Epistatic Suppression of Systemic Lupus Erythematosus: Fine Mapping of Sles1 to Less Than 1 Mb1
http://www.100md.com
免疫学杂志 2005年第14期
Abstract
Sle is a susceptibility locus for systemic autoimmunity derived from the lupus-prone NZM2410 mouse. The New Zealand White-derived suppressive modifier Sles1 was identified as a specific modifier of Sle1 and prevents the development of IgG anti-chromatin autoantibodies mediated by Sle1 on the C57BL/6 (B6) background. Fine mapping of Sles1 with truncated congenic intervals localizes it to a 956-kb segment of mouse chromosome 17. Sles1 completely abrogates the development of activated T and B cell populations in B6.Sle1. Despite this suppression of the Sle1-mediated cell surface activation phenotypes, B6.Sle1 Sles1 splenic B cells still exhibit intrinsic ERK phosphorylation. Classic genetic complementation tests using the nonautoimmmune 129/SvJ mouse suggests that this strain possesses a Sles1 allele complementary to that of New Zealand White, as evidenced by the lack of glomerulonephritis, splenomegaly, and antinuclear autoantibody production seen in (129 x B6.Sle1 Sles1)F1s. These findings localize and characterize the suppressive properties of Sles1 and implicate 129 as a useful strain for aiding in the identification of this elusive epistatic modifier gene.
Introduction
Genetic predisposition to the autoimmune disorder systemic lupus erythematous (SLE)3 is mediated by complex interactions between multiple susceptibility loci and environmental factors (1). Numerous SLE susceptibility loci have been mapped in a variety of mouse strains that spontaneously develop lupus, and genetically engineered models have implicated several additional genes in lupus predisposition (reviewed in Refs. 2 and 3). In addition, recent work has demonstrated suppressive modifier alleles that can significantly influence both the likelihood of lupus development and disease, as well as its severity and pathogenicity (4, 5, 6, 7).
We have used congenic dissection of loci identified via linkage analyses of the NZM2410 murine lupus model to better understand the underlying genetic mechanisms responsible for disease susceptibility. Characterization of the genetic and immunological properties of the different loci, in isolation and when genetically reassembled on the C57BL/6J (B6) background, have revealed that each locus contributes unique phenotypes, together recapitulating what is seen in the parental NZM2410 (8, 9, 10, 11, 12, 13). Of relevance to this study, B6.Sle1 mice develop with age a relatively benign autoimmune phenotype characterized by a loss in tolerance to chromatin and IgG autoantibody production, an increase in the proportion of activated B and T lymphocytes and mild splenomegaly (9, 14). Addition of other susceptibility loci to Sle1 has shown that this locus interacts epistatically with other loci, such as yaa, lpr, and Sle3, to mediate systemic autoimmunity and fatal lupus nephritis (10, 13, 15).
Sle1 and Sle3 are both derived from the New Zealand White (NZW) parent of NZM2410, which is a nonautoimmune inbred strain, suggesting that the NZW genome contains modifier loci that could prevent the development of autoimmunity normally driven by these potent susceptibility alleles. Linkage analyses of the NZW strain confirmed the presence of such loci, termed the Sles (SLE suppressor) loci. The most potent of these, Sles1 on mouse chromosome 17, specifically interacts with Sle1 to prevent the loss of immune tolerance to chromatin, but has no effect on the component phenotypes of the other NZM2410-derived loci, Sle2 and Sle3/5, on the B6 background (4). Because Sle1 has been shown to be key for the initiation of systemic autoimmunity in different murine lupus models, identification of Sles1 and the mechanism(s) by which it specifically suppresses Sle1 could provide insights useful for the development of strategies toward therapeutic intervention.
In this report, we demonstrate that the suppression mediated by Sles1 maps to a 956-kb interval on chromosome 17, beginning at the proximal end of the murine MHC and terminating just distal to the complement cluster. This analysis excludes TNF- and genes telomeric to it as possible candidates for Sles1. The majority of the splenic cell surface phenotypes observed in aged B6.Sle1 mice are abrogated in the presence of Sles1. In addition, using a classic genetic complementation test, we present evidence indicating that the nonautoimmune 129/SvJ (129) strain also harbors a complementary Sles1-like suppressive allele.
Materials and Methods
Mice
B6 mice were originally obtained from The Jackson Laboratory. The derivation of B6 congenic mice carrying the NZM2410-derived Sle1 interval (B6.Sle1), as well as bicongenic mice carrying Sle1 and Sles1, have been described previously (4, 9). Recombinant Sles1 mice were obtained by intercrossing (B6.Sle1 Sles1 x B6.Sle1)F1 progeny and PCR screening F1 intercross progeny for informative meioses within the Sles1 interval. Recombinant chromosomes were rescued by backcrossing to B6.Sle1 and intercrossing Sles1 recombinant progeny to produce homozygous B6.Sle1 Sles1 recombinant lines. For the F1 studies involving 129, B6, B6.Sle1, and B6.Sle1 Rec.4 z/z were crossed to 129 to generate (129 x B6), (129 x B6.Sle1), and (129 x B6.Sle1 Sles1)F1 progeny, respectively. All mice described in this study were females, bred and housed in our colony in the University of Texas Southwestern Medical Center Animal Resources Center’s specific pathogen-free facility under the supervision of Dr. J. Casco and with the approval of the Institutional Animal Care and Use Committee of the University of Texas Southwestern Medical Center.
PCR genotyping
Tail clips were obtained at weaning and used to prepare tail lysates. Briefly, 250 μl of tail lysis buffer (50 mM Tris (pH 8.0), 50 mM KCl, 2.5 mM EDTA, 0.45% Nonidet P-40, and 0.45% Tween 20) and 0.4 mg/ml proteinase K solution (Roche) were added per tail clip and kept for at least 20 h in a 55°C water bath. Tail lysates were vortexed and spun down at 14,000 rpm for 10 min. For a 20-μl PCR, 0.5 μl of tail lysate supernatant was used. PCR products were subsequently run out on 5–6.25% agarose gels to resolve the polymorphic bands. The fine mapping of the recombinant chromosomes spanning the Sles1 interval was performed using a series of 20 microsatellite markers within the original 25 cM Sles1 confidence interval. Those relevant to the results of this study include D17MIT100, D17MIT60, D17MIT146, D17MIT175, D17MIT16, D17MIT62, D17MIT28, D17MIT34, D17MIT83, D17MIT13, and TNF. To map the ends of the minimal Sles1 interval (Rec.6), a sequence between the proximal and distal NZW and B6 breakpoints (D17MIT62-D17MIT28 and D17MIT83-D17MIT13, respectively) was obtained from the Ensembl database (www.ensembl.org/Mus_musculus/). Using Sequencer software (Gene Codes), these sequences were analyzed for the presence of new microsatellite repeats for which flanking primers were designed. These novel microsatellite markers were then tested for B6 and NZW size polymorphisms, and those polymorphic between these two strains were used to map the breakpoints. Relevant polymorphic proximal and distal breakpoint primer sequences and PCR annealing temperatures are listed in Table I.
Renal pathology
Mice were terminated at 12 mo, and a longitudinal section of each kidney was fixed in 10% neutral buffered formalin (Sigma-Aldrich), paraffin embedded, cut into 3-μm sections, and stained with H&E with periodic acid-Schiff. The sections were examined in a blind manner for evidence of pathologic changes in the glomeruli, tubules, or interstitial areas. The glomeruli were screened for evidence of hypertrophy, proliferative changes, crescent formation, hyaline deposits, fibrosis/sclerosis, and basement membrane thickening. The glomerulonephritis (GN) severity was graded on a scale of 0–4, where 0 = normal, 1 = mild increase in mesangial cellularity and matrix, 2 = moderate increase in mesangial cellularity and matrix with thickening of glomerular basement membrane, 3 = focal endocapillary hypercellularity with obliteration of capillary lumina and a substantial increase in the thickness and irregularity of the glomerular basement membrane, and 4 = diffuse endocapillary hypercellularity, segmental necrosis, crescents, and hyalinized end-stage glomeruli. Similarly, the severity of tubulointerstitial nephritis was graded on a 0–4 scale based on the extent of tubular atrophy, inflammatory infiltrates, and interstitial fibrosis, as described previously (16). The number of infiltrating polymorphonuclear leukocytes was directly enumerated based on their typical polymorphonuclear morphology by examining 50 glomeruli/kidney/section/mouse.
Serology
Mice were bled at 5, 7, 9, and 12 mo and sera stored at –20°C. ELISA detection of IgM and IgG autoantibodies directed against chromatin and dsDNA were performed as described previously (9). All coating steps were for 30 min at 37°C followed by two washes in PBS (Sigma-Aldrich). Briefly, for both the anti-dsDNA and anti-chromatin ELISAs, Immunlux HB (Dynatech Laboratories) plates were precoated with 50 μl/well methylated BSA and then coated with 50 μl/well of 50 μg/ml dsDNA (dissolved in PBS and filtered; Sigma-Aldrich). For the anti-chromatin ELISAs, an additional coating step with 50 μl/well of 10 μg/ml total histones (Roche) was used. Following the last respective coating step and wash, ELISA plates were incubated overnight at 4°C with 200 μl/well ELISA blocking buffer (PBS, 0.1% gelatin, 3% BSA, and 3 mM EDTA). Test sera were added at a final dilution of 1/200 and 1/800 (serial) for anti-dsDNA and anti-chromatin ELISAs, respectively, in serum diluent (PBS, 0.1% gelatin, 2% BSA, 3 mM EDTA, and 0.05% Tween 20) and incubated for 2 h at room temperature. Bound IgM or IgG was detected using alkaline phosphatase-conjugated goat anti-mouse IgG or IgM (Roche) and p-nitrophenyl phosphate (Sigma-Aldrich) as the substrate. OD450 was measured by an Elx800 Automated Microplate Reader (Bio-Tek Instruments) and the raw optical densities for anti-chromatin ELISAs converted to arbitrary normalized units using a six-point standard curve generated by a mAb derived from a NZM2410 mouse (17). A 1/250 dilution of this supernatant was set at 1000 U/ml. For anti-dsDNA ELISAs, the standard curve was set using six serial dilutions of sera from a female (NZW x B6.Sle1)F1, with a 1/250 dilution of this sera set at 1000 U/ml.
Cell preparation and culture
Following the measurement of spleen weight, single-cell splenocyte suspensions were prepared, passed through nylon mesh, and depleted of RBCs using RBC lysis buffer (0.1 mM EDTA, 0.83% NH4Cl, and 0.1% KHCO3). These cells were then used for flow cytometric analyses (described below).
Flow cytometric analysis and Abs
In each FACS experiment, at least one mouse per genotype (B6, B6.Sle1, B6.Sle1 Sles1) was included. For all FACS analyses, B6.Sle1 Sles1 indicates that the B6.Sle1 Rec.1 z/z line was used. Cells were blocked with 2.4G2 (American Type Culture Collection) on ice. Cells (1.5 x 106/Ab mixture) were then stained on ice using optimal amounts of FITC, PE, PE-Texas Red, Cy-Chrome, PerCPCy5.5, allophycocyanin, or biotin-conjugated Abs at predetermined dilutions. Four-color combinations of the following Abs, obtained from BD Biosciences-Pharmingen were used for the various analyses: CD21/35 (7G6); CD23 (B3B4); IgM (R6-60.2); CD86 (GL1); CD45R (RA36B2); CD5 (53-7.3); CD11b (M1/70); NK1.1 (PK126); CD25 (PC61); CD69 (H1.2F3); CD4 (H129.19); CD8 (53-6.7); CD3 (145–2C11); CD19 (1D3); CD43 (S7); CD1d (1B1); CD62L (MEL-14); CD44 (1M7); and CD138 (281-2). After two washes, biotin-conjugated Abs were revealed using BD Pharmingen streptavidin-allophycocyanin. Appropriate isotype controls for all Ab combinations were included in all analyses for all samples. Stained cells were acquired on a FACSCalibur with CellQuest software (BD Biosciences). Dead cells were excluded on the basis of forward and side scatter properties, and 40,000 events within the lymphocyte gate were acquired per sample. Flow cytometric data was analyzed using FlowJo (Tree Star).
SDS-PAGE and immunoblotting
Splenic B lymphocytes were isolated using Dynal anti-B220 magnetic beads according to the manufacturer’s protocol (Dynal Biotech). Isolated cells were lysed in lysis buffer (300 mM NaCl, 50 mM Tris-Cl (pH 7.6), 0.5% Triton X-100, 10 μg/ml leupeptin, 10 μg/ml aprotinin, 1 mM PMSF, 1 mM sodium orthovanadate, 1 mM sodium molybdate, and 1 mM NaF), incubated on ice for 30 min, and then cleared by centrifuging at 12,000 x g for 20 min. Equivalent amounts of total cellular protein extract (4 μg) were fractionated on 10% SDS-PAGE and electrotransferred to polyvinylidene difluoride membrane using a Bio-Rad Transblot apparatus. The membrane was blocked overnight with TBST buffer (10 mM Tris-Cl (pH 8.0), 0.9% NaCl, and 0.1% Tween 20) plus 4% BSA (for phosphorylated proteins) or 5% nonfat dry milk and incubated for 1 h with primary mAb in TBST buffer plus 4% BSA (for phosphorylated proteins) or 5% nonfat dry milk. Following three washes in TBST, membranes were incubated with a secondary Ab for 1 h. The membranes were again washed three times in TBST and then developed using an ECL Plus kit (Amersham Biosciences) according to the manufacturer’s instructions. Protein bands were quantified by densitometric analysis using a computerized densitometer (Molecular Dynamics) and ImageQuant software (Molecular Dynamics). Rabbit polyclonal Abs specific for P44/42 MAPK were purchased from Cell Signaling Technology. Rabbit anti-phospho-ERK1/2 MAPKs polyclonal Abs were purchased from Promega. The secondary Abs (goat anti-rabbit IgG) were purchased from Santa Cruz Biotechnology.
Statistical analysis
Data was analyzed using parametric and nonparametric ANOVAs as determined by InStat3 (Graphpad Software), unless otherwise indicated. Error bars represent SEMs unless data are not normally distributed, in which case bars reflect the interquartile ranges.
Results
Sles1 maps to a 956-kb interval that excludes TNF-
Our previous work demonstrated the presence of a recessive NZW-derived epistatic modifier on mouse chromosome 17, termed Sles1, which specifically suppressed the break in tolerance to chromatin mediated by Sle1 on the B6 background. Consequently, congenic recombinant fine mapping was performed by intercrossing B6.Sle1 Sles1 het mice and identifying recombinant chromosomes in the Sles1 interval by PCR screening tail DNA from testcross progeny with polymorphic microsatellite markers flanking the original Sles1 interval. Once recombinant chromosomes were identified, additional markers were used to localize the recombinant breakpoint. The recombinant mice were backcrossed to B6.Sle1 to "rescue" the recombinant Sles1 chromosomes and generate new Sles1 recombinant lines. The new recombinant lines found to retain the recessive suppression phenotype of Sles1 were used to generate more recombinant strains. As shown in Fig. 1A, a series of six truncated recombinant chromosome lines, spanning the Sles1 interval and termed B6.Sle1 Rec.1-6, were identified via the analysis of 1329 testcross progeny. Interestingly, there appeared to be a recombinational hot spot between the markers D17MIT146 and D17MIT175, as evidenced by the greater than expected number of recombinants obtained between those two markers (2 = 22.3; p < 0.005).
Progeny from these lines, homozygous and heterozygous for the recombinant chromosomes, were aged and assayed for IgG anti-chromatin auto antibody production by ELISA at 5, 7, 9, and 12 mo of age. At the 12-mo terminal sacrifice, spleens were weighed to assess splenomegaly, a phenotype observed in aged B6.Sle1 mice (4, 9, 18).
As depicted in Fig. 2A, aged B6.Sle1 female mice have a significant increase in their mean spleen weights when compared with age- and sex-matched B6 (233.3 mg, interquartile range 163.9–280.2 mg vs 107.4 mg, interquartile range 99.7–127.1 mg); p < 0.001, nonparametric ANOVA). Comparison of spleen weights from the various homozygous recombinant lines with B6 and B6.Sle1 indicated that B6.Sle1 Rec.2 z/z and B6.Sle1 Rec.3 z/z mice differed significantly from B6 (p < 0.001; nonparametric ANOVA), whereas B6.Sle1 Rec.1 z/z, B6.Sle1 Rec.4 z/z, B6.Sle1 Rec.5 z/z, and B6.Sle1 Rec.6 z/z mice were indistinguishable from B6 (p > 0.05) and different from B6.Sle1 (p < 0.001 for all strains; nonparametric ANOVA). Importantly, mice heterozygous for all six recombinant chromosomes revert to the 2-fold increase in spleen weights seen in B6.Sle1, which is consistent with the recessive inheritance of Sles1 observed in the original bi-congenic analyses (4). These results map the suppression of splenomegaly into the genomic interval contained within Rec.1 and Rec.4-6.
Suppression of antinuclear autoantibody (ANA) production by Sles1 was assayed by anti-chromatin ELISAs at various time points. As shown in Fig. 2B, in which titers from 9- to 12-mo-old female mice of the various genotypes are presented, B6.Sle1 Rec.1 z/z, B6.Sle1 Rec.4 z/z, B6.Sle1 Rec.5 z/z, and B6.Sle1 Rec.6 z/z all suppress anti-chromatin ANA production (p < 0.001 for all strains compared with B6.Sle1 females; nonparametric ANOVA), hence mapping Sles1 to these genomic segments. In contrast, B6.Sle1 Rec.2 z/z and B6.Sle1 Rec.3 z/z failed to suppress anti-chromatin ANA production (p < 0.001 for all strains compared with B6; nonparametric ANOVA). These results correlate with the lack of splenomegaly in the different congenic lines, consistent with a single gene mediating both effects.
These data place Sles1 in the 956-kb genomic interval contained in Rec.6 (Fig. 1A). This interval includes the microsatellite markers D17MIT28 and D17MIT83 but excludes D17MIT62 and D17MIT13 at the proximal and distal ends respectively and contains 65 genes (Fig. 1). Using public sequence available from the Ensembl database (National Center for Biotechnology Information m33), we designed primers surrounding microsatellite repeats at the proximal (D17MIT62-D17MIT28) and distal breakpoints (D17MIT83-D17MIT13) to further map the ends of this interval. The 64-kb proximal and 27-kb distal breakpoints (between the primer pairs ms3-ms4 and ms7-ms8, respectively) harbor two additional genes each (www.ensembl.org/Mus_musculus/). As shown in Fig. 1B, this brings the maximum number of candidate genes in the minimal Sles1 interval to 69. This result is consistent with the gene-rich nature of this region of the genome, as has been recently reviewed (19, 20). In addition, given the recognized immunological functions of many of these genes, determining candidacy for Sles1 remains a challenge.
Sles1 suppresses the activation of T cells mediated by Sle1
The breach in tolerance to chromatin mediated by the Sle1 locus is accompanied by an increase in the activation phenotypes in splenic T and B lymphocytes (4, 9, 14). We performed a comprehensive analysis of the various splenic cell populations in ages 9- to 12-mo-old female B6, B6.Sle1, and B6.Sle1 Sles1 z/z (B6.Sle1 Rec.1 z/z) mice to better characterize these changes and determine their association with the suppression mediated by the Sles1 interval.
As shown in Table II, there were numerous quantitative differences in the percentages of various splenic subpopulations when B6.Sle1 mice were compared with age and gender-matched B6 controls. A small but significant decrease (p < 0.05; ANOVA) in the percentage of CD3+ T cells in B6.Sle1 mice was restored to B6 levels in the presence of Sles1 (Table II). As has been shown previously, B6.Sle1 mice had a highly significant increase in the percentage of CD3+ T cells expressing the early activation marker CD69, when compared with B6 controls (9). This activation phenotype was suppressed by Sles1 (Fig. 3A), as shown by the decrease in the percentage of CD69+ CD3 T cells in B6.Sle1 Sles1 vs B6.Sle1 mice (p < 0.001; ANOVA). The overall cell surface levels of CD69 on the CD3+ population, as determined by comparing median fluorescence intensities (MFIs), was also significantly higher in B6.Sle1 mice (p < 0.001; ANOVA), which also was restored to B6 levels by Sles1 (Fig. 3B; Table II).
We also investigated the splenic B1 and B2 B cell populations in these strains to look for potential changes in the splenic B1 population, as observed in some lupus models (27, 28). As shown in Table II, there were no significant differences in the splenic B1 lineages in any of the groups examined, but there was a decrease in the B2 population (CD5–CD23+B220+) in B6.Sle1 mice compared with B6 controls (29.4 ± 1.7 vs 38.3 ± 1.0%; p < 0.01; ANOVA), similar to the decrease seen in the follicular B cell population discussed above. Again, there was no change in B6.Sle1 Sles1 mice compared with both B6 and B6.Sle1.
Similar to what was observed in the T cell compartment, there was also an increase in the percentage of B cells in B6.Sle1 mice expressing the activation marker CD69. In Fig. 4B, the percentage of B220 B cells that expressed CD69 is shown for the various strains, which increased from 2.7 ± 0.3% in B6 to 6.8 ± 1.1% in B6.Sle1 mice (p < 0.001; ANOVA). This increase in the activation status of the B cell compartment was also reflected in the increased cell surface levels of the costimulatory molecule CD86 (B7.2) on follicular B cells in B6.Sle1 mice, as determined by comparisons of MFIs (Table I; p < 0.05; repeated measures ANOVA). Neither change was seen in B6.Sle1 Sles1 mice, in which the percentage of B220 B cells expressing CD69 is only 1.5 ± 0.2% (p < 0.001 vs B6.Sle1; ANOVA), and the MFI of CD86 on follicular B cells was correspondingly decreased (Table I). These results indicate that the suppression of the activation status of lymphocytes introduced by Sles1 affects both the T and B cell lineages, such that their splenic cell surface profiles are comparable to B6.
We recently demonstrated that ex vivo B cells from B6.Sle1 mice display an intrinsic increase in phosphorylated ERK2 ex vivo by 4 mo of age (26). Based on this result, we compared splenic B cells from B6.Sle1 Sles1 mice to determine whether the intrinsic ERK2 phosphorylation observed in B6.Sle1 was affected by the presence of Sles1. As shown in Fig. 4C, ex vivo splenic B cells from 10-mo-old B6.Sle1 Sles1 mice show an increase in the levels of phosphorylated ERK2 comparable to that seen in age- and gender-matched B6.Sle1 B cells. The persistence of this B cell activation phenotype could suggest that B cells in B6.Sle1 Sles1 mice are not receiving adequate T cell help for driving class switch to IgG anti-chromatin autoantibody production. However, ELISA analyses for both IgM and IgG anti-chromatin autoantibody production by B6.Sle1Sles1 mice at different ages indicate otherwise, as neither isotype anti-chromatin autoantibodies are present in B6.Sle1 Sles1 (data not shown). These findings suggest that the suppressive Sles1 allele acts downstream of these activated molecular pathways and that hyperphosphorylation of ERK2 can be uncoupled from the humoral autoimmunity elicited by Sle1.
Genetic complementation suggests a 129 Sles1 allele
Recent work by our group demonstrated that the break in tolerance to chromatin mediated by the Sle1 subcongenic locus, Sle1b, is associated with functional polymorphisms in the SLAM/CD2 family of immune receptor genes and that these polymorphisms are found in the majority of inbred strains examined, including the nonautoimmune 129/SvJ (129) (29). Furthermore, introgressing this chromosome 1 interval from 129 onto B6 background was sufficient to cause a loss in tolerance in chromatin, resulting in the production of serum ANAs (29, 30). This raised the question of why the 129 strain remains nonautoimmune, despite the presence of a potent Sle1-related SLAM/CD2 haplotype on chromosome 1. We hypothesized that, similar to what we previously observed in the NZW strain, epistatic modifiers might play a role in the lack of autoimmunity seen in 129.
To test whether one of the modifiers in 129 was a Sles1-like allele, we performed a classic genetic complementation test, comparing the autoimmunity elicited in (129 x B6.Sle1)F1s to that seen in (129 x B6.Sle1 Sles1)F1s, by crossing 129 to B6.Sle1 and B6.Sle1 Rec.4 z/z (B6.Sle1 Sles1). These comparisons are similar to the original studies performed with NZW in which it was shown that NZW homozygosity at Sles1 is sufficient to prevent systemic autoimmunity in the (NZW x B6.Sle1)F1 hybrid system.
The progeny of the described crosses with 129 are both homozygous for the autoimmunity-promoting SLAM/CD2 haplotype 2 on chromosome 1, whereas the remainder of the genome is heterozygous for 129 and B6 alleles (29). However, these two crosses differ at Sles1 in that the (129 x B6.Sle1 Sles1)F1s carry one copy of the NZW-derived Sles1 allele (31). Because heterozygosity of the Sles1 allele does not lead to suppression (see Fig. 2), we expect that if the 129- and B6-derived Sles1 loci are equivalent in their ability to potentiate autoimmunity mediated by the SLAM/CD2 family haplotype 2, then both sets of F1 intercross progeny would have comparable autoimmunity. However, if the 129 strain has a protective Sles1 allele, complementary to that of NZW, then the latter F1 intercross progeny would be expected to have abrogated autoimmunity.
As shown in Fig. 5A, the degree of moderate to severe GN, defined as being a GN score of 3, in 9-mo-old female (129 x B6.Sle1)F1s was significantly higher than that observed in both (129 x B6.Sle1 Sles1)F1s and (129 x B6)F1s (p < 0.01 and p < 0.001, respectively; ANOVA). In addition, the (129 x B6.Sle1 Sles1)F1 group was the only group with no evidence of lymphocyte infiltration of the kidneys (data not shown). Similarly, when spleen weights from the different mice were compared (Fig. 6B), the (129 x B6.Sle1)F1s exhibited significant splenomegaly (p < 0.001 vs B6; nonparametric ANOVA) whereas (129 x B6.Sle1 Sles1)F1s did not. In fact, as clearly shown in Fig. 5B, the only congenic lines with significant differences in spleen weights from B6 were those that have one or more copies of the B6 allele at the Sles1 locus, when Sle1 is homozygous. These data strongly support the presence of a complementary Sles1 modifier in the 129 genome.
The quantification of IgG autoantibody levels in the different strains further support this conclusion. As shown in Fig. 6, A and B, only the (129 x B6.Sle1)F1s had significant levels of serum anti-dsDNA IgG autoantibodies by 9 mo of age (vs B6, p < 0.01). In contrast, the penetrance of anti-dsDNA autoantibodies in (129 x B6.Sle1 Sles1)F1s did not differ significantly from age- and gender-matched B6 and 129 controls.
We also assayed for the more nonspecific anti-chromatin autoantibodies in the various strains. As shown in Fig. 6, C and D, at 9 mo of age both the titers and penetrance of serum IgG ANAs in the (129 x B6.Sle1)F1s were highly increased when compared with any of the control groups (p < 0.001; ANOVA). Interestingly, unlike anti-dsDNA autoantibodies, the (129 x B6)F1s themselves had an increased (50%) penetrance of anti-chromatin autoantibodies. In contrast to either of the other F1 groups, the (129 x B6.Sle1 Sles1)F1s had minimal serum anti-chromatin autoantibodies at 9 mo of age (Fig. 6, C and D). Significant differences in the titers and penetrance of anti-chromatin autoantibodies were discernible as early as 5 mo of age in (129 x B6.Sle1)F1s (data not shown).
These data indicate that the 129-derived Sles1 allele specifically complements the NZW Sles1 allele in preventing both the more pathogenic phenotypes, such as GN development and anti-dsDNA autoantibodies, as well as the development of the more benign serum anti-chromatin autoantibodies. Overall, these results are extremely similar to those originally obtained in the analysis of the (NZW x B6.Sle1 Sles1)F1 hybrid system (4). However, the presence of Sles1 in 129 is especially intriguing because classic analyses of the H2 in 129 have defined it as an H2b haplotype very closely related, but not identical, to the H2b haplotype carried by B6.
Discussion
The present study extends our previous work characterizing the suppression of autoimmunity mediated by Sles1 in the bicongenic B6.Sle1 Sles1 system. Our congenic fine mapping localizes Sles1 into an interval on chromosome 17 that encompasses 956 kb of the proximal portion of the murine MHC. Our current minimal Sles1 interval still contains 69 genes, consistent with the gene-rich nature of this segment of the genome (19, 20). Many of these positional candidates are known to possess immunological functions and are also highly polymorphic between the B6 and NZW strains (31, 32). Our genetic fine-mapping analysis definitively excludes TNF-, a known lupus susceptibility candidate gene and one implicated in other murine lupus models, as well as the genes distal to it, as possible Sles1 candidates (33, 34, 35, 36). As shown in the schematic in Fig. 7, this leaves many other potentially interesting candidates, including the class II genes H2-Ab1, Aa, Eb, and Ea, the complement genes proximal to TNF-, as well as other relevant genes such as Daxx, Notch4, and Hspa1a. Interestingly, unlike many other loci implicated in lupus and other autoimmune disease susceptibility, fine mapping of Sles1 did not reveal the presence of multiple subloci contributing to suppression (14, 37, 38).
In contrast to other autoimmune disorders, such as insulin-dependent diabetes mellitus and multiple sclerosis, the association of the class II MHC genes with susceptibility to SLE has proven to be very complicated and highly dependent on epistatic interactions with other susceptibility loci, as demonstrated in a variety of linkage studies in both human SLE patients and murine lupus models. Earlier studies using the classic (NZB x NZW]F1 (BWF1)) model, from which NZM2410 is derived, in various backcross analyses have suggested that the NZW allele at the H2 (H2z) confers a high risk of developing renal disease, and in particular, it is the NZW class II alleles that are the dominant contributors to disease in these models (39). Many of these studies have documented that heterozygosity at the H2 gene complex is crucial for lupus susceptibility and is associated with many of the phenotypic characteristics of lupus, such as GN development, autoantibody production, and splenomegaly (40). However, many such backcross studies have also indicated that homozygosity of the NZW alleles at the H2 region is protective. These somewhat contradictory data have led to a variety of hypotheses such as an increased self-recognition in heterozygotes due to mixed haplotype or mixed isotype class II molecules, suppressive effects caused by increased class II I-E expression, and competition by peptides derived from I-Ea resulting in decreased self-peptide presentation by class II I-A molecules during thymic development leading to increased autoimmune susceptibility (41, 42, 43, 44, 45). A major limitation in interpreting the role of H2 loci in these backcross studies is that the remainder of the genome also contains susceptibility alleles that are randomly segregating. Considering the importance of epistatic interactions in the development of SLE, this heterogeneous segregation complicates the unambiguous association of phenotypes with a specific locus.
The use of transgenics for the various NZW class II alleles (I-Ez and I-Az) on nonautoimmune backgrounds have also been used in various backcross studies to determine the contributions of these alleles to lupus susceptibility and have shown that transgenic expression of the class II molecules does not recapitulate what is seen when congenics for the entire NZW MHC are used in similar crosses. This suggests that nonclass II molecules in the MHC, in tight linkage disequilibrium with class II genes, may be the ultimate culprits in modulating disease susceptibility (46, 47). Recently, Zhang et al. (48) used intra-MHC recombinants to try and elucidate the contribution of variances in the level of I-E expression, when the I-A molecule haplotype is held constant, in a series of F1 crosses. They concluded that expression of the class II I-E molecule offers protection in a dose-dependent manner and postulate that this may be due to mechanisms including thymic deletion of a self-reactive TCR repertoire, determinant capture, differences in cytokine balances and competition for signal transduction moities. However, the recombinant breakpoint in their intra-MHC recombinant, H2-g2r, is defined only as occurring somewhere between H2-Ea and TNF-, a region containing >60 genes (www.ensembl.org/Mus_musculus/) (Ref. 48 ; Fig. 1B). It could hence also be concluded that a gene within this undefined interval, which partially overlaps with our minimal Sles1 interval, is responsible for the dose-dependent protection seen in their studies.
In addition to class II genes, the complement genes within the MHC have long been implicated in lupus susceptibility. Deficiencies in the complement components C1 and C4 are in fact the only known single gene deficiencies that are so strongly associated with human SLE (1). Furthermore, during lupus disease progression, acquired deficiencies in complement components are common and are presumed to result from the immune complex use of complement. There are two main, nonmutually exclusive theories that have been proposed to explain the link between complement deficiency and susceptibility to lupus. One involves the role of complement in the clearance of apoptotic cells and immune complexes and the other invokes the known ability of complement to determine activation thresholds in B and T lymphocytes (49). Interestingly, ablation of the C4 gene, located at the peak marker used in the original linkage studies describing Sles1, has been demonstrated to alter negative selection of immature self-reactive B cells in both the double-transgenic hen egg lysosome and the lpr null models (4, 50). Furthermore, unlike other complement deficiencies (C1qa, C3, and Cr1/2), the lack of C4 appears less dependent on a mixed 129 x B6 background for the manifestation of some degree of autoimmunity (7). However, serum levels of this complement component are equivalent in young, preautoimmune B6, B6.Sle1 and B6.Sle1 Sles1 mice (data not shown).
Other candidate genes in the Sles1 interval includes the MHC proximal gene Daxx, known to play an important role in the IFN--mediated inhibition of B lymphopoiesis, which is interesting considering the large number of recent studies implicating type I IFNs in lupus susceptibility (51, 52, 53, 54, 55, 56, 57). Heat shock proteins are known to have effects on the maturation of APCs in a manner reminiscent of cytokines, and heat shock protein 70, encoded by Hspa1a on the distal end of the minimal Sles1 interval, has been shown to be capable of converting T cell tolerance to autoimmunity in vivo (58, 59). Notch4 is also contained within the Sles1 interval, and given the recent wave of papers implicating members of the Notch family of receptors and ligands in regulating B and T lymphocyte development and function, it remains an attractive candidate for Sles1 (60, 61, 62, 63, 64, 65, 66, 67, 68, 69).
Our phenotypic analyses of the cell surface characteristics of the splenic immune cells from B6, B6.Sle1, and B6.Sle1 Sles1 mice reveal that many, but not all, of the differences seen in 9- to 12-mo-old B6.Sle1 splenocytes are suppressed in the presence of Sles1. Aged B6.Sle1 mice, despite their relatively benign autoimmune phenotype of ANA production, have significantly increased populations of activated B (B220+CD69+) and T (CD4+CD62L–CD44high; CD3+CD69+) cells, similar to those observed in much more severe lupus models (15, 21, 22, 23, 25, 70). It has been postulated that the spontaneous lymphocyte activation seen in these different lupus models is due to differences in thresholds for activation and immune regulation caused by the underlying genetic polymorphisms in relevant susceptibility loci (29, 71, 72). Intriguingly, both T and B cells from B6.Sle1 Sles1 mice have a significantly reduced activation phenotype, indistinguishable from B6, suggesting that Sles1 is impacting both lymphocyte compartments. This suggests that a lack of T cell help to the B cells is not the sole mechanism of suppression. This is further supported by the observation that B6.Sle1 Sles1 mice do not produce serum IgM autoantibodies either, indicating that lack of class-switching is not the reason these mice do not produce high titers of serum IgG autoantibodies. However, ex vivo B cells from B6.Sle1 Sles1 mice demonstrate a comparable increase in phosphorylated ERK2 to those from B6.Sle1, indicating that this activated molecular pathway can be uncoupled from the humoral autoimmunity elicited by Sle1.
Given the density and polymorphic nature of the genes within the Sles1 interval, definitive identification of the Sles1 gene has proven to be quite a challenge. However, our F1 analyses with the nonautoimmune 129 strain provide an alternative approach that may significantly accelerate this analysis. Because B6 and 129 have very closely related H2 haplotypes, most notably identity for class II I-A and I-E alleles, the degree of variability between the B6 and 129 alleles will be much reduced in comparison to B6 vs NZW. Consequently, we will seek variations in structure/function of candidate genes between B6 and 129, which should enhance our ability to quickly focus on relevant polymorphisms. Furthermore, this result excludes the candidacy of MHC class II molecules for Sles1.
The use of congenic recombinants has now been used to fine-map both susceptibility and suppressive loci in efforts to identify the underlying genes (14, 29). These congenic systems allow individual key susceptibility and modifier loci to be specifically combined so that phenotypes resulting from their genetic interactions can be studied. Hence, the ability of Sles1 to specifically target Sle1 could help elucidate the key epistatic interactions involved in the generation of severe systemic autoimmunity in lupus models where Sle1 has been shown to be necessary for initiation and development (13, 15, 23). Identification of Sles1, using both recombinant fine-mapping strategies and comparisons of B6, 129, and NZW alleles, will have important implications for our understanding of and ability to therapeutically intervene in SLE pathogenesis.
Acknowledgments
We thank Nisha Limaye, Alice Chan, and Katherine Belobrajdic for critical reading and helpful discussions. We also express gratitude to Dr. Jose Casco for excellent management of the mouse colony.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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.
1 This work was supported by a grant from the National Institutes of Health Grant 5R37AI04519607 (to E.K.W.).
2 Address correspondence and reprint requests to Dr. Edward K. Wakeland, Center for Immunology, University of Texas Southwestern Medical Center, 6000 Harry Hines Boulevard, Dallas, TX 75390-9093. E-mail address: edward.wakeland@utsouthwestern.edu
3 Abbreviations used in this paper: SLE, systemic lupus erythematous; NZW, New Zealand White; ANA, antinuclear autoantibody; GN, glomerulonephritis; MFI, median fluorescence intensity; MZ, marginal zone.
Received for publication March 7, 2005. Accepted for publication May 5, 2005.
References
Wakeland, E. K., K. Liu, R. R. Graham, T. W. Behrens. 2001. Delineating the genetic basis of systemic lupus erythematosus. Immunity 15: 397-408.
Nguyen, C., N. Limaye, E. K. Wakeland. 2002. Susceptibility genes in the pathogenesis of murine lupus. Arthritis Res. 4:(Suppl. 3): S255-S263.
Raman, K., C. Mohan. 2003. Genetic underpinnings of autoimmunity: lessons from studies in arthritis, diabetes, lupus, and multiple sclerosis. Curr. Opin. Immunol. 15: 651-659.
Morel, L., X. H. Tian, B. P. Croker, E. K. Wakeland. 1999. Epistatic modifiers of autoimmunity in a murine model of lupus nephritis. Immunity 11: 131-139.[
Santiago-Raber, M. L., B. R. Lawson, W. Dummer, M. Barnhouse, S. Koundouris, C. B. Wilson, D. H. Kono, A. N. Theofilopoulos. 2001. Role of cyclin kinase inhibitor p21 in systemic autoimmunity. J. Immunol. 167: 4067-4074.
Bolland, S., Y. S. Yim, K. Tus, E. K. Wakeland, J. V. Ravetch. 2002. Genetic modifiers of systemic lupus erythematosus in FcRIIB–/– mice. J. Exp. Med. 195: 1167-1174.
Paul, E., O. O. Pozdnyakova, E. Mitchell, M. C. Carroll. 2002. Anti-DNA autoreactivity in C4-deficient mice. Eur. J. Immunol. 32: 2672-2679.(Srividya Subramanian*, Yo)
Sle is a susceptibility locus for systemic autoimmunity derived from the lupus-prone NZM2410 mouse. The New Zealand White-derived suppressive modifier Sles1 was identified as a specific modifier of Sle1 and prevents the development of IgG anti-chromatin autoantibodies mediated by Sle1 on the C57BL/6 (B6) background. Fine mapping of Sles1 with truncated congenic intervals localizes it to a 956-kb segment of mouse chromosome 17. Sles1 completely abrogates the development of activated T and B cell populations in B6.Sle1. Despite this suppression of the Sle1-mediated cell surface activation phenotypes, B6.Sle1 Sles1 splenic B cells still exhibit intrinsic ERK phosphorylation. Classic genetic complementation tests using the nonautoimmmune 129/SvJ mouse suggests that this strain possesses a Sles1 allele complementary to that of New Zealand White, as evidenced by the lack of glomerulonephritis, splenomegaly, and antinuclear autoantibody production seen in (129 x B6.Sle1 Sles1)F1s. These findings localize and characterize the suppressive properties of Sles1 and implicate 129 as a useful strain for aiding in the identification of this elusive epistatic modifier gene.
Introduction
Genetic predisposition to the autoimmune disorder systemic lupus erythematous (SLE)3 is mediated by complex interactions between multiple susceptibility loci and environmental factors (1). Numerous SLE susceptibility loci have been mapped in a variety of mouse strains that spontaneously develop lupus, and genetically engineered models have implicated several additional genes in lupus predisposition (reviewed in Refs. 2 and 3). In addition, recent work has demonstrated suppressive modifier alleles that can significantly influence both the likelihood of lupus development and disease, as well as its severity and pathogenicity (4, 5, 6, 7).
We have used congenic dissection of loci identified via linkage analyses of the NZM2410 murine lupus model to better understand the underlying genetic mechanisms responsible for disease susceptibility. Characterization of the genetic and immunological properties of the different loci, in isolation and when genetically reassembled on the C57BL/6J (B6) background, have revealed that each locus contributes unique phenotypes, together recapitulating what is seen in the parental NZM2410 (8, 9, 10, 11, 12, 13). Of relevance to this study, B6.Sle1 mice develop with age a relatively benign autoimmune phenotype characterized by a loss in tolerance to chromatin and IgG autoantibody production, an increase in the proportion of activated B and T lymphocytes and mild splenomegaly (9, 14). Addition of other susceptibility loci to Sle1 has shown that this locus interacts epistatically with other loci, such as yaa, lpr, and Sle3, to mediate systemic autoimmunity and fatal lupus nephritis (10, 13, 15).
Sle1 and Sle3 are both derived from the New Zealand White (NZW) parent of NZM2410, which is a nonautoimmune inbred strain, suggesting that the NZW genome contains modifier loci that could prevent the development of autoimmunity normally driven by these potent susceptibility alleles. Linkage analyses of the NZW strain confirmed the presence of such loci, termed the Sles (SLE suppressor) loci. The most potent of these, Sles1 on mouse chromosome 17, specifically interacts with Sle1 to prevent the loss of immune tolerance to chromatin, but has no effect on the component phenotypes of the other NZM2410-derived loci, Sle2 and Sle3/5, on the B6 background (4). Because Sle1 has been shown to be key for the initiation of systemic autoimmunity in different murine lupus models, identification of Sles1 and the mechanism(s) by which it specifically suppresses Sle1 could provide insights useful for the development of strategies toward therapeutic intervention.
In this report, we demonstrate that the suppression mediated by Sles1 maps to a 956-kb interval on chromosome 17, beginning at the proximal end of the murine MHC and terminating just distal to the complement cluster. This analysis excludes TNF- and genes telomeric to it as possible candidates for Sles1. The majority of the splenic cell surface phenotypes observed in aged B6.Sle1 mice are abrogated in the presence of Sles1. In addition, using a classic genetic complementation test, we present evidence indicating that the nonautoimmune 129/SvJ (129) strain also harbors a complementary Sles1-like suppressive allele.
Materials and Methods
Mice
B6 mice were originally obtained from The Jackson Laboratory. The derivation of B6 congenic mice carrying the NZM2410-derived Sle1 interval (B6.Sle1), as well as bicongenic mice carrying Sle1 and Sles1, have been described previously (4, 9). Recombinant Sles1 mice were obtained by intercrossing (B6.Sle1 Sles1 x B6.Sle1)F1 progeny and PCR screening F1 intercross progeny for informative meioses within the Sles1 interval. Recombinant chromosomes were rescued by backcrossing to B6.Sle1 and intercrossing Sles1 recombinant progeny to produce homozygous B6.Sle1 Sles1 recombinant lines. For the F1 studies involving 129, B6, B6.Sle1, and B6.Sle1 Rec.4 z/z were crossed to 129 to generate (129 x B6), (129 x B6.Sle1), and (129 x B6.Sle1 Sles1)F1 progeny, respectively. All mice described in this study were females, bred and housed in our colony in the University of Texas Southwestern Medical Center Animal Resources Center’s specific pathogen-free facility under the supervision of Dr. J. Casco and with the approval of the Institutional Animal Care and Use Committee of the University of Texas Southwestern Medical Center.
PCR genotyping
Tail clips were obtained at weaning and used to prepare tail lysates. Briefly, 250 μl of tail lysis buffer (50 mM Tris (pH 8.0), 50 mM KCl, 2.5 mM EDTA, 0.45% Nonidet P-40, and 0.45% Tween 20) and 0.4 mg/ml proteinase K solution (Roche) were added per tail clip and kept for at least 20 h in a 55°C water bath. Tail lysates were vortexed and spun down at 14,000 rpm for 10 min. For a 20-μl PCR, 0.5 μl of tail lysate supernatant was used. PCR products were subsequently run out on 5–6.25% agarose gels to resolve the polymorphic bands. The fine mapping of the recombinant chromosomes spanning the Sles1 interval was performed using a series of 20 microsatellite markers within the original 25 cM Sles1 confidence interval. Those relevant to the results of this study include D17MIT100, D17MIT60, D17MIT146, D17MIT175, D17MIT16, D17MIT62, D17MIT28, D17MIT34, D17MIT83, D17MIT13, and TNF. To map the ends of the minimal Sles1 interval (Rec.6), a sequence between the proximal and distal NZW and B6 breakpoints (D17MIT62-D17MIT28 and D17MIT83-D17MIT13, respectively) was obtained from the Ensembl database (www.ensembl.org/Mus_musculus/). Using Sequencer software (Gene Codes), these sequences were analyzed for the presence of new microsatellite repeats for which flanking primers were designed. These novel microsatellite markers were then tested for B6 and NZW size polymorphisms, and those polymorphic between these two strains were used to map the breakpoints. Relevant polymorphic proximal and distal breakpoint primer sequences and PCR annealing temperatures are listed in Table I.
Renal pathology
Mice were terminated at 12 mo, and a longitudinal section of each kidney was fixed in 10% neutral buffered formalin (Sigma-Aldrich), paraffin embedded, cut into 3-μm sections, and stained with H&E with periodic acid-Schiff. The sections were examined in a blind manner for evidence of pathologic changes in the glomeruli, tubules, or interstitial areas. The glomeruli were screened for evidence of hypertrophy, proliferative changes, crescent formation, hyaline deposits, fibrosis/sclerosis, and basement membrane thickening. The glomerulonephritis (GN) severity was graded on a scale of 0–4, where 0 = normal, 1 = mild increase in mesangial cellularity and matrix, 2 = moderate increase in mesangial cellularity and matrix with thickening of glomerular basement membrane, 3 = focal endocapillary hypercellularity with obliteration of capillary lumina and a substantial increase in the thickness and irregularity of the glomerular basement membrane, and 4 = diffuse endocapillary hypercellularity, segmental necrosis, crescents, and hyalinized end-stage glomeruli. Similarly, the severity of tubulointerstitial nephritis was graded on a 0–4 scale based on the extent of tubular atrophy, inflammatory infiltrates, and interstitial fibrosis, as described previously (16). The number of infiltrating polymorphonuclear leukocytes was directly enumerated based on their typical polymorphonuclear morphology by examining 50 glomeruli/kidney/section/mouse.
Serology
Mice were bled at 5, 7, 9, and 12 mo and sera stored at –20°C. ELISA detection of IgM and IgG autoantibodies directed against chromatin and dsDNA were performed as described previously (9). All coating steps were for 30 min at 37°C followed by two washes in PBS (Sigma-Aldrich). Briefly, for both the anti-dsDNA and anti-chromatin ELISAs, Immunlux HB (Dynatech Laboratories) plates were precoated with 50 μl/well methylated BSA and then coated with 50 μl/well of 50 μg/ml dsDNA (dissolved in PBS and filtered; Sigma-Aldrich). For the anti-chromatin ELISAs, an additional coating step with 50 μl/well of 10 μg/ml total histones (Roche) was used. Following the last respective coating step and wash, ELISA plates were incubated overnight at 4°C with 200 μl/well ELISA blocking buffer (PBS, 0.1% gelatin, 3% BSA, and 3 mM EDTA). Test sera were added at a final dilution of 1/200 and 1/800 (serial) for anti-dsDNA and anti-chromatin ELISAs, respectively, in serum diluent (PBS, 0.1% gelatin, 2% BSA, 3 mM EDTA, and 0.05% Tween 20) and incubated for 2 h at room temperature. Bound IgM or IgG was detected using alkaline phosphatase-conjugated goat anti-mouse IgG or IgM (Roche) and p-nitrophenyl phosphate (Sigma-Aldrich) as the substrate. OD450 was measured by an Elx800 Automated Microplate Reader (Bio-Tek Instruments) and the raw optical densities for anti-chromatin ELISAs converted to arbitrary normalized units using a six-point standard curve generated by a mAb derived from a NZM2410 mouse (17). A 1/250 dilution of this supernatant was set at 1000 U/ml. For anti-dsDNA ELISAs, the standard curve was set using six serial dilutions of sera from a female (NZW x B6.Sle1)F1, with a 1/250 dilution of this sera set at 1000 U/ml.
Cell preparation and culture
Following the measurement of spleen weight, single-cell splenocyte suspensions were prepared, passed through nylon mesh, and depleted of RBCs using RBC lysis buffer (0.1 mM EDTA, 0.83% NH4Cl, and 0.1% KHCO3). These cells were then used for flow cytometric analyses (described below).
Flow cytometric analysis and Abs
In each FACS experiment, at least one mouse per genotype (B6, B6.Sle1, B6.Sle1 Sles1) was included. For all FACS analyses, B6.Sle1 Sles1 indicates that the B6.Sle1 Rec.1 z/z line was used. Cells were blocked with 2.4G2 (American Type Culture Collection) on ice. Cells (1.5 x 106/Ab mixture) were then stained on ice using optimal amounts of FITC, PE, PE-Texas Red, Cy-Chrome, PerCPCy5.5, allophycocyanin, or biotin-conjugated Abs at predetermined dilutions. Four-color combinations of the following Abs, obtained from BD Biosciences-Pharmingen were used for the various analyses: CD21/35 (7G6); CD23 (B3B4); IgM (R6-60.2); CD86 (GL1); CD45R (RA36B2); CD5 (53-7.3); CD11b (M1/70); NK1.1 (PK126); CD25 (PC61); CD69 (H1.2F3); CD4 (H129.19); CD8 (53-6.7); CD3 (145–2C11); CD19 (1D3); CD43 (S7); CD1d (1B1); CD62L (MEL-14); CD44 (1M7); and CD138 (281-2). After two washes, biotin-conjugated Abs were revealed using BD Pharmingen streptavidin-allophycocyanin. Appropriate isotype controls for all Ab combinations were included in all analyses for all samples. Stained cells were acquired on a FACSCalibur with CellQuest software (BD Biosciences). Dead cells were excluded on the basis of forward and side scatter properties, and 40,000 events within the lymphocyte gate were acquired per sample. Flow cytometric data was analyzed using FlowJo (Tree Star).
SDS-PAGE and immunoblotting
Splenic B lymphocytes were isolated using Dynal anti-B220 magnetic beads according to the manufacturer’s protocol (Dynal Biotech). Isolated cells were lysed in lysis buffer (300 mM NaCl, 50 mM Tris-Cl (pH 7.6), 0.5% Triton X-100, 10 μg/ml leupeptin, 10 μg/ml aprotinin, 1 mM PMSF, 1 mM sodium orthovanadate, 1 mM sodium molybdate, and 1 mM NaF), incubated on ice for 30 min, and then cleared by centrifuging at 12,000 x g for 20 min. Equivalent amounts of total cellular protein extract (4 μg) were fractionated on 10% SDS-PAGE and electrotransferred to polyvinylidene difluoride membrane using a Bio-Rad Transblot apparatus. The membrane was blocked overnight with TBST buffer (10 mM Tris-Cl (pH 8.0), 0.9% NaCl, and 0.1% Tween 20) plus 4% BSA (for phosphorylated proteins) or 5% nonfat dry milk and incubated for 1 h with primary mAb in TBST buffer plus 4% BSA (for phosphorylated proteins) or 5% nonfat dry milk. Following three washes in TBST, membranes were incubated with a secondary Ab for 1 h. The membranes were again washed three times in TBST and then developed using an ECL Plus kit (Amersham Biosciences) according to the manufacturer’s instructions. Protein bands were quantified by densitometric analysis using a computerized densitometer (Molecular Dynamics) and ImageQuant software (Molecular Dynamics). Rabbit polyclonal Abs specific for P44/42 MAPK were purchased from Cell Signaling Technology. Rabbit anti-phospho-ERK1/2 MAPKs polyclonal Abs were purchased from Promega. The secondary Abs (goat anti-rabbit IgG) were purchased from Santa Cruz Biotechnology.
Statistical analysis
Data was analyzed using parametric and nonparametric ANOVAs as determined by InStat3 (Graphpad Software), unless otherwise indicated. Error bars represent SEMs unless data are not normally distributed, in which case bars reflect the interquartile ranges.
Results
Sles1 maps to a 956-kb interval that excludes TNF-
Our previous work demonstrated the presence of a recessive NZW-derived epistatic modifier on mouse chromosome 17, termed Sles1, which specifically suppressed the break in tolerance to chromatin mediated by Sle1 on the B6 background. Consequently, congenic recombinant fine mapping was performed by intercrossing B6.Sle1 Sles1 het mice and identifying recombinant chromosomes in the Sles1 interval by PCR screening tail DNA from testcross progeny with polymorphic microsatellite markers flanking the original Sles1 interval. Once recombinant chromosomes were identified, additional markers were used to localize the recombinant breakpoint. The recombinant mice were backcrossed to B6.Sle1 to "rescue" the recombinant Sles1 chromosomes and generate new Sles1 recombinant lines. The new recombinant lines found to retain the recessive suppression phenotype of Sles1 were used to generate more recombinant strains. As shown in Fig. 1A, a series of six truncated recombinant chromosome lines, spanning the Sles1 interval and termed B6.Sle1 Rec.1-6, were identified via the analysis of 1329 testcross progeny. Interestingly, there appeared to be a recombinational hot spot between the markers D17MIT146 and D17MIT175, as evidenced by the greater than expected number of recombinants obtained between those two markers (2 = 22.3; p < 0.005).
Progeny from these lines, homozygous and heterozygous for the recombinant chromosomes, were aged and assayed for IgG anti-chromatin auto antibody production by ELISA at 5, 7, 9, and 12 mo of age. At the 12-mo terminal sacrifice, spleens were weighed to assess splenomegaly, a phenotype observed in aged B6.Sle1 mice (4, 9, 18).
As depicted in Fig. 2A, aged B6.Sle1 female mice have a significant increase in their mean spleen weights when compared with age- and sex-matched B6 (233.3 mg, interquartile range 163.9–280.2 mg vs 107.4 mg, interquartile range 99.7–127.1 mg); p < 0.001, nonparametric ANOVA). Comparison of spleen weights from the various homozygous recombinant lines with B6 and B6.Sle1 indicated that B6.Sle1 Rec.2 z/z and B6.Sle1 Rec.3 z/z mice differed significantly from B6 (p < 0.001; nonparametric ANOVA), whereas B6.Sle1 Rec.1 z/z, B6.Sle1 Rec.4 z/z, B6.Sle1 Rec.5 z/z, and B6.Sle1 Rec.6 z/z mice were indistinguishable from B6 (p > 0.05) and different from B6.Sle1 (p < 0.001 for all strains; nonparametric ANOVA). Importantly, mice heterozygous for all six recombinant chromosomes revert to the 2-fold increase in spleen weights seen in B6.Sle1, which is consistent with the recessive inheritance of Sles1 observed in the original bi-congenic analyses (4). These results map the suppression of splenomegaly into the genomic interval contained within Rec.1 and Rec.4-6.
Suppression of antinuclear autoantibody (ANA) production by Sles1 was assayed by anti-chromatin ELISAs at various time points. As shown in Fig. 2B, in which titers from 9- to 12-mo-old female mice of the various genotypes are presented, B6.Sle1 Rec.1 z/z, B6.Sle1 Rec.4 z/z, B6.Sle1 Rec.5 z/z, and B6.Sle1 Rec.6 z/z all suppress anti-chromatin ANA production (p < 0.001 for all strains compared with B6.Sle1 females; nonparametric ANOVA), hence mapping Sles1 to these genomic segments. In contrast, B6.Sle1 Rec.2 z/z and B6.Sle1 Rec.3 z/z failed to suppress anti-chromatin ANA production (p < 0.001 for all strains compared with B6; nonparametric ANOVA). These results correlate with the lack of splenomegaly in the different congenic lines, consistent with a single gene mediating both effects.
These data place Sles1 in the 956-kb genomic interval contained in Rec.6 (Fig. 1A). This interval includes the microsatellite markers D17MIT28 and D17MIT83 but excludes D17MIT62 and D17MIT13 at the proximal and distal ends respectively and contains 65 genes (Fig. 1). Using public sequence available from the Ensembl database (National Center for Biotechnology Information m33), we designed primers surrounding microsatellite repeats at the proximal (D17MIT62-D17MIT28) and distal breakpoints (D17MIT83-D17MIT13) to further map the ends of this interval. The 64-kb proximal and 27-kb distal breakpoints (between the primer pairs ms3-ms4 and ms7-ms8, respectively) harbor two additional genes each (www.ensembl.org/Mus_musculus/). As shown in Fig. 1B, this brings the maximum number of candidate genes in the minimal Sles1 interval to 69. This result is consistent with the gene-rich nature of this region of the genome, as has been recently reviewed (19, 20). In addition, given the recognized immunological functions of many of these genes, determining candidacy for Sles1 remains a challenge.
Sles1 suppresses the activation of T cells mediated by Sle1
The breach in tolerance to chromatin mediated by the Sle1 locus is accompanied by an increase in the activation phenotypes in splenic T and B lymphocytes (4, 9, 14). We performed a comprehensive analysis of the various splenic cell populations in ages 9- to 12-mo-old female B6, B6.Sle1, and B6.Sle1 Sles1 z/z (B6.Sle1 Rec.1 z/z) mice to better characterize these changes and determine their association with the suppression mediated by the Sles1 interval.
As shown in Table II, there were numerous quantitative differences in the percentages of various splenic subpopulations when B6.Sle1 mice were compared with age and gender-matched B6 controls. A small but significant decrease (p < 0.05; ANOVA) in the percentage of CD3+ T cells in B6.Sle1 mice was restored to B6 levels in the presence of Sles1 (Table II). As has been shown previously, B6.Sle1 mice had a highly significant increase in the percentage of CD3+ T cells expressing the early activation marker CD69, when compared with B6 controls (9). This activation phenotype was suppressed by Sles1 (Fig. 3A), as shown by the decrease in the percentage of CD69+ CD3 T cells in B6.Sle1 Sles1 vs B6.Sle1 mice (p < 0.001; ANOVA). The overall cell surface levels of CD69 on the CD3+ population, as determined by comparing median fluorescence intensities (MFIs), was also significantly higher in B6.Sle1 mice (p < 0.001; ANOVA), which also was restored to B6 levels by Sles1 (Fig. 3B; Table II).
We also investigated the splenic B1 and B2 B cell populations in these strains to look for potential changes in the splenic B1 population, as observed in some lupus models (27, 28). As shown in Table II, there were no significant differences in the splenic B1 lineages in any of the groups examined, but there was a decrease in the B2 population (CD5–CD23+B220+) in B6.Sle1 mice compared with B6 controls (29.4 ± 1.7 vs 38.3 ± 1.0%; p < 0.01; ANOVA), similar to the decrease seen in the follicular B cell population discussed above. Again, there was no change in B6.Sle1 Sles1 mice compared with both B6 and B6.Sle1.
Similar to what was observed in the T cell compartment, there was also an increase in the percentage of B cells in B6.Sle1 mice expressing the activation marker CD69. In Fig. 4B, the percentage of B220 B cells that expressed CD69 is shown for the various strains, which increased from 2.7 ± 0.3% in B6 to 6.8 ± 1.1% in B6.Sle1 mice (p < 0.001; ANOVA). This increase in the activation status of the B cell compartment was also reflected in the increased cell surface levels of the costimulatory molecule CD86 (B7.2) on follicular B cells in B6.Sle1 mice, as determined by comparisons of MFIs (Table I; p < 0.05; repeated measures ANOVA). Neither change was seen in B6.Sle1 Sles1 mice, in which the percentage of B220 B cells expressing CD69 is only 1.5 ± 0.2% (p < 0.001 vs B6.Sle1; ANOVA), and the MFI of CD86 on follicular B cells was correspondingly decreased (Table I). These results indicate that the suppression of the activation status of lymphocytes introduced by Sles1 affects both the T and B cell lineages, such that their splenic cell surface profiles are comparable to B6.
We recently demonstrated that ex vivo B cells from B6.Sle1 mice display an intrinsic increase in phosphorylated ERK2 ex vivo by 4 mo of age (26). Based on this result, we compared splenic B cells from B6.Sle1 Sles1 mice to determine whether the intrinsic ERK2 phosphorylation observed in B6.Sle1 was affected by the presence of Sles1. As shown in Fig. 4C, ex vivo splenic B cells from 10-mo-old B6.Sle1 Sles1 mice show an increase in the levels of phosphorylated ERK2 comparable to that seen in age- and gender-matched B6.Sle1 B cells. The persistence of this B cell activation phenotype could suggest that B cells in B6.Sle1 Sles1 mice are not receiving adequate T cell help for driving class switch to IgG anti-chromatin autoantibody production. However, ELISA analyses for both IgM and IgG anti-chromatin autoantibody production by B6.Sle1Sles1 mice at different ages indicate otherwise, as neither isotype anti-chromatin autoantibodies are present in B6.Sle1 Sles1 (data not shown). These findings suggest that the suppressive Sles1 allele acts downstream of these activated molecular pathways and that hyperphosphorylation of ERK2 can be uncoupled from the humoral autoimmunity elicited by Sle1.
Genetic complementation suggests a 129 Sles1 allele
Recent work by our group demonstrated that the break in tolerance to chromatin mediated by the Sle1 subcongenic locus, Sle1b, is associated with functional polymorphisms in the SLAM/CD2 family of immune receptor genes and that these polymorphisms are found in the majority of inbred strains examined, including the nonautoimmune 129/SvJ (129) (29). Furthermore, introgressing this chromosome 1 interval from 129 onto B6 background was sufficient to cause a loss in tolerance in chromatin, resulting in the production of serum ANAs (29, 30). This raised the question of why the 129 strain remains nonautoimmune, despite the presence of a potent Sle1-related SLAM/CD2 haplotype on chromosome 1. We hypothesized that, similar to what we previously observed in the NZW strain, epistatic modifiers might play a role in the lack of autoimmunity seen in 129.
To test whether one of the modifiers in 129 was a Sles1-like allele, we performed a classic genetic complementation test, comparing the autoimmunity elicited in (129 x B6.Sle1)F1s to that seen in (129 x B6.Sle1 Sles1)F1s, by crossing 129 to B6.Sle1 and B6.Sle1 Rec.4 z/z (B6.Sle1 Sles1). These comparisons are similar to the original studies performed with NZW in which it was shown that NZW homozygosity at Sles1 is sufficient to prevent systemic autoimmunity in the (NZW x B6.Sle1)F1 hybrid system.
The progeny of the described crosses with 129 are both homozygous for the autoimmunity-promoting SLAM/CD2 haplotype 2 on chromosome 1, whereas the remainder of the genome is heterozygous for 129 and B6 alleles (29). However, these two crosses differ at Sles1 in that the (129 x B6.Sle1 Sles1)F1s carry one copy of the NZW-derived Sles1 allele (31). Because heterozygosity of the Sles1 allele does not lead to suppression (see Fig. 2), we expect that if the 129- and B6-derived Sles1 loci are equivalent in their ability to potentiate autoimmunity mediated by the SLAM/CD2 family haplotype 2, then both sets of F1 intercross progeny would have comparable autoimmunity. However, if the 129 strain has a protective Sles1 allele, complementary to that of NZW, then the latter F1 intercross progeny would be expected to have abrogated autoimmunity.
As shown in Fig. 5A, the degree of moderate to severe GN, defined as being a GN score of 3, in 9-mo-old female (129 x B6.Sle1)F1s was significantly higher than that observed in both (129 x B6.Sle1 Sles1)F1s and (129 x B6)F1s (p < 0.01 and p < 0.001, respectively; ANOVA). In addition, the (129 x B6.Sle1 Sles1)F1 group was the only group with no evidence of lymphocyte infiltration of the kidneys (data not shown). Similarly, when spleen weights from the different mice were compared (Fig. 6B), the (129 x B6.Sle1)F1s exhibited significant splenomegaly (p < 0.001 vs B6; nonparametric ANOVA) whereas (129 x B6.Sle1 Sles1)F1s did not. In fact, as clearly shown in Fig. 5B, the only congenic lines with significant differences in spleen weights from B6 were those that have one or more copies of the B6 allele at the Sles1 locus, when Sle1 is homozygous. These data strongly support the presence of a complementary Sles1 modifier in the 129 genome.
The quantification of IgG autoantibody levels in the different strains further support this conclusion. As shown in Fig. 6, A and B, only the (129 x B6.Sle1)F1s had significant levels of serum anti-dsDNA IgG autoantibodies by 9 mo of age (vs B6, p < 0.01). In contrast, the penetrance of anti-dsDNA autoantibodies in (129 x B6.Sle1 Sles1)F1s did not differ significantly from age- and gender-matched B6 and 129 controls.
We also assayed for the more nonspecific anti-chromatin autoantibodies in the various strains. As shown in Fig. 6, C and D, at 9 mo of age both the titers and penetrance of serum IgG ANAs in the (129 x B6.Sle1)F1s were highly increased when compared with any of the control groups (p < 0.001; ANOVA). Interestingly, unlike anti-dsDNA autoantibodies, the (129 x B6)F1s themselves had an increased (50%) penetrance of anti-chromatin autoantibodies. In contrast to either of the other F1 groups, the (129 x B6.Sle1 Sles1)F1s had minimal serum anti-chromatin autoantibodies at 9 mo of age (Fig. 6, C and D). Significant differences in the titers and penetrance of anti-chromatin autoantibodies were discernible as early as 5 mo of age in (129 x B6.Sle1)F1s (data not shown).
These data indicate that the 129-derived Sles1 allele specifically complements the NZW Sles1 allele in preventing both the more pathogenic phenotypes, such as GN development and anti-dsDNA autoantibodies, as well as the development of the more benign serum anti-chromatin autoantibodies. Overall, these results are extremely similar to those originally obtained in the analysis of the (NZW x B6.Sle1 Sles1)F1 hybrid system (4). However, the presence of Sles1 in 129 is especially intriguing because classic analyses of the H2 in 129 have defined it as an H2b haplotype very closely related, but not identical, to the H2b haplotype carried by B6.
Discussion
The present study extends our previous work characterizing the suppression of autoimmunity mediated by Sles1 in the bicongenic B6.Sle1 Sles1 system. Our congenic fine mapping localizes Sles1 into an interval on chromosome 17 that encompasses 956 kb of the proximal portion of the murine MHC. Our current minimal Sles1 interval still contains 69 genes, consistent with the gene-rich nature of this segment of the genome (19, 20). Many of these positional candidates are known to possess immunological functions and are also highly polymorphic between the B6 and NZW strains (31, 32). Our genetic fine-mapping analysis definitively excludes TNF-, a known lupus susceptibility candidate gene and one implicated in other murine lupus models, as well as the genes distal to it, as possible Sles1 candidates (33, 34, 35, 36). As shown in the schematic in Fig. 7, this leaves many other potentially interesting candidates, including the class II genes H2-Ab1, Aa, Eb, and Ea, the complement genes proximal to TNF-, as well as other relevant genes such as Daxx, Notch4, and Hspa1a. Interestingly, unlike many other loci implicated in lupus and other autoimmune disease susceptibility, fine mapping of Sles1 did not reveal the presence of multiple subloci contributing to suppression (14, 37, 38).
In contrast to other autoimmune disorders, such as insulin-dependent diabetes mellitus and multiple sclerosis, the association of the class II MHC genes with susceptibility to SLE has proven to be very complicated and highly dependent on epistatic interactions with other susceptibility loci, as demonstrated in a variety of linkage studies in both human SLE patients and murine lupus models. Earlier studies using the classic (NZB x NZW]F1 (BWF1)) model, from which NZM2410 is derived, in various backcross analyses have suggested that the NZW allele at the H2 (H2z) confers a high risk of developing renal disease, and in particular, it is the NZW class II alleles that are the dominant contributors to disease in these models (39). Many of these studies have documented that heterozygosity at the H2 gene complex is crucial for lupus susceptibility and is associated with many of the phenotypic characteristics of lupus, such as GN development, autoantibody production, and splenomegaly (40). However, many such backcross studies have also indicated that homozygosity of the NZW alleles at the H2 region is protective. These somewhat contradictory data have led to a variety of hypotheses such as an increased self-recognition in heterozygotes due to mixed haplotype or mixed isotype class II molecules, suppressive effects caused by increased class II I-E expression, and competition by peptides derived from I-Ea resulting in decreased self-peptide presentation by class II I-A molecules during thymic development leading to increased autoimmune susceptibility (41, 42, 43, 44, 45). A major limitation in interpreting the role of H2 loci in these backcross studies is that the remainder of the genome also contains susceptibility alleles that are randomly segregating. Considering the importance of epistatic interactions in the development of SLE, this heterogeneous segregation complicates the unambiguous association of phenotypes with a specific locus.
The use of transgenics for the various NZW class II alleles (I-Ez and I-Az) on nonautoimmune backgrounds have also been used in various backcross studies to determine the contributions of these alleles to lupus susceptibility and have shown that transgenic expression of the class II molecules does not recapitulate what is seen when congenics for the entire NZW MHC are used in similar crosses. This suggests that nonclass II molecules in the MHC, in tight linkage disequilibrium with class II genes, may be the ultimate culprits in modulating disease susceptibility (46, 47). Recently, Zhang et al. (48) used intra-MHC recombinants to try and elucidate the contribution of variances in the level of I-E expression, when the I-A molecule haplotype is held constant, in a series of F1 crosses. They concluded that expression of the class II I-E molecule offers protection in a dose-dependent manner and postulate that this may be due to mechanisms including thymic deletion of a self-reactive TCR repertoire, determinant capture, differences in cytokine balances and competition for signal transduction moities. However, the recombinant breakpoint in their intra-MHC recombinant, H2-g2r, is defined only as occurring somewhere between H2-Ea and TNF-, a region containing >60 genes (www.ensembl.org/Mus_musculus/) (Ref. 48 ; Fig. 1B). It could hence also be concluded that a gene within this undefined interval, which partially overlaps with our minimal Sles1 interval, is responsible for the dose-dependent protection seen in their studies.
In addition to class II genes, the complement genes within the MHC have long been implicated in lupus susceptibility. Deficiencies in the complement components C1 and C4 are in fact the only known single gene deficiencies that are so strongly associated with human SLE (1). Furthermore, during lupus disease progression, acquired deficiencies in complement components are common and are presumed to result from the immune complex use of complement. There are two main, nonmutually exclusive theories that have been proposed to explain the link between complement deficiency and susceptibility to lupus. One involves the role of complement in the clearance of apoptotic cells and immune complexes and the other invokes the known ability of complement to determine activation thresholds in B and T lymphocytes (49). Interestingly, ablation of the C4 gene, located at the peak marker used in the original linkage studies describing Sles1, has been demonstrated to alter negative selection of immature self-reactive B cells in both the double-transgenic hen egg lysosome and the lpr null models (4, 50). Furthermore, unlike other complement deficiencies (C1qa, C3, and Cr1/2), the lack of C4 appears less dependent on a mixed 129 x B6 background for the manifestation of some degree of autoimmunity (7). However, serum levels of this complement component are equivalent in young, preautoimmune B6, B6.Sle1 and B6.Sle1 Sles1 mice (data not shown).
Other candidate genes in the Sles1 interval includes the MHC proximal gene Daxx, known to play an important role in the IFN--mediated inhibition of B lymphopoiesis, which is interesting considering the large number of recent studies implicating type I IFNs in lupus susceptibility (51, 52, 53, 54, 55, 56, 57). Heat shock proteins are known to have effects on the maturation of APCs in a manner reminiscent of cytokines, and heat shock protein 70, encoded by Hspa1a on the distal end of the minimal Sles1 interval, has been shown to be capable of converting T cell tolerance to autoimmunity in vivo (58, 59). Notch4 is also contained within the Sles1 interval, and given the recent wave of papers implicating members of the Notch family of receptors and ligands in regulating B and T lymphocyte development and function, it remains an attractive candidate for Sles1 (60, 61, 62, 63, 64, 65, 66, 67, 68, 69).
Our phenotypic analyses of the cell surface characteristics of the splenic immune cells from B6, B6.Sle1, and B6.Sle1 Sles1 mice reveal that many, but not all, of the differences seen in 9- to 12-mo-old B6.Sle1 splenocytes are suppressed in the presence of Sles1. Aged B6.Sle1 mice, despite their relatively benign autoimmune phenotype of ANA production, have significantly increased populations of activated B (B220+CD69+) and T (CD4+CD62L–CD44high; CD3+CD69+) cells, similar to those observed in much more severe lupus models (15, 21, 22, 23, 25, 70). It has been postulated that the spontaneous lymphocyte activation seen in these different lupus models is due to differences in thresholds for activation and immune regulation caused by the underlying genetic polymorphisms in relevant susceptibility loci (29, 71, 72). Intriguingly, both T and B cells from B6.Sle1 Sles1 mice have a significantly reduced activation phenotype, indistinguishable from B6, suggesting that Sles1 is impacting both lymphocyte compartments. This suggests that a lack of T cell help to the B cells is not the sole mechanism of suppression. This is further supported by the observation that B6.Sle1 Sles1 mice do not produce serum IgM autoantibodies either, indicating that lack of class-switching is not the reason these mice do not produce high titers of serum IgG autoantibodies. However, ex vivo B cells from B6.Sle1 Sles1 mice demonstrate a comparable increase in phosphorylated ERK2 to those from B6.Sle1, indicating that this activated molecular pathway can be uncoupled from the humoral autoimmunity elicited by Sle1.
Given the density and polymorphic nature of the genes within the Sles1 interval, definitive identification of the Sles1 gene has proven to be quite a challenge. However, our F1 analyses with the nonautoimmune 129 strain provide an alternative approach that may significantly accelerate this analysis. Because B6 and 129 have very closely related H2 haplotypes, most notably identity for class II I-A and I-E alleles, the degree of variability between the B6 and 129 alleles will be much reduced in comparison to B6 vs NZW. Consequently, we will seek variations in structure/function of candidate genes between B6 and 129, which should enhance our ability to quickly focus on relevant polymorphisms. Furthermore, this result excludes the candidacy of MHC class II molecules for Sles1.
The use of congenic recombinants has now been used to fine-map both susceptibility and suppressive loci in efforts to identify the underlying genes (14, 29). These congenic systems allow individual key susceptibility and modifier loci to be specifically combined so that phenotypes resulting from their genetic interactions can be studied. Hence, the ability of Sles1 to specifically target Sle1 could help elucidate the key epistatic interactions involved in the generation of severe systemic autoimmunity in lupus models where Sle1 has been shown to be necessary for initiation and development (13, 15, 23). Identification of Sles1, using both recombinant fine-mapping strategies and comparisons of B6, 129, and NZW alleles, will have important implications for our understanding of and ability to therapeutically intervene in SLE pathogenesis.
Acknowledgments
We thank Nisha Limaye, Alice Chan, and Katherine Belobrajdic for critical reading and helpful discussions. We also express gratitude to Dr. Jose Casco for excellent management of the mouse colony.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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.
1 This work was supported by a grant from the National Institutes of Health Grant 5R37AI04519607 (to E.K.W.).
2 Address correspondence and reprint requests to Dr. Edward K. Wakeland, Center for Immunology, University of Texas Southwestern Medical Center, 6000 Harry Hines Boulevard, Dallas, TX 75390-9093. E-mail address: edward.wakeland@utsouthwestern.edu
3 Abbreviations used in this paper: SLE, systemic lupus erythematous; NZW, New Zealand White; ANA, antinuclear autoantibody; GN, glomerulonephritis; MFI, median fluorescence intensity; MZ, marginal zone.
Received for publication March 7, 2005. Accepted for publication May 5, 2005.
References
Wakeland, E. K., K. Liu, R. R. Graham, T. W. Behrens. 2001. Delineating the genetic basis of systemic lupus erythematosus. Immunity 15: 397-408.
Nguyen, C., N. Limaye, E. K. Wakeland. 2002. Susceptibility genes in the pathogenesis of murine lupus. Arthritis Res. 4:(Suppl. 3): S255-S263.
Raman, K., C. Mohan. 2003. Genetic underpinnings of autoimmunity: lessons from studies in arthritis, diabetes, lupus, and multiple sclerosis. Curr. Opin. Immunol. 15: 651-659.
Morel, L., X. H. Tian, B. P. Croker, E. K. Wakeland. 1999. Epistatic modifiers of autoimmunity in a murine model of lupus nephritis. Immunity 11: 131-139.[
Santiago-Raber, M. L., B. R. Lawson, W. Dummer, M. Barnhouse, S. Koundouris, C. B. Wilson, D. H. Kono, A. N. Theofilopoulos. 2001. Role of cyclin kinase inhibitor p21 in systemic autoimmunity. J. Immunol. 167: 4067-4074.
Bolland, S., Y. S. Yim, K. Tus, E. K. Wakeland, J. V. Ravetch. 2002. Genetic modifiers of systemic lupus erythematosus in FcRIIB–/– mice. J. Exp. Med. 195: 1167-1174.
Paul, E., O. O. Pozdnyakova, E. Mitchell, M. C. Carroll. 2002. Anti-DNA autoreactivity in C4-deficient mice. Eur. J. Immunol. 32: 2672-2679.(Srividya Subramanian*, Yo)