TNF is well characterized as a mediator of inflammatory responses. TNF also facilitates organization of secondary lymphoid organs, particularly B cell follicles and germinal centers, a hallmark of T-dependent Ab responses. TNF also mediates defense against tumors. We examined the role of TNF in the development of inflammatory autoimmune disorders resembling systemic lupus erythematosus and Sjögren’s syndrome induced by excess B cell-activating factor belonging to the TNF family (BAFF), by generating BAFF-transgenic (Tg) mice lacking TNF. TNF−/− BAFF-Tg mice resembled TNF−/− mice, in that they lacked B cell follicles, follicular dendritic cells, and germinal centers, and have impaired responses to T-dependent Ags. Nevertheless, TNF−/− BAFF-Tg mice developed autoimmune disorders similar to that of BAFF-Tg mice. Disease in TNF−/− BAFF-Tg mice correlates with the expansion of transitional type 2 and marginal zone B cell populations and enhanced T-independent immune responses. TNF deficiency in BAFF-Tg mice also led to a surprisingly high incidence of B cell lymphomas (>35%), which most likely resulted from the combined effects of BAFF promotion of neoplastic B cell survival, coupled with lack of protective antitumor defense by TNF. Thus, TNF appears to be dispensable for BAFF-mediated autoimmune disorders and may, in fact, counter any proneoplastic effects of high levels of BAFF in diseases such as Sjögren’s syndrome, systemic lupus erythematosus, and rheumatoid arthritis.

Most ligands of the TNF family play important roles within the immune system (1), but in some cases also contribute to autoimmune diseases if their expression becomes dysregulated (2, 3). A recent example is B cell-activating factor belonging to the TNF family (BAFF,3 also called BLyS, zTNF4, THANK, TALL-1, and TNFSF13b), which plays an essential role in B cell survival and maturation (4, 5, 6, 7, 8, 9, 10). However, when overexpressed in transgenic (Tg) mice, BAFF induces autoimmune disorders with features of systemic lupus erythematosus (SLE) and Sjögren’s syndrome (SS) (11, 12, 13). As a potent B cell survival factor, BAFF is thought to promote autoimmunity via inappropriate survival of autoreactive B cells (reviewed in Refs. 8 ,9 ,14 , and 15). Elevated serum BAFF levels have been detected in patients with rheumatoid arthritis (RA), SS, and SLE (16, 17, 18, 19). Also, some patients with RA have elevated levels of BAFF in synovial fluid that often exceed serum levels (16). Interestingly, elevated BAFF levels have been detected in patients with non-Hodgkin’s lymphoma (NHL), and BAFF promotes survival of both NHL-derived B lymphoma cells and B cell chronic lymphocytic leukemia (B-CLL) cells (20, 21). Thus, BAFF may well promote survival and growth of certain B cell cancers.

BAFF acts at a critical immune checkpoint during peripheral B cell maturation (4). In the spleen, newly formed bone marrow-derived immature B cells, termed immature transitional type 1 (T1) B cells, differentiate into transitional type 2 (T2) B cells before final maturation (22). BAFF promotes the survival of T2 and mature B cells (23, 24, 25). Consequently, B cell maturation beyond the T1 B cell stage is impaired in BAFF−/− mice (26, 27). It is at the transitional immature B cell stage that autoreactive B cells are thought to undergo apoptosis following recognition of self Ags (28), and excessive BAFF production may prevent this censoring mechanism (4, 29). In addition, BAFF promotes survival of peripheral B cells, in particular plasmablasts (24, 30) and germinal center (GC) B cells (31). However, it is still unclear which of these survival functions are critical for the development of autoimmunity in BAFF-Tg mice. The observation that the T2 B cell subset is particularly expanded in BAFF-Tg mice suggested an effect of excessive BAFF production on B cell maturation as the possible origin of the autoimmune disorder (4, 9, 15, 23). Previous work has suggested that maturing T2 B cells could be precursors for marginal zone (MZ) B cells (22). The mechanism for this alternative B cell differentiation pathway is unclear, and may depend on specific characteristics of the B cell receptor on maturing B cells, leading to an accumulation of poly-reactive/self-reactive B cells in the MZ (32, 33, 34, 35). Interestingly, the MZ B cell compartment of BAFF-Tg mice is also dramatically expanded (23). Therefore, a connection between autoreactive T2 B cells, MZ B cells, and autoimmunity in BAFF-Tg mice is a possibility (23). Modification of the MZ B cell compartment often correlates with the emergence of autoimmune features in mice (36, 37), and infiltrating MZ-like B cells have been detected in salivary glands of BAFF-Tg mice (17) and the thyroid gland of patients with Graves’ disease (38). However, a direct connection between this B cell compartment and autoimmune disorders has never been confirmed in BAFF-Tg mice, because several other aspects of B cell function are modified in these animals. For instance, B cell activation and spontaneous GC formation are prominent in BAFF-Tg mice (12), and failure of immunological tolerance following B cell activation and Ig hypermutation in the GC reaction is another possibility.

TNF was originally described as an antitumor agent that delivered cytotoxic signals to tumor cells, yet, paradoxically, also contributed to cancer progression (39). TNF seems essential for promoting cytotoxic T cell immune responses against B cell lymphomas (40, 41). TNF also promotes inflammatory reactions, and controls several immune functions such as the organization of lymphoid tissues and GC formation (reviewed in Ref. 42). TNF-deficient mice have abnormal splenic B cell zones that are not organized into discrete follicles, but instead form a ring encircling the T cell zones (reviewed in Refs. 42 and 43). These mice also lack follicular dendritic cells (FDC), fail to mount GC reactions, and have impaired immune responses (43). Moreover, Fas ligand-deficient mice with generalized lymphoproliferative disorder (gld) that lack TNF have reduced splenomegaly and mild as well as delayed autoimmune disorders (44), suggesting that TNF participates in autoimmune processes linked to inappropriate lymphocyte survival. Importantly, anti-TNF treatment has proven very effective in treating RA in humans (45, 46) due, presumably, to reduction of inflammation and possibly inhibition of autoimmune responses and GC formation (45, 47). In view of the efficacy of anti-TNF therapies in treating some autoimmune conditions such as RA, we tested whether lack of TNF influenced BAFF-mediated autoimmune disorders, in particular via the prevention of GC formation in BAFF-Tg mice, by crossing BAFF-Tg mice onto TNF-deficient mice (TNF−/−). The resulting TNF−/− BAFF-Tg mice were not protected from autoimmune disease, and, indeed, some aspects were even more pronounced than in BAFF-Tg mice. Significantly, >35% of BAFF-Tg mice that lacked TNF developed lymphomas.

BAFF-Tg mice have been described previously (12). BAFF produced and secreted by hepatocytes is detectable at high levels in blood (Fig. 1 A). BAFF-Tg mice were maintained as heterozygotes for the transgene by backcrossing the animals onto C57BL/6 mice. We used BAFF-Tg mice after at least 10 backcrossing steps onto C57BL/6. At this point, BAFF-Tg mice were interbred to generate homozygous BAFF-Tg mice. TNF-deficient mice, with a C57BL/6 genetic background, were provided by J. Sedgwick (DNAX, Palo Alto, CA) (48). TNF−/− BAFF-Tg mice were established by breeding homozygous BAFF-Tg and TNF−/− mice and subsequently interbreeding the F1 generation. Animals chosen for experimental analysis were homozygous for the mutant TNF allele and homozygous for the BAFF transgene. The genotype of BAFF-Tg mice and TNF−/− BAFF-Tg was determined using PCR and Southern blot assays on genomic DNA obtained from 5-mm tail snips. TNF-α-specific primers were: sense-U (5′-ATC CGC GAC GTG GAA CTG GCA GAA-3′) and antisense-L (5′-CTG CCC GGA CTC CGC AAA GTC TAA-3′) (48). The optimal annealing temperature was 63°C, and extension time was 2 min. Before further breeding, mice appearing homozygous by Southern blot were further subjected to test breeding by mating with C57BL/6 to confirm by PCR that the entire resulting F1 progeny was positive for the BAFF transgene, indicating that the BAFF-Tg parent had copies of the transgene on both alleles. Littermates, wild type (wt; Tg negative/TNF+/+), BAFF-Tg/TNF+/+, and TNF−/−/Tg negative, were kept and used as controls in all experiments. The presence of protein in mouse urine was measured using Multistix 10 SG reagent strips for urinalysis (Bayer, Diagnostics Division, Elkhart, IN) (12).

FIGURE 1.

Serum BAFF levels, spleen weights, and CD4+ T cell numbers in BAFF-Tg and TNF−/− BAFF-Tg mice. A, Detection of mouse BAFF by ELISA (n ≥ 4 animals per group, 5 mo of age), including wt (squares), TNF−/− (diamonds), BAFF-Tg (circles), and TNF−/− BAFF-Tg mice (triangles) (mean indicated by horizontal bar). B, Splenic weights, wt littermates (▪), TNF−/− (□), BAFF-Tg (▦), and TNF−/− BAFF-Tg mice ( ) at various ages, as indicated. C, Numbers of CD4 (left panel) and CD8 (right panel) T cells in 12-mo-old wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice (coded as in B). Results in B and C are shown as the mean and SD of at least eight animals per group. D, Fresh splenocytes isolated from 6-mo-old wt, BAFF-Tg, TNF−/−, and TNF−/− BAFF-Tg littermates were stained in parallel with a mixture of anti-IgM/anti-CD23/anti-CD21/anti-IgD or anti-CD5/anti-B220/anti-IgM/anti-IgD or anti-B220/anti-Ly-6D Abs and analyzed by flow cytometry. Various B cell subsets were gated according to their respective phenotype: mature B cells (CD23+, IgMdull, IgD+, CD21intermediate(int)), T1 B cells (CD23, IgMhigh, IgD, CD21), T2 B cells (CD23+, IgMhigh, IgDhigh, CD21high), MZ B cells (CD23, IgMhigh, IgD, CD21high), B-1a B cells (B220int, IgMhigh, CD5+), B1-b (B220int, IgMhigh, CD5), and plasmablasts (B220low, Ly-6Dhigh). Percentages for all B cell subsets were calculated by the flow cytometer, and mean absolute numbers per spleen plus SD for five to nine animals are shown. Statistical analysis was done using ANOVA, and significance with relation to wt is indicated as follows: p < 0.001 (∗∗∗), p < 0.005 (∗∗), p < 0.05 (∗).

FIGURE 1.

Serum BAFF levels, spleen weights, and CD4+ T cell numbers in BAFF-Tg and TNF−/− BAFF-Tg mice. A, Detection of mouse BAFF by ELISA (n ≥ 4 animals per group, 5 mo of age), including wt (squares), TNF−/− (diamonds), BAFF-Tg (circles), and TNF−/− BAFF-Tg mice (triangles) (mean indicated by horizontal bar). B, Splenic weights, wt littermates (▪), TNF−/− (□), BAFF-Tg (▦), and TNF−/− BAFF-Tg mice ( ) at various ages, as indicated. C, Numbers of CD4 (left panel) and CD8 (right panel) T cells in 12-mo-old wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice (coded as in B). Results in B and C are shown as the mean and SD of at least eight animals per group. D, Fresh splenocytes isolated from 6-mo-old wt, BAFF-Tg, TNF−/−, and TNF−/− BAFF-Tg littermates were stained in parallel with a mixture of anti-IgM/anti-CD23/anti-CD21/anti-IgD or anti-CD5/anti-B220/anti-IgM/anti-IgD or anti-B220/anti-Ly-6D Abs and analyzed by flow cytometry. Various B cell subsets were gated according to their respective phenotype: mature B cells (CD23+, IgMdull, IgD+, CD21intermediate(int)), T1 B cells (CD23, IgMhigh, IgD, CD21), T2 B cells (CD23+, IgMhigh, IgDhigh, CD21high), MZ B cells (CD23, IgMhigh, IgD, CD21high), B-1a B cells (B220int, IgMhigh, CD5+), B1-b (B220int, IgMhigh, CD5), and plasmablasts (B220low, Ly-6Dhigh). Percentages for all B cell subsets were calculated by the flow cytometer, and mean absolute numbers per spleen plus SD for five to nine animals are shown. Statistical analysis was done using ANOVA, and significance with relation to wt is indicated as follows: p < 0.001 (∗∗∗), p < 0.005 (∗∗), p < 0.05 (∗).

Close modal

Rat anti-mouse BAFF mAbs (clones 1C9, 9B11, and 5A8) were obtained after spleen and lymph node (LN) fusion from Wistar rats immunized with surface BAFF-expressing RBL-2H3 rat mast cells (CRL-2256; American Type Culture Collection, Manassas, VA) obtained after transfection with full-length murine BAFF cDNA sequence inserted in a PCR-3 expression vector (Invitrogen, Carlsbad, CA), as previously detailed (49). Abs were purified on protein G-Sepharose for fast flow (Amersham Biosciences, Uppsala, Sweden). Abs 9B11 and 1C9 were labeled with 10× molar excess of EZ link sulfo-N-hydroxysuccinimide biotin (Pierce, Rockford, IL), while incubating at room temperature for 30 min. The biotinylation reaction was stopped with 150 mM glycine. The sample was applied onto a desalting column to remove the free biotin (Amersham Pharmacia, Uppsala, Sweden). The rat anti-mouse transmembrane activator and calcium modulator and cyclophilin ligand interactor (TACI) Ab (clone 8F10) was produced after immunization of Wistar rats with rTACI protein, prepared as follows. The extracellular domain of murine TACI (aa 2–121) was amplified by RT-PCR using cDNA derived from FACS-sorted splenic B cells and the following primers: sense, 5′-CCGCTCGAGGCTATGGCATTCTGCCCCAAA-3′, and antisense, 5′ CGCGGATCCTCAGCTACTTAGCCTCAATCCTGG-3′. The amplified TACI fragment was subcloned into the XhoI and BamHI sites of the expression vector pET-19b (Novagen, Madison, WI), resulting in the plasmid pET-19b-TACI containing an NH2-(His)10 tag. Correct TACI amplification and cloning were confirmed by automated sequencing. BL21(DE3) Escherichia coli strain (Novagen) was transformed with pET-19b-TACI vector and cultured at 37°C to an OD600 nm of 0.6. Expression of the protein was induced at 30°C by the addition of isopropyl β-D-thiogalactopyranoside to a final concentration of 1 mM, with continued incubation of the culture for 3 h. Cells were harvested by centrifugation, then lysed on ice for 1 h in a buffer containing 1% Triton X-100, 0.3 mg/ml lysozyme, and 1 mM PMSF. After cell debris removal by centrifugation, 10 mM imidazole and 300 mM NaCl were added to the cleared supernatant, which was loaded onto a HiTrap chelating high performance column precharged with 100 mM NiSO4 (Amersham Biosciences). After several washes of the column with buffer containing 10 mM imidazole, bound His-tagged TACI was eluted stepwise with elution buffer containing 50, 100, and finally 200 mM imidazole.

T-dependent Ab responses in control littermates as well as TNF−/− BAFF-Tg mice were tested using two different Ag, soluble nitrophenyl (NP)-coupled OVA, NP-OVA (Biosearch Technology, Novato, CA), and particulate SRBC (Institute of Medical and Veterinary Sciences, Adelaide, Australia). Mice were immunized i.p. with 150 μg of NP-OVA emulsified in CFA (Sigma-Aldrich, St. Louis, MO), followed by a second immunization with the same dose of Ag in IFA, 21 days later. Animals were bled before the initial immunization and 7 days after each immunization to determine Ab responses. Another group of mice was immunized i.p. with 100 μl of a 10% SRBC suspension (equivalent to 108-5 × 108 SRBC per mouse). Mice were bled on days 7 and 14 postimmunization, and hemagglutination assays were done, as previously described (50). To test the immune responses to T-independent type 2 (TI-2) Ag, mice were immunized i.p. with 30 μg NP-coupled Ficoll (NP59-aminoethyl carboxymethyl-Ficoll; Biosearch Technology). Approximately 100 μl of blood was collected 1 day before the initial immunization and 7 and 14 days following the immunization to measure the NP-specific Ab response.

Lymphocytes from LN, spleen, bone marrow, and thymus were isolated by mechanical disruption of the organs, followed by 5-min incubation in RBC lysis buffer (0.156 M ammonium chloride, 0.01 M sodium bicarbonate, and 1 mM EDTA) on ice. For multicolor flow cytometric analysis, cells were incubated in the presence of fluorochrome- and biotin-conjugated mAb against CD21/CD35 (7G6), CD23 (B3B4), B220 (RA3-6B2), CD4 (L3T4), CD8 (53-6.7), CD44 (IM7), L-selectin (MEL-14), CD5 (53-7.3), Ly-6D (49-H4), CD45R (B220), CD5 (Ly-1), CD1 (1B1) (BD PharMingen, San Diego, CA), and Cy5-conjugated goat anti-mouse IgM Ab (Jackson ImmunoResearch Laboratories, West Grove, PA). Where biotinylated Abs were used, streptavidin CyChrome, streptavidin PerCP (BD PharMingen), or streptavidin Cy5 (Jackson ImmunoResearch Laboratories) allowed detection. Staining with anti-TACI rat mAb 8F10 was detected using PE-conjugated anti-rat IgG, Fcγ chain specific (Jackson ImmunoResearch Laboratories), after blocking nonspecific binding with 10 μg/ml human Ig. Rat IgG2a (BD PharMingen) was used as isotype control. All Abs were diluted in FACS buffer (PBS, 1% BSA, and 0.02% NaN3). A FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ) with CellQuest software was used for data acquisition and analysis.

The concentration of soluble BAFF in mouse serum was determined by coating 384-well ELISA plates (Nalge Nunc International, Rochester, NY) with 5 μg/ml purified rat anti-mouse BAFF (clone 5A8) in 50 mM sodium bicarbonate buffer (pH 9.6) at 4°C overnight. ELISA plates were washed twice with PBS/0.1% Tween 20 (Sigma-Aldrich) and blocked with 1% BSA in PBS at 37°C for 2 h. To reduce interference from rheumatoid factor (RF) in the serum, Ig were cleared by treating serum dilutions with 0.05 vol of protein G-Sepharose (Amersham Biosciences) twice for 1 h at 4°C before pelleting the beads and collecting supernatants (precleared sera). A total of 20 μl/well serum or standard mouse rBAFF serial dilutions in ELISA buffer (PBS, 0.05% Tween, 1% BSA) was added to the plates and incubated for 1 h at 37°C. The plates were washed three times before adding 10 μg/ml biotinylated anti-mouse BAFF (clone 1C9), followed by HRP-labeled streptavidin (DAKO, Glostrup, Denmark). The plates were washed four times before the enzymatic reaction was conducted using the tetramethylbenzidine substrate kit (BD PharMingen). The reaction was stopped by adding 2 M H2SO4, and the OD was measured at 450 nm using Tecan SpectraImage microtiter plate reader (Bio-Tek Instruments, Winooski, VT). NP-specific Abs were detected by coating 384-well plates with 10 μg/ml NP2-BSA (Biosearch Technology) in 50 mM sodium bicarbonate buffer (pH 9.6) at 4°C overnight. A total of 20 μl/well serum serial dilutions (untreated or precleared) was incubated in NP2-BSA-coated 384-well plates, and NP-specific Abs bound to the plate were detected using HRP-labeled goat anti-mouse Ig (Jackson ImmunoResearch Laboratories). The incubation and washing steps as well as the enzymatic reaction were conducted, as described above. The titer (log base 2) for anti-NP Abs was defined as the serum dilution giving an OD 3 times higher than that of background (1 = 1/100 dilution). ELISA for the detection of RF and anti-dsDNA and anti-ssDNA autoantibodies were done as previously detailed (12). The titer (log base 2) for RF and anti-ds/ssDNA autoantibody response is defined as the serum dilution giving an OD 3 times higher than that of background (1 = 1/100 dilution for RF; 1 = 1/50 dilution for anti-DNA autoantibody). Anti-NP Abs produced by animals immunized with NP-Ficoll were detected in 384-well ELISA plate coated with 10 μg/ml NP23-BSA (Biosearch Technologies). Serial dilutions of the serum were added to the plate. Anti-NP Abs were detected using 1 μg/ml alkaline phosphatase-conjugated goat anti-mouse Ig (H + L), IgM, IgG1, IgG2a, IgG3, or IgA (Southern Biotechnology Associates, Birmingham, AL). Coating, incubation, and washing steps were conducted, as described above, and P-NP phosphate substrate (Sigma-Aldrich) was used for visualization. The titer (log base 2) is defined as the serum dilution giving an OD 3 times higher than that of background (where 1 = 1/100 dilution).

Splenocytes were obtained from wt, BAFF-Tg, TNF−/−, and TNF−/− BAFF-Tg (8–12 wk old) and prepared, as described above, for flow cytometry. Lymphocytes (2 × 106/ml) were then stimulated with 5 μg/ml goat anti-mouse μ-chain Ab (Jackson ImmunoResearch Laboratories) in complete RPMI culture medium. Cultured cells were collected at 8, 24, and 48 h for FACS analysis.

Various tissues (lymphoid tissues, kidneys, salivary glands, liver, lung, gut, and tumors) were collected from C57BL/6, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice. Tissues were either frozen in OCT compound (Tissue-Tek; Sakura, Finetek, CA) or fixed in 10% buffered Formalin and embedded in paraffin. Tissue sections, 5 μm thick, were stained with H&E for histologic examination or used for immunohistochemical staining, as previously described (17).

Both BAFF-Tg mice and TNF−/− BAFF-Tg mice had high BAFF levels in serum (Fig. 1,A). The splenic weight in TNF−/− BAFF-Tg and control littermates (wt, BAFF-Tg, and TNF−/− mice), measured at ages ranging from 2 to 13 mo (Fig. 1,B), revealed splenomegaly in TNF−/− BAFF-Tg mice in all age groups. Splenomegaly was not as severe as in BAFF-Tg mice, yet was significantly greater than that of wt and TNF−/− littermates. The spleens of TNF−/− mice were consistently smaller than spleens from wt mice (Fig. 1 B).

Analysis of the T cell compartments in TNF−/− BAFF-Tg mice revealed a clear disproportion between CD4 and CD8 T cells (Fig. 1,C). The absolute numbers of CD4 T cells in 12-mo TNF−/− BAFF-Tg mice were considerably greater than in age-matched wt littermates (Fig. 1,C), whereas CD4 T cell numbers in BAFF-Tg and TNF−/− mice were normal, as previously described (12, 48) (Fig. 1,C). In contrast, CD8 T cell numbers were slightly, but significantly fewer in TNF−/− BAFF-Tg mice than in control mice (Fig. 1 C), and this reduction was already significant in 6-mo animals (data not shown).

The B220+ B cell compartment was less expanded in TNF−/− BAFF-Tg mice when compared with littermate BAFF-Tg mice (Fig. 1,D, B220-positive cells), and this was not explicable by a reduced production of BAFF in TNF−/− BAFF-Tg mice (Fig. 1,A). However, a similar 4-fold increase in B cell numbers was observed in BAFF-Tg vs wt littermates, and in TNF−/− BAFF-Tg vs TNF−/− littermates (Fig. 1,D). The B cell compartment of TNF−/− mice was significantly smaller than in wt littermates and was reflected by reduced numbers of T1, T2, and mature B cells, but not MZ B cells (Fig. 1 D).

Analysis of all B cell subsets in TNF−/− BAFF-Tg mice revealed differences compared with BAFF-Tg mice (Fig. 1,D). The absolute numbers of T2 and MZ B cells, but not mature B cells, were significantly higher in TNF−/− BAFF-Tg mice compared with wt littermates (Fig. 1,D). In contrast to previous studies (12, 17, 23), BAFF-Tg mice in the present experiments were homozygous for the BAFF transgene. Further to studies with heterozygous BAFF-Tg mice (23), homozygous BAFF-Tg mice had significantly more T1 and B-1 B cells in the spleen than did wt littermates (Fig. 1,D). The numbers of B-1 cells were increased in older heterozygous BAFF-Tg mice (17) and, most likely, a dosage effect from the BAFF transgene may have accelerated this phenomenon in homozygous BAFF-Tg mice. However, B-1 B cells still represent a small proportion of all B cells in the spleen of homozygous BAFF-Tg mice used in these experiments (Fig. 1,D). The expansion of the T1 B cell compartment in homozygous BAFF-Tg mice was unclear. However, due to variations, no significant or consistent expansion of the T1 and B-1 B cell compartments was observed in the spleen of TNF−/− BAFF-Tg mice homozygous for the BAFF transgene (Fig. 1,D), and similar inconclusive results were obtained analyzing B-1 B cells in peritoneal lavages (data not shown). Interestingly, there were fewer mature B cells in the spleens of TNF−/− mice, perhaps reflecting the absence of B cell proliferation generated within the small residual GCs that are usually seen in wt mice under normal housing conditions, but are lacking in TNF−/− mice (Fig. 1,D). The number of plasma cells in all groups of mice was variable, but appeared higher in both BAFF-Tg and TNF−/− BAFF-Tg mice when compared with wt or TNF−/− littermates (Fig. 1 D). In conclusion, splenomegaly was less pronounced in BAFF-Tg mice lacking TNF, yet was significant and associated with the selective expansion of the T2, MZ B cell, and CD4 T cell compartments, but not that of mature B cells.

TNF−/− mice lack several important splenic features (51), in particular, B cell follicles and a clear T and B cell demarcation, which is also true for LN and Peyer’s patches. These mice also lack FDC networks, mucosal addressin cell adhesion molecule-1 expression on endothelial cells of MZ sinuses and on FDC, and GC formation is impaired (51). Overexpression of BAFF in TNF−/− BAFF-Tg mice did not alter the splenic, LN, and Peyer’s patch defects seen in TNF−/− (data not shown).

T-dependent immune responses are impaired in TNF−/− mice (52). As BAFF is a potent stimulator of B cell activation and Ab production (24, 49), we tested whether immune responses in TNF−/− BAFF-Tg mice were more effective than in TNF−/− mice. TNF−/− BAFF-Tg mice and control wt, BAFF-Tg, and TNF−/− littermates were immunized with SRBC, a T-dependent and particulate Ag. As expected, SRBC-specific IgG responses, measured 14 days after immunization (Fig. 2,A), were higher in BAFF-Tg mice compared with wt animals (Fig. 2,A). The initial IgM response on day 7 was lower in both TNF−/− BAFF-Tg and TNF−/− mice when compared with wt littermates (Fig. 2,A), and SRBC-specific IgM were no longer detected in these mice at day 14. This result indicated that the initial production of plasma cells producing SRBC-specific IgM was not sustained in TNF−/− BAFF-Tg mice similar to TNF−/− littermates. The IgG response to SRBC was similarly impaired in both TNF−/− and TNF−/− BAFF-Tg mice (Fig. 2,A). Thus, BAFF overexpression did not improve the Ab response to SRBC in TNF−/− BAFF-Tg mice compared with the response detected in TNF−/− mice. Similar results were obtained when mice were immunized with the soluble Ag NP-OVA with adjuvant, and no difference in anti-NP Ab response was noted between TNF−/− BAFF-Tg mice and TNF−/− control littermates (Fig. 2 B). These results indicate that despite being a strong stimulator of B cell activation and Ab production (24, 49), BAFF was unable to stimulate T-dependent Ab responses when overexpressed in TNF−/− mice.

FIGURE 2.

T-dependent immune responses are impaired in TNF−/− BAFF-Tg mice, while TI-2 responses are increased. A, Mice were immunized with SRBC, and SRBC-specific IgM (upper panel) and IgG (lower panel) Ab responses were determined 7 and 14 days postimmunization using SRBC hemagglutination assays in wt (□), TNF−/− (⋄⋄), BAFF-Tg (○), and TNF−/− BAFF-Tg (▵) littermates. Six animals were analyzed per group. Anti-SRBC IgM concentrations after 14 days (upper right panel) and IgG (lower panels) responses were significantly reduced in TNF−/− and TNF−/− BAFF-Tg mice compared with wt or BAFF-Tg littermates (p < 0.05 by ANOVA). B, NP-specific Ab response in NP-OVA-immunized wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice as indicated with the same symbols as in A, 7 days after a secondary immunization with NP-OVA, as detailed in Materials and Methods. At least seven animals per group were analyzed. The mean Ab production for each panel is shown with horizontal bars in A and B. No difference was observed between TNF−/− and TNF−/− BAFF-Tg or wt and BAFF-Tg mice (p > 0.05). C, NP-specific Ab response in wt (▪), TNF−/− (□), BAFF-Tg (▦), and TNF−/− BAFF-Tg ( ) mice 7 days after immunization with TI-2 NP-Ficoll. Six animals per group were used; statistical analysis was done using ANOVA, and significance with relation to wt is indicated as follows: p < 0.001 (∗∗∗), p < 0.005 (∗∗), and p < 0.05 (∗). All immunizations were given to 6-mo-old mice.

FIGURE 2.

T-dependent immune responses are impaired in TNF−/− BAFF-Tg mice, while TI-2 responses are increased. A, Mice were immunized with SRBC, and SRBC-specific IgM (upper panel) and IgG (lower panel) Ab responses were determined 7 and 14 days postimmunization using SRBC hemagglutination assays in wt (□), TNF−/− (⋄⋄), BAFF-Tg (○), and TNF−/− BAFF-Tg (▵) littermates. Six animals were analyzed per group. Anti-SRBC IgM concentrations after 14 days (upper right panel) and IgG (lower panels) responses were significantly reduced in TNF−/− and TNF−/− BAFF-Tg mice compared with wt or BAFF-Tg littermates (p < 0.05 by ANOVA). B, NP-specific Ab response in NP-OVA-immunized wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice as indicated with the same symbols as in A, 7 days after a secondary immunization with NP-OVA, as detailed in Materials and Methods. At least seven animals per group were analyzed. The mean Ab production for each panel is shown with horizontal bars in A and B. No difference was observed between TNF−/− and TNF−/− BAFF-Tg or wt and BAFF-Tg mice (p > 0.05). C, NP-specific Ab response in wt (▪), TNF−/− (□), BAFF-Tg (▦), and TNF−/− BAFF-Tg ( ) mice 7 days after immunization with TI-2 NP-Ficoll. Six animals per group were used; statistical analysis was done using ANOVA, and significance with relation to wt is indicated as follows: p < 0.001 (∗∗∗), p < 0.005 (∗∗), and p < 0.05 (∗). All immunizations were given to 6-mo-old mice.

Close modal

Trinitrophenyl (TNP)-specific IgM response following immunization with TI-2 Ag TNP-Ficoll has been shown to be normal in TNF−/− mice, while the TNP-specific IgG responses were normal at day 7 and slightly elevated after 14 days (52). Interestingly, TI-2 Ab response to NP-Ficoll is induced in both BAFF-Tg and TNF−/− BAFF-Tg (Fig. 2,C). NP-specific IgA response was particularly increased, but also total Ig (Fig. 2 C) and IgG2a (data not shown). Interestingly, NP-specific IgG3 response appeared significantly higher in TNF−/− BAFF-Tg mice when compared with BAFF-Tg, TNF−/−, and wt controls. NP-specific IgM response appeared slightly reduced in TNF−/− mice compared with the other three mouse groups, which showed similar IgM response to NP-Ficoll. Slight differences seen in the TNF−/− mice used in this study compared with TNF−/− mice used in previous work (52) may be attributable to the difference in genetic background.

We tested whether BAFF-induced emergence of autoreactive B cells in BAFF-Tg mice requires TNF expression by measuring levels of autoantibodies in the serum of TNF−/− BAFF-Tg mice and control wt, TNF−/−, and BAFF-Tg littermates. Production of RF at 12 mo of age was slightly higher in TNF−/− BAFF-Tg than in BAFF-Tg mice, and a similar observation was made for anti-dsDNA autoantibodies (p < 0.05, Fig. 3 A). In contrast, anti-ssDNA autoantibody responses appeared slightly higher in BAFF-Tg mice when compared with TNF−/− BAFF-Tg mice, yet both groups had significantly more anti-ssDNA autoantibodies than wt and TNF−/−, respectively (p < 0.01). Therefore, autoantibody production in BAFF-Tg mice does not require TNF and can be driven by mechanisms independent of conventional T-dependent GC immune responses.

FIGURE 3.

TNF−/− BAFF-Tg mice produce high levels of RF and anti-DNA autoantibodies, and develop nephritis and Sjögren’s-like pathologies. A, RF levels in serum of 12-mo-old animals (left panel), anti-dsDNA (middle panel), and anti-ssDNA (right panel), as determined by specific ELISA in wt (squares), TNF−/− (diamonds), BAFF-Tg (circles), and TNF−/− BAFF-Tg (triangles) mice. The mean autoantibody production for each panel is shown with horizontal bars. B, H&E-stained 5-μm kidney tissue sections from 12-mo-old wt (upper left), TNF−/− (upper right), BAFF-Tg (lower left), and TNF−/− BAFF-Tg mice (lower right). Arrows indicate glomeruli. C, H&E staining of submaxillary gland sections of 12-mo-old wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice, as indicated. Arrows show the presence of normal ducts and acinar cells. B and C, Representative of over 30 mice dissected per group.

FIGURE 3.

TNF−/− BAFF-Tg mice produce high levels of RF and anti-DNA autoantibodies, and develop nephritis and Sjögren’s-like pathologies. A, RF levels in serum of 12-mo-old animals (left panel), anti-dsDNA (middle panel), and anti-ssDNA (right panel), as determined by specific ELISA in wt (squares), TNF−/− (diamonds), BAFF-Tg (circles), and TNF−/− BAFF-Tg (triangles) mice. The mean autoantibody production for each panel is shown with horizontal bars. B, H&E-stained 5-μm kidney tissue sections from 12-mo-old wt (upper left), TNF−/− (upper right), BAFF-Tg (lower left), and TNF−/− BAFF-Tg mice (lower right). Arrows indicate glomeruli. C, H&E staining of submaxillary gland sections of 12-mo-old wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice, as indicated. Arrows show the presence of normal ducts and acinar cells. B and C, Representative of over 30 mice dissected per group.

Close modal

BAFF-Tg mice reaching 1 year of age develop glomerulonephritis (12). To determine whether TNF deficiency protected against disease, we examined kidneys from TNF−/− BAFF-Tg mice and control mice for signs of kidney disease (Fig. 3,B). At 12 mo of age, the kidneys of BAFF-Tg mice had advanced tissue damage, with abnormal glomeruli showing clear signs of global mesangial proliferation and obliteration of capillary lumina (Fig. 3,B). Kidneys of 12-mo-old TNF−/− BAFF-Tg mice showed similar abnormalities as in BAFF-Tg mice (Fig. 3,B). We also detected Ig deposits in the glomeruli of both BAFF-Tg and TNF−/− BAFF-Tg by immunofluorescent staining with FITC-labeled anti-mouse Ig Abs (data not shown). Therefore, progression of nephritis in BAFF-Tg mice is independent of TNF. SS-like features were also detected in BAFF-Tg mice (17), and were characterized by acinar cell atrophy in structures lining ducts. Salivary ducts were abnormal in the salivary glands of both BAFF-Tg and TNF−/− BAFF-Tg mice (Fig. 3 C).

BAFF is a factor that promotes survival of B lymphoma cells (20, 21). Moreover, we previously showed that the occasional aged BAFF-Tg mouse developed lymphoid masses due to B cell hyperplasia (17). However, tumor-like masses in BAFF-Tg mice were a rare event, 0.03% in heterozygous BAFF-Tg mice (data not shown) and 2.9% in homozygous BAFF-Tg mice (Table I). In contrast, >35% of TNF−/− BAFF-Tg mice at 12 mo of age had developed very large lymphoid masses at various locations, most frequently in cervical LN, mesenteric LN (MLN), inguinal LN, and the small intestine (Table I and Fig. 4, A–C). Fig. 4,B shows a cervical lymphoid mass from a TNF−/− BAFF-Tg mouse, compared with the usual enlarged cervical LN found in a BAFF-Tg mouse. We also observed a lesion of the small intestine in TNF−/− BAFF-Tg mice (Fig. 4,C) that distended the intestinal wall, and histological analysis showed a complete invasion and destruction of the mucosa by lymphoid cells (Fig. 4 D), predominantly by B220+ cells (data not shown). These features were those of an invasive extranodal mucosal-associated lymphoid tissue-like lymphoma, associated with lymphoepithelial lesions (53).

Table I.

Incidence and location of tumor formation in TNF−/− BAFF-Tg

GenotypeMice DissectedTumor FormationLocation (No. of cases)aIncidence (%)
wt 20 None 
TNF−/− 15 None 
BAFF-Tg 34 CLN 2.9 
TNF−/− BAFF-Tg 63 24 CLN (11) 38.1 
   MLN (7)  
   ALN (1)  
   SI (1) p < 0.0001b 
   SI + MLN (2)  
   CLN + MLN (1)  
GenotypeMice DissectedTumor FormationLocation (No. of cases)aIncidence (%)
wt 20 None 
TNF−/− 15 None 
BAFF-Tg 34 CLN 2.9 
TNF−/− BAFF-Tg 63 24 CLN (11) 38.1 
   MLN (7)  
   ALN (1)  
   SI (1) p < 0.0001b 
   SI + MLN (2)  
   CLN + MLN (1)  
a

CLN, Cervical LN; ALN, axillary LN; SI, small intestine.

b

The p value was calculated using the hypergeometric distribution test.

FIGURE 4.

Lymphomas in TNF−/− BAFF-Tg mice. A, Cervical lymphoid masses associated with submaxillary glands in two TNF−/− BAFF-Tg mice (right panels) and indicated with arrows are compared with BAFF-Tg mice salivary glands (left panel). B, Cervical lymphoid tumor in a TNF−/− BAFF-Tg mouse (right) compared with a typically enlarged BAFF-Tg cervical LN (left). C, Lymphoma of the small intestine of a TNF−/− BAFF-Tg mouse (above) compared with unaffected small intestine (below). The dotted line indicates the orientation of a cross section for histology. D, H&E staining of wt small intestine (left panel) compared with the small intestine lesion shown in C from a TNF−/− BAFF-Tg mouse (right panel). Respective boxed regions (upper panel) are also shown at higher magnification (lower panels). Arrow indicates residual mucosa and lymphoepithelial damage in TNF−/− BAFF-Tg mouse small intestine lesion (lower right panel). E, H&E staining of wt MLN (left panel) and a mesenteric lymphoid mass found in a TNF−/− BAFF-Tg mouse (right panel). The arrow indicates a follicle within the wt MLN (upper left), while no lymphoid microarchitecture was recognizable in the MLN mass found in the TNF−/− BAFF-Tg mouse (upper right). Lower panels, Show higher magnification of these tissues. Note that cells in the mesenteric lymphoid mass found in a TNF−/− BAFF-Tg mouse (lower right) display neoplastic features such as irregular nucleus, clumped chromatin, and increased cytoplasmic volume.

FIGURE 4.

Lymphomas in TNF−/− BAFF-Tg mice. A, Cervical lymphoid masses associated with submaxillary glands in two TNF−/− BAFF-Tg mice (right panels) and indicated with arrows are compared with BAFF-Tg mice salivary glands (left panel). B, Cervical lymphoid tumor in a TNF−/− BAFF-Tg mouse (right) compared with a typically enlarged BAFF-Tg cervical LN (left). C, Lymphoma of the small intestine of a TNF−/− BAFF-Tg mouse (above) compared with unaffected small intestine (below). The dotted line indicates the orientation of a cross section for histology. D, H&E staining of wt small intestine (left panel) compared with the small intestine lesion shown in C from a TNF−/− BAFF-Tg mouse (right panel). Respective boxed regions (upper panel) are also shown at higher magnification (lower panels). Arrow indicates residual mucosa and lymphoepithelial damage in TNF−/− BAFF-Tg mouse small intestine lesion (lower right panel). E, H&E staining of wt MLN (left panel) and a mesenteric lymphoid mass found in a TNF−/− BAFF-Tg mouse (right panel). The arrow indicates a follicle within the wt MLN (upper left), while no lymphoid microarchitecture was recognizable in the MLN mass found in the TNF−/− BAFF-Tg mouse (upper right). Lower panels, Show higher magnification of these tissues. Note that cells in the mesenteric lymphoid mass found in a TNF−/− BAFF-Tg mouse (lower right) display neoplastic features such as irregular nucleus, clumped chromatin, and increased cytoplasmic volume.

Close modal

Histological analysis of an enlarged MLN found in TNF−/− BAFF-Tg mice also showed abnormal features (Fig. 4,E). In contrast to normal MLN, MLN masses found in TNF−/− BAFF-Tg mice had no recognizable lymphoid structure, and immunohistochemical staining showed that the lymphoid population in these MLN masses was dominated by B220+ cells (data not shown). Higher power microscopic observation of the cells in these MLN masses also revealed abnormal features (Fig. 4 E). Although normal lymphocytes in wt MLN were small cells characterized by a very dark nucleus and minimal cytoplasm, cells in the MLN masses collected from TNF−/− BAFF-Tg mice displayed neoplastic features such as a medium size with round and irregular nucleus, clumped chromatin, presence of a small nucleolus, and increased cytoplasm size often with a villous appearance. Similar observations were made with tissues from enlarged cervical LN (data not shown). These pathologic features indicated intense mitotic activity and are reminiscent of neoplastic cells in MZ cell lymphomas in humans (53, 54). Clonality of lymphoma cells is another feature often tested, which will require extensive cloning, sequencing of V regions of Ig genes from tumor cells, as well as Southern blot experiments. Clonality, however, is thought to be less reliable than morphological changes of cells for the diagnosis of such lymphomas (55).

TACI, one of the BAFF receptors, has been shown to be a negative regulator of B cell growth and activation (56, 57). In addition, TACI-deficient mice develop lymphomas (58). Therefore, we examined whether the high incidence of lymphoma we see in TNF−/− BAFF-Tg mice could be associated with defective TACI expression on B cells in these mice. TACI expression has been shown to be up-regulated following B cell activation (58). We generated a mouse anti-TACI mAb and tested up-regulation of TACI expression on B cells from wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg littermates. Splenocytes from these mice were cultured and activated with goat anti-μ Abs for up to 48 h. At various time points, cells were harvested and stained with anti-TACI Ab, and TACI expression was analyzed by flow cytometry on gated B220+ cells. As described by others (58), B cell activation augmented TACI expression on wt B cells, but also on TNF−/− and BAFF-Tg B cells in culture, with a maximum expression level at 24 h postactivation (Fig. 5). TACI expression on resting TNF−/− BAFF-Tg B cells was similar to that of control B cells, but anti-μ activation only weakly up-regulated TACI expression on these cells (Fig. 5). We were unable to show any modulation of TACI expression with rTNF on wt splenocytes in vitro (data not shown), suggesting that a secondary effect resulting from the combination of excess BAFF and lack of TNF resulted in the development of TNF−/− BAFF-Tg B cells intrinsically unable to up-regulate TACI expression upon activation in vitro. As signaling through TACI is critical to protect mice against lymphomas, defect of TACI expression on B cells from TNF−/− BAFF-Tg may have contributed to the higher lymphoma incidence in these mice.

FIGURE 5.

Reduced TACI up-regulation in B cells from TNF−/− BAFF-Tg mice. Splenocytes from wt, BAFF-Tg, TNF−/−, and TNF−/− BAFF-Tg mice were activated for up to 48 h with goat anti-μ Abs. Mean fluorescence intensity (MFI) of TACI expression over time following activation is shown on gated B220+ B cells from wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice, as indicated, and was assessed by flow cytometry using an anti-mouse TACI Ab. These results are representative of three separate experiments.

FIGURE 5.

Reduced TACI up-regulation in B cells from TNF−/− BAFF-Tg mice. Splenocytes from wt, BAFF-Tg, TNF−/−, and TNF−/− BAFF-Tg mice were activated for up to 48 h with goat anti-μ Abs. Mean fluorescence intensity (MFI) of TACI expression over time following activation is shown on gated B220+ B cells from wt, TNF−/−, BAFF-Tg, and TNF−/− BAFF-Tg mice, as indicated, and was assessed by flow cytometry using an anti-mouse TACI Ab. These results are representative of three separate experiments.

Close modal

This study showed that TNF is clearly dispensable for the inflammatory sequela developing within the kidneys or salivary glands of BAFF-Tg mice. TNF undoubtedly contributes to certain autoimmune pathologies, illustrated by studies in animal models of rheumatoid arthritis, as well as the impressive benefit of anti-TNF therapies in human RA and inflammatory bowel disease (45, 46, 59). In addition, T-dependent mechanisms such as those controlled by CD40L have been shown to be critical for the pathogenesis of lupus in autoimmune-prone mice (60). Obviously, TNF-independent inflammatory mechanisms have led to tissue destruction in BAFF-Tg mice, and it is likely that at least a proportion of patients with SLE or SS will similarly have a disease process independent of TNF-mediated inflammation, or pathogenic events derived from the T-dependent GC reaction.

BAFF overexpression leads to intense splenic activity, in particular to spontaneous GC formation, as well as abnormalities of B cell maturation (12, 23). One aim of this study was to ascertain the relative importance of BAFF-mediated corruption of B cell tolerance in lymphoid GC (and thereafter), vs that which has been hypothesized to occur in the spleen during B cell maturation (4). During the GC reaction, harmful autoreactive B cells die in situ and are cleared by specialized tingible body macrophages (61). However, if these cells receive inappropriate survival signals (62), they may persist and trigger autoimmune disorders. In this study, we clearly show that autoimmunity in BAFF-Tg does not require conventional T-dependent immune responses and GC formation, and we conclude that critical immune tolerance checkpoints during B cell maturation in the spleen are the likely stages for the breakdown of immune tolerance. Interestingly, the mature/follicular B cell compartment remained normal in TNF−/− BAFF-Tg mice, in contrast to the T2 and MZ B cell compartments, which were enlarged, possibly reflecting the absence of proliferating mature GC B cells and impaired B cell proliferation from T-dependent responses in TNF−/− BAFF-Tg mice, a BAFF-stimulated feature that would normally contribute to mature B cell expansion in BAFF-Tg mice (12).

TNF−/− BAFF-Tg mice were clearly different from gld mice lacking TNF, because in the latter splenomegaly was not observed and signs of disease were considerably reduced and delayed (44). The gld mice develop SLE-like features similar to those of BAFF-Tg mice, yet the cause is linked to inappropriate survival of both B and T cells (63). Fas ligand is important in promoting activation-induced cell death and controlling immune responses (63); this ligand is highly expressed in GCs, participates in GC B cell apoptosis (62), and may also contribute to eliminate self-reactive cells at this site. However, with an excess of BAFF, impaired T-dependent immune responses and absence of GC did not influence the manifestation of autoimmunity in TNF−/− BAFF-Tg mice, pointing to the dysregulation of a separate pathway of B cell activation.

Interestingly, Ab responses to the TI-2 Ag NP-Ficoll were higher in both BAFF-Tg and TNF−/− BAFF-Tg, in particular IgA responses. This result parallels previous observations showing increased IgA response to T-independent Ags following BAFF injections in mice (24). Both MZ and B-1 B cells respond to T-independent Ags (64); however, in the case of NP-Ficoll immunization, only MZ B cells, but not B-1 B cells, participate in the response (65). BAFF has been shown to induce IgA switching in B cells (66), and elevated IgA levels are detected in the serum of BAFF-Tg mice (11). Mucosal immune responses are normally a major source of IgA production, and both T-dependent and T-independent mechanisms have been shown to trigger IgA production in the gut (67). Because B-1 B cells are unresponsive to NP-Ficoll immunization, elevated numbers of MZ B cells seem to be the likely explanation for the induced response to NP-Ficoll. LPS-activated MZ B cells produce more IgA than mature B cells in the presence of BAFF in vitro (our observation), but whether this is true in vivo and at which site is not clear. In addition, whileNP-specific IgA may be MZ B cell derived following NP-Ficoll immunization of BAFF-Tg mice, we cannot exclude that B-1 B cells may also produce elevated levels of IgA with other specificities in these mice. Abnormal IgA production can be associated with kidney pathologies such as IgA nephropathy (68). IgA deposits have been detected in the kidneys of BAFF-Tg mice (J. Gommerman, personal communication). Therefore, more investigations are needed to determine whether B-1 and/or MZ B cells are involved in T-independent mechanisms leading to potential IgA-dependent renal disease in BAFF-Tg mice.

The MZ B cell compartment is consistently enlarged in both BAFF-Tg and TNF−/− BAFF-Tg mice. MZ B cells display some degree of poly- and self-reactivity and are suspected to participate in autoimmune diseases (32, 34, 69). In addition, injection of prolactin or estrogen in mice led to lupus-like autoimmune disorders associated with the expansion of the MZ B cell compartment (36, 37). The size of the MZ B cell compartment, however, is normally variable and strain dependent, and its expansion does not necessarily correlate with disease (32, 70). Interestingly, treatment of BAFF-Tg mice with a lymphotoxin-β-R-Ig fusion protein that neutralized lymphotoxin-α/β specifically eliminated MZ B cells and protected these mice against autoimmune kidney disease (J. Gommerman, personal communication). Moreover, while MZ-like B cells are normally sessile and found solely in the spleen of wt mice, such B cells were detected in LN, blood, and salivary glands of BAFF-Tg mice (17, 23), suggesting a potential role in autoimmune inflammation. This observation is telling, because sequestration of autoreactive B cells in the MZ is essential for maintenance of immune tolerance (35). Therefore, collectively, these observations suggest that location and function, rather than size, of the MZ B cell population may be key factors determining whether these cells contribute to autoimmune mechanisms. The role of MZ B cells in autoimmunity is, however, very difficult to address due to the mixed nature of this population, containing both naive as well as memory B cells (32), and their pathogenic role has not been substantiated. In addition, the Yaa mutation that causes murine lupus is also responsible for the absence of MZ B cells (71). Our work has not established the exact role of MZ B cells in the autoimmune disease developing in BAFF-Tg mice, but it has narrowed the field of investigation to B cell maturation (T2 and MZ B cells) and the CD4 T cell compartment, and has eliminated GC and conventional T-dependent responses as major avenues to autoimmunity in these mice.

Although MZ B cells may participate in tissue damage seen in the salivary gland of BAFF-Tg mice (17) and potentially participate in some of the lymphomas developing in TNF−/− BAFF-Tg similarly to observations made in patients with SS (72), our studies gave us no indication that MZ B cells directly participate in kidney damages in BAFF-Tg mice. In fact, we never detected MZ B cells in inflamed kidneys of BAFF-Tg mice (data not shown), suggesting that pathology in the kidney may arise from an indirect or separate pathological mechanism.

Interestingly, high numbers of activated CD4 T cells were found in TNF−/− BAFF-Tg mice. Although TNF is known to be a potent mitogenic factor for activated T cells (73), other reports have shown that TNF can attenuate TCR signaling, and thereby protect locally against the emergence of autoreactive T cells (74). Increased numbers of CD4 T cells in TNF−/− BAFF-Tg mice may reflect the lack of this regulatory mechanism, which could explain exacerbation of disease in TNF−/− BAFF-Tg mice. In addition, BAFF can stimulate T cell responses and production of inflammatory cytokines (75). However, the exact role of T cells in autoimmune disease of BAFF-Tg mice may be difficult to assess because mice lacking TCR-α/β+ T cells have been shown to also develop lupus-like disorders (76), in keeping with a controlling role of T cells over B cell responses. Further analysis of BAFF receptor expression among various T cell subsets should help clarify this issue.

One of the most striking observations in TNF−/− BAFF-Tg mice was the emergence of conspicuous lymphoid tumors in >35% of 1-year-old animals. These tumors were located mostly in the cervical LN near the salivary glands, but also in the gut, similar to extranodal MZ cell lymphomas or mucosal-associated lymphoid tissue-type lymphomas. Such lymphomas are a well-recognized complication in SS (77), a disease associated with elevated BAFF levels (17, 18). We have previously described occasional instances of BAFF-Tg mice developing lymphoid masses and B cell hyperplasia (17), but the actual incidence was generally quite low. However, this complication is dramatically exacerbated in TNF−/− BAFF-Tg mice, indicating that TNF plays a critical role in protecting BAFF-Tg mice against B cell lymphoma. Interestingly, BAFF promotes the survival of B-CLL and NHL (20, 21), and patients with NHL have elevated levels of serum BAFF (20). B-CLL cells produce BAFF, which protects them against apoptosis and may function in an autocrine manner to promote survival and expansion of B-CLL tumors (21). In contrast, TNF sensitizes B cell lymphomas to T cell-mediated cytotoxicity (40, 74), so that lack of TNF possibly contributes to the increased risk of developing lymphomas. TACI−/− mice develop lymphomas, suggesting a regulatory role for TACI in B cells (58). TACI is present on the surface of naive TNF−/− BAFF-Tg B cells, but its expression failed to augment upon activation in vitro. Whether this anomaly is associated with increased tumor development in TNF−/− BAFF-Tg mice in addition to lack of TNF-mediated tumor surveillance is not clear. The observation of elevated tumor incidence in TNF−/− BAFF-Tg mice is relevant to clinical observations in humans, because, recently, a risk for development of lymphomas from anti-TNF therapies has emerged, with the identification of 26 cases of lymphomas in a cohort of patients treated with anti-TNF reagents (78). The BAFF-Tg mice used in this study generally express much higher levels of BAFF in the blood than most autoimmune patients, although certain individuals with SS, SLE, or RA patients do express very high BAFF levels in serum, or within inflammatory lesions (16, 17). A logical extension of this study will be to correlate BAFF serum levels in anti-TNF-treated patients with incidence of lymphoma.

We thank Jeffrey L. Browning, Jennifer Gommerman, Susan Kalled, Ken Ho, William Sewell, Robert Brink, and Stuart Tangye for their critical comments on the manuscript. We thank Eric Schmied and the staff at the Biological Testing Facility (Garvan Institute, Sydney, Australia) for assistance with animal care.

1

This work was supported by a Wellcome Trust senior fellowship, the Cooperative Research Center for Asthma, and a program grant from the National Health and Medical Research Council.

3

Abbreviations used in this paper: BAFF, B cell-activating factor belonging to the TNF family; B-CLL, B cell chronic lymphocytic leukemia; FDC, follicular dendritic cell; GC, germinal center; gld, generalized lymphoproliferative disorder; int, intermediate; LN, lymph node; MLN, mesenteric LN; MZ, marginal zone; NHL, non-Hodgkin’s lymphoma; NP, nitrophenyl; RA, rheumatoid arthritis; RF, rheumatoid factor; SLE, systemic lupus erythematosus; SS, Sjögren’s syndrome; T1/2, transitional type 1/2; Tg, transgenic; TI-2, T-independent type 2; TNP, trinitrophenyl; wt, wild type; TACI, transmembrane activator and calcium modulator and cyclophilin ligand interactor.

1
Locksley, R. M., N. Killeen, M. J. Lenardo.
2001
. The TNF and TNF receptor superfamilies: integrating mammalian biology.
Cell
104
:
487
.
2
Flavell, R. A..
2002
. The relationship of inflammation and initiation of autoimmune disease: role of TNF super family members.
Curr. Top. Microbiol. Immunol.
266
:
1
.
3
Mackay, F., S. L. Kalled.
2002
. TNF ligands and receptors in autoimmunity: an update.
Curr. Opin. Immunol.
14
:
783
.
4
Mackay, F., P. Schneider, P. Rennert, J. L. Browning.
2003
. BAFF and APRIL: a tutorial on B cell survival.
Annu. Rev. Immunol.
21
:
231
.
5
Do, R. K. G., S. Chen-Kiang.
2002
. Mechanism of BLyS action in B cell immunity.
Cytokine Growth Factor Rev.
13
:
19
.
6
Rolink, A., J. Tschopp, P. Schneider, F. Melchers.
2002
. BAFF is a survival and maturation factor for mouse B cells.
Eur. J. Immunol.
32
:
2004
.
7
Kalled, S. L..
2002
. BAFF: a novel therapeutic target for autoimmunity.
Curr. Opin. Investig. Drugs
3
:
1005
.
8
Mackay, F., C. Ambrose.
2003
. The TNF family members BAFF and APRIL: the growing complexity.
Cytokine Growth Factor Rev.
14
:
311
.
9
Mackay, F., J. L. Browning.
2002
. BAFF: a fundamental survival factor for B cells.
Nat. Rev. Immunol.
2
:
465
.
10
MacLennan, I. C. M., C. G. Vinuesa.
2002
. Dendritic cells, BAFF, and APRIL: innate players in adaptive antibody responses.
Immunity
17
:
235
.
11
Khare, S. D., I. Sarosi, X.-Z. Xia, S. McCabe, K. Miner, I. Solovyev, N. Hawkins, M. Kelley, D. Chang, D. Chang, et al
2000
. Severe B cell hyperplasia and autoimmune disease in TALL-1 transgenic mice.
Proc. Natl. Acad. Sci. USA
97
:
3370
.
12
Mackay, F., S. A. Woodcock, P. Lawton, C. Ambrose, M. Baetscher, P. Schneider, J. Tschopp, J. L. Browning.
1999
. Mice transgenic for BAFF develop lymphocytic disorders along with autoimmune manifestations.
J. Exp. Med.
190
:
1697
.
13
Gross, J. A., J. Johnston, S. Mudri, R. Enselman, S. R. Dillon, K. Madden, W. Xu, J. Parrish-Novak, D. Forster, C. Lofton-Day, et al
2000
. TACI and BCMA are receptors for a TNF homologue implicated in B-cell autoimmune disease.
Nature
404
:
995
.
14
Rolink, A. G., F. Melchers.
2002
. BAFFled B cells survive and thrive: roles of BAFF in B-cell development.
Curr. Opin. Immunol.
14
:
266
.
15
Mackay, F., C. R. Mackay.
2002
. The role of BAFF in B-cell maturation, T-cell activation and autoimmunity.
Trends Immunol.
23
:
113
.
16
Cheema, G. S., V. Roschke, D. M. Hilbert, W. Stohl.
2001
. Elevated serum B lymphocyte stimulator levels in patients with systemic immune-based rheumatic diseases.
Arthritis Rheum.
44
:
1313
.
17
Groom, J., S. L. Kalled, A. H. Cutler, C. Olson, S. A. Woodcock, P. Schneider, J. Tschopp, T. G. Cachero, M. Batten, J. Wheway, et al
2002
. Association of BAFF/BLyS overexpression and altered B cell differentiation with Sjögren’s syndrome.
J. Clin. Invest.
109
:
59
.
18
Mariette, X., S. Roux, J. Zhang, D. Bengoufa, F. Lavie, T. Zhou, R. Kimberly.
2003
. The level of BLyS (BAFF) correlates with the titre of autoantibodies in human Sjögren’s syndrome.
Ann. Rheum. Dis.
62
:
168
.
19
Zhang, J., V. Roschke, K. P. Baker, Z. Wang, G. S. Alarcon, B. J. Fessler, H. Bastian, R. P. Kimberly, T. Zhou.
2001
. A role for B lymphocyte stimulator in systemic lupus erythematosus.
J. Immunol.
166
:
6
.
20
Briones, J., J. M. Timmerman, D. M. Hilbert, R. Levy.
2002
. BLyS and BLyS receptor expression in non-Hodgkin’s lymphoma.
Exp. Hematol.
30
:
135
.
21
Novak, A. J., R. J. Bram, N. E. Kay, D. F. Jelinek.
2002
. Aberrant expression of B-lymphocyte stimulator by chronic lymphocytic leukemia cells: a mechanism for survival.
Blood
100
:
2973
.
22
Loder, F., B. Mutschler, R. J. Ray, C. J. Paige, P. Sideras, R. Torres, M. C. Lamers, R. Carsetti.
1999
. B cell development in the spleen takes place in discrete steps and is determined by the quality of B cell receptor-derived signals.
J. Exp. Med.
190
:
75
.
23
Batten, M., J. Groom, T. G. Cachero, F. Qian, P. Schneider, J. Tschopp, J. L. Browning, F. Mackay.
2000
. BAFF mediates survival of peripheral immature B lymphocytes.
J. Exp. Med.
192
:
1453
.
24
Do, R. K., E. Hatada, H. Lee, M. R. Tourigny, D. Hilbert, S. Chen-Kiang.
2000
. Attenuation of apoptosis underlies B lymphocyte stimulator enhancement of humoral immune response.
J. Exp. Med.
192
:
953
.
25
Thompson, J. S., P. Schneider, S. L. Kalled, L. Wang, E. A. Lefevre, T. G. Cachero, F. Mackay, S. A. Bixler, M. Zafari, Z.-Y. Liu, et al
2000
. BAFF binds to the TNF receptor-like molecule BCMA and is important for maintaining the peripheral B cell population.
J. Exp. Med.
192
:
129
.
26
Schiemann, B., J. L. Gommerman, K. Vora, T. G. Cachero, S. Shulga-Morskaya, M. Dobles, E. Frew, M. L. Scott.
2001
. An essential role for BAFF in the normal development of B cells through a BCMA-independent pathway.
Science
293
:
2111
.
27
Gross, J. A., S. R. Dillon, S. Mudri, J. Johnston, A. Littau, R. Roque, M. Rixon, O. Schou, K. P. Foley, H. Haugen, et al
2001
. TACI-Ig neutralizes molecules critical for B cell development and autoimmune disease: impaired B cell maturation in mice lacking BLyS.
Immunity
15
:
289
.
28
Hardy, R. R., K. Hayakawa.
2001
. B cell development pathways.
Annu. Rev. Immunol.
19
:
595
.
29
Phan, T. G., M. Amesbury, S. Gardam, J. Crosbie, J. Hasbold, P. D. Hodgkin, A. Basten, R. Brink.
2003
. B cell receptor-independent stimuli trigger immunoglobulin switch recombination and production of IgG autoantibodies by anergic self-reactive B cells.
J. Exp. Med.
197
:
845
.
30
Balazs, M., F. Martin, T. Zhou, J. F. Kearney.
2002
. Blood dendritic cells interact with splenic marginal zone B cells to initiate T-independent immune responses.
Immunity
17
:
341
.
31
Yan, M., S. A. Marsters, I. S. Grewal, H. Wang, A. Ashkenazi, V. M. Dixit.
2000
. Identification of a receptor for BLyS demonstrates a crucial role in humoral immunity.
Nat. Immun.
1
:
37
.
32
Martin, F., J. F. Kearney.
2002
. Marginal-zone B cells.
Nat. Rev. Immunol.
2
:
323
.
33
Martin, F., J. F. Kearney.
2000
. Positive selection from newly formed to marginal zone B cells depends on the rate of clonal production, CD19 and btk.
Immunity
12
:
39
.
34
Li, Y., H. Li, M. Weigert.
2002
. Autoreactive B cells in the marginal zone that express dual receptors.
J. Exp. Med.
195
:
181
.
35
Li, Y., H. Li, D. Ni, M. Weigert.
2002
. Anti-DNA B cells in MRL/lpr mice show altered differentiation and editing pattern.
J. Exp. Med.
196
:
1543
.
36
Grimaldi, C. M., D. J. Michael, B. Diamond.
2001
. Expansion and activation of a population of autoreactive marginal zone B cells in a model of estrogen-induced lupus.
J. Immunol.
167
:
1886
.
37
Peeva, E., D. Michael, J. Cleary, J. Rice, X. Chen, B. Diamond.
2003
. Prolactin modulates the naive B cell repertoire.
J. Clin. Invest.
111
:
275
.
38
Segundo, C., C. Rodriguez, A. Garcia-Poley, M. Aguilar, I. Gavilan, C. Bellas, J. A. Brieva.
2001
. Thyroid-infiltrating B lymphocytes in Graves’ disease are related to marginal zone and memory B cell compartments.
Thyroid
11
:
525
.
39
Balkwill, F..
2002
. Tumor necrosis factor or tumor promoting factor?.
Cytokine Growth Factor Rev.
13
:
135
.
40
Wallgren, A., R. Festin, C. Gidlof, M. Dohlsen, T. Kalland, T. H. Totterman.
1993
. Efficient killing of chronic B-lymphocytic leukemia cells by superantigen-directed T cells.
Blood
82
:
1230
.
41
Lens, S. M., K. Tesselar, B. F. den Drijver, M. H. van Oers, R. A. van Lier.
1996
. A dual role for both CD40-ligand and TNF-α in controlling human B cell death.
J. Immunol.
156
:
507
.
42
Matsumoto, M., Y.-X. Fu, H. Molina, D. D. Chaplin.
1997
. Lymphotoxin-α-deficient and TNF receptor-I-deficient mice define developmental and functional characteristics of germinal centers.
Immunol. Rev.
156
:
137
.
43
Kosco-Vilbois, M. H., J.-Y. Bonnefoy, Y. Chvatchko.
1997
. The physiology of murine germinal center reactions.
Immunol. Rev.
156
:
127
.
44
Korner, H., E. Cretney, P. Wilhelm, J. M. Kelly, M. Roellinghoff, J. D. Sedgwick, M. J. Smyth.
2000
. Tumor necrosis factor sustains the generalized lymphoproliferative disorder (gld) phenotype.
J. Exp. Med.
191
:
89
.
45
Feldmann, M., R. N. Maini.
2001
. Anti-TNFα therapy of rheumatoid arthritis: what have we learned?.
Annu. Rev. Immunol.
19
:
163
.
46
Lee, D. M., M. E. Weinblatt.
2001
. Rheumatoid arthritis.
Lancet
358
:
903
.
47
Weyand, C. M., P. J. Kurtin, J. J. Goronzy.
2001
. Ectopic lymphoid organogenesis.
Am. J. Pathol.
159
:
787
.
48
Korner, H., M. Cook, D. S. Riminton, F. A. Lemckert, R. M. Hoek, B. Ledermann, F. Kontgen, B. Fazekas de St Groth, J. D. Sedgwick.
1997
. Distinct roles for lymphotoxin-α and tumor necrosis factor in organogenesis and spatial organization of lymphoid tissue.
Eur. J. Immunol.
27
:
2600
.
49
Schneider, P., F. Mackay, V. Steiner, K. Hofmann, J. L. Bodmer, N. Holler, C. Ambrose, P. Lawton, S. Bixler, H. Acha-Orbea, et al
1999
. BAFF, a novel ligand of the tumor necrosis factor (TNF) family, stimulates B-cell growth.
J. Exp. Med.
189
:
1747
.
50
Mackay, F., G. R. Majeau, P. Lawton, P. S. Hochman, J. L. Browning.
1997
. Lymphotoxin but not tumor necrosis factor functions to maintain splenic architecture and humoral responsiveness in adult mice.
Eur. J. Immunol.
27
:
2033
.
51
Matsumoto, M., S. Mariathasan, M. H. Nahm, F. Baranyay, J. J. Peschon, D. D. Chaplin.
1996
. Role of lymphotoxin and the type I TNF receptor in the formation of germinal centers.
Science
271
:
1289
.
52
Pasparakis, M., L. Alexopoulou, V. Episkopou, G. Kolias.
1996
. Immune and inflammatory responses in TNFα-deficient mice: a critical requirement for TNFα in the formation of primary B cell follicles, follicular dendritic cell networks and germinal centers, and in the maturation of the humoral immune response.
J. Exp. Med.
184
:
1397
.
53
Maes, B., C. De Wolf-Peeters.
2002
. Marginal zone cell lymphoma: an update on recent advances.
Histopathology
40
:
117
.
54
Franco, V., A. M. Florena, E. Iannitto.
2003
. Splenic marginal zone lymphoma.
Blood
101
:
2464
.
55
Calvert, R. J., P. A. Evans, J. A. Randerson, A. S. Jack, G. J. Morgan, M. F. Dixon.
1996
. The significance of B-cell clonality in gastric lymphoid infiltrates.
J. Pathol.
180
:
26
.
56
Von Bulow, G.-U., J. M. van Deursen, R. J. Bram.
2001
. Regulation of the T-independent humoral response by TACI.
Immunity
14
:
573
.
57
Yan, M., H. Wang, B. Chan, M. Roose-Girma, S. Erickson, T. Baker, D. Tumas, I. S. Grewal, V. M. Dixit.
2001
. Activation and accumulation of B cells in TACI-deficient mice.
Nat. Immun.
2
:
638
.
58
Seshasayee, D., P. Valdez, M. Yan, V. M. Dixit, D. Tumas, I. S. Grewal.
2003
. Loss of TACI causes fatal lymphoproliferation and autoimmunity, establishing TACI as an inhibitory BLyS receptor.
Immunity
18
:
279
.
59
D’Haens, G..
2003
. Anti-TNF therapy for Crohn’s disease.
Curr. Pharm. Des.
9
:
289
.
60
Kalled, S. L., A. H. Cutler, S. K. Datta, D. W. Thomas.
1998
. Anti-CD40 ligand antibody treatment of SNF1 mice with established nephritis: preservation of kidney function.
J. Immunol.
160
:
2158
.
61
MacLennan, I. C. M..
1994
. Germinal centers.
Annu. Rev. Immunol.
12
:
117
.
62
Van Eijk, M., T. Defrance, A. Hennino, C. de Groot.
2001
. Death-receptor contribution to the germinal-center reaction.
Trends Immunol.
22
:
677
.
63
Siegel, R. M., F. K. Chan, H. J. Chun, M. J. Lenardo.
2000
. The multifaceted role of Fas signaling in immune cell homeostasis and autoimmunity.
Nat. Immun.
1
:
469
.
64
Martin, F., J. F. Kearney.
2000
. B-cell subsets and the mature preimmune repertoire: marginal zone and B1 B cells as part of a “natural immune memory.”.
Immunol. Rev.
175
:
70
.
65
Berland, R., H. H. Wortis.
2002
. Origins and functions of B-1 cells with notes on the role of CD5.
Annu. Rev. Immunol.
20
:
253
.
66
Litinskiy, M., B. Nardelli, B. M. Hilbert, H. Bing, A. Schaffer, P. Casali, A. Cerutti.
2002
. DCs induce CD40-independent immunoglobulin class switching through BLyS and APRIL.
Nat. Immun.
3
:
822
.
67
Fargarasan, S., T. Honjo.
2003
. Intestinal IgA synthesis: regulation of front-line body defenses.
Nat. Rev. Immunol.
3
:
63
.
68
Donadio, J. V., J. P. Grande.
2002
. IgA nephropathy.
N. Engl. J. Med.
347
:
738
.
69
Zeng, D., M.-K. Lee, J. Tung, A. Brendolan, S. Strober.
2000
. A role for CD1 in the pathogenesis of lupus in NZB/NZW mice.
J. Immunol.
164
:
5000
.
70
Schuster, H., T. Martin, L. Marcellin, J. C. Garaud, J. L. Pasquali, A. S. Korganov.
2002
. Expansion of the marginal zone B cells is not sufficient for the development of renal disease in NZB × NZW F1 mice.
Lupus
11
:
277
.
71
Amano, H., E. Amano, T. Moll, D. Marinkovic, N. Ibnou-Zekri, E. Martinez-Soria, I. Semac, T. Wirth, L. Nitschke, S. Izui.
2003
. The Yaa mutation promoting murine lupus causes defective development of marginal zone B cells.
J. Immunol.
170
:
2293
.
72
Morse, H. C., III, J. F. Kearney, P. G. Isaacson, M. Carroll, T. N. Fredrickson, E. S. Jaffe.
2001
. Cells of the marginal zone: origins, function and neoplasia.
Leuk. Res.
25
:
169
.
73
Tartaglia, L. A., D. V. Goeddel, C. Reynolds, I. S. Figari, R. F. Weber, B. M. Fendly, M. A. Palladino, Jr.
1993
. Stimulation of human T-cell proliferation by specific activation of the 75-kDa tumor necrosis factor receptor.
J. Immunol.
151
:
4637
.
74
Kollias, G., D. Kontoyiannis.
2002
. Role of TNF/TNFR in autoimmunity: specific TNF receptor blockade may be advantageous to anti-TNF treatments.
Cytokine Growth Factor Rev.
13
:
315
.
75
Huard, B., P. Schneider, D. Mauri, J. Tschopp, L. E. French.
2001
. T cell costimulation by the TNF ligand BAFF.
J. Immunol.
167
:
6225
.
76
Wen, L., S. J. Roberts, J. L. Viney, F. S. Wong, C. Mallick, R. C. Findly, Q. Peng, J. E. Craft, M. J. Owen, A. C. Hayday.
1994
. Immunoglobulin synthesis and generalized autoimmunity in mice congenitally deficient in αβ+ T cells.
Nature
369
:
654
.
77
Abbondanzo, S. L..
2001
. Extranodal marginal-zone B-cell lymphoma of the salivary gland.
Ann. Diagn. Pathol.
5
:
246
.
78
Brown, S. L., M. H. Greene, S. K. Gershon, E. T. Edwards, M. M. Braun.
2002
. Tumor necrosis factor antagonist therapy and lymphoma development.
Arthritis Rheum.
46
:
3151
.