Effects of Acute and Chronic Inflammation on B-Cell Development and Differentiation

Early B-cell Development

In adult mice and humans, the bone marrow (BM) is the primary site of B lymphopoiesis and the screening of newly formed B lymphocytes for self-reactivity. Many of the characteristic events of B-cell development, including the rearrangement of the Igh and Igκ/λ gene loci, correlate with expression of surface markers, allowing the definition of specific stages in B-cell development by multiparameter flow cytometry (Osmond et al., 1998). That several phenotypic definitions have been used by different laboratories to describe B-cell development has led to slightly different models/nomenclature for B lymphopoiesis (Osmond et al., 1998). To avoid confusion, and for the sake of simplicity, in this review we shall focus on the molecular events that define B-cell development.

B cells derive from multipotent, self-renewing hematopoietic stem cells (HSC) and transit through a series of maturational steps and developmental checkpoints in the bone marrow (Hardy et al., 2007; Kondo et al., 2003). The initial developmental checkpoint that defines the B lineage is the lymphoid/myeloid branch point which is thought to occur at or after HSC differentiate into non-renewing multipotent progenitors (MPP) that express the adhesion molecule VCAM-1 (Fig. 1) (Kondo et al., 2003). Functional analyses suggest that MPP first lose the potential for megakaryocyte/erythroid (MegE) differentiation, followed by a diminished capacity to generate granulocytes and macrophages (GM), and the eventual commitment to the lymphoid lineages (Adolfsson et al., 2005; Lai and Kondo, 2006). B lineage specification begins as early as the VCAM-1⁻ MPP stage of development, also known as the early lymphoid progenitor (ELP), where cells express B-cell specific transcription factors together with other lineage specific molecules (Igarashi et al., 2002; Lai et al., 2005). The more differentiated, common lymphoid progenitors (CLP) give rise to pre-pro-B cells (also known as CLP-2) and completely lose T-and NK cell potential upon reaching the pro-B stage (Fig. 1) (Tudor et al., 2000).

Extracellular factors are necessary for the expression of B-cell specific transcription factors, including EBF (early B-cell factor), LRF (liver regeneration factor), and Pax5 (paired box). In mice, IL-7 receptor (IL-7Rα) signaling is necessary for the expression of EBF, at least at the transition of CLP to pre-pro-B cells/CLP-2, as B-cell development is blocked at the pre-pro-B stage in IL-7 and IL-7Rα deficient mice as a consequence of low EBF expression (Dias et al., 2005; Kikuchi et al., 2005). Similarly, lack of the transcriptional repressor LRF (also known as Pokemon) arrests B-cell development at the pre-pro-B/CLP-2 stage and results in extrathymic T-cell development in bone marrow (Maeda et al., 2007), presumably as the result of altered Notch signaling (Pui et al., 1999; Radtke et al., 1999). Another B-lineage transcription factor, Pax5, is also expressed in CLP, even though one of its targets, the B-lineage signal transducer, CD19, is not (Hsu et al., 2006). Presumably, Pax5 levels are insufficient in CLP to induce this critical target and establish B-cell commitment (Hsu et al., 2006). B-cell specific transcription factors form transcriptional networks along with more generally expressed transcription factors including PU.1 and E2A (Nutt and Kee, 2007). In pro-B cells, Pax5 and Stat5 interact and critically regulate the accessibility of the Igh gene loci to the RAG1/2 recombinase and thereby control the V_H-to-D_HJ_H rearrangements necessary for the production of μH polypeptides (Bertolino et al., 2005; Fuxa et al., 2004). Pax5 and E2A are also important in regulating the expression of the SLC polypeptides, Vpre-B and λ5 (Melchers, 2005); in addition, E2A plays a crucial role in the regulation of Igk rearrangement (Lazorchak et al., 2006) and consequently, the generation of immature B cells that express mature BCR capable of interacting with exogenous and self-antigens.

Roughly 90% of the cells that commit to the B lineage do not reach the immature B cell stage (Rolink et al., 1998). Many, perhaps as many as 75% of developing B cells express BCR that recognize self-antigens (Wardemann et al., 2003). Immature B cells that encounter their antigen ligands in the bone marrow undergo apoptosis (Nemazee and Buerki, 1989), or are rendered non-functional by the process known as anergy (Goodnow et al., 1988), or reactivate Ig gene rearrangement to replace the self-reactive BCR (Gay et al., 1993; Radic et al., 1993; Tiegs et al., 1993). This last tolerance mechanism, receptor editing, is unusual in that it operates on the autoreactive receptor rather than the autoreactive cell (Nemazee and Weigert, 2000). Once past this initial tolerance checkpoint, the residual immature B cell compartment begins a series of maturation steps as they emigrate from the bone marrow to the spleen or other peripheral lymphoid tissues where they complete their development.

Peripheral B-cell development

The final stages of B-cell development take place in the spleen where transitional B cells (Carsetti et al., 1995), essentially intermediate forms between the immature and mature phenotypes, undergo selection into the peripheral B-cell compartments (Loder et al., 1999). Type 1 transitional (T1) B cells continue to mature through the transitional 2 (T2) B cell stage, which in turn completes B-cell development/maturation as mature follicular or marginal zone B cells. Tolerance continues to operate during the T1 and T2 stages of B-cell development, as the frequency of self-reactive clones continues to decrease through the immature, T1, and T2 stages of B-cell development (Wardemann et al., 2003). However, unlike immature B cells, T1 B cells do not appear able to undergo receptor editing on exposure to self-antigen ligands (Wang et al., 2007). Instead, T1 cells undergo apoptosis following BCR ligation whereas BCR ligation on T2 B cells induces proliferation and differentiation to antibody secretion similar to mature B cells (Petro et al., 2002). Though the mechanism is still unclear, studies of the frequencies of V_H genes in immature and mature B cell populations have uncovered biases in V_H gene usage, suggesting that positive selection also occurs in the transitional stages of B-cell development (Gu et al., 1991; Levine et al., 2000).

In addition to tonic BCR signaling, extracellular factors are required for transitional B cells to complete their development. B-cell activating factor of the TNF family (BAFF) is an important regulator of transitional B-cell maturation, and controls the T1→T2 transition. Mice deficient for BAFF have severely reduced numbers of B cells beyond the T1 stage of development but normal numbers of developing B cells in the BM (Gorelik et al., 2003; Schiemann et al., 2001; Shulga-Morskaya et al., 2004). Three receptors for BAFF have been identified, transmembrane activator and calcium-modulator and cyclophilin-ligand interactor (TACI), B-cell maturation antigen (BCMA), and BAFF receptor (BAFF-R) (Mackay and Browning, 2002). However, BAFF-R is the principal receptor for controlling peripheral B-cell development, as mice deficient in this receptor exhibit a phenotype comparable to BAFF deficient mice (Sasaki et al., 2004). BAFF-R signaling activates several transcription factors, including NF-κB (nuclear factor of κ B cells) (Kanakaraj et al., 2001), as well as expression of anti-apopototic genes, including Bcl-2, Bcl-x_L, and Mcl-1 (Do et al., 2000; Hatada et al., 2003; Hsu et al., 2002; Woodland et al., 2008). Signaling through BAFF-R also upregulates the expression of CD21 and CD23, B-cell associated receptors (for complement and IgE, respectively) that distinguish T1 B cells from more mature T2 and mature B cells (Gorelik et al., 2004).

A recent study identified a subset of T2 B cells that are highly sensitive to proliferative signals through BAFF-R, suggesting that BAFF targets this population as part of the homeostatic regulation of peripheral B-cell development (Meyer-Bahlburg et al., 2008). Further evidence for BAFF as a homeostatic regulator of peripheral B lymphopoiesis is based on observations that patients exhibit significant increases in serum BAFF concentrations during B cell depletion therapy (Cambridge et al., 2006; Pers et al., 2007), suggesting that BAFF concentrations mediate the size of the peripheral B-cell pool. Interestingly, transgenic mice that express high levels of BAFF exhibit large increases in the numbers of T2 and mature B cells, and typically show evidence of systemic autoimmune disease (Batten et al., 2000; Khare et al., 2000; Mackay et al., 1999). Presumably, BAFF overexpression allows for the survival of T1 B cells that would normally be purged by apoptosis (Thien et al., 2004). Consistent with this hypothesis, transgenic B cells bearing self-reactive receptors are more dependent on BAFF for survival than are alloreactive B cells (Lesley et al., 2004).

Elevated serum BAFF has also been associated with several autoimmune diseases, including lupus, rheumatoid arthritis, and Sjogren’s syndrome (Becker-Merok et al., 2006; Cheema et al., 2001; Groom et al., 2002; Jonsson et al., 2005; Zhang et al., 2001), garnering much attention for BAFF as a potential therapeutic target.

Though BAFF production is typically attributed to myeloid cells, experiments with reciprocal BM chimeras demonstrated that BAFF production by radiation-resistant (presumably non-hematopoietic) cells is necessary for peripheral B-cell development and sufficient for humoral immune responses (Gorelik et al., 2003). However, BAFF-deficient mice reconstituted with BAFF-sufficient bone marrow mount antibody responses in response to immunization with T-cell dependent antigens, indicating that hematopoietic cell-derived BAFF contributes significantly to B-cell maintenance and their differentiation to antibody-secreting cells. Hematopoietic sources of BAFF include dendritic cells (Schneider et al., 1999), macrophages (Craxton et al., 2003), neutrophils (Scapini et al., 2003), T cells (Yoshimoto et al., 2006), and even B cells themselves (Chu et al., 2007; Daridon et al., 2007).

BAFF is required for normal peripheral B-cell development, but is it necessary for mature B cells during an immune response? A role for BAFF in the maintenance of germinal centers has been proposed based on the observation that BAFF-deficient and BAFF-R-deficient mice have attenuated germinal center responses (Rahman et al., 2003; Vora et al., 2003). However, a recent study found that freshly isolated germinal center B cells have no detectable BAFF bound to their surface receptors, though they express receptors capable of binding exogenous BAFF (Darce et al., 2007). Follicular B cells, on the other hand, have BAFF bound to their surface receptors. Together these findings suggest that the germinal center is a BAFF-poor environment, consistent with the concept that the survival of germinal center B cells is determined by competition for antigen (Kelsoe, 1996). It seems possible, therefore, that attenuated germinal center responses in BAFF-deficient mice may be explained better by the indirect effects of BAFF on B-cell development. We note that the few B cells that reach maturity in BAFF-deficient mice do not express CD21, a receptor for the complement activation products C3d and C3b (Carroll, 2004). The binding of CD21 to C3-decorated antigen strongly amplifies BCR signaling (Fearon and Carroll, 2000) and mice deficient in CD21 exhibit decreased antibody responses to T-dependent antigens (Ahearn et al., 1996; Molina et al., 1996), at least in part due to the inability of CD21 deficient B cells to persist in germinal centers (Fischer et al., 1998).

Although a direct role for BAFF in germinal centers can be debated, other B-cell compartments may rely on BAFF to generate effective humoral responses. BAFF has been implicated in T-independent antibody responses to infection and has been shown to promote class-switching independently of T-cell derived CD40 ligand(Castigli et al., 2005; Litinskiy et al., 2002), observations that suggest a direct link between B cells and innate immune cells in T-independent humoral responses. Finally, BAFF production by blood dendritic cells also appears to be required for the differentiation of marginal zone B cells into plasmacytes secreting low-affinity antibody during the early humoral response to bacterial infections (Balazs et al., 2002).

Inflammation and B lymphopoiesis

While infection and inflammation have long been known to affect leukocyte production (Apperley et al., 1989; Nagaoka et al., 2000; Young et al., 1989), it has only been in recent years that the molecular and cellular details of inflammatory hematopoiesis have been explored (Fig. 2). Several recent studies demonstrated that inflammation alters hematopoietic output of bone marrow by favoring myelopoiesis, especially granulopoiesis, over lymphopoiesis (Nagai et al., 2006; Nagaoka et al., 2000; Ueda et al., 2005; Ueda et al., 2004). Our current understanding of how inflammation affects B lymphopoiesis is based on three mechanisms: i) diversion of lymphoid progenitors from the B lineage; ii) induction of apoptosis in developing B cells; and iii) mobilization of developing B cells to the periphery.

Remarkably, microbial products can have direct effects on the differentiation of hematopoietic progenitors. Nagai and his colleagues (Nagai et al., 2006) have demonstrated that hematopoietic progenitors express at least some Toll-like receptors (TLR) and that TLR ligation induces their differentiation into innate immune cells. CLP incubated with TLR ligands differentiate into dendritic cells in vitro. Redirected differentiation of CLP into dendritic cells as a result of TLR ligation could reduce the B-cell progenitor pool, limiting B lymphopoiesis during infection; whether or not this potential reduction is manifest during infections and/or chronic inflammation is unknown but may represent a significant aspect of clinical management. Hopefully, we will soon understand how signaling through TLR overrides the fate decisions that are already established in CLP, such as the expression of Pax5 and the RAG1/2 recombinase.

Hematopoiesis is also affected by inflammatory cytokines. TNFα, perhaps the premiere pro-inflammatory cytokine (Beutler and Cerami, 1989), is known to play a major role in modulating B lymphopoiesis. For example, the bone marrow of mice infected with influenza have significantly lower numbers of developing B cells due to the anti-proliferative effects of TNFα and lymphotoxin (LTα) (Sedger et al., 2002). However, this study did not demonstrate whether the proliferative suppression by TNFα and LTα was direct or acted indirectly by altering the bone marrow environment.

More recently, TNFα has been shown to affect B lymphopoiesis by regulating the expression of chemokines and growth factors in the bone marrow. The CXCL12/CXCR4 axis is crucial to hematopoiesis (Nagasawa, 2006; Nagasawa et al., 1996) and is a target of inflammatory signals (Fedyk et al., 2001). The importance of this chemokine in B lymphopoiesis is evident in CXCL12-deficient mice, which have severely reduced B-cell numbers (Nagasawa et al., 1996). Similarly, blocking CXCR4 with antagonists in vivo mobilizes B cells from the bone marrow into the circulation (Martin et al., 2003). On a finer scale, CXCL12 gradients in the bone marrow may guide developing B cells to different microenvironments; Martin and colleagues proposed that CXCL12 is a necessary component of the stepwise progression of developing B cells through specialized bone marrow “niches” (Tokoyoda et al., 2004). They observed that early B lineage cells (pre-pro-B/CLP-2 cells) associate with CXCL12-expressing reticular cells whereas B-lineage cells in the subsequent, more mature developmental stage (i.e., pro-B cells) associate with reticular cells that express IL-7 but not CXCL12.

Mice immunized with sterile adjuvants such as alum also exhibit significant decreases in the numbers of developing B cells in the bone marrow. This loss of developing B cells correlates with a reduction in bone marrow CXCL12 (Ueda et al., 2004); indeed adjuvant-induced reductions in CXCL12 expression and the concomitant mobilization of B cells from the bone marrow to the blood circulation are attenuated in TNFα deficient mice, demonstrating that TNFα can modulate B lymphoiesis by reducing the ability of the BM to retain developing B cells. This premature emigration may reflect the physiologic process for removing transitional B cells into the periphery (Fedyk et al., 1999) but results in a transient loss of central B lymphopoiesis (Ueda et al., 2005; Ueda et al., 2004).

Inflammation also affects the expression of stem cell factor (SCF) in the bone marrow. The receptor for SCF, c-Kit, is expressed by hematopoietic stem cells (Ikuta et al., 1991; Ogawa et al., 1991), myeloid progenitors (Akashi et al., 2000), common lymphoid progenitors (Kondo et al., 1997), and early B-cell progenitors (Hunte et al., 1998). Mice with inactive c-Kit exhibit age-dependent blockades in B-cell development (Waskow et al., 2002), indicating that B lymphopoiesis in mice more than 10 days old requires SCF. Adjuvant immunization decreases SCF expression in BM, which, like CXCL12 expression, coincides with the mobilization of developing B cells to the circulation (Ueda et al., 2005). However, the inflammatory mediator(s) that alters SCF production in BM is not known.

The reduction in B-cell production during an inflammatory response is accompanied by increased production of neutrophils, a process referred to as “emergency granulopoiesis” (Basu et al., 2000; Hirai et al., 2006). The reciprocal production of B cells and granulocytes suggests that the progenitors of each compartment utilize a common developmental niche (Fig. 2). This hypothesis is supported by the observation that developing B cells and granulocytes co-localize beneath BM stromal cells in vitro to form mixed “cobblestone clusters” (Ueda et al., 2005). Interestingly, the intravenous administration of recombinant TNFα reduces CXCL12 in the bone marrow and mobilizes developing B cells to the periphery, but has only a modest effect on granulopoiesis. On the other hand, administration of IL-1 to mice, a proinflammatory cytokine that has been linked to granulopoiesis (Hestdal et al., 1992; Moore and Warren, 1987; Stork et al., 1988; Ueda et al., 2005), modestly increases BM neutrophil numbers but does not affect B cells. However, co-administration of TNFα and IL-1β recapitulates all of the effects of adjuvant-induced inflammation, mobilizing BM B cells to the periphery and significantly enhancing granulopoiesis, suggesting that these two proinflammatory cytokines synergize to redirect hematopoiesis in favor of granulopoiesis (Ueda et al., 2005).

Inflammation-induced reductions in BM B cells are accompanied by increases in the number of developing B cells in the spleen, indicating that inflammation mobilizes developing B cells from the BM to the periphery (Nagaoka et al., 2000; Ueda et al., 2005; Ueda et al., 2004). The fact that inflammation mobilizes developing B cells instead of inducing apoptosis leads to an important question: do developing B cells that have mobilized to the periphery continue their development outside of the bone marrow? The primary implication of extramedullary lymphopoiesis is that B cells could develop in a “selection-light” environment (Sandel et al., 2001). Development in such an environment might allow for the survival of self-reactive B cells that would normally be deleted in the bone marrow, thereby raising the potential for autoimmune pathology. This hypothesis provides a potential mechanism to explain the strong linkage between inflammation and the induction of systemic autoimmune disease.

On the other hand, the possibility of immature B cells playing a role in immune responses to pathogens cannot be excluded. Immature B cells are prominent in the spleens of adjuvant-immunized mice, placing them in an ideal position to interact with blood-borne antigens. Support for this model was provided in a recent report that focused on antibody production by immature/T1 B cells (Ueda et al., 2007). In this study, the authors observed that the number of splenic T1 B cells significantly increased during inflammation as a result of extramedullary lymphopoiesis. Though signaling through the BCR alone induces apoptosis in T1 B cells (Petro et al., 2002), LPS stimulation in vitro induces proliferation and secretion of class-switched antibody, an effect significantly augmented by the inclusion of BAFF. This observation indicates that T1 B cells are competent to respond to positive BCR signals and participate in T-independent humoral responses. The production of class-switched antibody requires the activity of activation-induced cytosine deaminase (AID). Surprisingly enough, T1 B cells, but not T2 B cells, have been found to express functional levels of AID as part of their developmental program, rather than in response to TLR ligation or T-cell help (Ueda et al., 2007). In addition, T1 B cells express low but significant amounts of Blimp-1 (B-lymphocyte maturation promoting), a transcription factor required for differentiation into ASC (Angelin-Duclos et al., 2000; Turner et al., 1994). Together, these observations suggest that T1 B cells, like marginal zone B cells, may be poised to respond rapidly to antigens as part of the “innate” humoral response. This notion is supported by the recent observation that human transitional B cells differentiate into plasmacytes upon encounter with TLR ligands (Capolunghi et al., 2008).

What is the physiologic role of antigen-responsive T1 B cells? Since T1 B cells have not completed developmental selection, they are enriched for self-reactive clones compared to mature follicular B cells (Wardemann et al., 2003). Normally, the production of low-affinity, polyreactive antibody during the early immune responses is attributed to B-1 and marginal zone B cells (Martin et al., 2001) but the ability of T1 B cells to respond quickly to antigen (Ueda et al., 2007) suggests that these developmentally immature cells may also contribute to early humoral defense.

The observation that LPS-stimulated T1 B cells are dependent on BAFF for survival (Ueda et al., 2007) raises the possibility that BAFF serves as the limiting factor in the recruitment of T1 B cells into immune responses. In this model, the amount of BAFF available in normal, uninfected individuals would be sufficient for B-cell development, maturation, and homeostasis, but not high enough to allow the survival of T1 B cells that encounter self- or exogenous antigens. Conversely, increased BAFF production during infection would promote the survival of T1 cells that encounter antigen, allowing their participation in immune responses. Indeed, the observation that dendritic cells upregulate BAFF production in response to inflammatory cytokines suggests that modulated BAFF production during infection may represent a physiological and protective response to abet humoral immunity (Litinskiy et al., 2002).

Chronic inflammation and B lymphopoiesis

If acute inflammation, even the modest inflammation that results from immunization with adjuvant, affects central hematopoiesis, what are the effects of persistent inflammation? Generally, inflammation is regarded as a beneficial, self-limiting, healing process, allowing for the clearance of the initiating stimuli via localized recruitment of immune effector cells. However, when the inflammatory stimulus persists, the resulting chronic inflammation can be protective or pathogenic depending on the circumstances. For example, granulomas can wall off persistent inflammatory agents when the normal mechanisms of antigen clearance fail or are otherwise insufficient (Chattopadhyay, 1994).

We and many others have used pristane (2,6,10,14-tetramethylpentadecane), a natural, saturated 19-C alkane, to induce chronic peritoneal inflammation (Potter, 2003; Richards et al., 1999). When injected into the peritoneal cavity, pristane is either phagocytosed by macrophages or surrounded by infiltrating leukocytes to form oil-cell complexes. These complexes adhere to peritoneal surfaces and preferentially to the mesenteric membranes (Potter and Maccardle, 1964). With time, the infiltrating leukocyte complexes expand to form structures known as oil granulomas.

Chronic inflammation induces tertiary lymphoid tissues via the same cellular and molecular signals that direct the organization and formation of secondary lymphoid tissues (Drayton et al., 2006). Mice deficient for TNFα (TNFα^−/−), LTα (LTα^−/−), or mature B cells (μMT) exhibit impaired abilities to form normal secondary lymphoid tissues including lymph nodes and Peyer’s patches (De Togni et al., 1994; Fu et al., 1998; Gonzalez et al., 1998; Pasparakis et al., 1996). The generation of organized oil granulomas under conditions of chronic inflammation led us to question whether these granulomas might be considered a sort of tertiary lymphoid tissue (Nacionales et al., 2006; Potter and Maccardle, 1964). If so, mutations that affect the development of secondary lymphoid tissues might also be expected to interfere with oil granuloma development and/or organization. To investigate this potential relationship, we injected C57BL/6 mice that were genetically deficient for genes [TNFα (Pasparakis et al., 1996), LTα (De Togni et al., 1994), μMT (Kitamura et al., 1991), and LAT (Zhang et al., 1999)] known to be important in the development and organization of secondary lymphoid tissues. Remarkably, the mutation with the strongest effect on oil granuloma formation was μMT, an IgCμ mutation that prevents complete B cell development (Kitamura et al., 1991). The μMT mutation is effective in interfering with the development of secondary lymphoid tissues - but these tissues contain large numbers of B cells (Ngo et al., 2001; Tumanov et al., 2004). We note that the μMT mutation also blocks the formation of oil granulomas even though very few B cells are present in the granuloma structure (manuscript in preparation). TNFα^−/− mice exhibit modestly impaired granuloma formation after pristane whereas oil granuloma development in LTα deficient mice is comparable to normal controls (manuscript in preparation). From these observations, we conclude that pristane induced granulomas likely represent an inflammatory tissue that is organized, at least in part, by the genetic pathways that also organize secondary and tertiary lymphoid tissues. B cells, though rare in the oil granulomas of C57BL/6 mice, appear to exert a powerful effect on the development of these tissues.

Chronic inflammation may also contribute to the development of B-cell tumors. Pristane-induced oil granulomas contain macrophages and fibroblasts that secrete IL-6 (Fagarasan et al., 2000; Potter, 2003), an inflammatory cytokine that promotes the differentiation of B cells into plasmacytes (Roldan and Brieva, 1991). Mice that overexpress IL-6 develop spontaneous myelomas and lymphomas at much higher rates than control mice (Kovalchuk et al., 2002). Plasmacytes bearing the chromosomal translocations t(12;15) or t(6;15) proliferate and survive within the IL-6-rich environment of pristane-induced oil granulomas, resulting in the development of plasmacyte tumors called plasmacytomas or myelomas (Fagarasan et al., 2000; Potter, 2003). Thus, granulomatous tissues associated with chronic inflammation promote specific B-cell tumors by providing an environment where neoplastic B cells or plasmacytes escape normal regulatory cues.

Interestingly, the development of pristane-induced plasmacytomas is strain dependent; BALB/c mice are susceptible whereas C57BL/6 mice are resistant, indicating that the genetic differences between the strains affect both granuloma development and predisposition for plasmacytomas. Indeed, oil granulomas in C57BL/6 and BALB/c mice differ in their anatomical location on the mesentery; in C57BL/6 mice, pristane elicits oil granulomas at the gut/mesentery boundary whereas in BALB/c mice granulomas are centripetally distributed on the mesentery [(Potter and Maccardle, 1964) and manuscript in preparation]. The identification of genes that regulate oil granuloma formation is required to understand better the differential susceptibilities to pristane-induced plasmacytomas. At least two genes that confer resistance to plasmacytomas elicited by pristane have been identified (Potter et al., 1994), but their roles in oil granuloma development are unknown.

Chronic inflammation is also associated with the induction or exacerbation of autoimmunity. Pristane-induced autoimmunity is observed in many strains of mice (Mizutani et al., 2005; Satoh et al., 2003; Satoh et al., 2000) and has been linked to the induction of apoptosis by peritoneal cells, which provides the autoantigen substrate necessary to break tolerance (Calvani et al., 2005; Kirou et al., 2005; Potter and Maccardle, 1964; Zhuang et al., 2005). Sequestration of pristane by granuloma formation may act to inhibit autoimmunity; C57BL/6 mice wall off injected pristane much more efficiently than do BALB/c mice and C57BL/6 mice are less prone to develop severe autoimmune disease in response to pristane than are BALB/c mice (Nacionales et al., 2006; Potter and Maccardle, 1964; Wooley et al., 1989).

Conclusions

There is much more to hematopoiesis and B-cell development than the normal, steady-state hematopoietic pathways that have been described in naïve mice and humans. Inflammation redirects central hematopoiesis by altering the bone marrow environment to favor myelopoiesis and by mobilizing developing B cells to the blood and peripheral lymphoid sites where exposure to antigens and inflammatory agents are common. The utility, consequences, and potential dangers of extramedullary B lymphopoiesis are not yet known but deserve additional study and consideration, especially in the light of the persistent inflammation associated with chronic infection and autoimmunity. In the periphery, B-cell tolerance may be less efficient and the capacity of AID⁺ T1 B cells to respond to antigenic stimuli suggests a pathway for generating and expanding clones of B cells reactive with self-antigens. A particularly attractive test of this hypothesis may be found in the epitope spreading associated with the desmoglein antibodies responsible for pemphigoid disease (Amagai et al., 1999; Futei et al., 2000; Ishii et al., 1997; Miyagawa et al., 1999; Tsunoda et al., 2002).