle and bottom) a spleen at (middle) 24 hours and (bottom) 72 hours after transfer of MD4 B cells (red) loaded with bHEL (green). Tracks are shown in white dotted lines. Data are representative of three independent experiments. (B) Serial block face scanning electron microscopy of MD4 B cells containing high levels of bHEL (red) after 24 hours of culture with IL-4 and CD40 stimulation. (Left) Three-dimensional reconstruction showing the cell membrane (cyan) and nucleus (green). (Right) Quantification of number of beads per compartment (bottom) and number of compartments per cell (top) from two experiments (n = 38 cells, mean ± SEM). (C and D) CTV-labeled MD4 B cells (purple) loaded with bHEL (red) analyzed by IS after 72 hours of culture with IL-4 and CD40 stimulation. Polarization is quantified by comparing the centroid positions of the bright-field (BF) images and the particular fluorescent species in two independent experiments (P, polarized; U, unpolarized). (C) Undivided B cells. (Left) Descending images show B cells with increased antigen polarization. (Middle) Distribution of delta centroid values of the bright-field images compared with CTV fluorescence (purple) or bHEL (red). (Right) Quantification of antigen polarization (n = 557 cells). (D) Divided B cells. (Left) Descending images show B cells after successive rounds of division (d) and (right) quantification of antigen polarization [division 1 (d1), n = 3087; d2, n = 3792; d3, n = 898]. (E) MD4 B cells pre-loaded with bHEL (red) cultured for 48 hours with IL-4 and CD40 stimulation, stained with anti-histone H3 (blue) and γ-tubulin [microtubulin-organizing center (MTOC); green]. (Left) Confocal images. (Right) Quantification of antigen polarization in metaphasic B cells (two MTOCs, n =17 cells) from two independent experiments. (F) MD4 B cells loaded with bHEL (red) cultured for 44 hours and treated for 4 hours with cytochalasin B stained with α-tubulin (green) and B220 (B cell membrane; blue). (Left) Confocal images. (Right) Quantification of antigen polarization (n = 35 cells) from two independent experiments. Scale bars in (A), 30 μm; in (C) and (D), 2 μm; in (E) and (F), 1 μm.

To examine what happens to the polarized distribution of antigen as the B cell undergoes division, we have developed a strategy using an imaging flow cytometer [ImageStream (IS)] to simultaneously visualize the distribution of bHEL antigen and division in B cells (23) (fig. S2A). In line with our in vivo observations (Fig. 1A), although some cells contain bHEL that is not polarized, around 75% of undivided cells exhibit a polarized distribution of antigen after 3 days in culture with CD40 stimulation and IL-4 (Fig. 1C). After polarization was established, the distribution of antigen was not altered after treatment with nocodazole, suggesting that the microtubule network is not important for maintaining antigen polarization (fig. S2B). We observed that a similar polarization of bHEL was preserved in cells that had undergone up to three rounds of division (Fig. 1D and fig. S2C). Independent of cell division, the fluorescence corresponding to the bead itself or to the HEL antigen was colocalized, establishing that the bHEL fluorescence represents an accurate readout for the presence of antigen (fig. S2D). These findings indicate that B cells maintain a polarized distribution of antigen even after B cell division.

Because the IS system allows the separation of cells on the basis of multiple parameters, we took advantage of this strategy to investigate antigen polarization in B cells as they progressed through mitosis (fig. S3A). We observed that B cells maintain a polarized distribution of accumulated bHEL, but not the ancestral polarity network member protein kinase C–ζ or cell surface proteins such as CD86 and MHCII, as they enter prophase and as they progress through metaphase and anaphase (figs. S3, B to D, and S4, A to D). To verify these findings at high resolution, we visualized the antigen distribution in dividing B cells that were pre-loaded with bHEL, using confocal imaging. In line with the IS data, we observed that antigen was indeed polarized in the majority of B cells, both as they entered mitosis (Fig. 1E and fig. S5A) and if they were trapped in metaphase by incubation with cytochalasin B before fixation (Fig. 1F and fig. S5, B and C). Thus it appears that B cells maintain antigen in a polarized distribution both over time and during division.

The observed polarization raises the possibility that B cell division results in asymmetric antigen segregation and the generation of progeny with unequal antigen inheritance. To investigate this possibility, we constructed two absolute in silico models to predict the impact of cell division on antigen distribution, considering a population initially containing high levels of antigen (23) and labeled with a division marker that is equally segregated on division. The first model assumes that antigen is segregated symmetrically on division, giving rise to a linear plot in which the intensities of antigen and division marker are directly correlated (Fig. 2A). On the other hand, the second model assumes that antigen is always segregated asymmetrically, so that one daughter receives all of the accumulated antigen while the other receives none (Fig. 2A). Although the first model predicts that five rounds of division are required to generate cells lacking antigen, that population emerges after one division in the asymmetric model (Fig. 2A). Moreover, the latter model is also characterized by a population retaining high levels of antigen even at later rounds of division. But which of these two division profiles most accurately recapitulates the features of antigen segregation during B cell division in vivo?

Fig. 2

Antigen can be asymmetrically segregated on B cell division. (A) Schematic and fluorescence dot plots for in silico models examining the an-tigen (bHEL) content of dividing B cells relative to an equally divided cell marker such as CTV. (B) CTV-labeled CD45.2 MD4 B cells loaded with high levels of green fluorescent bHEL-OVA were (left) sorted before cotransfer with CD4+ OT-II T cells into CD45.1 recipient mice. (Middle) Flow cytometry profile of CTV and bHEL fluorescence of CD45.2 splenic B cells 3 days after transfer. B cells containing high (red) and low (blue) levels of bHEL after successive rounds of division (d1, d2, …) are marked. (Right) Quantification of the total number (black), bHEL-high (red), and bHEL-low (blue) B cells at each round of division. Data are representative of two independent experiments. (C) In silico mixed model assuming that 25% of B cells undergo fully asymmetric division, while the remaining cells undergo fully symmetric division. (Left) Fluorescence dot plot generated for the mixed model as compared with the experimental data. Comparison of (middle) CTV and (right) bHEL fluorescence predicted in the mixed model versus collected data in the right panel of fig. S6A. (D) A20 B cells expressing HEL-specific D1.3 BCR and histone H2B (H2B)–enhanced green fluorescent protein (eGFP) (green) were loaded with flash red fluorescent bHEL (red) and imaged by low-light microscopy. (Top and middle) Images of B cells at indicated times (hours:minutes:seconds) undergoing (top) asymmetric and (bottom) symmetric segregation of bHEL as they divide. Dashed white lines in the images represent the cell boundary. Scale bars, 10 μm. (Bottom) Quantification of B cells that undergo asymmetric and symmetric division in 21 mitosis events, during four independent experiments.

To address this question experimentally, a homogeneous population of B cells sorted to contain high levels of both the cell division marker, CellTracker Violet (CTV), and of antigen was transferred into recipient mice; in this case, the antigen also contained ovalbumin (OVA; bHEL-OVA) to allow B cells to receive necessary help for proliferation from cotransferred OVA-specific OT-II T cell receptor transgenic T cells. After 3 days, B cells had undergone up to four rounds of division, and we observed two characteristic features of the asymmetric model: the rapid emergence of a substantial population containing no bHEL-OVA and the maintenance of B cells retaining high levels of antigen after several rounds of division (Fig. 2B). However, although CTV was symmetrically divided between progeny, the distribution of antigen did not fully recapitulate either of our in silico models. The asymmetric segregation of antigen appears to be intrinsic to dividing B cells, because similar profiles were obtained for B cells cocultured with OT-II T cells, CD40 stimulation, and IL-4 or lipopolysaccharide for 3 days (fig. S6A).

Our data strongly suggest that although B cell division can occur symmetrically, it often results in the asymmetric segregation of polarized antigen. We tested this hypothesis by comparing our experimental observations (fig. S6A) with a third, combined model, merging the previous two models (Fig. 2C) and incorporating parameters from our in vitro data (23) (fig. S6B). We found that to recapitulate the predominant features of antigen segregation in vitro, it was necessary to include at least 25% fully asymmetric division in the combined model (Fig. 2C). In line with this third model, we observed that a population of dividing A20 B cells was capable of segregating loaded bHEL antigen both asymmetrically and symmetrically between daughter B cells (Fig. 2D; movies S4 and S5; and fig. S6, C and D). Because the combined model assumes that the segregation of antigen between daughter cells is either fully symmetric or fully asymmetric, this probably represents an underestimate of the contribution of asymmetric B cell division both in vitro and in vivo. Taken together, our theoretical and experimental approaches establish that the asymmetric segregation of antigen occurs on B cell division.

We moved on to explore the functional significance that the differential stores of antigen confer on progeny. Because BCR-mediated internalization delivers antigen into compartments containing MHC molecules (fig. S7A), we examined the influence that the retention of antigen has on the capacity to present antigen to cognate T cells. B cells containing high levels of bHEL were cultured with CD40 stimulation and IL-4 for 3 days, and the extent of antigen presentation on the surface was detected by staining with C4H3 antibody (24). We observed that the amount of HEL-loaded MHC was consistently higher for B cells that contained more bHEL (Fig. 3A), although the extent of antigen presentation decreased on progression through cell division. Moreover, the size of the intracellular antigen store was related to the level of bHEL and to progress through division in a similar manner (fig. S7B). As such, the amount of antigen retained in the cell is indeed related to the capacity to present antigenic peptide in complex with MHC on the B cell surface. To examine the corresponding impact on T cell activation, the extent of IL-2 secretion and proliferation of cocultured T cells was measured. We observed that the amount of bHEL-OVA retained by B cells was related to the extent of IL-2 secretion by cocultured OT-II cells (Fig. 3B). Moreover, the ability of B lymphoma cells expressing HEL-specific D1.3 BCR (25) to present antigen to HEL-specific 2G7 T cells was also directly related to the amount of antigen internalized (fig. S8A). In a similar manner, the quantity of antigen stored in B cells was related to the extent of T cell proliferation induced: Whereas B cells containing high levels of bHEL-OVA stimulated almost 60% of T cells to divide, proliferation was severely impaired in T cells cocultured with B cells containing low levels of bHEL-OVA (fig. S8B).

Fig. 3

The amount of antigen stored determines the capacity for B cell presentation to activate cognate T cells. (A) CTV-labeled B cells from MD4xB10.BR F1 mice were loaded with green fluorescent bHEL and cultured with IL-4 and CD40 stimulation for 3 days. B cells containing high (red) and low (blue) levels of bHEL after successive rounds of division (d1, d2, …) are marked. Surface staining for HEL-loaded IAk complexes using C4H3 antibody is shown. (Left) Flow cytometry profile. (Middle) Staining in B cells from bHEL-high (red) and bHEL-low (blue) B cells from the d1 population. Nonspecific staining was measured after loading of MD4 B cells with microspheres coated with antibody to IgM and is shown in solid gray profile. (Right) Quantification of relative staining across successive rounds of division compared with undivided B cells and corrected for the extent of MHC II expression at each round. Data are representative of three independent experiments. (B and C) CTV-labeled HyHEL-10 B cells containing high levels (green) of bHEL-OVA were sorted and cultured in medium supplemented with (B) IL-4 and CD40 stimulation or (C) CD4+ OT-II T cells. (B) HyHEL10 B cells were cultured for 3 days before sorting according to levels of bHEL-OVA (high and low) and were cultured for a further day with OVA-specific CD4+ OT-II T cells. IL-2 secretion as a readout of antigen presentation was measured by enzyme-linked immunosorbent assay. Means ± SEM are shown. Data are representative of two independent experiments. (C) CTV-labeled MD4 B cells after 3 days in culture were sorted according to levels of bHEL-OVA (high, red; low, blue). 105 divided and sorted B cells were cultured for a further 3 days with carboxyfluorescein diacetate succinimidyl ester (CFSE)–labeled CD4+ OT-II T cells, and T cell proliferation was (left) measured by flow cytometry and (right) quantified at the indicated times. Data are representative of three independent experiments.

Finally, we wanted to establish that B cells containing differential amounts of stored antigen as a result of asymmetric segregation were capable of inducing T cell activation to different extents. Indeed, whereas B cells containing high levels of bHEL-OVA triggered considerable proliferation in OT-II T cells after 2 days, this was dramatically diminished when low levels of antigenic beads were present in the B cells (Fig. 3C). Although viability issues prevent the monitoring of B cell activation markers directly, before coculture with OT-II cells, the two B cell populations did not express significantly different levels of activation markers, surface MHC, or costimulatory molecules (fig. S9). Thus, the difference in T cell proliferation observed appears to be directly due to the level of antigenic peptide displayed on the B cell surface, and the differential levels of antigen generated by asymmetric segregation directly influence the ability of B cells to induce functional responses in cognate T cells.

Our observation of asymmetry is both interesting and surprising, although perhaps on reflection not unexpected, considering that the partitioning of the endosomal compartment during mitosis occurs in a stochastic manner (26). However, because BCR stimulation is associated with the fusion of endosomal vesicles (27) and formation of a large autophagosome-like compartment (28), we postulate that internalization by this route promotes asymmetric antigen segregation on B cell division. In contrast to the unequal division of proteins in T cells (29), the asymmetric segregation of antigen in B cells does not appear to depend on prolonged contact with antigen-presenting cells and the associated formation of the immunological synapse. But what is the functional significance of this asymmetric antigen segregation? We postulate that daughter B cells receiving high amounts of antigen might allow for more effective competition for limited T cell help both at the B-T cell border (14, 15) and in the GC (30), facilitating somatic hypermutation and class switch. Asymmetric cell division also allows progeny to be cleansed of accumulated protein (31), rapidly generating antigen-negative B cells with a diminished capacity to present antigen. The survival of these cells depends on further accumulation of antigen, allowing “proofreading” of newly mutated BCR and the retention of B cells with higher-affinity BCR for further selection during affinity maturation. Thus, the unequal partitioning of polarized antigen after division generates distinct B cell populations that either retain or rapidly eliminate antigen that together may be critical for the effective induction of humoral immune responses and the production of high-affinity antibodies.

Supporting Online Material

Materials and Methods

Figs. S1 to S9

References (3243)

Movies S1 to S5

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. Acknowledgments: We thank D. Davies, Biomedical Research Council, Guy’s and St Thomas’ National Health Service Foundation Trust, and King’s College London for assistance with flow cytometry and H. Armer for technical assistance with electron microscopy. We also thank the members of the Lymphocyte Interaction Laboratory for critical reading of the manuscript. This work was funded by Cancer Research UK. F.D.B. also receives support from the Royal Society Wolfson Merit Award, and O.T. receives financial support from Sociète Française de Transplantation, Sociète Française de Nephrologie, and Fondation pour la Recherche Medicale. The authors declare that they have no competing financial interests.
View Abstract

Navigate This Article