Systemic lupus erythematosus is characterized by loss of tolerance to DNA and other nuclear Ags. To understand the role of T cells in the breaking of tolerance, an anti-DNA site-specific transgenic model of spontaneous lupus, B6.56R, was studied. T cells were eliminated by crossing B6.56R with CD4−/− or TCRβ−/−δ−/− mice, and the effects on anti-dsDNA serum levels, numbers of anti-dsDNA Ab-secreting cells, and isotypes of anti-dsDNA were analyzed. In addition, the development and activation of B cells in these mice were examined. Surprisingly, the presence of T cells made little difference in the development and character of the serum anti-dsDNA Ab in B6.56R mice. At 1 mo of age, anti-dsDNA Abs were somewhat lower in mice deficient in αβ and γδ T cells. Levels of Abs later were not affected by T cells, nor was autoantibody class switching. B cell activation was somewhat diminished in T cell-deficient mice. Thus, in the B6 background, the presence of an anti-dsDNA transgene led the production of autoantibodies with a specificity and isotype characteristic of murine systemic lupus erythematosus with little influence from T cells. TLR9 also did not appear to play a role. Although we do not yet understand the mechanism of this failure of immunoregulation, these results suggest that similar processes may influence autoimmunity associated with clinical disease.
Systemic lupus erythematosus (SLE)3 is a multigene, multiorgan autoimmune disease. The clinical manifestations are accompanied by production of autoantibodies targeting self-proteins and nucleic acids (1, 2). The mechanisms of the breakdown of B cell tolerance are central to the pathogenesis of SLE.
The maintenance of tolerance to dsDNA occurs at several stages or checkpoints, both centrally and in the periphery. Investigation of the mechanism of each stage will give us a better understanding of the development of SLE (3, 4, 5). Transgenic mice with rearranged specific autoantibody genes have provided important insights into the mechanisms for breaking of tolerance. One of these models, 3H9/56R (56R) is a site-directed H chain transgene that is derived from the 3H9 H chain with a replacement of aspartate with arginine at position 56 in CDR2. Because of the arginine replacement, 56R Ab has a higher affinity to dsDNA than the 3H9 Ab. The 3H9 IgH variable region was originally cloned from an anti-dsDNA hybridoma derived from an MRL/lpr mouse. With the 56R mutation, this H chain creates a DNA binding Ab with nearly every L chain in the mouse repertoire, with the exception of a handful of “editors,” such as Vκ38. The knock-in (KI) gene 3H9/56R was generated by replacing the JH locus with a VDJ segment from 3H9/56R. Because the KI model permits the transgene locus to undergo editing, isotype switching, and somatic mutation, it enables us to study a specific subset of the anti-dsDNA B cell population in its transition from tolerance to autoimmunity (6, 7, 8, 9).
Autoreactive B cells exist in the circulating repertoire of normal animals, but they remain quiescent because of a lack of T cell help or because of regulatory mechanisms. A large number of studies have focused on the pathogenic role of T cells in autoimmune disease (10, 11, 12). The experimentally induced chronic graft-versus-host (cGVH) model studied in our laboratory is clearly induced by alloreactive CD4 T cells. Early studies showed neonatal thymectomy abrogated lymphoproliferation and prevented autoimmune disease in MRL/lpr mice (13). In several experimental models of SLE, mAb treatment to eliminate T cells reduced autoimmunity (14, 15, 16). CD4-deficient MRL/lpr mice produced few anti-dsDNA Abs, despite enhanced lymphoproliferation (17). MHC class II (MHC II)-deficient MRL/lpr mice manifested a reduction of autoantibodies (18). In addition, TCR-deficient mouse studies showed that αβ and γδ T cells played different roles in murine autoimmune disease. Although αβ T cells helped the production of autoantibodies and immunopathogenesis, γδ T cells mediated both propagation and regulation of autoantibody production (19, 20). MRL/lpr mice deficient in both T cell lineages had severely attenuated disease. Thus, T cells appear to play an important role at one or more tolerance checkpoints for anti-dsDNA B cells.
When the 56R transgene is bred to the C57BL/6 (B6) background (B6.56R), as opposed to a BALB/c background, the mice produce anti-dsDNA spontaneously. Several genetic and experimental manipulations can enhance the levels of anti-dsDNA found in B6.56R mice, including cGVH (21, 22). In the present work, we have focused on the mechanisms involved in the spontaneous loss of tolerance to DNA in B6.56R. We found that the development of anti-DNA autoantibody in 56R mice is due to an intrinsic deregulation of B cells. These B cells are fully functional without the help of T cells: they differentiate, class switch, and give rise to cells producing IgG anti-dsDNA.
Materials and Methods
Site-directed H chain-transgenic 3H9H/56R BALB/c mice were provided by M. Weigert (University of Chicago, Chicago, IL). C57BL/6.56R+/− (B6.56R) mice were generated by backcrossing the 3H9H/56R (56R) locus onto the C57BL/6 (B6) background (13 times) in this laboratory.
C57BL/6-CD4tmlMak (B6.CD4−/−) and B6.129P2-TCRbtm1/MomTCRdtm1Mom (B6.TCRβ−/−δ−/−) mice were purchased from The Jackson Laboratory. These two strains of mice were bred with B6.56R to obtain B6.56R+/−-CD4−/− (B6.56R-CD4−/−) and B6.56R+/−-TCRβ−/−TCRδ−/− (B6.56R-TCRβ−/−δ−/−) offspring.
All mice were bred and maintained in the animal facility of the School of Medicine at the University of Pennsylvania. The animal breeding, maintenance, and experimental procedures were conducted by the protocols approved by the Institutional Animal Care and Use Committee of the University.
Mouse genotypes were determined by PCR amplification of genomic DNA from mouse tails. A standard protocol of 0.25 mM dNTP, 1.5 mM MgCl2, 0.2 μM 5′ primer, 0.2 μM 3′ primer, and 0.2 U of Platinum Taq polymerase (Invitrogen) was used for each PCR.
The conditions of the 56R PCR, resulting in a 330-bp DNA fragment, are one cycle of 5 min at 94°C, 35 cycles of 30 s at 94°C, 30 s at 62°C, and 30 s at 72°C, followed by 7 min at 72°C. The 5′ primer sequence is 5′-CTG TCA GGA ACT GCA GGT AAG G and the 3′ primer sequence is 5′-CAT AAC ATA GGA ATA TTT ACT CCT CGC.
The wild-type and disrupted CD4 genes were detected, following protocols (Cd4tm1, tm1Knw, tm1Mak) from The Jackson Laboratory.
The disrupted TCRβ gene was identified by TCRβ-neo PCR, resulting in an ∼1.5-kb DNA fragment. The 5′ primer sequence is 5′-GTC TTC TTT GGT AAA GGA AC and the 3′ sequence is 5′-CTT GGG TGG AGA GGC TAT TC. The wild-type TCRβ gene was identified by TCRβ PCR, resulting in an ∼450-bp DNA fragment. The 5′ primer sequence is 5′-GTC TTC TTT GGT AAA GGA AC and the 3′ sequence is 5′-TTC CAG AAC ACA GCC TCC. Both PCRs were performed under the same conditions: one cycle of 5 min at 94°C, 30 cycles of 30 s at 94°C, 30 s at 55°C, and 90 s at 72°C, followed by 3 min at 72°C.
Disrupted TCRδ genes were identified by TCRδ-neo PCR, resulting in an ∼2-kb DNA fragment. The 5′ primer sequence is 5′-ACC TTG AGG TGC TGA GAA G and the 3′ primer sequence is 5′-GTC GAG TGC ACA GTT TCA C. The wild-type TCRδ gene was identified by TCRδ PCR, resulting in an ∼800-bp DNA fragment. The 5′ primer sequence is 5′-ACC TTG AGG TGC TGA GAA G and the 3′ sequence is 5′-AAT AGC CTC TCC ACC CAA G. Both PCRs were performed under the same condition: one cycle of 5 min at 94°C, 30 cycles of 30 s at 94°C, 30 s at 58°C, and 2 min at 72°C followed by 3 min at 72°C.
Serum samples were collected by monthly tail bleeding and stored at −20°C with 0.1% sodium azide.
Spleen cells were isolated by disrupting the splenic capsule between frosted ends of glass slides in HBSS (Mediatech). Erythrocytes were lysed by 150 mM ammonium chloride solution. Single-cell suspensions were filtered through 70-μm cell strainers (BD Falcon). Viable cell counts were determined by trypan blue exclusion.
Anti-dsDNA Ab detection
Serum levels of anti-dsDNA Abs were detected by solid-phase ELISAs. Calf thymus DNA (Sigma-Aldrich) was treated with S1 nuclease at 37°C for 45 min to remove the single-stranded segments of DNA and then extracted by phenol-chloroform. Polyvinylchloride flat-bottom microtiter plates (Thermo Electron) were treated with 0.01% poly-l-lysine (Sigma-Aldrich) for 1 h at room temperature. After washing plates with double-distilled H2O and letting them air dry, the plates were coated with 100 μl of 2.5 μg/ml calf thymus dsDNA at 4°C overnight. The dsDNA was diluted in 02. M borate-buffered solution (pH 8.2). Coated plates were blocked with 0.5% BSA and 0.4% Tween 80 in borate-buffered solution (BBT) at room temperature for 2 h. Serum samples and controls were diluted in BBT, added to plates, and incubated at 4°C overnight. Alkaline phosphatase-labeled detecting Abs were diluted in BBT to 50 ng/ml, added to plates, and incubated at room temperature for 1 h before the addition of 1 mg/ml p-nitrophenylphosphate (Sigma-Aldrich) in 10 mM diethanolamine.
Anti-dsDNA total IgG was detected with alkaline phosphatase-labeled goat anti-mouse IgG F(ab′)2, γ-specific (Jackson ImmunoResearch Laboratories). Allotype anti-dsDNA Abs were detected with 1 ng/ml preadsorbed rabbit anti-mouse IgG2aa or IgG2ab (also known as IgG2c; Accurate Chemical & Scientific) and then alkaline phosphatase-labeled donkey anti-rabbit IgG F(ab′)2 (Jackson ImmunoResearch Laboratories). The isotypes of anti-dsDNA Abs were detected with 100 ng/ml biotinylated goat anti-mouse IgM, IgG, IgG1, IgG2a, IgG2b, or IgG3 (Southern Biotechnology Associates), followed by 10 μg/ml avidin-alkaline phosphatase (Sigma-Aldrich).
A laboratory pooled serum standard from 5- to 9-mo-old MRL/lpr mice with a high anti-dsDNA Ab titer was used as experimental reference and positive control. One-month-old B6 mouse serum without detectable anti-dsDNA was used as a negative control. The standard serum was subjected to 16 two-fold dilutions, starting from 1/500 for IgG and IgG2aa detection, 1/250 for IgG2ab detection. The serum samples and controls were diluted 1/1000 for IgG and IgG2aa detection and 1/500 for IgG2ab detection.
OD at 405 nm of each well was determined using a microplate reader (E-Max; Molecular Devices) and analyzed with Soft-Max version 4.5.
The levels of serum autoantibody IgG2aa and IgG2ab were reported as equivalent dilution factor (EDF) to MRL/lpr serum standards. The levels of serum autoantibody isotypes were reported as OD.
Ab-forming cell detection
IgG and anti-dsDNA Ab-forming cells were detected by enzyme-linked immunosorbent spot-forming cell assays (ELISPOT). Ninety-six well MultiScreen-IP sterile white plates (Millipore) were coated with 50 μl of 1 μg/ml rat anti-mouse IgG (Jackson ImmunoResearch Laboratories) or 50 μg/ml calf thymus dsDNA in PBS (10 mM PBS, pH 7.0) at room temperature overnight. Coated plates were blocked with freshly prepared 5% nonfat milk in PBS. For IgG or anti-dsDNA-producing cell detection, 4,000 or 10,000 cells were added to each well, respectively. The plates were incubated at 37°C for 2 h and the cells were washed away with PBS. Alkaline phosphatase-labeled rat anti-mouse κ (clone 187.1) (100 ng/ml in 1% BSA/PBS) was added to the wells and incubated at 4°C overnight. Specific Ab-binding spots were visualized with 5-bromo-4-chloro-3-indolyl phosphate/NBT (Sigma-Aldrich) and scanned by an ImmunoSpot scanner (Cellular Technology).
T cell and B cell phenotypic analysis
Four-color flow cytometric analysis (FACS) was conducted to determine cell phenotypes. All biotin- or fluorochrome-labeled Abs were purchased from BD Pharmingen/BD Biosciences.
Cells were stained using combinations of the following fluorochrome- or biotin-conjugated anti-mouse Abs: CD3ε-allophycocyanin (clone 145-2C11), CD4-FITC (clone RM4-5), CD8a-PE (clone 53-6.7), TCRβ-CyChrome (clone H57-597), TCRγ/δ-PE (clone GL3), CD19-allophycocyanin (clone 1D3), CD21/CD35-allophycocyanin (clone 7G6), CD22.2-FITC (clone Cy34.1), CD23-PE (clone B3B4), CD24-PE (clone M1/69), CD40-biotin (clone 3/23), CD80-FITC (clone 16-10A1), CD86-PE (clone GL1), IgDa-FITC (clone AMS9.1), IgDb-biotin (clone 217-170), IgMa-FITC (DS-1), IgMb- biotin (clone AF6-78), IgG1-PE (clone R19-15), IgG2a/2b-FITC (clone R2-40), IgG3- biotin (clone R40-82), κ-chain-PE (clone 187.1), λ1, λ2, and λ3-FITC (clone R26-46). Biotin-conjugated Abs were visualized by secondary reagent streptavidin-PerCP.
FACS staining was conducted on ice using HBSS supplemented with 3% BSA (HyClone). One and a half million cells were used for each staining sample. Cells were treated with 10 μl of rat anti-mouse FcγR (clone 2.4G2) tissue culture supernatant for 15 min. Biotinylated Abs were added to the cells and incubated for 20 min. After washing twice (200 μl of HBSS, centrifuged at 200 × g for 3 min at 4°C), fluorochrome- conjugated Abs or streptavidin were added to the cells and incubated for 20 min. After washing twice more, cells were suspended in 200 μl of 1% formaldehyde in PBS and stored in the dark at 4°C.
Flow cytometric data were acquired on a FACSCalibur (BD Bisociences) cytometer and analyzed using FlowJo software version 6.3 (Tree Star).
ELISA, ELISPOT, and cell number data are presented as geometric means from each group of samples, error bars representing SE of each group of data. Statistical significance (p) was determined by Student’s t test to indicate the differences between the experimental group and control group (B6.56R-CD4−/− vs B6.56R or B6.56R-TCRb−/−δ−/− vs B6.56R).
Genotypes of experimental mice
All mice used in this study were on the C57BL/6 (B6) background. The 56R transgene locus was backcrossed onto the B6 background for 13 generations. Both B6.56R+/−-CD4−/− or B6.56R+/−-TCRβ−/−δ−/− mice were from the F2 generation of B6.56R+/− crossed with B6.CD4−/− or B6.TCRβ−/−δ−/−, respectively. Tables I and II depict the genotypes and designated names of all mouse groups in this study.
|Group .||No. of Mice .||56R .||CD4 .|
|Group .||No. of Mice .||56R .||CD4 .|
|Group .||No. of Mice .||56R .||TCRβ .||TCRδ .|
|Group .||No. of Mice .||56R .||TCRβ .||TCRδ .|
Anti-dsDNA Ab production in B6.56R-CD4−/− mice
To determine the role of CD4 T cells in the breaking of B cell tolerance, we measured serum anti-dsDNA Abs in mice lacking CD4 T cells (CD4−/−). Because the 56R transgene encodes a-allotype anti-dsDNA Abs, while endogenous genes of B6 mice encode b-allotype Abs, we used an allotype-specific anti-dsDNA ELISA to distinguish whether the autoantibodies were from the 56R transgene or whether they originated from endogenous genes.
Fig. 1 shows allotype-specific anti-dsDNA Ab levels from B6, B6.56R, B6.CD4−/−, and B6.56R-CD4−/− mice. It was clear that the production of anti-dsDNA Abs by the 56R-transgenic mice did not require CD4 T cells. It was noteworthy, however, that CD4−/− mice produced less anti-dsDNA IgG2aa than the mice with CD4 T cells at 3 and 6 mo, but not at 12 mo. (p = 0.009, 0.008, and 0.245, respectively). Two serum samples from each group were tested for dsDNA using a Crithidia luciliae indirect immunofluorescent assay and the binding of serum to kinetoplast corresponded to the ELISA results. The serum anti-dsDNA levels did not vary between males and females (data not shown).
Isotype switching in B6.56R-CD4−/− mice
Through cognate B cell-T cell interactions and provision of cytokines, CD4 T cells regulate B cell maturation to favor Ig isotype switching (23, 24). Some Ig isotypes are classically T cell-dependent, such as IgE, IgG1, IgG2a, and IgG2b, and some are not, such as IgM, and IgG3 (25, 26, 27). We therefore quantitated the isotypes of anti-dsDNA Abs in B6.56R-CD4KO mice.
Fig. 2 shows the isotypes of anti-dsDNA Abs in 6-mo-old mice. Elimination of CD4 T cells reduced IgM anti-dsDNA levels in mice with or without the 56R transgene. IgM anti-dsDNA levels were lower in B6.56R-CD4−/− mice compared with B6.56R mice and in B6.CD4−/− mice compared with B6 mice. Total anti-dsDNA IgM, IgG, IgG2a, and IgG2b levels were lower in B6.56R-CD4−/− mice compared with B6.56R mice. Very low amounts of anti-dsDNA IgG1 or IgG3 were detected and only in B6.56R sera.
Anti-dsDNA Ab production in B6.56R-TCRβ−/−δ−/− mice
The data regarding anti-dsDNA Ab production in B6.56R-CD4−/− suggested that T cells may accelerate the loss of tolerance to DNA in B6.56R mice, but that they are not absolutely required. When we tested 4-mo-old B6.56R-TCRβ−/− sera for anti-dsDNA IgG, the autoantibody levels of those mice were in fact 2.8 times higher than those of the B6.56R control group (p = 0.017; data not shown). We therefore tested mice that were more fully T cell deficient, i.e., TCRβ−/−TCRδ−/−.
Fig. 3 shows allotype-specific assays of anti-dsDNA of four experimental mouse groups. Serum samples were collected at the age of 1, 2, and 3 mo. The serum anti-dsDNA levels did not vary between males and females (data not shown). The B6.56R- TCRβ−/−δ−/− mice produced less anti-dsDNA IgG2aa than B6.56R mice at 1 mo, (p = 0.001), but anti-dsDNA Ab production in B6.56R-TCRβ−/−δ−/− mice increased from 1 to 3 mo of age, while B6.56R mouse serum anti-dsDNA Abs slightly decreased during this age interval. The dynamic change of anti-dsDNA production in B6.56R-TCRβ−/−δ−/− mice resulted in higher anti-dsDNA levels at the age of 3 mo, compared with B6.56R (p < 0.01). There were no significant amounts of IgG2ab anti-dsDNA Abs in any of the four groups of mice.
Anti-dsDNA autoantibody-producing cells in B6.56R-TCRβ−/−δ−/− mice
To confirm serum anti-dsDNA Ab results, spleen autoantibody-forming cells (AFC) were detected by anti-dsDNA ELISPOT in 2- to 4 mo-old mice. Whether expressed as AFC per 106 spleen cells (Fig. 4,a) or as AFC per spleen (Fig. 4 b), more anti-dsDNA-producing cells were detected in the mice without functional TCRs. This increase was consistent in three experiments.
Isotype switching in B6.56R-TCRβ−/−δ−/− mice
In CD4-deficient B6.56R mice, not only IgM, but also IgG2a and IgG2b anti-dsDNA Abs were detected. To test further for a regulatory role for T cells in isotyping switching in this system, we performed isotype-specific ELISAs on serum samples from 3-mo-old mice that were completely T cell deficient (Fig. 5). Surprisingly, mice lacking TCRβδ cells had increased levels in all isotypes tested, (IgM, IgG, IgG1, IgG2a, IgG2b, and IgG3). As seen before, the IgG2a and IgG2b isotypes were most prominent. IgG1 was still quite low, but the 56R mice without T cells had a striking increase in IgG3 anti-dsDNA Abs.
B6.56R-TCRβ−/−δ−/−mouse B cell numbers
Mice that carried the 56R transgene had smaller numbers of spleen cells than their counterparts, i.e., B6.56R mice had 1.6-fold fewer spleen cells than B6 mice and B6.56R-TCRβ−/−δ−/− mice had 1.9-fold fewer spleen cells than B6.TCRβ−/−δ−/− mice. Since the TCRβ−/−δ−/− mice obviously lack T cells, we determined the absolute number of B cells (CD19+) in each strain. Similar to the number of spleen cells, B6.56R mice had 2.4-fold fewer CD19+ cells than B6 mice, and B6.56R-TCRβ−/−δ−/− mice had 2.9-fold fewer spleen cells than B6.TCRβ−/−δ−/− mice. B6.56R-TCRβ−/−δ−/− mice had 1.2-fold more CD19+ cells than B6.56R mice, but the difference was not significant statistically (data not shown).
The lower numbers of B cells in 56R transgene-positive mice might be due to negative selection by deletion of self-reactive B cells. If so, this process did not appear to be T cell dependent. Substitution of the endogenous H chain allele for the transgenic one probably also represents a tolerance mechanism (receptor editing). We therefore stained splenic B cells for IgM and IgD of a and b allotypes in four independent assays, each assay consisting of one B6, B6.56R, B6.TCRβ−/−δ−/−, and B6.56R-TCRβ−/−δ−/−, and staining control consisted of a mixture with an equal number of B6.C20 (IgMa/IgDa) and B6 (IgMb/IgDb) spleen cells. IgMa and IgMb double-staining cells were seen in both B6.56R and B6.56R-TCRβ−/−δ−/− mice; the staining was not due to flow cytometry compensation error because no such effect was shown in control cells. These cells most likely express the IgMa with a κ L chain editor that is cross-reactive with AF6-78 anti-IgMb (28). Analysis of 56R mice showed that ∼50% of CD19+ cells expressed the transgene (IgMa), while 15–20% of CD19+ cells expressed the endogenous μ-chain (IgMb) (Fig. 6,a). There was no significant difference in transgene H chain usage between B6.56R (a:b = 2.55) and B6.56R- TCRβ−/−δ−/− (a:b = 2.66) mice. Therefore, receptor editing also does not seem to be grossly affected by the absence of T cells. Similar results were shown for IgD allotype staining (Fig. 6 b); ∼40–60% of CD19+ cells in 56R mice expressed transgene (IgDa), while 15–17% of CD19+ cells expressed the endogenous d chain (IgDb).
Marginal zone B cells in B6.56R-TCRβ−/−δ−/− mice
Cross-linking of BCRs mediates signal transduction events that direct immature B cells into the mature follicular or marginal zone B cell subsets (26, 27). Although the T-dependent follicular B cells are clearly a source of autoantibodies, T-independent marginal zone B cells may also contribute to autoantibody development. We therefore investigated the population of marginal zone B cells.
FACS analysis showed a marked increase of marginal zone B cells (IgM+CD21highCD23low) in 56R transgene-positive mice, as has been previously reported (9, 29). In addition, there was a slight increase of marginal zone B cells in B6.56R-TCRβ−/−δ−/− mice, compared with that in B6.56R mice, (Fig. 7), but the difference was not statistically significant (data not shown.) The absence of T cells did not appear to have a major effect on the distribution of B cells in the marginal zone vs follicular subsets.
B cell activation in B6.56R-TCRβ−/−δ−/− mice
The expression of B cell activation markers by FACS indicated an active B cell population in 56R transgene-positive mice. Among four experiment groups, there was no difference seen in MHC II, Fas, CD40, B7.1, and B7.2 expression between B6 and B6.TCRβ−/−δ−/− groups (data not shown). Both B6.56R and B6.56R-TCRβ−/−δ−/− mice showed a more activated phenotype. However, the expression of several B cell markers was different between B6.56R and B6.56R-TCRβ−/−δ−/− mice. Although CD40 and B7.1 had similar expression levels in B6.56R and B6.56R-TCRβ−/−δ−/− mice, MHC II, Fas, and B7.2 levels in B6.56R-TCRβ−/−δ−/− mice were not increased, but rather more resembled those in B6 mice (Fig. 8). Thus, the activation of B cells that accompanies the loss of tolerance in B6.56R mice appears to affect the expression of only certain markers.
Systemic autoimmune disease is the result of a series of interactions within the immune system that ultimately leads to the loss of self-tolerance to nuclear Ags. Normally, autoreactive cells, both B and T, are controlled at various checkpoints that contribute to the maintenance of self-tolerance. The mechanisms of tolerance include deletion, receptor editing, anergy, suppression (active regulation), and ignorance. It is likely that each mechanism is <100% efficient, but the ultimate effect of the multiple sequential checkpoints is the avoidance of autoimmune disease. If one checkpoint is seriously compromised or if several are modestly ineffective, then at least some self-reactive cells can reach the point of full maturity and produce potentially injurious self-recognition. In this process, the loss of T cell and B cell tolerance may interplay in a self-reinforcing process.
In the present study, we have begun to elucidate the mechanism for the spontaneous loss of tolerance to DNA by B cells in B6 mice that bear the 56R site-directed anti-DNA transgene. Our work mainly focused on the role of T cells in autoantibody production. We chose B6.56R mice because 3H9/56R is a well-defined and well-studied transgene, which can be subjected to somatic genetic alterations, including secondary rearrangement, class switching, and somatic mutation (9, 30, 31, 32, 33). Earlier work in our group has demonstrated spontaneous anti-dsDNA Ab production in B6.56R mice. In addition, T-dependent abnormal B cell help provided by the cGVH reaction produced higher titers of anti-dsDNA Abs by the knock-in transgene, as well as by endogenous Igh genes (21). Furthermore, others (34) have reported the breakdown of tolerance in 56R-transgenic mice under the influence of additional SLE genes. Most previous studies addressing the role of T cells in the breakdown of B cell tolerance have shown that αβ and/or γδ T cells are required for autoimmunity. The αβ T cells play a central role in the pathogenesis of the disease and autoantibody production, whereas the γδ T cells are implicated in both propagation and regulation of the autoimmune response (35, 36, 37, 38, 39). Since these studies have used complex mouse models, such as MRL/lpr, there may be multiple factors that influenced the outcome of the results.
Our present data showed that anti-dsDNA Abs appear in all B6.56R-congenic strains examined. That is, the absence of CD4, of the TCRαβ receptor or of both the TCRαβ and TCRγδ did not affect the development of anti-dsDNA autoantibodies in a major way. Since the double-TCR knockout mice have no T cells, these results indicate that the loss of tolerance to DNA can occur in the complete absence of T cell help. Nevertheless, certain modest differences were found which suggest that T cell help may accelerate the loss of tolerance in this model. Specifically, the serum level of anti-dsDNA Abs was lower in 3- and 6-mo-old B6.56R-CD4−/− mice compared with the B6.56R mice in the same age group. In addition, anti-dsDNA Ab titers were lower in B6.56R-TCRβ−/−δ−/− mice than B6.56R mice at the age of 1 mo, although by the age of 3 mo, these autoantibody levels were actually higher in the mice without T cells. The levels seen in this spontaneous model were frequently >50% of that detected in aged MRL/lpr serum, chosen as a positive control. Furthermore, to distinguish whether the increase of the anti-dsDNA level was caused by more Ab-producing cells or by the cells that produced more Abs, anti-dsDNA Ab-forming cells were detected by ELISPOT. The results showed more anti-dsDNA Ab-producing cells in B6.56R-TCRβ−/−δ−/− mice, even though the number of CD19+ cells in B6.56R- TCRβ−/−δ−/− mice was almost the same as that in B6.56R mice. Finally, a spot check of sera that were positive by ELISA for anti-dsDNA showed that they were also positive by indirect immunofluorescence with C. luciliae. Thus, the T-independent loss of tolerance in these mice truly recognizes the SLE-specific Ag-native DNA. Nevertheless, we cannot determine at this point whether the autoreactive B cells that are activated in the absence of T cells are the same as the ones that respond in the presence of autologous T cells or when stimulated by allogeneic help in the cGVH.
Given the basic T independence of the spontaneous anti-dsDNA response in B6.56R mice, it was quite surprising that the major isotypes of this autoantibody detected were generally IgG2a and IgG2b, in addition to IgM. Although IgM and IgG2b can be produced in some T-independent responses, IgG2a is generally T dependent, particularly through T1 cytokines such as IFN-γ or IL-12 (39, 40, 41, 42, 43, 44). In addition, it is the major isotype of anti-dsDNA autoantibodies in several models of spontaneous or experimentally induced SLE in mice (45, 46, 47). The isotype distribution did not change markedly in the various 56R transgene-positive, knockout strains examined, except the B6.56R-TCRβ−/−δ−/− mice uniquely had greatly increased levels of IgG3 anti-dsDNA. Little IgG1 was seen in any of the mice. This latter finding parallels the frequent paucity of this isotype in the autoantibodies of many (but not all) SLE models (48). Normally, IgG1 is the most prominent T-dependent IgG subclass against soluble protein Ags, although viruses and bacteria can provoke T-dependent, IgG2a-skewed responses (40, 41). A similar outcome was observed in BAFF-transgenic mice that were completely deficient in T cells. In these mice, elevated levels of IgG2c and IgG2b in response to T-dependent Ag were detected (49).
The expression of B cell developmental and activation markers showed no differences between B6 and B6.TCRβ−/−δ−/− mice. In both of these strains with the anti-dsDNA KI transgene, B cells showed evidence of increased activation by up-regulation of CD40 and B7.1 expression. In addition, the degree of receptor editing indicated by the expression of the endogenous Igh allele was not altered by the absence of T cells. That is, IgMa and IgMb staining revealed no transgenic and endogenous IgM differences between B6.56R-TCRβ−/−δ−/− and B6.56 mice; IgDa and IgDb staining showed the same results. Finally, the expansion of marginal zone B cells previously noted in 56R mice was also not apparently T cell dependent.
Following the first evidence of the involvement of TLR, especially TLR9, in the activation of autoantibody-producing B cells (50, 51), our previous work showed that the lack of TLR9 down-regulated somewhat the production of anti-dsDNA in B6.56R mice (22). In this study, we have also compared the serum levels of anti-dsDNA Abs in B6.56R, B6.56R-TCRβ−/−δ−/− and B6.56R-TCRβ−/−δ−/−TLR9−/− mice at the age of 1, 2, 3, and 6 mo. ELISA results showed no significant difference of serum anti-dsDNA levels among all three groups of male mice, while the levels of B6.56R-TCRβ−/−δ−/− and B6.56R-TCRβ−/−δ−/−TLR9−/− were increased in the 6-mo-old female mice (data not shown).
Overall then, the loss of tolerance in this model appears to proceed largely by a T cell-independent mechanism. Since the end result is a distribution of autoantibody isotypes similar to that found in more robust models of murine SLE, it is likely that the mechanisms involved in anti-dsDNA production in B6.56R mice reflect some of the events important for systemic autoimmunity in general.
We thank Drs. Eline T. Luning Prak and Anita P. Kuan for helpful discussions.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by the Arthritis Foundation and the Alliance for Lupus Research, the Lupus Research Institute, the Lupus Foundation of South New Jersey, the Department of Veterans Affairs, and the National Institutes of Health (Grants R01-AR-34156, U19-AI-46358, R01-AI063626, and R01 DE017590).
Abbreviations used in this paper: SLE, systemic lupus erythematosus; KI, knock-in; cGVH, chronic graft-versus-host; MHC II, MHC class II; EDF, equivalent dilution factor; AFC, autoantibody-forming cell.