Identification of new quantitative trait loci in mice with collagen-induced arthritis.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 50, No. 11, November 2004, pp 3721–3728 DOI 10.1002/art.20624 © 2004, American College of Rheumatology Identification of New Quantitative Trait Loci in Mice With Collagen-Induced Arthritis Kristin Bauer, Xinhua Yu, Patrik Wernhoff, Dirk Koczan, Hans-Juergen Thiesen, and Saleh M. Ibrahim Objective. Collagen-induced arthritis (CIA) in the mouse is one of the most widely used autoimmune experimental models, with many features similar to rheumatoid arthritis. This study sought to identify potential genetic regulatory mechanisms of CIA in major histocompatibility complex–matched (H2-q) F2 hybrid mice. Methods. We used 126 polymorphic markers to perform simple sequence-length polymorphism analysis on 290 F2 hybrids of arthritis-susceptible (DBA/1J) and arthritis-resistant (FVB/N) inbred mouse strains. The major clinical traits (disease severity and onset) were assessed, and serum antibodies specific to type II collagen (CII) were determined by enzyme-linked immunosorbent assay in 270 F2 mice. Lymph nodes from 94 F2 mice were used to test the ratio of CD4 to CD8 by fluorescence-activated cell sorter analysis, and cell proliferation was determined by XTT test. Results. Two quantitative trait loci (QTLs) identified in previous studies were confirmed; these were severity-controlling Cia2 and onset-controlling Cia4 on chromosome 2. Moreover, we identified 5 new QTLs, 1 for CII-specific IgG2a antibodies on chromosome 5, 2 controlling the CII-specific IgG1 antibody response on chromosomes 10 and 13, 1 for the CD4:CD8 ratio on chromosome 2, and 1 for cell proliferation (measured by XTT test) on chromosome 16. Complement component C5 was identified as the probable main candidate gene for the QTLs Cia2 and Cia4. F2 mice carrying a 2-basepair deletion of C5, the FVB/N allele, had low incidence and less severe disease as compared with those carrying the DBA/1J allele. Conclusion. This genome scan provides additional evidence confirming the role of C5 as a probable candidate gene for Cia2 and Cia4 loci, and identifies new QTLs controlling new traits in autoimmune arthritis. Rheumatoid arthritis (RA) is a complex, multifactorial polygenic disease that is influenced by age and sex as well as hormonal and environmental factors (1). Susceptibility to RA is controlled by genetic factors, as indicated by many studies showing higher rates of disease concordance in monozygotic twins than in dizygotic twins, and higher incidence in offspring of RA patients (2). The genetic contribution to RA susceptibility is estimated to be as much as 60%, of which the HLA– DRB1 locus is thought to account for 30–50% (3,4). Identification of genetic loci regulating RA is complicated by genetic heterogeneity and incomplete penetrance, as well as environmental factors (1). Thus, genetic analysis of well-defined experimental models of autoimmune arthritis has the potential to markedly accelerate the genetic investigation of RA. Experimental models of arthritis, such as collagen-induced arthritis (CIA) in mice, have been used extensively for studying the roles of autoimmunity and inflammation in the pathogenesis of arthritis. As is the case in humans, most studies indicate that both non– major histocompatibility complex (MHC) genes and MHC genes are associated with the susceptibility to CIA. Indeed, the main quantitative trait locus (QTL) identified, Cia1, is located on chromosome 17 and includes the MHC. However, susceptibility/severity of the disease varies significantly among inbred strains and even among those sharing the same MHC haplotype. For example, inbred DBA/1J mice are markedly suscep- Supported by grants from EUROME (EU FP5) and BMBF (FKZ 01ZZ0108). Kristin Bauer, MSc, Xinhua Yu, MSc, Patrik Wernhoff, PhD, Dirk Koczan, PhD, Hans-Juergen Thiesen, MD, Saleh M. Ibrahim, MD, PhD: Institute for Immunology, University of Rostock, Rostock, Germany. Ms Bauer and Mr. Yu contributed equally to this work. Address correspondence and reprint requests to Saleh M. Ibrahim, MD, PhD, Institute of Immunology, Rostock University, Schillingallee 70, 18055 Rostock, Germany. E-mail: saleh.ibrahim@ med.uni-rostock.de. Submitted for publication February 12, 2004; accepted in revised form August 10, 2004. 3721 3722 BAUER ET AL tible to CIA, whereas FVB/N mice are completely resistant; both strains carry the H2-q haplotype. The conservative estimate for the quantification of the non-MHC susceptibility genes is suggested to be more than 20. Recently, a few studies localized QTLs that regulate CIA in mice, mostly in crosses involving the DBA/1J strain (5,6). These QTLs (Cia2, Cia3, Cia4, Cia6, Cia7, and Cia8) are located on chromosomes 2, 6, 7, and 10. Other QTLs identified in crosses involving other strains are located on chromosomes 1, 3, 5, 8, 13, 14, 17, 18, 19, and X (7–9). We assume, along with others, that there are additional QTLs contributing to the susceptibility to CIA and to additional disease traits that have not been identified. We therefore set out to identify these QTLs in a genome scan of the F2 generation between the FVB/N and DBA/1J inbred mouse strains. Herein we report the identification of QTLs controlling the severity of disease, onset of disease, autoantibody response, CD4:CD8 ratio, and cell proliferation (measured by XTT test) in mice with CIA. MATERIALS AND METHODS Mice, immunization, and scoring. All animals used in this study were obtained from The Jackson Laboratory (Bar Harbor, ME) and were housed at the animal facility at the University of Rostock. All procedures and assays were preapproved by the local Animal Care Committee. CIA was induced in control and experimental animals according to established protocols. In brief, DBA/1J, FVB/N, (DBA/1J ⫻ FVB/N)F1, and (DBA/1J ⫻ FVB/N)F2 progeny were immunized at 8–12 weeks of age at the base of the tail with 125 g of bovine type II collagen (CII; Chondrex, Redmond, WA) dissolved in 50 l of 0.1M acetic acid and mixed with an equal volume (50 l) of Freund’s complete adjuvant (Freund’s incomplete adjuvant with 4 mg/ml Mycobacterium tuberculi; Difco, Detroit, MI). Mice were followed up for 14 weeks postimmunization. The clinical scoring of arthritis was commenced from 18 days after immunization. Animals were monitored 3 times weekly for signs of CIA. An arthritis index was assigned to each mouse by using the following criteria: 0 ⫽ no signs of arthritis; 1 ⫽ swelling and redness in a single joint; 2 ⫽ inflammation in multiple joints; and 3 ⫽ severe swelling, joint erosion, and/or ankylosis. Each paw was scored on a scale of 0–3, with the index being the sum of the scores for all 4 paws. The severity trait was defined as the maximum arthritis index score observed in each individual mouse. The onset trait was calculated as described previously (6). In brief, the earliest day of observed onset after immunization was given a score of 50 and the latest day of observed onset was given a score of 0. The onset score for each of the remaining mice was calculated as follows: 50 ⫻ (day after latest observed onset day ⫺ the observed onset day)/(day after latest observed onset day ⫺ day after the earliest observed onset day). Healthy mice were given an onset score of 0. Genomic screening. For the genetic analysis, 126 informative microsatellite markers covering the genome to the Table 1. Summary of arthritis susceptibility in DBA/1J, FVB/N, F1, and F2 mice* Strain Incidence, no. (%) Onset, days Maximum severity score DBA† FVB† F1 F2 15/15 (100) 0/15 (0) 13/60 (21.7) 85/290 (29.3) 39.1 ⫾ 11.7 – 62.1 ⫾ 18.1 44.5 ⫾ 19.9 7.91 ⫾ 3.08 – 3.15 ⫾ 2.99 4.51 ⫾ 3.48 * Except where indicated otherwise, values are the mean ⫾ SD. † Only male mice were tested. extent of 92%, with an average intermarker distance of 11.5 cM, were used. The intermarker distance ranged from 5.1 cM to 27 cM for the different chromosomes according to the Mouse Genome Informatics marker positions. The accuracy of our loci order and interval maps was verified by comparing the genetic map calculated from our data with the Mouse Genome Informatics map. Microsatellite markers were chosen from The Jackson Laboratory mouse database (available at their Web site at http://www.jax.org) and were tested if they were informative for the DBA/1J and FVB/N strains. (A complete list of markers is available by request.) Primers for informative microsatellite markers were ordered from Metabion (PlaneggMartinsried, Germany). The genomic DNA used for genotyping the mice was isolated from a 1-cm tail clip by using standard isolation protocols (10). Genomic DNA from each animal was genotyped for 126 microsatellite loci (Mouse Genome Informatics, Bar Harbor, ME) by using polymerase chain reaction (PCR) amplification. Genomic DNA (20 ng) was amplified in a final volume of 10 l containing HotStar Taq polymerase (0.25 units; Qiagen, Chatsworth, CA), primers (0.1 M each), 50 mM KCl, 10 mM Tris, 2.5 mM MgCl2, 0.2 mM dNTP, and 0.02 M M13-IRD700 or M13-IRD800 (Li-Cor, Lincoln, NE). Amplification conditions were as follows: 95°C for 15 minutes, followed by 2 cycles of 94°C for 30 seconds, 57°C for 1 minute, 72°C for 1 minute, then another 37 cycles of 94°C for 30 seconds, 55°C for 1 minute, 72°C for 1 minute, and a final extension at 72°C for 7 minutes. The reactions were performed using a GeneAmp PCR System 9700 cycler (Applied Biosystems, Foster City, CA). The PCR products were resolved on denaturing polyacrylamide gels, and were detected by using a Li-Cor Model 4200L automated DNA sequencer. The genotypes were scored independently by at least 2 people (KB and XY), or using the Saga software supplied by Li-Cor. Measurement of antibodies, CD4:CD8 ratio, and cell proliferation by XTT test. Blood was obtained from mice at the thirtieth day after immunization; sera were prepared and used for antibody determination. Antibodies to CII were determined by enzyme-linked immunosorbent assay (ELISA). Maxisorp 96-well plates (Nunc, Hanover Park, IL) were coated with bovine CII (0.25 g of CII per well). Anti-IgG1 or anti-IgG2a secondary antibodies (for Th2- and Th1-supported isotypes, respectively) were used to determine the collagenspecific IgG1 or IgG2a antibody response. A serial dilution (1:50 to 1:50,000) of the pooled sera of F2 mice was used for choosing the dilution factor for the test, the concentration of which provided the median optical density value used for positive sample determination. Pooled sera were used as standard to make the adjustment among the different plates. At day 90 after immunization, lymph nodes of the mice QUANTITATIVE TRAIT LOCI IN COLLAGEN-INDUCED ARTHRITIS Figure 1. Plots of quantitative trait loci (QTLs), on chromosomes 2, 5, and 16, containing non–major histocompatibility complex–linked loci for the disease traits of (collagen-induced) arthritis severity, CD4:CD8 ratio, IgG2a response level, and cell proliferation (by XTT test). Log-likelihood values were determined by using R/qtl software. Logarithm of odds (LOD) scores are presented in relation to positions of the marker loci along the chromosome. Horizontal bar indicates the median LOD score as the cutoff for a positive association. a–c show newly identified QTLs, while d shows a previously reported QTL that was confirmed in this study. a, Trmq3 (T cell ratio modifier QTL 3) on chromosome 2, controlling the CD4:CD8 ratio trait. b, Cia27 on chromosome 5, controlling the IgG2a phenotype. c, Lp1 (lymphocyte proliferation 1) on chromosome 16, by XTT, controlling the cell proliferation trait. d, Cia2 on chromosome 2, controlling the severity phenotype. were used to make single-cell suspensions. CD4 and CD8 cells were determined by fluorescence-activated cell sorter (FACS) analysis as described before (11). Briefly, cells (1 ⫻ 106) were washed twice with FACS buffer, then incubated with anti-CD8–fluorescein isothiocyanate and anti-CD4– phycoerythrin for 20 minutes in the dark, followed by washing twice. FACScan and FACSCalibur (Becton Dickinson, Mountain View, CA) were used for these analyses. Lymphocytes were gated using forward and side scatter, and 10,000 lymphocytes were counted. The percentage of CD4 cells was divided by the percentage of CD8 cells to determine the CD4:CD8 ratio. Lymph nodes from the mice at the ninetieth day after 3723 immunization were also used to make single-cell suspensions for the XTT cell proliferation test, which was done according to the manufacturer’s protocol (Roche Applied Science, Indianapolis, IN). Briefly, 4 ⫻ 105 lymph node cells were cultured in 100 l medium per well in 96-well plates and were incubated at 37°C and 5% CO2 with concanavalin A (1 g) stimulation. After 24 hours, a mixture of 0.1 ml PMS (an electron-coupling reagent) and 5 mg/5 ml XTT was added to each well with a volume of 50 l and the cells were incubated for an additional 16 hours for XTT formazan production. The spectrophotometric absorbance of the samples was determined with an ELISA reader at 450 nm. Linkage analysis. All linkage analyses were made with QTX Map Manager software (12) and the imputation model in R/qtl software package (13,14). The order of the loci was obtained from the Mouse Genome Informatics map (available at their Web site at http://www.informatics.jax.org). The results for severity of arthritis, onset of arthritis, collagen-specific antibody responses, lymph node CD4:CD8 ratio, and XTT cell proliferation were used as phenotypes. Continuous trait values were checked for normal distribution, and logarithmic values were used when necessary. With regard to the significant and suggestive linkage threshold values, we followed the guidelines resulting from the permutation test (number of permutations ⫽ 1,000). The associations between marker and phenotype were also tested by F statistic (analysis of variance [ANOVA]) using KaleidaGraph software). The 2-locus interaction analyses were performed using the scan-two function in R/qtl software, with Harley-Knott regression. The scan-two functions perform a 2-dimensional genome scan with a 2-QTL model. The logarithm of odds (LOD) combined values (LODjoint) represent a comparison of a full model with covariates (y ⫽ ⫹ ␤q1 ⫹ ␤q2 ⫹ ␤q1 ⫻ q2 ⫹ A␥ ⫹ Z␦q1 ⫹ Z␦q2 ⫹ Z␦q1 ⫻ q2 ⫹ ⑀) with a null model (y ⫽ ⫹ A␥ ⫹ ⑀). The genome-wide significance levels for the scan-two results were established using permutation tests (number of permutations ⫽ 1,000). RESULTS Arthritis penetrance and variance. Table 1 presents the results of the experiment determining the incidence and severity of CIA in DBA/1J control mice Table 2. Summary of the quantitative trait loci (QTLs) identified* QTL Chromosome Peak LOD score Significance level† Flanking markers Position, cM Confidence interval, cM Trait Cia2‡ Cia4‡ Cia27 Cia28 Cia29 Lp1 Trmq3 2 2 5 13 10 16 2 D2Mit81 D2Mit81 D5Mit277 D13Mit144 D10Mit96 D16Mit138 D2Mit398 12.08§ 10.57§ 4.28§ 3.2¶ 3.63# 3.8# 5.03§ 3.4 3.5 3.4 3.4 3.4 3.5 3.42 D2Mit81–D2Mit367 D2Mit81–D2Mit367 D5Mit201–D5Mit425 D13Mit13–D13Mit144 D10Mit261–D10Mit14 D16Mit60–D16Mit139 D2Mit61–D2Mit395 2 4 60 52 52 31 65 6 10 38 15 28 38 14 Severity Onset IgG2a IgG1 IgG1 Cell proliferation by XTT CD4:CD8 * LOD ⫽ logarithm of odds; Lp1 ⫽ lymphocyte proliferation 1; Trmq3 ⫽ T cell ratio modifier QTL 3. † Significance level (P ⬍ 0.05) determined according to permutation test (n ⫽ 1,000). ‡ Reference. § P ⬍ 0.01 (highly significant). ¶ P ⬍ 0.1 (suggestive). # P ⬍ 0.05 (significant). 3724 BAUER ET AL Table 3. Analysis of variance for chromosomes with evidence of linkage* Phenotype, loci, marker Severity Chromosome 2 Cia2 D2Mit81 D2Mit367 D2Mit61 D2Mit398 D2Mit395 D2Mit145 D2Mit148 Onset Chromosome 2 Cia4 D2Mit81 D2Mit367 D2Mit61 D2Mit398 D2Mit395 D2Mit145 D2Mit148 IgG2a Chromosome 5 Cia27 D5Mit348 D5Mit387 D5Mit54 D5Mit201 D5Mit277 D5Mit425 D5Mit31 D5Mit143 IgG1 Chromosome 13 Cia28 D13Mit16 D13Mit275 D13Mit19 D13Mit13 D13Mit144 IgG1 Chromosome 10 Cia29 D10Mit248 D10Mit126 D10Mit20 D10Mit261 D10Mit96 D10Mit14 Cell proliferation by XTT Chromosome 16 Lp1 D16Mit131 D16Mit101 D16Mit60 D16Mit138 D16Mit139 D16Mit153 D16Mit52 CD4:CD8 Chromosome 2 Trmq3 D2Mit81 D2Mit367 D2Mit61 D2Mit398 D2Mit395 D2Mit145 D2Mit148 Position, cM F† P 13.5 26.2 34 57.9 66.9 98.5 105 29.66 25.18 14.11 4.06 4.82 1.48 0.32 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.018 0.0087 0.227 0.722 13.5 26.2 34 57.9 66.9 98.5 105 21.51 21.84 15.97 3.72 3.03 1.93 0.1 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.026 0.049 0.146 0.902 8 15 28 42 58 65 78 86 4.28 4.57 4.27 8.42 9.73 9.58 4.1 1.24 0.014 0.011 0.014 0.00028 ⬍0.0001 ⬍0.0001 0.017 0.29 10 16 24 35 48 1.12 0.08 1.26 3.2 5.88 0.329 0.923 0.285 0.042 0.003 7 21 35 47 56 65 3.68 1.16 4.13 5.17 5.28 1.21 0.026 0.313 0.017 0.006 0.005 0.3 4.3 17 23.4 31 43.1 56.8 66.8 0.95 3.04 5.68 9.51 6.76 7.48 7.01 0.391 0.052 0.0047 0.00018 0.0018 0.00098 0.00147 13.5 26.2 34 57.9 66.9 98.5 105 1.58 3.21 2.39 14.47 9.83 6.72 4.37 0.212 0.044 0.097 ⬍0.0001 0.00014 0.00189 0.015 * See Table 2 for definitions. † Test of rank, 2 degrees of freedom. and (DBA/1J ⫻ FVB/N)F1 and F2 experimental animals. As expected, collagen-immunized DBA/1J mice had an arthritis incidence of 100% (15 of 15) with a mean ⫾ SD arthritis index of 7.91 ⫾ 3.08 and a mean ⫾ SD time to arthritis onset of 39.1 ⫾ 11.7 days. However, FVB/N mice did not develop arthritis. The (DBA/1J ⫻ FVB/ N)F1 progeny had an arthritis incidence of 21.7% (13 of 60), an arthritis index of 3.15 ⫾ 2.99, and time to arthritis onset of 62.1 ⫾ 18.1 days. The (DBA/1J ⫻ FVB/N)F2 animals had an arthritis incidence of 29.3% (85 of 290) with an arthritis index of 4.51 ⫾ 3.48 and time to arthritis onset of 44.5 ⫾ 19.9 days (Table 1). Identification of arthritis QTLs. The linkage analysis was performed by combining the genotype data from the genome scan with 126 informative microsatellite markers and the phenotype data on the antibody responses, CD4:CD8 ratio, and cell proliferation (determined by XTT test). Two hundred seventy mice were used to determine the collagen-specific IgG1 and IgG2a levels. The IgG1 values ranged from 0.063 to 2.702 (mean ⫾ SD 0.414 ⫾ 0.518). For IgG2a, the optical density values at 405 nm among the samples ranged from 0.068 to 0.963 (mean ⫾ SD 0.472 ⫾ 0.198). The CD4:CD8 ratio values ranged from 0.81 to 3.83 (mean ⫾ SD 1.77 ⫾ 0.536). In addition, the cell proliferation as measured by XTT was also defined as a phenotype. These absorbance values ranged from 0.17 to 2.65 (mean ⫾ SD 1.32 ⫾ 0.64). Seven QTLs were found in the screen, located on chromosomes 2, 5, 10, 13, and 16 (Figure 1 and Table 2). Two of the identified QTLs had been discovered in previous studies and were confirmed herein (5). Cia2, controlling the severity phenotype on chromosome 2, had highly significant evidence of linkage to marker D2Mit81 (LOD score of 12.1). The second locus, Cia4, controlling the onset phenotype and located on chromosome 2, also had highly significant evidence of linkage to marker D2Mit81 (LOD score of 10.6). Novel QTLs were identified on chromosomes 2, 5, 10, 13, and 16. Cia27 on chromosome 5 was identified as a QTL controlling the CII-specific IgG2a phenotype; it had significant evidence of linkage to marker D5Mit277 (LOD score of 4.3). We identified 2 loci controlling the CII-specific IgG1 antibody phenotype: Cia28 on chromosome 13, having suggestive evidence of linkage to marker D13Mit144 (LOD score of 3.2), and Cia29, showing significant evidence of linkage to marker D10Mit96 (LOD score of 3.6). Trmq3 (T cell ratio modifier QTL 3), controlling the CD4:CD8 ratio, was found on chromosome 2 with significant evidence of linkage to marker D2Mit398 (LOD score of 5). Finally, QTL Lp1 (lymphocyte proliferation 1) on chromosome 16, controlling the cell proliferation measured by XTT test, had significant QUANTITATIVE TRAIT LOCI IN COLLAGEN-INDUCED ARTHRITIS Table 4. 3725 Two-locus interaction analysis* Phenotype Position 1, marker 1 (loci) Position 2, marker 2 (loci) LODjoint LOD significance level† IgG1 IgG2a Onset Onset Onset Onset Onset Severity Severity D1Mit111 (Cia9) D5Mit425 (Cia27) C5 (Cia4) C5 (Cia4) C5 (Cia4) C5 (Cia4) C5 (Cia4) C5 (Cia2) C5 (Cia2) D13Mit144 (Cia28) D10Mit96 (Cia29) D2Mit395 (Trmq3) D10Mit261 (Cia8) D11Mit4 (C1qbp) D13Mit144 (Cia28) D16Mit138 (Lp1) D5Mit425 (Cia27) D10Mit261 (Cia8) 9.12 10.34 12.14 13.55 13.23 10.95 12.4 14.94 15.98 8.63 8.57 9.37 9.37 9.37 9.37 9.37 10.63 10.63 * Logarithm of odds (LOD) combined values were calculated by 2-locus analysis in R/qtl (13), with Haley-Knott regression. See Table 2 for definitions. † Significance levels (P ⬍ 0.05) determined according to permutation tests (number of permutations ⫽ 1,000). evidence of linkage to marker D16Mit138 (LOD score of 3.8). To confirm the existence of these loci, a 2-point analysis (F statistic, ANOVA) was performed. The data from this analysis (Table 3) yielded similar results, thus demonstrating an association between the linkage analysis results and the findings by ANOVA. Cia2 was associated with marker D2Mit81 (F ⫽ 29.66, P ⬍ 0.0001), Cia4 with D2Mit81 (F ⫽ 21.51, P ⬍ 0.0001), Cia27 with D5Mit277 (F ⫽ 9.73, P ⬍ 0.0001), Cia28 with D13Mit144 (F ⫽ 5.88, P ⫽ 0.003), Cia29 with D10Mit96 (F ⫽ 5.28, P ⫽ 0.005), Trmq3 with D2Mit398 (F ⫽ 14.47, P ⬍ 0.0001), and Lp1 with D16Mit138 (F ⫽ 9.51, P ⫽ 0.00018). Genome-wide 2-locus interaction analyses were performed on all phenotypes using R and R/qtl software (13,14). The disease phenotypes for onset and severity showed significant interactions with the Cia2 and Cia4 loci, marker C5, on chromosome 2, together with Cia8, Cia27, Cia28, C1qbp, Lp1, and Trmq3 (Table 4). Two significant interactions were found for the phenotypes controlling antibody levels for IgG1 and IgG2a, respectively. The IgG1 levels showed a linkage between the Cia9 locus on chromosome 1 and Cia28 on chromosome 13 (LODjoint ⫽ 9.12). The Cia27 locus on chromosome 5 and Cia29 on chromosome 10 controlled the IgG2a levels (LODjoint ⫽ 10.34). C5 as a candidate gene for CIA. In the linkage analysis, we confirmed 2 loci on chromosome 2, Cia2 and Cia4. Hemolytic complement (Hc), or complement component C5, is a gene that has been mapped to the region of this genomic interval. A 2-basepair deletion in exon 5 of the C5 gene, which leads to deficiency of the molecule, has been previously described (15). The association of C5 deficiency with resistance against CIA has also been reported previously (5,7,16–20). Therefore, we investigated whether the resistance of the FVB/N strain to the induction of CIA could be the result of C5 deficiency. Using genotyping on polyacrylamide gels, we showed that the 2-basepair deletion in the C5 gene is present in the FVB/N mouse strain. We then genotyped all F2 mice for this deletion and analyzed the incidence and severity of arthritis among different groups genotyped for C5. The analyses showed that both incidence and severity were clearly decreased in mice deficient in C5. The incidence in C5⫺/⫺ mice was highly significantly lower than in C5⫹/⫹ mice (incidence 8.7% in C5⫺/⫺ versus 53.2% in C5⫹/⫹; P ⬍ 0.001 by chi-square test) and C5-deficient Figure 2. Chromosomes with previously published Cia quantitative trait loci (QTLs) (thick, striped vertical bars) (positions determined according to original reference and the Mouse Genome Informatics database) (see refs. 5–9) were compared with all QTLs that were identified in this study (Cia27–Cia29, Lp1, and Trmq3). Thin, solid vertical bars represent chromosomal regions to which linkage of microsatellite markers was found. Cia2 and Cia4 on chromosome 2 were previously identified and were confirmed. Cia27 on chromosome 5, Cia28 on chromosome 13, Cia29 on chromosome 10, Trmq3 on chromosome 2, and Lp1 on chromosome 16 are newly identified QTLs. The application used for visualization is available (under Fischer G, et al) at the Web site http://www.ensembl.pzr.uni-rostock.de/Mus_ musculus/expressionview. 3726 BAUER ET AL mice also had highly significantly lower disease severity than did mice with sufficient C5 (mean ⫾ SD arthritis index 0.11 ⫾ 0.39 in C5⫺/⫺ versus 3.0 ⫾ 3.9 in C5⫹/⫹; P ⬍ 0.001 by Mann-Whitney test). Although a few C5-deficient mice get disease, the severity is very low. This suggests that the resistance to CIA in these mice is mainly caused by C5 deficiency. However, because C5heterozygous mice also still develop disease (incidence 27.8% in C5⫹/⫺ versus 53.2% in C5⫹/⫹ [P ⬍ 0.001 by chi-square test]; mean ⫾ SD arthritis index 1.1 ⫾ 2.3 in C5⫹/⫺ versus 3.0 ⫾ 3.9 in C5⫹/⫹ [P ⬍ 0.001 by Mann-Whitney test]), the contribution of an additional gene cannot be completely excluded. DISCUSSION In this study, we identified 7 QTLs controlling the pathogenesis of CIA, of which 2 had been identified before and the other 5 are new QTLs (Table 2 and Figure 2). These QTLs were confirmed by ANOVA test (Table 3). The susceptibility and resistance of DBA/1J and FVB/N mouse strains, respectively, to the induction of CIA is well documented. As observed in previous studies, the DBA/1J mice in these experiments had an arthritis incidence of 100%, whereas the FVB/N mice were completely resistant (0% incidence). As expected and previously seen in similar crosses, the (DBA/1J ⫻ FVB/N)F1 progeny were susceptible to CIA (Table 1). However, there was incomplete penetrance of the susceptible phenotype (21% incidence). A comparison of the severity (arthritis) indexes between the F2 progeny and the parental DBA/1J showed significant differences (P ⬍ 0.0001), suggesting the existence of protective FVB/N loci. Two genomic intervals (QTLs) were associated with CIA severity and onset (Figure 2), overlapping with known loci previously linked to CIA, i.e., Cia2 and Cia4 located on chromosome 2. This locus seems to be a common locus for CIA, as confirmed by our data. Despite having many candidate genes in the locus on chromosome 2 (Cia2), such as Ptgs1, Dpp4, and Ssb (5), it is most likely that the main susceptibility gene is complement C5 (Hc). This gene is an important mediator of inflammation, and its role in the disease process of CIA is well documented (16–20). The implication of C5 in the resistance exhibited by SWR/J mice (5) and NOD mice (7) to CIA induction has been described before, and it has also been identified as a possible candidate gene for serum/antibody-transferred arthritis (21). Nevertheless, in both, the SWR ⫻ DBA/1J cross and our current FVB/N ⫻ DBA/1J cross, there is a clear effect on disease susceptibility and severity in the het- erozygous form of the gene. This suggests that there is a dose effect of the gene, although there is evidence that levels of C5 are comparable in heterozygous mice and homozygous C57BL/10.Q mice (7). Another explanation is that there are 2 susceptibility genes operating within this locus, a possibility that will be explored with the use of congenic mice, which are currently being generated. The first new QTL identified by us, Cia29, is located on chromosome 10 and controls the collagenspecific IgG1 antibody trait. When considering the peak (52 cM) of this QTL and taking into account the confidence interval, it extends quite close to the peak of the known QTL Cia8 (40.4 cM) (Figure 2), a locus identified previously as a disease severity QTL (6). This locus appears to be a common locus for autoimmune diseases, including CIA (6), experimental allergic encephalomyelitis (EAE) (22), and systemic lupus erythematosus (23). Cia29 could be the same locus as Cia8, although in our cross, the linkage with severity was weak. Assuming that this is the same locus, then Cia8/29 is indeed a major locus contributing to autoimmune inflammation. There are a number of candidate genes found in this region that have been implicated in the modulation of CIA susceptibility, such as the macrophage migration inhibitory factor Mif (24), the autoimmune regulator aire (25), and perforin (Bauer K, et al: unpublished data). The autoantibody-linked Pgia26 locus, identified in proteoglycan-induced arthritis, was also mapped to chromosome 10, but this was at a locus more at the beginning of the chromosome, at 16 cM (26). Perhaps the most interesting finding from this new genome scan has been the identification of Cia27 on chromosome 5, a QTL controlling the collagen-specific IgG2a level. It is the first time that this genomic interval on chromosome 5 has been identified in association with autoimmune antibodies and arthritis. We were able to show that F2 mice carrying the DBA allele in this interval produced more IgG2a than did F2 mice with the FVB allele in this area (the latter had levels corresponding to the IgG2a levels in the parental mice) (data not shown). We also observed in our experiments that the IgG2a level was associated with the severity of the disease (data not shown). Previous studies also showed that IgG2a is one of the important isotypes contributing to the pathogenesis of arthritis (27). Candidate genes found in the region of the peak marker that would be predicted to have an influence on the inflammation process of the disease are CXCL9 and CXCL10. A recent report by Ruschpler and coworkers provides this assumption. They showed an up-regulated expression of CXCL9 as well as CXCL10 in inflammatory synovial tissues of patients with rheumatoid arthritis (28). Furthermore, the receptor of CXCL9 and CXCL10, QUANTITATIVE TRAIT LOCI IN COLLAGEN-INDUCED ARTHRITIS CXCR3, was also found to be highly expressed in synovial mast cells. Earlier studies assigned CXCR3 protein primarily to the Th1 lymphocytes (29–32). The presence of this T cell subset is consistent with the changes in the IgG2a level, because we actually choose this IgG isotype to detect the Th1 pathway, which is known to be required for the induction of CIA in DBA/1 mice (33). Previous studies also identified QTLs that control autoantibody production in arthritis models. With regard to CIA in a cross between Balb/c and DBA/1J strains, 4 QTLs have been found on chromosomes 3, 17, 18, and X, and these are known as mCia2, mCia1, mCia18, and mCia13, respectively (34). QTLs on chromosomes 1, 2, 7, 8, 10, 11, 16, and 18, controlling autoantibody production in proteoglycan-induced arthritis, have also been described before in a cross between Balb/c and DBA/2 (24). Furthermore, in the Lyme disease model, IgG-controlling loci have been identified in a C3H ⫻ C57BL/6 cross on chromosome 9 (Bb9) and chromosome 12 (Bb6) (35). In addition, different antibody QTLs have been described in rat arthritis models (36), such as Ciaa3, a locus that regulates the production of IgG anticollagen antibodies, identified in a linkage study of a (DA ⫻ BN)F2 cross (37). In the region of chromosome 5, in which we found Cia27, different loci linked to autoimmune disease traits have been mapped: for Lyme disease, Bb2 and Bb3 (35), for arthritis, Pgia16, Cia13, and Cia14 (34,38), and also recently one for EAE, Eae26. However, no locus with linkage to an antibody phenotype has been reported on chromosome 5. In our experiments, the CD4:CD8 ratio of lymph node cells always changed in different stages of the immunization. DBA mice had a higher CD4:CD8 ratio than did FVB mice (⬃4:2.5) without immunization; however, after immunization with CII, the difference in the ratio between the 2 mouse strains started to decrease with age (data not shown). The existence of the Trmq3 locus, which is linked to the CD4:CD8 phenotype and located on chromosome 2, suggests there is a gene controlling the development of T cells in this area. Trmq3 overlapped with Trmq2, a QTL associated with the CD4:CD8 ratio in a C57BL/6 ⫻ DBA/2 cross (39). In our study, we could show that this phenotype is probably associated with susceptibility to CIA. Interesting genes that seem to regulate CD4⫹ and CD8⫹ populations have been mapped to chromosome 2, such as notch1 and bcl-x (39,40). Increased expression of notch1 seems to enhance the size of the CD8⫹ population and bcl-x induction appears to be required for B7–CD28 interaction that again affects the CD4:CD8 ratio (39). Unexpectedly, the FVB, not the DBA allele, increased the 3727 ratio in our experiments, since the lymph node cells were obtained from mice at day 90 after immunization. One possible explanation could be the dynamic of the CD4: CD8 ratio during the immunization as we described above. Therefore, the gene in this locus regulating the CD4:CD8 ratio is probably affected by the immunization with CII. Another new locus that we identified in this study, Lp1, is located on chromosome 16 and linked to lymphocyte cell proliferation determined by XTT assay. This locus showed colocalization with Eae11, which is also associated with spleen cell proliferation as well as onset and incidence in EAE (41,42). The performed 2-locus analysis showed that several of the previously defined loci, e.g., Cia2, Cia4, Cia8, and Cia9, were involved in controlling the disease and antibody responses. The disease traits of onset and severity all involved the chromosome 2 marker C5 (Cia2 or Cia4). The antibody responses were controlled by Cia9 and Cia22 (IgG1) and Cia27 and Cia29 (IgG2a). When we analyzed the gene region of interest for the QTLs Cia27, Cia28, and Cia29, we could not reveal any explanation as to why there was an association with antibody levels. Our present analysis has identified 5 unique loci and confirmed 2 common loci for CIA susceptibility, collagen-specific antibody responses, the CD4:CD8 ratio, and cell proliferation (measured by XTT test). Using a combination of congenic mapping, positional cloning, and gene expression profiling, we are currently pursuing identification of the candidate genes within a few of these loci. ACKNOWLEDGMENTS The authors wish to thank Dr. G. Brockmann for advice on linkage analysis, as well as Ilona Klamfuss and Eva Lorbeer for help with the animals. REFERENCES 1. Silman AJ. Epidemiology of rheumatoid arthritis. APMIS 1994; 102:721–8. 2. Wordsworth P, Bell J. Polygenic susceptibility in rheumatoid arthritis. Ann Rheum Dis 1991;50:343–6. 3. MacGregor AJ, Snieder H, Rigby AS, Koskenvuo M, Kaprio J, Aho K, et al. Characterizing the quantitative genetic contribution to rheumatoid arthritis using data from twins. Arthritis Rheum 2000;43:30–7. 4. Wordsworth BP, Lanchbury JS, Sakkas LI, Welsh KI, Panayi GS, Bell JI. HLA-DR4 subtype frequencies in rheumatoid arthritis indicate that DRB1 is the major susceptibility locus within the HLA class II region. Proc Natl Acad Sci U S A 1989;86:10049–53. 5. McIndoe RA, Bohlman B, Chi E, Schuster E, Lindhardt M, Hood L. Localization of non-MHC collagen-induced arthritis susceptibility loci in DBA/1J mice. Proc Natl Acad Sci U S A 1999;96: 2210–4. 3728 6. Yang HT, Jirholt J, Svensson L, Sundvall M, Jansson L, Pettersson U, et al. Identification of genes controlling collagen-induced arthritis in mice: striking homology with susceptibility loci previously identified in the rat. J Immunol 1999;163:2916–21. 7. Johansson AC, Sundler M, Kjellen P, Johannesson M, Cook A, Lindqvist AK, et al. Genetic control of collagen-induced arthritis in a cross with NOD and C57BL/10 mice is dependent on gene regions encoding complement factor 5 and Fc␥RIIb and is not associated with loci controlling diabetes. Eur J Immunol 2001;31: 1847–56. 8. Jirholt J, Cook A, Emahazion T, Sundvall M, Jansson L, Nordquist N, et al. Genetic linkage analysis of collagen-induced arthritis in the mouse. Eur J Immunol 1998;28:3321–8. 9. Adarichev VA, Valdez JC, Bardos T, Finnegan A, Mikecz K, Glant TT. Combined autoimmune models of arthritis reveal shared and independent qualitative (binary) and quantitative trait loci. J Immunol 2003;170:2283–92. 10. Laird PW, Zijderveld A, Linders K, Rudnicki MA, Jaenisch R, Berns A. Simplified mammalian DANN isolation procedure. Nucleic Acids Res 1991;19:4293. 11. Myrick C, DiGuisto R, DeWolfe J, Bowen E, Kappler J, Marrack P, et al. Linkage analysis of variations in CD4:CD8 T cell subsets between C57BL/6 and DBA/2. Genes Immun 2002;3:144–50. 12. Manly KF, Cudmore RH Jr, Meer JM. Map Manager QTX, cross-platform software for genetic mapping. Mamm Genome 2001;12:930–2. 13. Broman KW, Wu H, Sen S, Churchill GA. R/qtl: QTL mapping in experimental crosses. Bioinformatics 2003;19:889–90. 14. Sen S, Churchill GA. A statistical framework for quantitative traitmapping. Genetics 2001;159:371–82. 15. Wetsel RA, Fleischer DT, Haviland DL. Deficiency of the murine fifth complement component (C5): a 2-base pair gene deletion in a 5⬘-exon. J Biol Chem 1990;265:2435–40. 16. Wang Y, Kristan J, Hao L, Lenkoski CS, Shen Y, Matis LA. A role for complement in antibody-mediated inflammation: C5-deficient DBA/1 mice are resistant to collagen-induced arthritis. J Immunol 2000;164:4340–7. 17. Spinella DG, Jeffers JR, Reife RA, Stuart JM. The role of C5 and T-cell receptor Vb genes in susceptibility to collagen-induced arthritis. Immunogenetics 1991;34:23–7. 18. Reife RA, Loutis N, Watson WC, Hasty KA, Stuart JM. SWR mice are resistant to collagen-induced arthritis but produce potentially arthritogenic antibodies. Arthritis Rheum 1991;34:776–81. 19. Watson WC, Brown PS, Pitcock JA, Townes AS. Passive transfer studies with type II collagen antibody in B10.D2/old and new line and C57BL/6 normal and beige (Chediak-Higashi) strains: evidence of important roles for C5 and multiple inflammatory cell types in the development of erosive arthritis. Arthritis Rheum 1987;30:460–5. 20. Mori L, de Libero G. Genetic control of susceptibility to collageninduced arthritis in T cell receptor ␤-chain transgenic mice. Arthritis Rheum 1998;41:256–62. 21. Ji H, Gauguier D, Ohmura K, Gonzales A, Duchatelle V, Danoy P, et al. Genetic influences on the end-stage effector phase of arthritis. J Exp Med 2001;194:321–30. 22. Butterfield RJ, Blankenhorn EP, Roper RJ, Zachary JF, Doerge RW, Teuscher C. Identification of genetic loci controlling the characteristics and severity of brain and spinal cord lesions in experimental allergic encephalomyelitis. Am J Pathol 2000;157: 637–45. 23. Vidal S, Kono DH, Theofilopoulos AN. Loci predisposing to autoimmunity in MRL-Fas lpr and C57BL/6-Faslpr mice. J Clin Invest 1998;101:696–702. 24. Mikulowska A, Metz CN, Bucala R, Holmdahl R. Macrophage migration inhibitory factor is involved in the pathogenesis of collagen type II-induced arthritis in mice. J Immunol 1997;158: 5514–7. 25. Anderson MS, Venanzi ES, Klein L, Chen Z, Berzins SP, Turley SJ, et al. Projection of an immunological self shadow within the thymus by the aire protein. Science 2002;298:1395–401. BAUER ET AL 26. Otto JM, Cs-Szabo G, Gallagher J, Velins S, Mikecz K, Buzas EI, et al. Identification of multiple loci linked to inflammation and autoantibody production by a genome scan of a murine model of rheumatoid arthritis. Arthritis Rheum 1999;42:2524–31. 27. Watson WC, Townes AS. Genetic susceptibility to murine collagen II autoimmune arthritis: proposed relationship to the IgG2 autoantibody subclass response, complement C5, major histocompatibility complex (MHC) and non-MHC loci. J Exp Med 1985; 162:1878–91. 28. Ruschpler P, Lorenz P, Eichler W, Koczan D, Hanel C, Scholz R, et al. High CXCR3 expression in synovial mast cells associated with CXCL9 and CXCL10 expression in inflammatory synovial tissues of patients with rheumatoid arthritis. Arthritis Res Ther 2003;5:R241–52. 29. Bonecchi R, Bianchi G, Bordignon PP, D’Ambrosio D, Lang R, Borsetti A, et al. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s. J Exp Med 1998;187:129–34. 30. Sallusto F, Mackay CR, Lanzavecchia A. Selective expression of the eotaxin receptor CCR3 by human T helper 2 cells. Science 1997;277:2005–7. 31. Sallusto F, Lenig D, Mackay CR, Lanzavecchia A. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J Exp Med 1998;187:875–83. 32. Garcia-Lopez MA, Sanchez-Madrid F, Rodriguez-Frade JM, Mellado M, Acevedo A, Garcia MI, et al. CXCR3 chemokine receptor distribution in normal and inflamed tissues: expression on activated lymphocytes, endothelial cells, and dendritic cells. Lab Invest 2001;81:409–18. 33. Luross JA, Williams NA. The genetic and immunopathological processes underlying collagen-induced arthritis. Immunology 2001;103:407–16. 34. Adarichev VA, Valdez JC, Bardos T, Finnegan A, Mikecz K, Glant TT. Combined autoimmune models of arthritis reveal shared and independent qualitative (binary) and quantitative trait loci. J Immunol 2003;170:2283–92. 35. Weis JJ, McCracken BA, Ma Y, Fairbairn D, Roper RJ, Morrison TB, et al. Identification of quantitative trait loci governing arthritis severity and humoral responses in the murine model of lyme disease. J Immunol 1999;162:948–56. 36. Griffiths MM, Remmers EF. Genetic analysis of collagen-induced arthritis in rats: a polygenic model for rheumatoid arthritis predicts a common framework of cross-species inflammatory/autoimmune disease loci. Immunol Rev 2001;184:172–83. 37. Griffiths MM, Wang J, Joe B, Dracheva S, Kawahito Y, Shepard JS, et al. Identification of four new quantitative trait loci regulating arthritis severity and one new quantitative trait locus regulating autoantibody production in rats with collagen-induced arthritis. Arthritis Rheum 2000;43:1278–89. 38. Otto JM, Chandrasekeran R, Vermes C, Mikecz K, Finnegan A, Rickert SE, et al. A genome scan using a novel genetic cross identifies new susceptibility loci and traits in a mouse model of rheumatoid arthritis. J Immunol 2000;165:5278–86. 39. Myrick C, DiGuisto R, DeWolfe J, Bowen E, Kappler J, Marrack P, et al. Linkage analysis of variations in CD4:CD8 T cell subsets between C57BL6 and DBA/2. Genes Immun 2002;3:144–50. 40. Deftos ML, He YW, Ojala EW, Bevan MJ. Correlating notch signaling with thymocyte maturation. Immunity 1998;9:777–86. 41. Karlsson J, Zhao X, Lonskaya I, Neptin M, Holmdahl R, Andersson A. Novel quantitative trait loci controlling development of experimental autoimmune encephalomyelitis and proportion of lymphocyte subpopulations. J Immunol 2003;170:1019–26. 42. Butterfield RJ, Blankenhorn EP, Roper RJ, Zachary JF, Doerge RW, Sudweeks J, et al. Genetic analysis of disease subtypes and sexual dimorphisms in mouse experimental allergic encephalomyelitis (EAE): relapsing/remitting and monophasic remitting/nonrelapsing EAE are immunogenetically distinct. J Immunol 1999; 162:3096–102.