A knowledge-driven interaction analysis reveals potential neurodegenerative mechanism of multiple sclerosis susceptibility.

Gene-gene interactions are proposed as an important component of the genetic architecture of complex diseases, and are just beginning to be evaluated in the context of genome-wide association studies (GWAS). In addition to detecting epistasis, a benefit to interaction analysis is that it also increases power to detect weak main effects. We conducted a knowledge-driven interaction analysis of a GWAS of 931 multiple sclerosis (MS) trios to discover gene-gene interactions within established biological contexts. We identify heterogeneous signals, including a gene-gene interaction between CHRM3 (muscarinic cholinergic receptor 3) and MYLK (myosin light-chain kinase) (joint P=0.0002), an interaction between two phospholipase C-ß isoforms, PLCß1 and PLCß4 (joint P=0.0098), and a modest interaction between ACTN1 (actinin alpha 1) and MYH9 (myosin heavy chain 9) (joint P=0.0326), all localized to calcium-signaled cytoskeletal regulation. Furthermore, we discover a main effect (joint P=5.2E-5) previously unidentified by single-locus analysis within another related gene, SCIN (scinderin), a calcium-binding cytoskeleton regulatory protein. This work illustrates that knowledge-driven interaction analysis of GWAS data is a feasible approach to identify new genetic effects. The results of this study are among the first gene-gene interactions and non-immune susceptibility loci for MS. Further, the implicated genes cluster within inter-related biological mechanisms that suggest a neurodegenerative component to MS.

Interferon regulatory factor 5 gene variants and pharmacological and clinical outcome of Interferonß therapy in multiple sclerosis.

Interferon-ß (IFNß) therapy is effective in approximately half of the patients with relapsing-remitting multiple sclerosis (RRMS). Clinical non-responders were characterized by an increased expression of IFN response genes before the start of therapy, and a lack of a pharmacologically induced increase in IFN response gene activity. Because Interferon Regulatory Factor 5 (IRF5) is a master regulator of IFN-activity, we carried out a candidate gene study of IRF5 gene variants in relation to the pharmacological and clinical response upon IFNß treatment. We found that patients with the IRF5 rs2004640-TT and rs47281420-AA genotype exerted a poor pharmacological response to IFNß compared with patients carrying the respective G-alleles (P=0.0006 and P=0.0023, respectively). Moreover, patients with the rs2004640-TT genotype developed more magnetic resonance imaging (MRI)-based T2 lesions during IFNß treatment (P=0.003). Accordingly, an association between MRI-based non-responder status and rs2004640-TT genotype was observed (P=0.010). For the rs4728142-AA genotype a trend of an association with more T2 lesions during IFNß treatment and MRI-based non-responder status was observed (P=0.103 and P=0.154, respectively). The clinical relevance of the rs2004640-TT genotype was validated in an independent cohort wherein a shorter time to first relapse was found (P=0.037). These findings suggest a role for IRF5 gene variation in the pharmacological and clinical outcome of IFNß therapy that might have relevance as biomarker to predict the response to IFNß in multiple sclerosis.

Paradoxical inhibition of T-cell function in response to CTLA-4 blockade; heterogeneity within the human T-cell population.

T-cell co-stimulation delivered by the molecules B7-1 or B7-2 through CD28 has a positive effect on T-cell activation, whereas engagement of cytotoxic T-lymphocyte antigen 4 (CTLA-4) by these molecules inhibits activation. In vivo administration to mice of blocking monoclonal antibodies or Fab fragments against CTLA-4 can augment antigen-specific T-cell responses and, thus, therapy with monoclonal antibody against CTLA-4 has potential applications for tumor therapy and enhancement of vaccine immunization. The effects of B7-1 and B7-2 co-stimulation through CD28 depend on the strength of the signal delivered through the T-cell receptor (TCR) and the activation state of T cells during activation. Thus, we sought to determine whether these factors similarly influence the effect of B7-mediated signals delivered through CTLA-4 during T-cell activation. Using freshly isolated human T cells and Fab fragments of a monoclonal antibody against CTLA-4, we demonstrate here that CTLA-4 blockade can enhance or inhibit the clonal expansion of different T cells that respond to the same antigen, depending on both the T-cell activation state and the strength of the T-cell receptor signal delivered during T-cell stimulation. Thus, for whole T-cell populations, blocking a negative signal may paradoxically inhibit immune responses. These results provide a theoretical framework for clinical trials in which co-stimulatory signals are manipulated in an attempt to modulate the immune response in human disease.

A non-synonymous SNP within membrane metalloendopeptidase-like 1 (MMEL1) is associated with multiple sclerosis.

Several single-nucleotide polymorphism (SNP) genome-wide association studies (GWASs) have been completed in multiple sclerosis (MS). Follow-up studies of the variants with the most promising rankings, especially when supplemented by informed candidate gene selection, have proven to be extremely successful. In this study we report the results of a multi-stage replication analysis of the putatively associated SNPs identified in the Wellcome Trust Case Control Consortium non-synonymous SNP (nsSNP) screen. In total, the replication sample consisted of 3444 patients and 2595 controls. A combined analysis of the nsSNP screen and replication data provides evidence implicating a novel additional locus, rs3748816 in membrane metalloendopeptidase-like 1 (MMEL1; odds ratio=1.16, P=3.54 × 10⁻6) in MS susceptibility.

Lymphokine regulation of CD45R expression on human T cell clones.

Whether the expression of higher molecular weight isoforms of the T-200 complex represents different lineages of T cells and/or a sequential stage of the differential pathway of T cells has been unclear. Understanding T cell expression of higher molecular weight isoforms of the T-200 complex (CD45R) may be important because of their association with regulation of immune responses. By direct single cell cloning, we observed a number of long-term T cell clones that expressed CD45RA (2H4). CD45RA expression could be further regulated by ionomycin or the cytokines IL-1 and IL-6, but not IL-2, IL-4, or IFN-gamma. These results indicate that CD45RA expression may define T cell lineages of activated T cells partially controlled by the cytokines IL-1 and IL-6. Further, these results may associate regulatory actions of IL-1 and IL-6 with their ability to increase CD45RA expression in subpopulations of human T cells.

Increased frequency of interleukin 2-responsive T cells specific for myelin basic protein and proteolipid protein in peripheral blood and cerebrospinal fluid of patients with multiple sclerosis.

Equal numbers of CD4+ T cells recognizing myelin basic protein (MBP) and proteolipid protein (PLP) are found in the circulation of normal individuals and multiple sclerosis (MS) patients. We hypothesized that if myelin-reactive T cells are critical for the pathogenesis of MS, they would exist in a different state of activation as compared with myelin-reactive T cells cloned from the blood of normal individuals. This was investigated in a total of 62 subjects with definitive MS. While there were no differences in the frequencies of MBP- and PLP-reactive T cells after primary antigen stimulation, the frequency of MBP or PLP but not tetanus toxoid-reactive T cells generated after primary recombinant interleukin (rIL-2) stimulation was significantly higher in MS patients as compared with control individuals. Primary rIL-2-stimulated MBP-reactive T cell lines were CD4+ and recognized MBP epitopes 84-102 and 143-168 similar to MBP-reactive T cell lines generated with primary MBP stimulation. In the cerebrospinal fluid (CSF) of MS patients, MBP-reactive T cells generated with primary rIL-2 stimulation accounted for 7% of the IL-2-responsive cells, greater than 10-fold higher than paired blood samples, and these T cells also selectively recognized MBP peptides 84-102 and 143-168. In striking contrast, MBP-reactive T cells were not detected in CSF obtained from patients with other neurologic diseases. These results provide definitive in vitro evidence of an absolute difference in the activation state of myelin-reactive T cells in the central nervous system of patients with MS and provide evidence of a pathogenic role of autoreactive T cells in the disease.

Direct ex vivo analysis of activated, Fas-sensitive autoreactive T cells in human autoimmune disease.

The frequency of clonally expanded and persistent T cells recognizing the immunodominant autoantigenic peptide of myelin basic protein (MBP)p85-99 was directly measured ex vivo in subjects with typical relapsing remitting multiple sclerosis (MS). T cells expressing mRNA transcripts encoding T cell receptor (TCR)-alpha and -beta chains found in T cell clones previously isolated from these subjects recognizing the MBPp85-99 epitope were examined. In contrast to frequencies of 1 in 10(5)-10(6) as measured by limiting dilution analysis, estimates of the T cell frequencies expressing MBPp85-99-associated TCR chain transcripts were as high as 1 in 300. These high frequencies were confirmed by performing PCR on single T cells isolated by flow cytometry. MBPp85-99 TCR transcripts were present in IL-2 receptor alpha-positive T cells which were induced to undergo Fas-mediated cell death upon antigen stimulation. These data demonstrate that at least a subpopulation of patients with MS can have a very high frequency of activated autoreactive T cells.

T cell receptor V alpha-V beta repertoire and cytokine gene expression in active multiple sclerosis lesions.

Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system with presumed autoimmune etiology. A recent study has suggested the presence of a restricted T cell receptor (TCR) V alpha repertoire in MS lesions. The presence of such a restricted TCR repertoire at the site of inflammation would have important consequences for the pathogenesis and the ultimate treatment of MS. To further characterize the TCR V alpha and V beta repertoire in MS plaque tissue, we examined a series of 26 histologically well-characterized central nervous system (CNS) tissue specimens from six MS patients as well as samples from five normal postmortem cases and a case of subacute sclerosing panencephalitis. RNA was extracted from frozen sections and cDNAs were amplified by polymerase chain reaction using primers for TCR V alpha (V alpha 1-18) and V beta (V beta 1-19) gene families. This analysis demonstrated a broad TCR V alpha-V beta repertoire in active lesions, while fewer TCR V genes were detected in chronic plaques and control samples. Even though a large number of TCR V alpha and V beta gene segments were present in the majority of active lesions, there were clear differences in the TCR repertoire between plaques from the same case, suggesting that local events influence the TCR repertoire at the level of T cell recruitment or T cell expansion. Examination of cytokine mRNAs demonstrated that IL-1 mRNA was present in the majority of acute and subacute plaques, while IL-2 and IL-4 mRNA were detected in only a few acute lesions. These data demonstrate that the TCR repertoire in MS plaques is polyclonal. However, autoreactive alpha/beta T cells thought to be critical in the initiation of the inflammatory process probably represent a minor fraction of T cells in active MS plaques and may use a limited number of TCR V gene segments for recognition of the autoantigen.

Early signaling defects in human T cells anergized by T cell presentation of autoantigen.

Major histocompatibility complex class II-positive human T cell clones are nontraditional antigen-presenting cells (APCs) that are able to simultaneously present and respond to peptide or degraded antigen, but are unable to process intact protein. Although T cell presentation of peptide antigen resulted in a primary proliferative response, T cells that had been previously stimulated by T cells presenting antigen were completely unresponsive to antigen but not to interleukin 2 (IL-2). In contrast, peptide antigen presented by B cells or DR2+ L cell transfectants resulted in T cell activation and responsiveness to restimulation. The anergy induced by T cell presentation of peptide could not be prevented by the addition of either autologous or allogeneic B cells or B7+ DR2+ L cell transfectants, suggesting that the induction of anergy could occur in the presence of costimulation. T cell anergy was induced within 24 h of T cell presentation of antigen and was long lasting. Anergized T cells expressed normal levels of T cell receptor/CD3 but were defective in their ability to release [Ca2+]i to both alpha CD3 and APCs. Moreover, anergized T cells did not proliferate to alpha CD2 monoclonal antibodies or alpha CD3 plus phorbol myristate acetate (PMA), nor did they synthesize IL-2, IL-4, or interferon gamma mRNA in response to either peptide or peptide plus PMA. In contrast, ionomycin plus PMA induced both normal proliferative responses and synthesis of cytokine mRNA, suggesting that the signaling defect in anergized cells occurs before protein kinase C activation and [Ca2+]i release.

Structural requirements for binding of an immunodominant myelin basic protein peptide to DR2 isotypes and for its recognition by human T cell clones.

Immunodominant T cell epitopes of myelin basic protein (MBP) may be target antigens for major histocompatibility complex class II-restricted, autoreactive T cells in multiple sclerosis (MS). Since susceptibility to MS is associated with the DR2 haplotype, the binding and presentation of the immunodominant MBP(84-102) peptide by DR2 antigens were examined. The immunodominant MBP(84-102) peptide was found to bind with high affinity to DRB1*1501 and DRB5*0101 molecules of the disease-associated DR2 haplotype. Overlapping but distinct peptide segments were critical for binding to these molecules; hydrophobic residues (Val189 and Phe92) in the MBP(88-95) segment were critical for peptide binding to DRB1*1501 molecules, whereas hydrophobic and charged residues (Phe92, Lys93) in the MBP(89-101/102) sequence contributed to DRB5*0101 binding. The different registers for peptide binding made different peptide side chains available for interaction with the T cell receptor. Although the peptide was bound with high affinity by both DRB1 and DRB5 molecules, only DRB1 (DRB1*1501 and 1602) but not DRB5 molecules served as restriction elements for a panel of T cell clones generated from two MS patients suggesting that the complex of MBP(84-102) and DRB1 molecules is more immunogenic for MBP reactive T cells. The minimal MBP peptide epitope for several T cell clones and the residues important for binding to DRB1*1501 molecules and for T cell stimulation have been defined.

Oligoclonal T lymphocytes in the cerebrospinal fluid of patients with multiple sclerosis.

We have investigated the T cell populations in the cerebrospinal fluid (CSF) of chronic progressive multiple sclerosis (MS) patients. Individual T cells from the CSF and blood were cloned before expansion and their clonotypes were defined by analysis of rearranged T cell receptor beta chain and gamma chain genes. 87 T cell clones from blood and CSF of two patients with chronic progressive MS were examined for common TCR gene rearrangement patterns. In one patient, 18 of 28 CSF-derived T cell clones demonstrated common TCR gene rearrangements indicating oligoclonal T cell populations; in the blood, two patterns were found twice among 26 T cell clones. In another patient, 5 of 27 CSF-derived clones had common TCR gene rearrangement patterns. In contrast, no common beta chain rearrangement pattern was found among 67 T cell clones derived from the blood or CSF of a patient with subacute sclerosing panencephalitis, among 20 clones from the CSF of a patient with herpes zoster meningoencephalitis, or among 66 clones from a normal subject. A subject with atypical, fatal MS of 8-mo duration was also studied and did not have oligoclonal T cells in the CSF or blood. These results demonstrate that distinct oligoclonal T cell populations can be found in the CSF immune compartment of subjects with nonmalignant inflammatory disease and they can create a new avenue for the investigation of the specificity of the T cell response within the central nervous system.

Human fetal liver gamma/delta T cells predominantly use unusual rearrangements of the T cell receptor delta and gamma loci expressed on both CD4+CD8- and CD4-CD8- gamma/delta T cells.

Substantial numbers of both alpha/beta and gamma/delta T cells are present in human fetal liver, which suggests a role of the fetal liver in T cell development. The diversity of fetal liver T cell receptor (TCR) gamma and delta chain rearrangements was examined among both CD4+CD8- and CD4-CD8- gamma/delta T cell clones. In addition, TCR delta chain transcripts from three fetal livers were sequenced after polymerase chain reaction amplification of TCR delta chains with V delta 1 or V delta 2 rearrangements. Five of six fetal liver gamma/delta T cell clones had a V delta 2-D delta 3-J delta 3 gene rearrangement with limited junctional diversity; three of these clones had an unusual CD4+CD8- phenotype. V delta 2-D delta 3-J delta 3 gene rearrangements were also common among both in-frame and out-of-frame transcripts from three fetal livers, indicating that they are the result of an ordered rearrangement process. TCR gamma chain sequences of the fetal liver gamma/delta T cell clones revealed V gamma 1-J gamma 2.3, V gamma 2-J gamma 1.2, and V gamma 3-J gamma 1.1 rearrangements with minimal incorporation of template-independent N region nucleotides. TCR gamma chain rearrangements found in these fetal liver T cell clones were different from those that have been observed among early thymic gamma/delta T cell populations, while similar TCR delta chain rearrangements are found among gamma/delta T cells from both sites. These data demonstrate that the fetal liver harbors gamma/delta T cell populations distinct from those found in the fetal thymus, suggesting that the fetal liver is a site of gamma/delta T cell development in humans. These unusual T cell populations may serve a specific function in the fetal immune system.

Expression of costimulatory molecules B7-1 (CD80), B7-2 (CD86), and interleukin 12 cytokine in multiple sclerosis lesions.

Resting autoreactive T cells are present in the circulation of normal individuals without pathologic consequences. In autoimmune animal models, stimulation of these self-reactive T cells in the presence of costimulatory molecules B7-1 results in T cell-mediated autoimmune disease, whereas B7-2 stimulation generates regulatory autoreactive T cells that abrogate disease severity. Thus, reactivation in the brain of myelin-autoreactive T cells by antigen with costimulatory molecules may be a critical event in the pathophysiology of multiple sclerosis (MS), a putative autoimmune disease of central nervous system (CNS) myelin. We investigated the expression of cytokines and costimulatory molecules in a panel of 41 histologically characterized CNS specimens from 15 MS and 10 control cases using semiquantitative reverse transcriptase-polymerase chain reaction and immunocytochemistry. In four cases, vascular CNS infarcts with inflammation were compared with MS plaques from the same brain. We observed increased expression of B7-1 and interleukin (IL) 12p40 in acute MS plaques, particularly from early disease cases but not in inflammatory infarcts. B7-1 staining was localized predominantly to the lymphocytes in perivenular inflammatory cuffs but not the parenchyma. In contrast, B7-2 was expressed predominantly on macrophages both in MS lesions of varied time duration and in inflammatory infarcts. These findings indicate that an early event in the initiation of MS involves upregulation of B7-1 and IL-12, resulting in conditions that maximally stimulate T cell activation and induction of T helper 1-type immune responses.

Functional gamma chain-associated T cell receptors on cerebrospinal fluid-derived natural killer-like T cell clones.

We have derived 33 independent T cell clones from the cerebrospinal fluid (CSF) of a patient with subacute sclerosing panencephalitis using a single T cell cloning method. 6% (2 of 33) of these clones express the T cell receptor gamma (TCR-gamma) protein and are called CSF TCR-gamma clones. Phenotypic analyses of the CSF TCR-gamma clones indicate that they are WT-31-, CD3+, CD4-, and CD8-. The TCR-gamma protein exists on the cell surface as part of an 85-kD disulphide-linked dimer noncovalently associated with the CD3 polypeptides. The CSF TCR-gamma clones have NK-like activity that can be inhibited by anti-CD3 mAbs. Both CSF TCR-gamma clones proliferated in response to anti-CD3 mAbs coupled to Sepharose beads and/or IL-2. Furthermore, stimulation of one of these clones with anti-CD3 mAbs results in a rapid rise in intracellular calcium. These data suggest that T cells bearing the CD3-TCR-gamma protein complex are functional and play a role in the human immune response.

Distinct molecular forms of human T cell receptor gamma/delta detected on viable T cells by a monoclonal antibody.

A second type of TCR molecule has been identified on human and murine T lymphocytes, which involves the protein products of the gamma and delta genes. T lymphocytes bearing this receptor may constitute a separate cell lineage with a distinct immune function. We have produced an mAb, which specifically detects human TCR-gamma/delta in native as well as denatured states, this in contrast to previously used anti-gamma chain peptide sera, which only reacted with denatured protein. The receptor occurs in different molecular forms, with or without interchain disulphide bonds, in which a delta chain may or may not be detected by cell surface iodination. The mAb is reactive with all these receptor forms. Therefore, this antibody could be used to determine the expression of TCR-gamma/delta on viable human T lymphocytes. In normal individuals, TCR-gamma/delta was found on a subset composing 2-7% of CD3+ lymphocytes in peripheral blood and 0.1-1.0% in thymus. The majority of these cells do not express the CD4 or CD8 antigens, although a significant percentage of CD8+ cells was found. TCR-gamma/delta+ cells in peripheral blood are resting lymphocytes, as judged by ultrastructural analysis. T cell clones with different receptor types can display MHC-nonrestricted cytolytic activity, which is shown to be induced by the culture conditions, most likely by growth factors such as IL-2. This strongly suggests that TCR-gamma/delta does not play a role in target cell recognition in MHC-nonrestricted cytotoxicity. The anti-TCR-gamma/delta antibody can specifically induce cytotoxic activity in clones expressing the receptor, but in addition inhibit growth factor induced cytotoxicity, which indicates a regulatory role of the TCR-gamma/delta/CD3 complex in MHC-nonrestricted cytotoxicity.

T cell receptor (TCR) usage determines disease susceptibility in experimental autoimmune encephalomyelitis: studies with TCR V beta 8.2 transgenic mice.

Experimental allergic encephalomyelitis (EAE) is an autoimmune disease that can be induced in laboratory animals by immunization with the major myelin proteins, myelin basic protein (MBP) and proteolipid protein (PLP). We analyzed the role of the T cell receptor (TCR) repertoire in susceptibility to EAE induced by these two autoantigens. Autoreactive T cells induced after immunization with MBP use a limited set of TCR. In contrast, we demonstrate that T cell clones that recognize the encephalitogenic PLP epitope (PLP 139-151) use diverse TCR genes. When the TCR repertoire is limited by introduction of a novel rearranged TCR V beta 8.2 chain in transgenic SJL mice, EAE could be induced in the transgenic mice by immunization with the encephalitogenic epitopes of PLP, but not with the encephalitogenic epitope of MBP. Thus, skewing the TCR repertoire affects the susceptibility to EAE by immunization with MBP but not with PLP. These data demonstrate the biological consequences of the usage of a more diverse T cell repertoire in the development of an autoimmune disease.

Regulatory T cell clones induced by oral tolerance: suppression of autoimmune encephalomyelitis.

Experimental autoimmune encephalomyelitis (EAE) is a cell-mediated autoimmune disease that serves as an animal model for multiple sclerosis. Oral administration of myelin basic protein (MBP) suppresses EAE by inducing peripheral tolerance. T cell clones were isolated from the mesenteric lymph nodes of SJL mice that had been orally tolerized to MBP. These clones were CD4+ and were structurally identical to T helper cell type 1 (TH1) encephalitogenic CD4+ clones in T cell receptor usage, major histocompatibility complex restriction, and epitope recognition. However, they produced transforming growth factor-beta with various amounts of interleukin-4 and interleukin-10 and suppressed EAE induced with either MBP or proteolipid protein. Thus, mucosally derived TH2-like clones induced by oral antigen can actively regulate immune responses in vivo and may represent a different subset of T cells.

Shared human T cell receptor V beta usage to immunodominant regions of myelin basic protein.

Multiple sclerosis (MS) may be an autoimmune disease mediated by T cells specific for a myelin protein. Investigations have demonstrated myelin basic protein (MBP)-reactive T cells that were activated in vivo in MS patients, suggesting that MBP may be a target antigen in MS. The variable (V) region of the T cell receptor (TCR) beta chain was examined among 83 T cell lines from both MS patients and healthy subjects that were reactive with the immunodominant region of human MBP (residues 84 to 102) or with a second immunodominant region of MBP (143 to 168). V beta 17 and to a lesser extent V beta 12 were frequently used in recognition of MBP(84-102) among different individuals. In contrast, V beta 17 was very infrequent among lines reactive with MBP (143-168). These data demonstrate shared TCR V beta gene usage for the recognition of immunodominant regions of the human autoantigen MBP. Such TCR structures may be used as targets for specific immunotherapy in MS.

Double-blind pilot trial of oral tolerization with myelin antigens in multiple sclerosis.

Multiple sclerosis (MS) is thought to be an autoimmune disease mediated by T lymphocytes that recognize myelin components of the central nervous system. In a 1-year double-blind study, 30 individuals with relapsing-remitting MS received daily capsules of bovine myelin or a control protein to determine the effect of oral tolerization to myelin antigens on the disease. Six of 15 individuals in the myelin-treated group had at least one major exacerbation; 12 or 15 had an attack in the control group. T cells reactive with myelin basic protein were reduced in the myelin-treated group. No toxicity or side effects were noted. Although conclusions about efficacy cannot be drawn from these data, they open an area of investigation for MS and other autoimmune diseases.

Effects of oral administration of type II collagen on rheumatoid arthritis.

Rheumatoid arthritis is an inflammatory synovial disease thought to involve T cells reacting to an antigen within the joint. Type II collagen is the major protein in articular cartilage and is a potential autoantigen in this disease. Oral tolerization to autoantigens suppresses animal models of T cell-mediated autoimmune disease, including two models of rheumatoid arthritis. In this randomized, double-blind trial involving 60 patients with severe, active rheumatoid arthritis, a decrease in the number of swollen joints and tender joints occurred in subjects fed chicken type II collagen for 3 months but not in those that received a placebo. Four patients in the collagen group had complete remission of the disease. No side effects were evident. These data demonstrate clinical efficacy of an oral tolerization approach for rheumatoid arthritis.