In Vitro Effects of Rituximab on the Proliferation, Activation and Differentiation of Human B Cells

Authors


Irma Joosten, I.Joosten@labgk.umcn.nl

Abstract

Rituximab is a chimeric anti-CD20 monoclonal antibody (mAb) used in B-cell malignancies, various autoimmune disorders and organ transplantation. Although administration of a single dose of rituximab results in full B-cell depletion in peripheral blood, there remains a residual B-cell population in secondary lymphoid organs. These nondepleted B cells might be altered by exposure to rituximab with subsequent immunomodulatory effects. Therefore, we analyzed in vitro the effects of rituximab on proliferation, activation and differentiation of CD19+ B cells by means of carboxyfluorescein succinimidyl ester (CFSE)-based multiparameter flow cytometry. Rituximab inhibited the proliferation of CD27 naïve, but not of CD27+ memory B cells. Interestingly, upon stimulation with anti-CD40 mAb and interleukin-21 in the presence of rituximab there was an enrichment of B cells that underwent only one or two cell divisions and displayed an activated naïve phenotype (CD27IgD+CD38−/+). The potency of prestimulated B cells to induce T-cell proliferation was increased by exposure of the B cells to rituximab. Of note, after stimulation with rituximab-treated B cells, proliferated T cells displayed a more Th2-like phenotype. Overall, these results demonstrate that rituximab can affect human B-cell phenotype and function, resulting in an altered outcome of B–T cell interaction.

Abbreviations: 
ADCC

antibody-dependent cellular cytotoxicity

BAFF-R

B-cell activating factor receptor

BCR

B-cell receptor

CCR

CC chemokine receptor

CDC

complement-dependent cytotoxicity

CFSE

carboxyfluorescein succinimidyl ester

CXCR

CXC chemokine receptor

IFNγ

interferon-γ

IL

interleukin

mAb

monoclonal antibody

PBMCs

peripheral blood mononuclear cells

RA

rheumatoid arthritis

RTX

rituximab

Th

T helper cells

TNF

tumor necrosis factor

W/O

without

Introduction

Rituximab (RTX) is a chimeric IgG1 monoclonal antibody (mAb) that recognizes human CD20, a cell-surface glycoprotein, expressed on B cells throughout differentiation and lost during maturation to plasma cells (1,2). RTX depletes B cells through three mechanisms: antibody-dependent cellular cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC) and apoptosis. Although the exact contribution of these mechanisms in vivo remains unclear, it is assumed that ADCC is the most dominant player, followed by CDC. Cell death through apoptosis is thought to be minimal (2–5).

Anti-B cell therapy has an important place in the treatment of patients with malignant B-cell lymphoma (5). Furthermore, RTX is used to reduce the level of autoantibodies in various autoimmune disorders (6,7). However, treatment has also beneficial effects on T-cell mediated diseases, such as rheumatoid arthritis (RA), multiple sclerosis and type 1 diabetes, which suggests that the therapeutic effects cannot be solely ascribed to effects on antibody production (7–10). In renal transplantation, RTX is used in ABO-incompatible transplantation, in desensitization protocols and for treatment of antibody-mediated rejection (11–13). Because the presence of CD20+ B cells in the graft is associated with poor outcome (14), RTX may be of additional value in the treatment of steroid resistant acute rejection with intragraft B-cell infiltration (15,16). The exact effects of RTX on B cells, and probably indirectly on T cells, in these various clinical conditions remains to be elucidated.

After activation by antigen recognition, B cells are able to proliferate, produce various cytokines and process antigen for presentation to T cells (17–19). T follicular helper cells provide help to B cells by expression of CD40L and production of interleukin-21 (IL-21; Ref. 20). Depending on the activating conditions, B cells will eventually differentiate into plasma cells or will acquire a memory phenotype, which allows for rapid responses upon subsequent encounter of cognate antigen (17,18,21,22). The cell surface molecules CD27 and IgD are used to identify the two main B-cell populations in peripheral blood: naïve B cells are CD27IgD+, whereas circulating memory B cells are CD27+IgD (23). Recently, a regulatory B-cell subset has been identified, which is predominantly found within the CD24highCD27+ B-cell population, and produces IL-10. This population is very small in healthy individuals but was found to be increased in patients with autoimmune disease (24,25).

In RTX-treated patients, the duration of B-cell depletion in the peripheral blood appears to depend on the disease in combination with the number of RTX infusions. After a single dose of RTX there is a nearly complete B-cell depletion in peripheral blood. However, it should be appreciated that there remains a residual B-cell population in secondary and tertiary lymphoid organs (26–28).

So far it is not clear what the exact functional capacity of these persisting but potentially modulated B cells might be, and how they could influence the immune response after treatment with RTX. To address this issue we set up an in vitro nondepleting B-cell stimulation model and investigated the effects of RTX treatment on the phenotypic and functional characteristics of B cells.

Materials and Methods

Cell isolation

Buffy coats from healthy donors, who gave written informed consent for scientific use of the buffy coats, were purchased from Sanquin Blood Bank, Nijmegen, The Netherlands. Peripheral blood mononuclear cells were isolated by density gradient centrifugation (Lymphoprep; Nycomed Pharma, Roskilde, Denmark). CD19+ B cells were positively selected using anti-CD19 magnetic microbeads (Miltenyi Biotec, Utrecht, The Netherlands). This resulted in a CD19+ B cell enrichment of more than 97%. CD19+CD27 and CD19+CD27+ B cells were isolated by high-purity fluorescence-activated cell sorting on an ALTRA cell sorter (Beckman-Coulter, Mijdrecht, The Netherlands). A rerun confirmed that the purity of the sorted cells was more than 98%.

CD4+CD25 T cells were obtained by sequential negative and positive selection of CD25 cells and CD4+ cells, respectively, with anti-CD25 and anti-CD4 magnetic microbeads (Miltenyi Biotec). This resulted in a 98% pure CD4+CD25 population.

Culture conditions

Freshly isolated B cells (5 × 104 cells per well) were cultured in 200 μL medium in the presence of anti-CD40 mAb (αCD40 mAb, 1 μg/mL, Bioceros, Utrecht, The Netherlands) and recombinant human IL-21 (25 ng/mL, ZymoGenetics, Seattle, WA, USA) for 6 days in 96-well round bottom plates (Greiner, Frickenhausen, Germany) in a 37°C, 95% humidity, 5% CO2 incubator. The culture medium consisted of RPMI-1640 supplemented with pyruvate (0.02 mM), glutamax (2 mM), penicillin (100 U/mL), streptomycin (100 μg/mL; all from Gibco, Paisley, United Kingdom), and 10% heat-inactivated pooled human serum (HPS). Where indicated, RTX was added in a concentration of 5 μg/mL.

In separate experiments, the effects of RTX and complement on viability of the B cells was assessed by incubating B cells (5 × 104 cells per well, 100-μL medium) for 30 min at 37°C with or without RTX (5 μg/mL), and subsequently with 10–30% of HLA-ABC/DR rabbit complement (Invitrogen, Breda, The Netherlands) or culture medium only. After an additional 60 min, the cell viability was assessed by flow cytometry using Annexin V-FITC (AnxV) and propidium iodide (PI, 5 μg/mL; Bender Medsystems, Vienna, Austria) according to the manufacturer's instructions.

In some experiments, the cells that remained after incubation with RTX and complement were washed and subsequently incubated for 6 days with medium containing αCD40 mAb, IL-21 and RTX, as described earlier.

Flow cytometry and CFSE labeling

For cell surface staining, the following fluorochrome-conjugated mAbs were used: CD19(SJ25C1)-PeCy7, CD70(Ki-24)-PE, CD80(L307.4)-PeCy5, CD86(233-(fun-1))-PE, IgD(IA6–2)-biotin, CC chemokine receptor (CCR4[1G1])-PeCy7, CCR6(11A9)-PE, HLA-DR(L243(G46–6))-PE (BD Biosciences, Erembodegem, Belgium), CD27(1A4-CD27)-PeCy5, CD62L(DREG56)-ECD, CD69(TP1.55.3)-ECD (Beckman-Coulter), CD38(HIT2)-PE (Immunotools, Friesoythe, Germany), CXC chemokine receptor (CXCR3[1C6/CXCR3])-PeCy5 (eBioscience, Uithoorn, The Netherlands), CD27(M-T271)-PE, CD95(DX2)-PE (Dako, Glostrup, Denmark), CD138(B-B4)-PeCy5 (IQ Products, Groningen, The Netherlands), CD25(4E3)-biotin (Miltenyi Biotec), CD27(M-T271)-biotin (Ancell, Bayport, MN, USA), IgM-PE (Southern Biotech, Birmingham, AL, USA), B-cell activating factor receptor (BAFF-R; 11C1; BioLegend, San Diego, CA, USA). The detection of biotinylated antibodies was performed with streptavidin conjugated to Quantum dots (Qdot 605, Invitrogen). Isotype-matched antibodies were used to define marker settings. Intracellular analysis of IL-2(MQ1–17H12)-PE, IL-6(AS12)-PE, IL-10(JES3–19F1)-PE, tumor necrosis factor (TNFα[Mab11])-PE (BD Biosciences), GATA3(TWAJ)-Alexa fluor 647, RORγt(AFKJS-9)-PE, FOXP3(PCH101)-Alexa fluor 647, IL-4(8D4–8)-PeCy7, IL-17(EBIO64DEC17)-Alexa fluor 647, IL-21(eBIO3A3-N2)-PE, interferon (IFNγ[4S.B3])-PeCy7 (eBioscience), Tbet(4B10)-PE (Santa Cruz Biotech, Heidelberg, Germany) and IL-22(142928)-PE (R&D Systems, Oxon, UK) was performed after fixation and permeabilization, using Fix and Perm reagent (eBioscience). Before intracellular cytokine measurement, the cells were stimulated for 4 h with PMA (12.5 ng/mL) plus ionomycin (500 ng/mL) in the presence of Brefeldin A (5 μg/mL; Sigma-Aldrich, Zwijndrecht, The Netherlands).

To study cell division by flow cytometry, 1–10 × 106 cells of interest were labeled with 0.5–1 μM carboxyfluorescein succinimidyl ester (CFSE) (Molecular Probes, Leiden, The Netherlands) before stimulation. The cell phenotype was analyzed by five-color flow cytometry (FC500, Beckman-Coulter) and data were analyzed using CXP software (Beckman-Coulter). In some experiments, cells were counted by flow cytometry using Flow-Count fluorospheres (Beckman-Coulter).

Cytokine detection

After 6 days of culture, B cells were labeled with CD27(M-T271)-PE. CFSEintCD27 and CFSElowCD27high were isolated on an ELITE cell sorter (Beckman-Coulter). Human IL-1β, IL-2, IL-4, IL-6, IL-10, IFNγ and TNFα were determined in the supernatant of these B-cell cultures using a human cytokine multiplex kit (Invitrogen) according to the manufacturer's instructions. Before cytokine measurement, B cells (2.5–5 × 104 cells per well) were stimulated overnight with PMA (12.5 ng/mL) plus ionomycin (500 ng/mL; Sigma-Aldrich). Supernatants were collected and stored at –20°C until analysis.

In vitro T-cell proliferation assay

B cells recovered at day 6 of culture were washed, counted and added to 5 × 104 CFSE-labeled allogeneic CD4+CD25 T cells in a 1:1 ratio for an additional 6 days of culture. CD4+ T cells were analyzed by flow cytometry.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 5.03 (GraphPad Software Inc., La Jolla, CA, USA). Data in box-and-whisker plots represent the median, lower and upper quartiles, and minimum and maximum values. Paired t-tests were used to compare results obtained with cells cultured in the presence or absence of RTX. The Wilcoxon matched-pairs signed rank test was used for nonnormally distributed data. The p values <0.05 were considered statistically significant and are indicated with asterisks.

Results

RTX inhibits B-cell proliferation without apoptosis induction

To mimic the in vivo situation in lymphoid organs, where B cells are exposed to RTX but not depleted, we first characterized a nondepleting in vitro B-cell stimulation model. Freshly isolated B cells were incubated in the presence or absence of RTX for 30 min, followed by an additional 60 min in the presence or absence of complement. Next, cell viability was assessed by flow cytometry using AnxV and PI staining. RTX alone did not increase the frequency of apoptotic (AnxV+PI) or necrotic (AnxV+PI+) cells (Figure 1A). The addition of complement to the RTX-treated cells resulted in a dose-dependent necrosis of B cells, whereas complement alone had a negligible effect on the percentage and absolute number of viable cells (AnxVPI; Figure 1B). Notably, the phenotype of the viable B cells that remained after exposure to RTX and complement was largely similar to that of the cells that were exposed to RTX only (Figure 1C).

Figure 1.

Characterization of an in vitro nondepleting system to establish the functional effects of RTX on B cells. (A) Freshly isolated CD19+ B cells were incubated with (bottom row) or without (top row) 20% rabbit HLA-ABC/DR complement in the presence (right panel) or absence (left panel) of 5 μg/mL RTX. Annexin V (AnxV) and propidium iodide (PI) staining were analyzed using flow cytometry. The percentage of viable (AnxVPI), apoptotic (AnxV+PI) and necrotic (AnxV+PI+) cells is indicated. One representative experiment of four is presented. (B) The percentages and absolute numbers of viable cells (AnxVPI) are shown after 30 min incubation in the absence (W/O; open) or presence of RTX (grey) with subsequent addition of various concentrations of complement for 60 min. Shown are the results for four different donors. (C) The proportions of IgD+, CD27+ and CD38+ B cells within the population of viable cells recovered after incubation with RTX and 20% complement. Shown are the results for four different donors. (D) CFSE-labeled CD19+ B cells were incubated first in the absence (W/O) or presence of RTX and complement. After washing, the cells were stimulated with αCD40 mAb and IL-21 in the presence or absence of RTX. Histograms show the CFSE division patterns at day 6 of total B-cell population (top) or the remaining B cells after RTX and complement mediated lysis (bottom). (E) Summary plot showing the percentages of proliferated B cells (data from four independent experiments, using four different blood donors); *p < 0.05; **p < 0.01; ns, not significant; W/O, without addition of RTX or complement.

To assess the effect of RTX on the proliferative capacity of B cells, fresh B cells were stimulated with αCD40 mAb and IL-21, in the presence or absence of RTX. Dose-response kinetics, using 0, 2.5, 5 and 10 μg/mL RTX revealed optimal inhibition of proliferation at a concentration of 5 μg/mL (data not shown). In the absence of RTX, about 70% of the cells had divided at day 6 of culture, whereas in the RTX-treated condition this percentage was significantly lower (60%, p < 0.01; Figures 1D and E, upper and middle panels). The same experiment was repeated with the small proportion of B cells that remained after incubation with RTX and complement as starting population. In this case a similar inhibition of proliferation by RTX was observed (Figures 1D and E, lower panels). As the effects of RTX on the total B-cell population and on the fraction remaining after exposure to RTX and complement appeared similar, we continued to use the nondepleted B-cell population in further experiments for practical reasons (cell numbers).

B-cell stimulation in the presence of RTX results in a population shift toward an activated naïve phenotype

After the finding that RTX inhibited B-cell proliferation, we studied the phenotype of the RTX-treated B cells by means of CFSE-based multiparameter flow cytometry. Typically, stimulation of B cells with αCD40 mAb and IL-21 for 6 days resulted in the loss of IgD and the gain of CD27 expression during sequential cycles of cell division (Figure 2A, middle panels). The inhibition of cell division by RTX resulted in a relative increase of IgD+CD27 cells. On the basis of the CFSE content and the expression of CD27, IgD and CD38, we defined four B-cell subsets after αCD40 mAb and IL-21 stimulation: undivided CFSE++CD27IgD+CD38 cells (from here on referred to as subset I), undivided CFSE++CD27+IgD−/+CD38 cells (subset II), cells that underwent one or two cell divisions (CFSE−/+CD27IgD+CD38−/+; subset III) and cells that underwent three or more cell divisions (CFSECD27++IgDCD38++; subset IV). A similar subset distribution was observed when the fraction of B cells remaining after incubation with RTX and complement was used as starting population for stimulation with αCD40 mAb and IL-21 (data not shown). The presence of RTX during B-cell stimulation resulted in a relative increase of both the percentage and absolute cell numbers in subset III and a decrease of cells in subset IV (Figures 2B and C).

Figure 2.

Effect of RTX on B-cell maturation. (A) A representative experiment showing flow cytometric analysis of IgD and CD27 expression of freshly isolated CD19+ B cells (start day 0). Subsequently CD19+ B cells were CFSE-labeled and stimulated with αCD40 mAb and IL-21 in the presence or absence of 5 μg/mL RTX. Plots show the IgD and CD27 expression at day 6 of culture against the CFSE expression. (B) Upper panel: Based on the CFSE content and cell surface expression of CD27, IgD, and CD38 on day 6 of culture with αCD40 mAb and IL-21, four B cell subsets can be defined: CFSE++CD27IgD+CD38 (subset I), CFSE++CD27+IgD−/+CD38 (subset II), CFSE−/+CD27IgD+CD38−/+ (subset III), and CFSECD27++IgDCD38++ (subset IV). Lower panel: Percentages of the different B cell subsets after stimulation in the presence or absence of 5 μg/mL RTX. One representative experiment of eight is presented. (C) Summary plot showing the percentages (n = 14) and absolute numbers (n = 6) of the different subsets within the total CD19+ B-cell population; *p < 0.05, **p < 0.01, ***p < 0.001. W/O, without addition of RTX.

Next, the phenotype of the four different subsets was assessed in more detail. Freshly isolated B cells were positive for BAFF-R and the majority of these cells expressed CD24 and CD62L. Approximately half of this population expressed IgD, around 25% expressed CD27 and IgM, whereas expression of CD38, CD138, CD25, CD69, CD95, CD70, CD80 and CD86 was virtually lacking (Figure 3, left panel). Upon stimulation, there was an increased expression of the activation marker CD25 in all subsets, and CD69 was expressed in subsets I, II and III (Figure 3). Subset II was characterized by a relatively low expression of IgD. CD138 was most strongly expressed by cells in subset III, whereas CD38 and CD95 were typically expressed by cells in subset IV.

Figure 3.

Phenotypic characterization of RTX-treated B cells. The expression of various markers on freshly isolated CD19+ B cells (start) and on B cells that were CFSE-labeled and stimulated with αCD40 mAb and IL-21 in the presence or absence of 5 μg/mL RTX. Data are shown from six to eight independent experiments, with different donors; *p < 0.05, **p < 0.01. W/O, without addition of RTX.

Addition of RTX resulted in a significantly higher percentage of IgD+ cells in all subsets, except for subset IV, which lacked IgD+ cells. Furthermore, there was a lower percentage of CD24+ cells in subsets I and II. The expression of other cell surface markers was less markedly changed in the presence of RTX (Figure 3).

RTX inhibits the proliferation of CD19+CD27 naïve but not of CD19+CD27+ memory B cells

As we noticed that the starting population of freshly isolated B cells consisted of both naïve CD27 and memory-type CD27+ cells, we examined which of these populations was most affected by RTX. Sorted CD19+CD27 and CD19+CD27+ B cells were labeled with CFSE and cultured with αCD40 mAb and IL-21 in the presence or absence of 5 μg/mL RTX for 6 days (Figure 4A). In this system, the CD27+ cells proliferated much stronger than the CD27 cells (Figure 4B). The same gating strategy as in Figure 2B was then used to define four B-cell subsets. Upon stimulation, part of the CD19+CD27 B cells divided and differentiated into subset III and to a lesser extent into subset IV type cells, whereas the majority of memory CD19+CD27+ B cells rapidly turned into subset IV-type cells. So, RTX inhibited the proliferation and differentiation of naïve CD19+CD27 but not of memory CD19+CD27+ B cells.

Figure 4.

Effects of RTX on sorted CD27 naïve and CD27+ memory B cells. (A) CD19+ B cells were sorted into a CD27 (naïve) and CD27+ (memory) B-cell fraction. (B) CFSE expression profile of these cells is presented after 6 days of culture with αCD40 mAb and IL-21 in the presence (grey filled histogram) or absence (black line histogram) of RTX. Percentages of the different B-cell subsets as defined in Figure 2B are shown. One representative experiment of six is presented. CFSE-labeled naïve CD19+CD27 (C) and memory CD19+CD27+ (D) B cells were cultured as described in (A). The expression at day 0 (start) and after 6 days of culture of the indicated molecules was assessed by flow cytometry. Data are shown from six to seven independent experiments, with different donors; *p < 0.05, **p < 0.01. W/O, without addition of RTX.

As for the total CD19+ population, we also examined the detailed phenotype of isolated CD19+CD27 and CD19+CD27+ B cells before and after stimulation, in the presence or absence of RTX (Figures 4C and D, respectively). The main finding was that in the presence of RTX, stimulation of CD19+CD27 B cells resulted in a larger fraction of IgD+ cells in subsets I and III, similarly as previously observed for the total CD19+ population. Similar to CD19+ B cells, subset III, differentiated from CD19+CD27 B cells, expressed significantly less CD69 and more CD95 after exposure to RTX, which resulted in a heterogeneous activation status. As expected, RTX had no effect on the phenotype of subset IV, differentiated either from CD19+CD27 or from CD19+CD27+ B cells. Again, RTX increased the percentage of IgD positive B cells in all subsets, except subset IV. Thus, RTX inhibits the proliferation of naïve CD19+CD27, but not memory of CD19+CD27+ B cells, with a population shift into a more naïve (IgD+) phenotype.

RTX has no major effects on the B-cell-cytokine profile

We wondered whether the RTX-induced B-cell population shift was associated with changes in cytokine production. At day 6 of culture, CFSEintCD27 (subset III) and CFSElowCD27high (subset IV) cells were sorted, restimulated with PMA/ionomycin and cytokines were measured in the supernatant. TNFα, IL-2, IL-6 and IL-10 were exclusively produced by CFSEintCD27 cells but not by CFSElowCD27high cells. RTX slightly increased the production of IL-6 by CFSEintCD27 cells but did not change the production of TNFα, IL-2 and IL-10 (Figure S1A).

We also measured the intracellular cytokine production of TNFα, IFNγ, IL-2, IL-4, IL-6 and IL-10 in the B-cell subsets. After exposure to RTX, the fraction of TNFα-producing cells in subsets I + II was slightly enlarged, whereas the fraction of IL-4 producing cells was smaller (Figure S1B). RTX had no effect on the percentage of IL-2 or IL-6 expressing B cells.

B cells exposed to RTX enhance CD4+ T-cell proliferation and induce a more Th2-like phenotype

Next, we studied the effect of RTX on the antigen-presenting function of B cells, after having established that RTX had no effect on HLA-DR expression (Figure S2). B cells stimulated with αCD40 mAb and IL-21 for 6 days in the presence or absence of RTX were added to CFSE-labeled allogeneic CD4+CD25 T cells for an additional 6 days of culture. B cells that were stimulated in the presence of RTX induced considerably stronger T-cell proliferation than B cells stimulated in the absence of RTX (Figures 5A and B).

Figure 5.

T-cell proliferation upon stimulation with B cells. (A) CFSE division patterns of CD4+CD25 T cells after coculture with αCD40 mAb and IL-21 prestimulated B cells in the presence or absence of RTX. Numbers denote the percentage of proliferated cells. (B) Summary plot showing the percentage of proliferated cells from eight different donors; **p < 0.01. (C) The surface expression of the indicated molecules, the intracellular expression of several transcription factors, and the intracellular cytokine staining within the proliferated T-cell population were analyzed on day 6 of culture. (D) The Th1/Th2 ratio of the proliferated T-cell population, represented as IL-2/IL-4 and T-bet/GATA3 ratios. Shown are the results of four to five different donors; *p < 0.05. W/O, without addition of RTX.

T cells that proliferated after stimulation with RTX-treated B cells contained a smaller fraction of CD27+ cells and a larger fraction of CD70+ cells, however there was no effect on T-cell activation (CD25 and CD69) or the ratio of naïve/memory T cells (CD62L; Figure 5C). Interestingly, after stimulation of T cells with RTX-treated B cells, the proliferated T cells showed an increased expression of the Th2-associated chemokine-receptor CCR4, transcription factor GATA3 and increased production of cytokine IL-4. The expression of Th1-associated (CXCR3, Tbet, IL-2, IFNγ, TNFα) and Th17-associated (CCR6, RORγt, IL-17) markers in these T cells was not influenced by exposure of the B cells to RTX before their use as T-cell stimulators. Consequently, the proliferated T cells displayed a decreased Th1/Th2 ratio, as represented by decreased IL-2/IL-4 and T-bet/GATA3 ratios (Figure 5D).

Discussion

In this study, we showed that RTX inhibited the proliferation of stimulated human B cells, which was associated with a relative increase of B cells with an activated naïve phenotype. Aside from this population shift, there were no major changes in phenotype or cytokine profile of the various B-cell subsets as such. B cells stimulated in the presence of RTX induced stronger T-cell proliferation, compared to B cells stimulated in the absence of RTX. Moreover, the resulting T-cell population showed a more Th2-like phenotype.

Studies on the immune processes involved in transplant rejection have been mainly focused on T-cell-mediated mechanisms. Accordingly, most immunosuppressive drugs especially target T cells. Interestingly, anti-B cell therapy appears to be of additional value in ABO-incompatible transplantation, desensitization protocols and for treatment of antibody-mediated rejection (11–13). Moreover, renal transplant patients treated with a single-dose of RTX as an induction therapy, together with standard immunosuppressive treatment, showed a tendency toward fewer and milder rejection episodes compared to the placebo-group (29). However, the relative contribution of B cells in these conditions and the exact mechanism of action of RTX remain to be elucidated. In renal transplant patients, a single dose of RTX effectively depleted all peripheral B cells and B cells in the renal allograft, but spared up to 50% of B cells in lymph nodes (26). Notably, RTX appeared to be bound to these B cells (26). Also in patients with RA, the number of synovial B cells was significantly decreased at 4 weeks after RTX-treatment but there was no complete depletion (30). These important observations suggest that treatment with RTX may affect the phenotype or function of a residual population of B cells. Our in vitro culture system mimicked this scenario, since B cells were exposed to RTX, but survived when no complement was added to the culture medium. Moreover, the small fraction of B cells that remained after complement mediated lysis resembled the total B-cell population with respect to phenotype and effects of RTX. To imitate the in vivo stimulation of B cells in the lymph nodes, we added αCD40 mAb and IL-21 to the culture medium. Furthermore, the culture medium was supplemented with pooled human serum, resulting in IgG1 concentrations of about 500–1000 μg/mL. We therefore consider the observed effects of RTX (used in a concentration of 5 μg/mL) specific for its CD20 binding property and not a consequence of nonspecific binding of the IgG1 molecule.

CD20 is a tetraspanin-like protein involved in lipid raft formation (31). Binding of RTX to CD20 interferes with B-cell receptor (BCR) signaling by preventing BCR relocalization into the lipid rafts and inhibition of the downstream BCR signaling cascade (32). These effects of RTX can explain the inhibition of B-cell proliferation that we observed. Interestingly, we found that the antiproliferative effect of RTX was limited to the CD27 naïve B-cell population. In agreement with these findings, it has been reported that RTX administered in vivo as part of a desensitization protocol decreased the number of splenic naïve B cells, but had no effect on the number of CD27+ memory B cells (33). These discordant effects of RTX on naïve and memory B cells do not seem to be directly related to the level of CD27 expression, because Franke et al. found no relationship between the RTX-induced changes in gene expression and the surface expression of CD27 in various B-cell lines (34). The antiproliferative effect of RTX during stimulation of the B cells was accompanied by a shift in the distribution among B-cell subsets, as defined by the CFSE content in combination with the expression of CD27, IgD and CD38. After 6 days of culture in the presence of RTX, there was an enrichment of CD27IgD+CD38−/+ cells representing an activated naïve phenotype. These cells underwent only one or two cell divisions (subset III), but did not go into apoptosis which can be explained by the lack of cross-linking of the Fc parts of RTX in our in vitro culture system (31).

A striking finding was that B cells stimulated with αCD40 mAb and IL-21 in the presence of RTX induced stronger T-cell proliferation than B cells stimulated in the absence of RTX. Because the phenotype and cytokine profile of the various B-cell subsets only marginally differed, we believe that the altered distribution among the B-cell subsets was mainly responsible for this observation. There was a relative increase of activated naïve B cells (subset III) and a decrease of plasmablast-like cells (subset IV). Plasmablasts (and plasma cells) are terminally differentiated B cells specialized in producing antibodies whereas activated B cells can have antigen presenting functions (35). A relative increase of these latter cells after B-cell stimulation in the presence of RTX can explain the stronger induction of T-cell proliferation.

The enhanced T-cell proliferation induced by RTX-treated B cells might implicate that anti-B cell therapy can also have deleterious effects, as has been reported for renal transplant patients (36). On the other hand, we observed that the proliferated T cells displayed a differentiation toward a Th2-like phenotype that is usually associated with graft acceptance rather than rejection (37). Interestingly, a recent study showed the accumulation of B cells with an inhibited profile in a rat model of allograft tolerance (38). This changed B-cell profile was accompanied by a deviation toward the Th2-related IgG1 isotype alloantibodies. A similar shift in the B-cell population has also been observed in renal transplant patients with stable kidney graft function in the absence of immunosuppression (39).

In summary, we have demonstrated that RTX can affect B-cell proliferation and differentiation, leading to an altered distribution among defined B-cell subsets. Exposure of B cells to RTX has effects on subsequent interaction between B and T cells in vitro. In organ transplantation, anti-B cell therapy should therefore be applied with caution, and preferably be accompanied by additional studies that give more insight into the effects of RTX on the alloimmune response in vivo.

Acknowledgments

The authors thank R. Woestenenk (Dept. of Laboratory Medicine – Laboratory of Hematology) for cell sorting. This study was supported by a grand of the Dutch Kidney Foundation (nr C09–2301).

Author contribution: E.K. designed the research, performed experiments, analyzed data and wrote the paper. H.K. designed the research, analyzed data and wrote the paper. L.B. developed and contributed vital reagents. L.H. designed the research, analyzed data and wrote the paper. I.J. designed the research, analyzed data and wrote the paper.

Disclosure

The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. L.H. has received research funds from Roche, the manufacturer of RTX. Roche had no role in study design, data collection, preparation of the manuscript and decision to publish. The other authors of this manuscript have no conflicts of interest to disclose.

Ancillary