Cells within the DC lineage of white blood cells are the sentinels of the immune system, the most potent of APCs; at various ports of entry, they specialize in capturing antigens and stimulating T cell–associated immunity (1–3). DCs regulate the activation of T cells not only by efficiently processing and presenting immunogenic peptides in association with self-MHC (1–4), but also via expression of coreceptor molecules such as CD80 and CD86 (5–7) and the production of cytokines such as IL-12 (8–11).
Although a role for DCs in the activation of T cells is well established, it is less clear to what extent DCs regulate other cells. It is known that DCs are essential for the development of antibody responses (12–15), but this has been thought to be simply a secondary consequence of the requirement for DCs to activate naive helper T cells. Now several recent findings suggest that DCs may directly regulate B cell maturation. In this issue of the Journal of Experimental Medicine Dubois et al. (16) report that small numbers (250–1,000 cells) of highly purified CD1a+ dendritic–Langerhans cells (D–Lc) can directly stimulate activated B cells both to proliferate and produce significant amounts of IgM or IgG. This dramatic effect is dependent on the B cells being stimulated through CD40 receptors. Of the B cell populations tested, IgG- and IgA-producing memory B cells are the most responsive to D–Lcs, but IgD+ B cells can also be stimulated by DCs to produce IgM in vitro.
How might DCs directly induce B cell maturation? DCs and B cells both constitutively express CD40 and class II, and both, in response to certain cytokines, LPS or CD40 ligation, upregulate surface CD54 (ICAM-1), CD80, and CD86 (4, 7, 10, 17–19). And both DCs and B cells can engage T cells in reciprocal dialogues (19), during which both the T cell and APC are activated via either cell–cell interaction molecules or locally secreted cytokines. Thus, it is likely that some type of reciprocal signaling also occurs between DCs and B cells.
Although the DC-dependent B cell proliferation described by Dubois et al. (16) could be mediated by a soluble factor, optimal induction of antibody production by DCs required both cell–cell contact, soluble factors, and activation of DCs via their CD40 receptors. Stimulation via CD40 has a number of effects on DCs including preventing them from dying (20); so one possibility is that some kind of rescue signal may be required to keep DCs alive in long-term 15-d cultures with B cells. CD40 ligation also induces DCs (and B cells) to express functional CD40L (21), and CD40L+ DCs can stimulate B cells directly to secrete IgG and IgA via a CD40L-dependent pathway (21). Thus, some or all of the cell–cell contact requirement for DC-dependent antibody production may be mediated by CD40L–CD40 signaling. The fact that DCs and B cells can be seen clustered in close contact (16) suggests that as much remains to be learned about DC–B cell and DC-T-B interactions as remained for DC-T interactions after DCs and T cells were first found clustered together (12–14). One intriguing possibility is that B cells in close proximity with DCs may compete for signals from T cells (such as CD40L) and thereby influence the type or degree of cytokines made by DCs (22).
Relatively little is known about the range of cytokines or chemokines made by cells in the DC lineage, let alone which of them may regulate B cells. Although IL-12 is secreted by DCs after CD40 ligation (9, 10), it probably has no direct effect on resting or activated B cells, which do not express IL-12 receptors (23). DCs or DC-related cell lines have been reported to make IL-1, IL-6, and TNF-α (e.g., 2, 24–27), all of which can promote B cell maturation (28). The types of chemokines DC lineage cells can make such as macrophage inflammatory protein-1 gamma (29) are just beginning to be defined. It will be important to compare CD1a+ versus CD1a− DCs for their ability to make known B cell-stimulating factors such as IL-11 and chemokines as well as to assess whether B cells produce factors that can regulate or attract DCs (30, 31).
The recent in vitro studies of DC–B cell interactions (16, 21) leave unanswered a critical question: Where might a DC stimulate a B cell in vivo? Dubois and coworkers point out that after challenge with antigen, T cell–dependent B cell activation and maturation occurs within extrafollicular regions of peripheral lymphoid tissues (32–34). So one likely sequence is that DCs process antigen, migrate to the T cell zones in peripheral lymphoid tissues and there both activate naive T cells and are activated by T cells (1, 4, 17). Antigen-specific B cells in these T cell regions then could form clusters with T cells and their associated DCs (13, 14) and receive signals from both cell types. Since both T cells and B cells enter the spleen through marginal zones (35) and most CD40L+ T cells are found within periarteriolar lymphoid sheaths (PALS) (36), it would be interesting to know whether or not during an immune response B cells actually associate with DCs in these regions, e.g., along the outer edge of the PALS as they migrate into primary follicles.
Another potential site where DCs might encounter and regulate B cells is the germinal center (GC). It had been thought that DCs in lymphoid tissues, such as interdigitating DCs, are relatively restricted to T cell zones. However, GCs also contain specialized CD4+ CD11c+ DCs, designated GCDCs (37), which as potent activators of T cells may function to sustain GC memory T cells or promote T–B interactions within germinal centers. Although Grouard et al. (37) suggest that GCDCs most likely function by stimulating GC T cells, they also show that GCDCs express both complement and Fc receptors opening the possibility that immune complexes on GCDCs may regulate GC memory B cells. While germinal centers persist for only ∼3 wk after immunization, memory B blasts continue to proliferate in follicles for months after the onset of T cell–dependent antibody responses (32). These cells are probably the source of plasma cells and memory cells required to maintain longterm antibody production and could well be regulated by GCDCs. Follicular B cells can be induced to differentiate into plasmablasts by signals such as IL-1 (32, 38), which DCs can produce. Furthermore, some plasmablasts and plasma cells express CD28 (39); thus, GCDCs (37) and other DCs that can express CD80/86 might be able to sustain or stimulate CD28+ plasmablasts through the CD28 pathway. Clearly, much remains to be done beyond these suggestive studies (16, 21, 37) to define the molecular mechanisms involved in DC–B cell interactions and their roles in human diseases involving dysregulated B cells such as lymphomas and autoimmune diseases.
This work was supported by National Institutes of Health grants GM37905 and RR00166.
I thank Carl June for a nice discussion.
Address correspondence to Edward A. Clark, Dept. Microbiology, Box 357242, University of Washington Medical Center, Seattle, WA 98195.