Using genetic linkage analysis of proteoglycan-induced arthritis (PGIA), a murine model for rheumatoid arthritis, we identified two loci, Pgia8 and Pgia9, on chromosome 15 (chr15) that appear to be implicated in disease susceptibility. Immunization of congenic strains carrying the entire chr15 and separately each of the two loci of DBA/2 arthritis-resistant origin in susceptible BALB/c background confirmed locations of two loci on chr15: the major Pgia9 and lesser Pgia8 locus. Distal part of chr15 (Pgia9) showed a major suppressive effect on PGIA susceptibility in females (40%, p < 0.001), whereas the effect of this locus in congenic males was still significant but weaker. Proximal part of chr15 (Pgia8) demonstrated mild and transient effect upon arthritis; this effect was PGIA-promoting in males and suppressive in females. Pgia8 and Pgia9 loci demonstrated an additive mode of inheritance, since when they were both incorporated in consomic chr15 strain, the total effect was a sum of the two loci. Using F2 population of the intercross of wild-type and chr15 consomic strain, we confirmed and refined quantitative trait locus positions and identified a strong correlation between disease susceptibility and lymphocyte-producing cytokines of TNF-α and IL-6. Both Pgia8 and Pgia9 loci on chr15 appear to control IL-6 production in spleen cultures of arthritic mice, providing an important link to the mechanism of autoimmune inflammation.
Rheumatoid arthritis (RA)3 is a chronic inflammatory autoimmune disease influenced by both genetic and environmental factors (1, 2). The genetic background of RA is complex and heterogeneous, being linked to different sets of chromosome regions in different ethnic populations (3, 4, 5, 6, 7, 8, 9). Meta-analysis of genome scans of RA-affected families has located the most common susceptibility loci on human chromosomes 1p, 2q, 6p, 6q, 8p, 12p, 16p, and 18q (10).
Progressive polyarthritis induced in susceptible BALB/c mice via systemic immunization with human cartilage proteoglycan (PG) aggrecan is a murine model for RA (11, 12, 13). PG-induced arthritis (PGIA) resembles RA in many ways, as indicated by clinical assessments, biochemical and immunologic tests, and histopathology studies of diarthrodial joints (12). Like RA, PGIA is an autoimmune disease, which is critically dependent on MHC (H-2 complex in mice); however, multiple chromosome loci were found to control disease in different crosses of murine strains (12). Genome-wide linkage analysis of MHC-matched (BALB/c × DBA/2)F2 mice revealed that different sets of non-MHC loci were linked to PGIA susceptibility, severity, or disease onset, while gender was a major modulator of the quantitative trait loci (QTLs) affecting either male- or female-restricted penetrance (14).
Women are affected with RA two to three times more frequently than men, and joint deformities are also more pronounced in women (15, 16); however, factors that account for the increased susceptibility of females are still not completely understood. Disparity of XY vs XX chromosomes in males vs females could be directly linked to higher female preponderance. Therefore, a direct effect of sex chromosomes upon arthritis phenotype is a compelling hypothesis. Indeed, several linkage studies have confirmed the involvement of chromosome X in patients with RA (3, 4, 5, 7), and of both chromosomes X and Y in a number of animal models of arthritis (14, 17, 18, 19, 20, 21, 22, 23, 24). Apparently gender is an important risk factor for arthritis, and sex chromosome loci may indeed influence arthritis susceptibility.
In addition to X and Y chromosomes, genetic loci on somatic chromosomes may also affect sex-related penetrance of the disease (20). For example, in PGIA, a number of QTLs showed association with gender when the MHC effect was excluded in BALB/c × DBA/2 intercross (14). Remarkably, chromosome 15 (chr15) carries two gender-effected loci: Pgia8 and Pgia9. To gain insight into the mechanisms by which these two QTLs on chr15 can differentially affect arthritis in males and females, we transferred the entire chr15, and each locus separately, from the arthritis-resistant DBA/2 strain to the PGIA-susceptible BALB/c background, then analyzed clinical phenotypes of these PG-immunized consomic and congenic mice. In addition, we produced an F2 population in which chr15 carried BALB/c, DBA/2, or both alleles, while all other chromosomes were BALB/c homozygous. Using this two-way approach, we were able to determine how the Pgia8 and Pgia9 loci affect arthritis susceptibility and severity in males and females via a T cell-mediated cytokine-controlled mechanism.
Materials and Methods
Marker-assisted congenic animal breeding
All animal experiments were approved by the Institutional Animal Care and Use Committee (Rush University Medical Center, Chicago, IL). Animals were maintained in a pathogen-free environment. PGIA-susceptible BALB/c female and PGIA-resistant DBA/2 male mice (National Cancer Institute, Kingston colony, NY) were mated, and the F1 males were backcrossed to parental BALB/c strain to obtain N2 generation (n = 150) (Fig. 1). Marker-assisted selection protocol for speed congenic breeding was used to generate QTL-specific congenic strains (25). N2 males were genotyped with 130 genomic markers selected for detectable length polymorphism between the parental strains using information from the Mouse Genome Informatics database (http:// www.informatics.jax.org). N2 males bearing DBA/2-type heterozygous QTLs on chr15 and the highest number of BALB/c alleles in the rest of the genome were selected for the next backcross. Approximately forty offspring males after each backcross were genotyped with ten chr15-specific markers, and additional markers were used for the genomic regions found to be heterozygous in the previous backcross. At backcross level N6, when the entire genome was BALB/c homozygous except chr15, N6 males and females were intercrossed, generating N6F1 population of the consomic strain B-15D (Fig. 1). Chromosome Y of DBA/2 origin was replaced with BALB/c through the mating of B-15D females with BALB/c males. Simultaneously, we selected mice with recombinations carrying either proximal (B-15pD) or distal (B-15dD) parts of chr15 from DBA/2 origin, thus, establishing either Pgia8- or Pgia9-specific congenic strains (Figs. 1 and 2 C).
Chr15-specific genetic linkage analysis
For chr15-specific genetic linkage analysis, wild-type BALB/c females were mated with B-15D males; (BALB/c × B-15D)F1 hybrids were then intercrossed to produce F2 hybrid mice (total n = 195; 114 females and 81 males) (Table I), which were all genotyped with chr15-specific genomic markers listed in Fig. 3. All chromosomes in these F2 hybrids were wild-type BALB/c type except for chr15, which was either homozygous BALB/c, homozygous DBA/2, or heterozygous.
|.||n .||Incidence .||Severity ± SEM .|
|Females||114||83.33||5.37 ± 0.38a|
|Males||81||87.65||6.85 ± 0.42|
|Total||195||85.13||6.01 ± 0.29|
|.||n .||Incidence .||Severity ± SEM .|
|Females||114||83.33||5.37 ± 0.38a|
|Males||81||87.65||6.85 ± 0.42|
|Total||195||85.13||6.01 ± 0.29|
Males developed more severe arthritis than females according to non-paired two-tails Student’s t test, p < 0.01.
Genetic linkage analysis, marker regression, and interval mapping for putative QTLs within chr15 were performed with Map Manager QTX software (26). Significant thresholds for linkage were calculated for each trait (arthritis and Ag-specific IL-6 production) using empirical 5000 permutations at 1-cM interval (26). Permutation test for arthritis score revealed threshold for suggestive linkage at likelihood ratio statistics (LRS) 4.2 and for significant linkage at LRS 11.5. Permutations for level of IL-6 production in lymphocytes trait established suggestive threshold at LRS 4.0 and significant threshold at LRS 11.3. QTL positions were confirmed using bootstrap re-sampling method incorporated in Map Manager QTX (26).
Ag and immunization
Articular cartilage from knee joints was obtained from osteoarthritic patients undergoing joint replacement surgery. The use of human cartilage for PG isolation was approved by the Institutional Review Board of Rush University Medical Center, Chicago, IL. PG isolation has been described in detail (27, 28). Mice were immunized with human PG at 12 wk of age using a standard immunization protocol (27, 28). Emulsion containing 100 μg of PG protein and 2 mg of dimethyldioctadecylammonium bromide adjuvant in 200 μl of PBS (pH 7.4) was injected i.p. on days 0, 21, 42, and 63. Arthritis severity was determined using a visual scoring system based on the extent of swelling and redness of the front and hind paws (11, 27). We performed earlier visual scoring under control of histopathological examination and found a strong correlation between visually observed parameters of arthritis and microscopy picture of joint tissues upon H&E staining (12). Animals were examined at least two times a week and inflammation was scored from 0 to 4 for each paw, thus, resulting in a cumulative arthritis score ranging from 0 to 16 for each animal (11, 27). Susceptibility to disease for each mouse was considered as a binary qualitative trait; i.e., susceptibility has only two values: either “1” for positive (arthritis-susceptible) or “0” for non-arthritic (resistant) animals. Incidence of disease was expressed as percent of arthritic mice to the total number of PG-immunized mice. Disease severity score (ranging from 1 to 16) was applied only to arthritic mice. Day of disease onset was used to characterize how quickly a mouse developed arthritis. Mice were sacrificed two weeks after the fourth injection.
Lymphocyte responses and cytokine production
Ag-specific lymphocyte responses were measured in spleen cell cultures in the presence of 25 μg/ml human PG Ag. IL-1β, IL-4, IL-6, IFN-γ, and TNF-α production were measured in cell culture supernatants of Ag (PG)-stimulated spleen cell cultures on day 4 using ELISA (BD Biosciences). Secreted cytokine concentrations were normalized and expressed as ng/million cells (12, 27, 29). Ag-specific IL-2 production was measured as a proliferation response of CTLL-2 cells to IL-2 in 48-h spleen cell supernatants (CTLL-2 bioassay) (27). Lymphocyte proliferation was assessed on day 5 by incorporation of [3H]thymidine (30). Ag-specific lymphocyte proliferation was expressed as stimulation index. Similarly, we expressed cytokine responses using “normalized change,” which was calculated as follow: [(cytokine concentration measured in PG-stimulated spleen cultures) − (cytokine concentration measured in non-stimulated cultures)]/(cytokine concentration measured in non-stimulated cultures).
Statistical analysis was performed using SPSS statistical software package (SPSS). Since the incidence of disease demonstrated a non-parametric distribution, we used the Mann-Whitney U test to examine differences between populations. To compare immunization curves, comparison between a group of wild-type and a group of congenic mice was performed for every day of scoring. Two-sample non-paired Student’s t test was used for comparison of means of two groups, where the data showed normal distribution (severity of the disease, cytokines). To determine associations between cytokine levels and arthritis, we used Spearman’s rank correlation coefficient ρ. Significance level for all statistical tests was routinely set at p < 0.05.
Arthritis in chr15 congenic mice
Previously, we performed genome-wide linkage analysis of PG-immunized (BALB/c × DBA/2)F2 hybrid mice and identified two gender-dependent loci on chr15 (14). To confirm positions of loci and their genetic effects upon arthritis phenotypes, we transferred the entire chr15D from arthritis-resistant DBA/2 to arthritis-susceptible BALB/c strain in a series of marker-assisted speed congenic backcrosses, thus, generating B-15D consomic females (Fig. 1). Additionally, we selected B-15pD and B-15dD congenic strains that carried either proximal or distal parts of chr15 corresponding to Pgia8 and Pgia9 loci (Figs. 1 and 2 C).
We immunized males and females of consomic B-15D, congenic B-15dD, and congenic B-15pD strains, and of wild-type BALB/c littermates (Fig. 2). The strongest effect upon PGIA was observed in congenic B-15dD mice carrying the distal part of chr15 covering Pgia9 locus (Fig. 2). The onset of arthritis in congenic males (Fig. 2,A) and females (Fig. 2,B) was delayed by at least 2 wk, and disease incidence in B-15dD congenic females was as little as 50% that in wild-type BALB/c female littermates (p < 0.001, Mann-Whitney U test) (Fig. 2 B). In B-15dD congenic males, suppression of arthritis was also evident, although the significance was weaker (p < 0.05).
Immunization of B-15pD congenic females carrying the proximal part of the chr15 also revealed a reduced PGIA susceptibility, albeit much weaker than in B-15dD females (Fig. 2,B). The proximal part of chr15, which incorporates Pgia8 locus, did not show any effect on arthritis severity in either males or females when compared with wild-type littermates (data not shown). We observed higher disease susceptibility and a more rapidly developing disease after the third immunization in B-15pD congenic males than in wild-type control male littermates, although this effect was transient and was sustained for only 10–14 days (Fig. 2,A). Therefore, proximal chr15 loci demonstrated opposing effects upon PGIA, exacerbating disease in males and suppressing it in females (Fig. 2, A vs B), although the effect was not very strong (p < 0.05) and could be detected only during a period of less than two weeks.
Consomic mice carry the entire chr15D; thus, both Pgia8 and Pgia9 loci are present. As one might expect, the immunization curve of B-15D mice is close to the arithmetic average of immunizations of B-15pD and B-15dD congenics. In B-15D consomic females, the arthritis-suppressive effect is significant (p < 0.01), but weaker than in B-15pD congenic females (p < 0.001) (Fig. 2,B). In B-15D consomic males, the PGIA-suppressive effect of the proximal part of chr15 is offset by the arthritis-promoting effect of the distal part of chr15, resulting in an insignificant genetic effect (Fig. 2 A). Severity of paw inflammation in B-15pD or B-15dD mice was similar to that assessed in wild-type BALB/c mice during the entire observation period (data not shown).
In a summary of results for congenic strains immunization, the major suppressive effect upon arthritis susceptibility was identified for Pgia9 locus on the distal part of chr15, while the effect of Pgia8 locus upon disease susceptibility was transient.
Genetic control of arthritis in chr15-specific F2 hybrid mice
Using congenic strains, we confirmed the genetic effects of genes located both within the proximal/centromeric and distal/middle parts of chr15 upon clinical and immunologic arthritis phenotypes (Fig. 2). The size of either Pgia8 or Pgia9 locus, however, was too large, and the identification of causative genes within these loci required a dramatic reduction in size of these chromosome intervals. For this purpose, we used one of the modification of selective phenotyping methods (31) and generated a chr15-specific (BALB/c × B-15D)F2 hybrid population. In these F2 mice, all genetic loci are BALB/c homozygous, except for chr15 alleles. Therefore, the overall genetic variance in this cross is dramatically reduced, and linkage analysis is focused for chr15 loci only.
We immunized (BALB/c × B-15D)F2 males and females (total n = 195) and scored them for PGIA (Table I). Males were slightly more susceptible to PGIA (incidence 87.7 vs 83.3%) and developed more severe arthritis than littermate females; the difference between genders was significant for arthritis severity (p < 0.01) (Table I). The entire F2 population was genotyped for chr15 alleles, and genetic linkage analysis identified two major loci linked to the clinical score of PGIA (Fig. 3,A). Interval mapping also found two loci controlling IL-6 production by PG-stimulated lymphocytes. Positions of these two IL-6-controlling loci (Fig. 3,B) were identical with clinical loci (Fig. 3 A); furthermore, they matched previously identified Pgia8 and Pgia9 intervals (14). Locus positions were further confirmed with bootstrap re-sampling. Pgia8 locus controlled arthritis in females and lymphocyte IL-6 response in males. Pgia9 locus controlled arthritis in both males and females and was equally responsible for IL-6 production in both genders.
Arthritis in (BALB/c × B-15D)F2 mice is associated with IL-6 and TNF-α production
Lymphocytes from F2 hybrid mice were assayed for Ag-induced production of TNF-α, IFN-γ, IL-1β, IL-2, IL-4, and IL-6. Pair-wise correlations between all measured cytokines and arthritis were calculated for all mice (Table II and Fig. 4). In basal conditions, when no PG-stimulation was applied to lymphocytes, arthritis score did not correlate with in vitro production of any cytokine (Table II “Basal”). However, the existence of an associative cytokine network is expected (12, 13, 14, 32), even though that seems to be irrelevant to the clinical phenotype, because cytokine levels correlated with each other. For example, analysis of association between lymphocyte proliferation and cytokine production by non-stimulated splenocytes indicated that the most potent growth factors were IL-2, IL-4, and especially IL-6. This is indicated by the strength of their coefficient of correlation (ρ) with cell proliferation index, which was highest for IL-6 (ρ 0.703) (Table II).
|.||IL-1β .||TNF-α .||IL-6 .||IL-4 .||IFN-γ .||IL-2 .||Proliferation .|
|.||IL-1β .||TNF-α .||IL-6 .||IL-4 .||IFN-γ .||IL-2 .||Proliferation .|
Basal pair-wise Spearman’s coefficients of correlation rho between arthritis susceptibility and cytokines in basal conditions, when no lymphocyte stimulation was applied, are shown. PG-stimulated, Ag-specific correlations between arthritis susceptibility and cytokines produced by lymphocyte cultures stimulated with cartilage PG. Both positive and negative significant coefficients are boldfaced:
, p < 0.05 and
, p < 0.01.
However, when lymphocytes were stimulated with Ag, correlation between arthritis severity and cytokine levels became obvious (Table II “PG-stimulated” and Fig. 4). Surprisingly, PGIA correlated significantly with only two cytokines, TNF-α and IL-6; moreover, these correlations were negative. Apparently, at the next level, TNF-α and IL-6 concentrations affected all other measured cytokines (Fig. 4). For PG-stimulated lymphocytes, again, IL-6 was the dominant cell growth inducer (ρ 0.48, p < 0.01), having even greater effect than TNF-α and IL-2. Thus, IL-6 and TNF-α appear to provide a link between the disease and the underlying immune mechanisms of inflammation.
To further understand the mechanism of disease, particularly in relation to Pgia8 and Pgia9 genetic loci and gender, we determined the allele-specificity of lymphocyte responses in (BALB/c × B-15D)F2 hybrid population. For each of two loci, normalized changes were calculated separately for mice carrying either homozygous (BALB/c or DBA/2) or heterozygous alleles at locus peak genomic marker (Fig. 5). DBA/2 alleles of both Pgia8 and Pgia9 loci increased Ag-specific production of TNF-α by 1.5- to 2.0-fold. In contrast, the effect of these DBA/2 alleles upon IL-6 production was opposing: Pgia9D increased IL-6 secretion 2-fold, but Pgia8D suppressed IL-6 level almost 3-fold. Since IL-6 and TNF-α were the two major proinflammatory cytokines in this F2 population (Table II), when they worked in the same direction, as happened for Pgia9D, the total effect upon T cell proliferation became significant (Fig. 5). In the case of Pgia8D, growth stimulation effects of IL-6 and TNF-α were mixed, and the overall result upon proliferation was not significant.
Two QTLs that control murine arthritis, Pgia8 and Pgia9 on chr15, were identified in genome screening studies of (BALB/c × C3H)F2 and (BALB/c × DBA/2)F2 hybrid populations (13, 14, 30). In a most recent analysis, we found that these two loci affected PGIA severity in a gender-specific mode (14). To further analyze the genetic effect of these two chr15 loci, we transferred the proximal part of chr15 containing Pgia8 locus and the distal part of chr15 bearing Pgia9 locus to the BALB/c genetic background.
Immunization of congenic strains confirmed the presence of arthritis-controlling genes on chr15, although the effects of these loci were of different magnitude and gender specificity. The major locus was found on a distal part of chr15, Pgia9. In females, this locus suppressed arthritis susceptibility by ∼50% (Fig. 2), and it seemed to have no effect upon severity of inflammation (data not shown). In males, the Pgia9 locus demonstrated a lesser but still significant arthritis suppression (∼10%). The genetic effect of the Pgia8 locus was smaller than that of Pgia9 and either suppressed or promoted arthritis susceptibility, depending on gender. B-15pD females were less susceptible to PGIA; however, arthritis incidence in B-15pD males was higher, albeit for only a limited period of time (Fig. 2). The two loci together in B-15D consomic mice exhibited an additive mode of inheritance for Pgia8 and Pgia9 loci, since PGIA susceptibility in these consomic mice appeared to be an average of disease susceptibilities of congenic mice bearing proximal and distal regions of chr15.
Genetic analysis of chr15-specific F2 hybrid population not only confirmed the presence of two loci on chr15, but also refined loci position. However, the major finding of this study was that the locus controlling murine arthritis (Pgia9) occupies the same genomic position as the locus governing production of IL-6 by spleen cells. IL-6 cytokine has a pleiotropic effect (33), but it is probably best known for its strong stimulation of growth of murine plasmacytomas and myelomas (34). More relevant to our data on the control of IL-6 production by chr15 loci, IL-6 has been demonstrated to decrease aggrecan and collagen type II production by chondrocytes; therefore, cytokine is associated with cartilage deterioration (35). Moreover, IL-6 supports osteoclast formation, which leads to irreversible cartilage and bone erosion in inflamed joints (36).
The hypothesis explaining Pgia9 etiology could be based on the neutrophil cytosolic factor (Ncf4) gene located within the locus. This gene is also part of rat pristane-induced arthritis locus 17 (Pia17), which overlaps with Pgia9 (Fig. 6). Several lines of evidence point to the Ncf4 gene. First, our data indicate that this gene is up-regulated in arthritis paws (37) (Fig. 6). Second, a recent case-control study identifies an association of the NCF4 encoding gene with RA (38) and Crohn’s disease (39). Third, an earlier study of PIA in rats identified Ncf1 as a causative arthritis factor within Pia4 locus on rat chr12 (40). Ncf4 is a close homologue to Ncf1; both proteins are part of NADPH-oxidase complex. We hypothesize that enhanced expression of Ncf4 in macrophages of B-15dD congenic females leads to increased production of reactive oxygen species by macrophages, which have been shown to suppresses T cell responses and arthritis development in mice (41).
Interestingly, reactive oxygen is one of the mechanisms for chromosome aberrations (42, 43). Moreover, TNF-α, which was another cytokine associated with arthritis in this study, is also an inducer of chromosome abnormalities independently of the reactive oxygen species pathway (44). From this point of view, localization of c-myc encoding gene right in the middle of the locus is rather remarkable, since c-myc is a well-known site for illegitimate translocation to IgH chain locus on murine chr12, thereby inducing T(12;15) plasmacytoma formation (45). The major growth factor promoting survival and proliferation of such plasmacytomas is IL-6 (45). As for another type of arthritis, PIA, the formation of plasmacytomas in BALB/c peritoneum, lymph nodes, lamina propria, and Payer’s patches has been demonstrated earlier (45). Interestingly, that correlation between PIA and plasmacytoma formation in immunized BALB/c females was negative (46), like the negative correlation between PGIA and lymphocyte IL-6 production in this study. These similarities further support our hypothesis about the mechanism of murine inflammation.
The Pgia8 and Pgia9 loci on chr15 correspond to multiple locations on rat and human genomes. The Pgia8 locus clusters with multiple murine loci associated with type II CIA (Cia30, Cia31, and Cia32), experimental allergic encephalomyelitis (Eae2), progression of autoimmune arthritis (Paam1), Borrelia burgdorferi-associated arthritis (Bbaa14), and Mycobacterium bovis-induced systemic lupus erythematosus (Mbis3) (Fig. 5). The murine Pgia8 locus corresponds to CIA-controlling locus Cia7 on rat chromosome 2. The Pgia8-syntenic region is located on the short arm of human chromosome 5, a segment linked to several immune deficiencies such as complement cascade-related C6, C7, and C9 deficiencies and severe combined immunodeficiency associated with a mutant IL-7 receptor (Fig. 5). The mouse Pgia9 locus clustered with murine Cia4, Eea32, Eae38 and Bbaa22 loci, and with rat Pia17 locus in PIA (Fig. 5). Homologous human loci include genes associated with Hashimoto’s thyroiditis and juvenile Paget’s disease on human chromosome 8q, and with juvenile RA and neutrophil immunodeficiency syndrome loci on human chromosome 22 (Fig. 5).
Clustering Pgia8 and Pgia9 loci with a dozen loci that control numerous autoimmune diseases in mice and rats might indicate a complex genetic structure within each locus, i.e., the presence of multiple genes that govern disease. Indeed, we detected an interaction between Pgia26, the major locus on mouse chromosome 3 (14), with Pgia8 on chr15 (47). Both Pgia26 and Pgia8 governed arthritis susceptibility, and both loci were female-specific (14). When locus-locus interaction was considered, Pgia26 and Pgia8 each appeared to be a complex multi-chromosomal locus. This phenomenon of locus-locus interaction in PG-immunized F2 hybrids (47) was similar to recent findings reported for CIA (48, 49). The Cia5/Eae3 locus on mouse chromosome 3 has recently been separated into three loci (48), and the Eae2 locus on chr15 into four (Cia26, Cia30, Cia31, and Cia32), when these genomic regions were analyzed simultaneously on chromosomes 3 and 15 (49). QTL locations were confirmed with congenic strains and advanced intercross techniques (48, 49).
In this study, we addressed a mechanism whereby somatic chromosome regions Pgia8 and Pgia9 loci on chr15 function in a gender-specific way and affect the susceptibility and progression of PG-induced autoimmune arthritis. Despite the absence of major PGIA loci on chr15, the fact is that Pgia8 and Pgia9 loci account for ∼15% of arthritis variance in BALB/c × DBA/2 cross, congenic strains and chromosome-specific genetic linkage analyses helped us to confirm that these genomic regions indeed carry genes affecting arthritis susceptibility.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by National Institutes of Health/National Institute of Arthritis and Musculoskeletal and Skin Diseases Grants AR051101, AR045652 and AR040310. V.A.A. is a recipient of the Sontag Foundation and the Arthritis National Research Foundation Awards.
Abbreviations used in this paper: RA, rheumatoid arthritis; CIA, collagen-induced arthritis; chr15, chromosome 15; LRS, likelihood ratio statistic; Ncf, neutrophil cytosolic factor; PG, proteoglycan; PGIA, PG-induced arthritis; PIA, pristane-induced arthritis; QTL, quantitative trait locus (loci).