B cell–activating factor belonging to the tumor necrosis factor family (BAFF; also known as B lymphocyte stimulator [BLyS]) is a 285–amino acid type II transmembrane protein member of the tumor necrosis factor ligand superfamily (1, 2). Constitutive overexpression of BAFF in BAFF-transgenic (Tg) mice that, otherwise, are not autoimmune-prone leads to systemic lupus erythematosus (SLE)–like features (3–5).
In humans, circulating levels of BAFF are elevated in many systemic rheumatic disorders, including SLE (6, 7). Importantly, overexpression of BAFF is not immutably associated with autoimmune disease. For example, patients with various hematologic malignancies or immunodeficiency disorders harbor elevated circulating levels of BAFF (8–11). Thus, BAFF may be incapable of driving clinically important autoimmunity under certain in vivo conditions. To test this possibility, we turned to 2 murine models in which resistance to autoimmunity has been conferred through discrete genetic manipulations.
The first model uses C57BL/6 (B6)–congenic mice that are genetically deficient in the I-Aβ chain (B6.Aβ−/− mice) and, therefore, do not express class II major histocompatibility complex (MHC). Circulating autoantibody levels in SLE-prone mice are markedly reduced in the absence of class II MHC (12, 13), highlighting the vital role of class II MHC molecules in the development of humoral autoimmunity.
The second model uses B6-congenic mice homozygous for the Sle1 region (B6.Sle1 mice) (14). These mice develop elevated titers of circulating IgG antichromatin autoantibodies but rarely develop renal disease (15, 16). Full-blown disease develops when other genetic “insults” are superimposed, such as the Sle2 or Sle3 regions (17, 18). Importantly, there are genetic modifiers, especially the Sles1 region, which abrogate the ability of B6.Sle1 mice to generate SLE-associated autoantibodies (19). Circulating autoantibody levels in B6.Sle1.Sles1 mice are identical to those in B6 wild-type mice, and B6.Sle1.Sles1 mice do not develop end-organ (kidney) disease, even when the Sle2 or Sle3 region is introgressed.
In this study, we demonstrated that constitutive overexpression of BAFF in BAFF-Tg B6.Aβ−/− mice overcomes the resistance to autoimmunity associated with class II MHC deficiency and promotes humoral autoimmunity. Nevertheless, the renal immunopathology is very modest, and clinical disease does not develop. BAFF-Tg B6.Sle1.Sles1 mice also develop humoral autoimmunity, but the constitutive overexpression of BAFF is incapable of reversing the Sles1-mediated inhibition of Sle1-dependent autoimmunity, and renal immunopathology is very limited. Thus, the consequences of BAFF overexpression on in vivo autoimmunity may be highly dependent upon the genetic background of the host, raising the possibility that BAFF overexpression, even when present, may not necessarily be driving disease in some SLE patients. Accordingly, the clinical effects of therapeutic BAFF neutralization may be widely disparate across the heterogeneous population of patients with SLE.
- Top of page
- MATERIALS AND METHODS
- AUTHOR CONTRIBUTIONS
Constitutive overexpression of BAFF in non–autoimmune-prone hosts is known to lead to B cell expansion, high titers of circulating IgG autoantibodies, and in some cases, Ig deposition in the kidneys (3–5). We report here that constitutive BAFF overexpression can overcome some, but not all, of a host's genetic resistance to SLE-like autoimmunity.
The first model we studied took advantage of the strong dependence on class II MHC expression that development of circulating autoantibodies and disease has in SLE-prone mice (12, 13), providing prima facie evidence that a class II MHC–deficient state creates an autoimmune-resistant environment. In comparison to their class II MHC–sufficient B6.Aβ+/− littermates, class II MHC–deficient B6.Aβ−/− mice harbored increased spleen B cells and Ig-secreting cells along with the expected marked decrease in CD4+ cells (Figures 1 and 2). Consistent with this decrease in CD4+ cells, total serum IgG levels were substantially lower in B6.Aβ−/− mice than in their B6.Aβ+/− littermates. Not surprisingly, serum levels of SLE-associated IgG autoantibodies were low in these non–autoimmune-prone B6.Aβ−/− and B6.Aβ+/− mice.
Introduction of a BAFF transgene into B6.Aβ+/− mice led, as expected, to increased numbers of spleen B cells and Ig-secreting cells and to increased levels of serum total IgG and IgG antichromatin antibody, although the increase in serum levels of neither IgM nor IgA antichromatin antibodies achieved statistical significance. Furthermore, although spleen B cell numbers did not increase in B6.Aβ−/−.Baff mice, spleen levels of Ig-secreting cells and serum levels of total IgG and IgG autoantibodies did, despite the paucity of CD4+ cells and the complete absence of class II MHC. Whether the inability of class II MHC–deficient B cells to expand in response to BAFF overexpression reflects an inherent defect of such B cells or whether it is consequent to the marked reduction in CD4+ cells remains to be determined. Definitive studies will require the generation of mice in which class II MHC deficiency is limited solely to B cells.
Regardless, the increased levels of IgG autoantibodies in B6.Aβ−/−.Baff mice proves that the effects of constitutive overexpression of BAFF on humoral autoimmunity do not require expression of H-2b, a haplotype that permits the development of SLE-associated IgG autoantibodies in BALB/c and C3H mice (29). Moreover, the divergent effects of BAFF overexpression on B cells and IgG autoantibody levels in B6.Aβ−/− mice highlight the in vivo separability of B cell numbers from B cell–derived products. Along this vein, we have previously documented serologic autoimmunity in BAFF-deficient NZM2328 mice despite their lifelong profound reductions in B cells (30).
The absence of BAFF-driven B cell expansion in B6.Aβ−/− mice sharply contrasts with the BAFF-driven B cell expansion observed in wild-type B6 mice or B6 mice completely devoid of T cells (24). It is possible, but unlikely, that class II MHC–deficient B cells inherently have impaired proliferative capacity in response to BAFF overexpression, since BAFF overexpression leads to increased numbers of Ig-secreting cells and increased production of IgG autoantibodies by such class II MHC–deficient cells. An alternative, non–mutually exclusive, explanation invokes CD4+ Treg cells, which are increased in BAFF-Tg mice (24, 31). CD4+FoxP3+ Treg cells in class II MHC–deficient hosts differ phenotypically from those in class II MHC–intact hosts (32), so their respective inhibitory effects on BAFF-driven B cell expansion may also differ. A third, non–mutually exclusive, explanation invokes CD8+ cells. In contrast to class II MHC–intact hosts in which most FoxP3+ cells are CD4+, the FoxP3+ cells in class II MHC–deficient hosts are evenly distributed between CD4+ and CD8+ subsets (32). Thus, the regulatory role of CD8+ Treg cells in class II MHC–deficient hosts may be considerable, and such cells may modulate what otherwise would be unfettered BAFF-driven expansion of B cells. Additional experimentation, including the genetic depletion of CD8+ cells, will be required to definitively determine the mechanisms underlying the absence of BAFF-driven B cell expansion in B6.Aβ−/− mice.
The increase in circulating IgG autoantibodies in B6.Aβ−/−.Baff mice was associated with no incremental glomerular deposition of IgG and only very modest incremental glomerular deposition of C3. Importantly, no changes in the renal histopathologic appearance were noted in B6.Aβ−/−.Baff mice (Figure 3). These findings must be reconciled with a previous report of substantial renal immunopathologic changes in BAFF-Tg, T cell–deficient B6 mice (24). It may be that CD4+ and/or CD8+ Treg cells that are present in class II MHC–deficient hosts but absent from T cell–deficient hosts play a protective immunomodulatory role. Alternatively, subtle differences in genetic backgrounds and/or local environments may have affected the immunopathologic outcomes. Further investigation in this area is warranted.
The second model we studied took advantage of the ability of the Sle1 region to confer development of IgG antichromatin autoantibodies upon B6 mice and the ability of the Sles1 region to abrogate such autoimmunity (15, 16, 19). Accordingly, a Sles1-bearing host is resistant to Sle1-dependent autoimmunity (although not necessarily resistant to Sle1-independent autoimmunity).
In Sle1-expressing mice, functional changes are independently expressed in both B cells and T cells (26, 33). The Sle1 locus is a complex of multiple susceptibility alleles (25), including Cr2 (34), 4 Slam family genes (Slam, Ly108, Cd84, and Cd48) (35), and Fcgr2b (36). Both Sle1 (NZW) alleles of Cr2 and Fcgr2b alter germinal center B cell function (34, 36), and the Sle1 (NZW) allele of Ly108 impedes negative selection of transitional B cells (37). Importantly, both B cells and T cells (including CD4+ memory cells) were increased in B6.Sle1 mice (Figure 4), and introduction of a BAFF transgene led to further expansions in these cells. Determination of how BAFF overexpression interacts with the individual susceptibility alleles within the Sle1 locus will require breeding of the BAFF transgene into the respective individual subcongenic strains. Regardless, expansion of CD4+ memory cells may be especially important, since expansion of this population in SLE-prone NZM2328 mice closely reflects the onset and severity of clinical disease (22, 38).
One remarkable finding was that BAFF-driven expansion of T cells was abrogated in B6.Sle1.Sles1.Baff mice, whereas BAFF-driven expansion of B cells was unaffected. Despite this lack of effect on BAFF-driven B cell expansion, the IgG autoantibody profile of B6.Sle1.Sles1.Baff mice resembled that of B6.Baff mice rather than B6.Sle1.Baff mice (Figure 5), raising the possibility that the Sle1-associated component of the IgG autoimmune response is largely driven by T cells. Indeed, the Sle1a and Sle1c subregions can promote the development of chromatin-specific autoreactive CD4+ cells (39). Since the Sles1 locus maps to a small region that contains the class II MHC genes and abrogates Sle1-driven T cell activation while leaving Sle1-driven ERK phosphorylation in B cells intact (40), it may inhibit autoreactive CD4+ cells in B6.Sle1.Sles1.Baff mice, and such inhibition may be resistant to BAFF overexpression. Nonetheless, the Sles1 region may also importantly affect Sle1-driven B cell activation (as opposed to B cell expansion) and/or autoantibody affinity maturation and thereby contribute to inhibition of Sle1-associated autoimmunity through T cell–independent means. Formal testing of these possibilities is needed, and B6.Sles1.Baff mice could be especially informative in establishing the effects of the Sles1 region on the biologic consequences of constitutive BAFF overexpression.
In any case, BAFF overexpression appears to be completely capable of promoting Sle1-independent IgG (and IgA) autoantibody production. The inability of a Sles1-bearing host to overcome the autoantibody-promoting effects of BAFF overexpression is reminiscent of the inability of a Sles1-bearing host to overcome the autoantibody-promoting effects of the Sle2 or Sle3 regions (19). Regardless, and consistent with their elevated circulating levels of IgG autoantibody, glomerular deposition of IgG and C3 in B6.Sle1.Sles1.Baff mice was greater than that in B6.Sle1.Sles1 mice. In contrast, no renal histopathology was noted in B6.Sle1.Sles1.Baff mice (Figure 6), and clinical disease did not develop.
Our collective findings in 2 disparate models of autoimmune resistance demonstrated a discordance among glomerular IgG deposition, glomerular C3 deposition, glomerular IgA deposition, and renal histopathology. Glomerular deposition of IgG and IgA in B6.Sle1.Baff mice was greater than that in B6.Sle1 mice, but glomerular deposition of C3 was similar. Conversely, glomerular C3 and IgA deposition in B6.Aβ−/−.Baff mice was greater than that in B6.Aβ−/− mice, but glomerular IgG deposition was similar. Furthermore, no renal histopathology developed in B6.Sle1.Sles1.Baff mice despite considerable glomerular deposition of IgG, C3, and IgA.
Indeed, the limited renal immunopathology and absence of clinical disease in 12-month-old B6.Baff mice indicate that longstanding overexpression of BAFF requires additional factors to promote bona fide disease. The similar renal immunopathology in the 12-month-old B6.Sle1.Baff mice in the present study and in the 3-month-old B6.Sle1.Baff mice previously described (28) suggests that the immunopathology promoted by BAFF overexpression reaches a plateau. Only in the presence of other genetic and/or environmental factors might this finite BAFF-driven immunopathology progress to clinical disease.
The differential ability of BAFF overexpression to promote or permit different facets of autoimmunity may have important ramifications for human disease. Although BAFF antagonists have been highly effective in treating murine SLE (5, 41–43), results in treating human SLE have been less impressive (44–46). It may be that in some patients, only certain autoimmune elements of SLE are promoted or permitted by BAFF. If clinical disease in any given patient is largely BAFF-independent, then that patient would not be expected to respond to a BAFF antagonist. Future studies will be needed to distinguish patients whose disease is largely BAFF-driven from those whose disease is not.