Section VIII - RHEUMATOID ARTHRITIS Chapter 64 - Etiology and Pathogenesis of Rheumatoid Arthritis Gary S. Firestein Rheumatoid arthritis (RA) is the most common inflammatory arthritis, affecting about 1 percent of the general population worldwide. Although the incidence is surprisingly constant across the globe regardless of geographic location and race, there are some exceptions. For instance, the incidence in China is somewhat lower (about 0.3 percent), whereas it is substantially higher in other groups, such as the Pima Indians in North America (about 5 percent). Because of its prevalence and the ready accessibility of joint samples for laboratory investigation, RA has, in many ways, served as a model for the study of inflammatory and immune-mediated diseases. As such, the information gleaned from these studies has also provided new and unique insights into the mechanisms of normal immunity. Although RA is properly considered a disease of the joints, it is important to recognize that it can exhibit a variety of extra-articular manifestations. These manifestations clearly show that RA has features of a systemic disease that is capable of involving a variety of major organ systems. In some cases, rheumatoid factor (RF) production with the formation of immune complexes that fix complement contribute to these extra-articular findings. Moreover, one of the great mysteries of RA is why the synovium is the primary target. Keys to understanding these phenomena lie in comprehension of the arthrotropism of antigens and inflammatory cells, and in learning what specific receptors and chemotactic gradients exist to focus the inflammation within joints. Despite intensive work, only modest progress has been achieved in determining the cause of RA. Although clues have been provided by detailed studies of immunogenetics of the class II major histocompatibility (MHC) loci and the usage of specific RF genes, the plain truth is that we still simply do not know what causes RA.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Section VIII - RHEUMATOID ARTHRITIS
Chapter 64 - Etiology and Pathogenesis of Rheumatoid Arthritis
Gary S. Firestein
Rheumatoid arthritis (RA) is the most common inflammatory arthritis, affecting about 1 percent of the general population worldwide. Although the incidence is surprisingly constant across the globe regardless of geographic location and race, there are some exceptions. For instance, the incidence in China is somewhat lower (about 0.3 percent), whereas it is substantially higher in other groups, such as the Pima Indians in North America (about 5 percent). Because of its prevalence and the ready accessibility of joint samples for laboratory investigation, RA has, in many ways, served as a model for the study of inflammatory and immune-mediated diseases. As such, the information gleaned from these studies has also provided new and unique insights into the mechanisms of normal immunity.
Although RA is properly considered a disease of the joints, it is important to recognize that it can exhibit a variety of extra-articular manifestations. These manifestations clearly show that RA has features of a systemic disease that is capable of involving a variety of major organ systems. In some cases, rheumatoid factor (RF) production with the formation of immune complexes that fix complement contribute to these extra-articular findings. Moreover, one of the great mysteries of RA is why the synovium is the primary target. Keys to understanding these phenomena lie in comprehension of the arthrotropism of antigens and inflammatory cells, and in learning what specific receptors and chemotactic gradients exist to focus the inflammation within joints.
Despite intensive work, only modest progress has been achieved in determining the cause of RA. Although clues have been provided by detailed studies of immunogenetics of the class II major histocompatibility (MHC) loci and the usage of specific RF genes, the plain truth is that we still simply do not know what causes RA.
It is in the area of pathogenesis that the most progress has been made since the early 1990s: The role of small molecule mediators of inflammation (e.g., arachidonic acid metabolites), cytokines, growth factors, chemokines, adhesion molecules, and metalloproteinases has been carefully defined. These products attract and activate cells from the peripheral blood and evoke proliferation and activation of synoviocytes. The proteases can subsequently lead to behavior similar to that of a localized tumor, invading and destroying articular cartilage, subchondral bone, tendons, and ligaments (Fig. 64-1) . New appreciation of these pathogenic mechanisms has increased awareness that irreversible loss of articular cartilage begins relatively early in the course of RA, and that therapies to suppress the synovitis must be effective early if joint destruction is to be avoided.
SUSCEPTIBILITY TO DEVELOPMENT OF RHEUMATOID ARTHRITIS
Although the etiology of RA remains a mystery, a variety of studies suggest that a blend of environmental and
Figure 64-1 High-power light microscopic view at the pannus-synovium junction. The space between them is shrinkage artifact. Tongues of proliferating tissue have invaded the residual cartilage shown at the upper left. Small blood vessels (BV) are seen just below the cartilage-pannus junction. Small, darkly staining cells, probably lymphocytes (L), are just below the invading surface.
genetic factors is responsible; a contribution of either one is necessary but not sufficient for full expression of the disease. The most compelling example is in monozygotic twins, in whom the concordance rate is perhaps 30 to 50 percent when one twin is affected compared with 1 percent for the general population. The risk for a fraternal twin of a patient with RA also high (about 2 to 5 percent), but this is not more than the rate for other first-degree relatives. Although the immunogenetics is at best incompletely understood, perhaps the most dominant risk factor (albeit not the only one) is the class II MHC haplotype of an individual.
Role of HLA-DR in the Susceptibility to and Severity of Rheumatoid Arthritis
The structure of class II surface molecules on antigen-presenting cells is of great importance in the susceptibility and severity of RA. Chapter 7 presents a detailed discussion of class I and II MHC antigens and of the insights provided by intensive study of the molecular genetics of RA. Initiation of a certain T cell immune response is dependent, in part, on the presence or absence of particular MHC (in this case, DR) allelic products. The MHC also helps determine the T cell repertoire in a given individual; those T cells that cannot recognize the endogenous MHC antigens are
eliminated in fetal development through thymic deletion (see Chapter 8) . At the same time, autoreactive T cells are eliminated. The end result of this process in the thymus is a fine balance in discrimination between self and non-self; it is not surprising in light of the complexity of this process that autoantibodies can occur to various antigens, to a variable extent, in all of us. The role of the MHC in shaping T cell receptor (TcR) gene usage is not a trivial point because, as discussed later, characterizing specific T cell repertoires in the joint or blood of RA patients has been a subject of intense scrutiny, and a bias in the usage of TcR genes has been assumed to be de facto evidence of ongoing T cell activation. Since HLA-DR haplotypes are not randomly distributed among RA patients compared with healthy controls, it is only natural to expect that the T cell repertoire in these patients is skewed as well.
A genetic link between HLA-DR was initially described in the 1970s with the observation that HLA-DR4 occurred in 70 percent of patients compared with 28 percent of controls, giving a relative risk of having RA to those with HLA-DR4 of approximately 4 to 5. [1] HLA-DR4 can be divided into at least five subtypes: Dw4, Dw10, Dw13, Dw14, and Dw15, with varied degrees of association with RA. Later, careful study of the MHC using complementary DNA (cDNA) probes directed against specific alpha- and beta-chains of the DR loci have revealed "susceptibility cassettes" or shared epitopes on the beta-chains of DR that predispose to the development of RA.
The susceptibility to RA is associated with the third hypervariable region of DRbeta chains, from amino acid 70 through 74. [2] [3] The susceptibility epitope is glutamine-leucine-arginine-alanine-alanine (QKRAA) or QRRAA, a sequence found in Dw4 and Dw14 (in which RA is more prevalent) in addition to some DR1beta chains. Current nomenclature attempts to clarify these ambiguities by including information on the specific DRbeta sequences. For instance, the DR4beta chains with the greatest association with RA are referred to as DRB*0401, DRB*0404, DRB*0101, and DRB*1402 ( Table 64-1 (Table Not Available) ). Individuals with DRbeta chains exhibiting other substitutions in this region (e.g., Dw10) have no increased susceptibility to RA. Once the structure of this sequence is considered, up to 96 percent of patients with RA exhibit the appropriate HLA-DR locus in some studies. [4] In certain ethnic and racial groups, the association with DR4 positivity is not nearly as prominent, including Greeks, Pakistanis, and African Americans. [5] [6] HLA-DRB1*0901 homozygosity in Japanese has been associated with RA whereas the heterozygotic state has not. [7]
The QKRAA epitope may also predict the severity of established RA. [8] In one study, 100 percent of patients who inherited two DRB*04 genes had rheumatoid nodules compared with 59 percent of patients with only one gene. Major organ system involvement was observed in 61 and 11 percent of these patients, respectively. These data suggest a "dose response" effect of the HLA genes and imply that severity, rather than susceptibility, is the major contribution of HLA-DR to the disease. This notion is not universally accepted, and the interpretation depends greatly on the inclusion of patients with transient, self-limited arthritis and varies with race and ethnic background. If such patients (who frequently lack HLA-DR4) are included in the analysis, then the correlations between DR4 and RA susceptibility are significantly weakened. [9]
What is so special about the susceptibility cassette? The dose effect of the QKRAA
epitope strongly argues against a role for binding of a specific rheumatoid antigen because DR surface density usually does not alter T cell responses. The crystal structure of HLA-DR1 has
TABLE 64-1 -- Nomenclature for HLA-DR Alleles and Associations with Rheumatoid Arthritis
(Not Available)
From Weyand CM, Hicok KC, Conn DL, Goronzy JJ: The influence of HLA-DRB1 genes on disease severity in rheumatoid arthritis. Ann Intern Med 177:801, 1992.
been solved, and one can infer the three-dimensional structure of other HLA-DR alleles. [10] The region that is so carefully associated with RA (QKRAA) primarily faces away from the antigen-binding cleft of the DR molecule, which determines the specificity of peptides that bind the DR molecule and can be presented to CD4+ helper T cells. This is at some odds with the notion that antigen presentation (and a specific arthrotropic etiologic agent) is responsible for RA, because only individuals with the appropriate antigen-binding groove would be able to mount an immune response that ultimately leads to arthritis. No specific DR-binding peptides (i.e., antigens) have been detected in haplotypes containing the susceptibility cassette. [11] There are several additional caveats: (1) other factors clearly must be involved, because many healthy individuals carry the QKRAA motif and do not develop RA; (2) the converse hypothesis is also plausible, that is, QKRAA might create a "hole" in the immune response due to a DR topography that prevents an arthrotropic agent from binding, therefore preventing an appropriate T cell-mediated response; (3) the association between the shared epitope and RA might have nothing to do with antigen recognition and might function by shaping the T cell repertoire in the thymus [12] ; and (4) specific DR sequences might alter intracellular MHC trafficking and antigen loading, thereby indirectly affecting antigen presentation in a nonspecific fashion. [13]
HLA-DQ Associations
DQw7 has also been cited as an important marker gene in RA. [14] The concept that the HLA-DRB1 locus is associated with protection from RA and that the actual arthritogenic peptide-presenting molecule is HLA-DQ has also been proposed. [15] This hypothesis was based on the observation that the presentation of collagen II peptide by the H-2Aq molecule (equivalent to DQB1) in mice determined the susceptibility to collagen arthritis, but polymorphic expression of the functional H-2Eb gene (equivalent to DRB1) conferred protection. Perhaps the DQ/DR haplotype is responsible for RA predisposition in humans and the polymorphism of the HLA-DR4 allele determines the degree of protection. This observation is supported by the fact that HLA-DQB1*0302 (DQw8) and HLA-DQB1*0301 (DQw7) are in linkage disequilibrium with most DR4 haplotypes and the majority of HLA-DR4+ RA patients express one of these alleles. [16] In other studies, DQB1*0301 was associated with an increased risk of IgM RF positivity. [17]
One possible mechanism for protection is through the presentation of DRB1 HV3
peptide by RA-associated DQ molecules. To investigate this, DQ8 transgenic mice were produced that express a functional DQB1*0302/DQA1*0301 dimeric molecule in the absence of endogenous mouse class II molecules. [18] These mice were immunized with a series of DRB1 HV3 peptides corresponding to most HLA-DRB1 alleles, and in vitro T cell responses against the immunizing peptides were assessed. Although HV3 peptides derived from nonassociated DRB1 molecules were highly immunogenic in DQ8 transgenic mice, the HV3 peptides derived from RA-associated DRB1 alleles failed to induce a DQ8-restricted T cell response.
Immunoglobulins
The class II MHC associations described previously primarily implicate cellular immune responses. There are some associations noted in the humoral system, albeit not as striking. For example, a particular immunoglobulin kappa genotype appears to confer a risk of RA. [19] Although it cannot be considered an immunogenetic determinant, strictly speaking, deficient galactosylation of immunoglobulin might also be a risk factor for the development of autoimmune diseases, including RA. [20] The immunoglobulin G (IgG) glycosylation defect is present before the onset of RA and is especially prominent on the IgG1 , IgG2 , and IgG4 isotypes of RF. [21] The cause might be reduced galactosyl transferase activity in RA B cells. Deficient galactosylation is also thought to be predictive in some settings for patients with early synovitis who will progress ultimately to full-blown RA. [22] Autoantibodies lacking galactose are more pathogenic than those with a normal glycosylation in murine collagen arthritis. Passive transfer of arthritis in T cell-primed mice is exacerbated by infusions of IgG fractions that lack galactose, whereas the galactosylated fraction is nonpathogenic. [23]
Cytokine Production and Promoter Polymorphisms
Given the importance of cytokines in RA (see later), some cytokine genotypes might be associated with RA. The most intriguing evidence relates to tumor necrosis factor-alpha (TNF-alpha) production in chronic inflammatory diseases (see also Chapter 20) . This proinflammatory factor is thought to be a major cytokine in the pathogenesis of RA, and the TNF genes are located in the MHC locus on chromosome 6 in humans. There are high- and low-TNF-producing strains of mice associated with specific restriction fragment polymorphisms, and the low-TNF phenotype correlates with susceptibility to lupus-like disease. [24] The low-TNF phenotype also appears to be associated with nephritis in human systemic lupus erythematosus (SLE). [25] In light of other known associations within the MHC locus in RA, it would clearly be of great interest to know the molecular basis of such phenotypes and whether they confer susceptibility to RA. For instance, there are several polymorphisms of the TNF-alpha promoter, including two at positions -238 and -308. Although there are no differences in terms of susceptibility, the age of onset of RA is associated with the former and the presence of rheumatoid nodules with the latter. [26] In a second study, TNF microsatellite polymorphisms were associated with RA. [27]
With regard to other cytokines, allelic variations have been described for interleukin 1alpha (IL-1alpha) [28] and interleukin 1 receptor antagonist (IL-1Ra), [29] a naturally occurring IL-1 antagonist. One of the IL-1alpha polymorphisms
is associated with juvenile RA. As with TNF-alpha-associated phenotypes in inflammatory diseases, additional studies are needed to determine the functional relevance of these observations. IL-10 promoter polymorphisms thus far do not appear to correlate with RA susceptibility.
Gender
RA is one of many chronic inflammatory diseases that predominate in women. The ratio of female to male patients (2:1 to 4:1) is significant yet not nearly as high as that found in Hashimoto's thyroiditis (25:1 to 50:1), systemic lupus erythematosus (9:1), or even autoimmune diabetes mellitus (type I, 5:1). The basis of the gender differences is not known but presumably is related to effects of the hormonal milieu on immune function.
Still incompletely understood are the mechanisms underlying the effect of pregnancy on RA. Pregnancy usually is associated with remission of the disease in the last trimester. [30] More than 75 percent of pregnant patients with RA improve, starting in the first or second trimester; but 90 percent of these experience a flare of disease associated with a rise in RF titers in the weeks or months after delivery. [31]
The mechanism of protection is not yet defined but might be related to the production of large amounts of the suppressive cytokines like IL-10 during pregnancy or alterations in cell-mediated immunity. [32] In murine proteoglycan-induced synovitis, for instance, the development of arthritis is related to the production of antiproteoglycan antibodies. Pregnancy decreases clinical arthritis in these animals even though titers of pathogenic antibody are unchanged. [33] A possible relationship between the alleviation of RA symptoms during the last trimester of pregnancy and immunogenetics may be supported by the observation that during pregnancy, alloantibodies in the maternal circulation develop against paternal HLA antigens. More recently, maternal-fetal disparity in HLA class II phenotypes was correlated with pregnancy-induced remission. Over three fourths of pregnant women with maternal-fetal disparity of HLA-DRB1, DQA, and DQB haplotypes had significant improvement, whereas disparity was only observed in one fourth of women whose pregnancy was characterized by continuous active arthritis. [34] This also suggested that suppression of maternal immune responses to paternal HLA haplotypes might be protective.
POSSIBLE DIRECT CAUSES OF RHEUMATOID ARTHRITIS
As reviewed earlier, one or multiple genetic factors may predispose an individual to develop RA. Attempts to identify specific environmental factors as etiologic agents have generally met with disappointment, in part because no single agent has consistently been implicated. A guess, based on available data, is that several environmental stimuli, possibly viruses or retroviruses, infect an individual with the appropriate genetic background and through some mechanism the inflammatory response is focused in joints ( Table 64-2 ). After gaining a toehold there, the synovitis persists even in the absence of the offending agent because of local autoimmunity and the cyclic automaticity that enables the disease to become self-perpetuating.
Infectious Agents
Bacteria, Mycobacteria, Mycoplasma, and Their Components
There is no convincing evidence that a pyogenic bacterium or mycobacterium causes RA, although animal models of arthritis in which immunization with bacterial cell walls (e.g., streptococcal cell wall arthritis in Lewis rats) or killed mycobacteria (e.g., adjuvant arthritis in Lewis rats) exhibit many clinical and histologic features of RA. Extensive searches for such an organism in synovial tissue or joint effusions have been negative. However, using sensitive polymerase chain reaction techniques to identify bacterial DNA in synovial tissue has demonstrated that a high percentage of RA and reactive arthritis patients contain bacterial nucleotide sequences. [35] [36] The bacteria identified are not unique and generally represent a cross-section of skin and mucosal bacteria, including Acinetobacter and Bacillus spp. It is possible that the synovium functions as an adjunct to the reticuloendothelial system in arthritis and local macrophages accumulate circulating bacterial products. Although the bacteria or their products might persist in the joint, it is likely a secondary phenomenon that can contribute to established synovitis. [37]
Similarly, considerable attention has been directed at a potential role for Mycoplasma and Chlamydia in arthritis. Mycoplasma-derived superantigens, such as from Mycoplasma arthritidis, can directly induce T cell-independent cytokine production by macrophages and can exacerbate or trigger arthritis in mice immunized with type II collagen. [38] Despite this and other circumstantial evidence, serious efforts to identify Mycoplasma organisms DNA in joint samples have been essentially negative, and there is no direct evidence to support these organisms as etiologic agents. [39] The situation is similar with Chlamydia: a synovitis can be explained by this organism in a small minority of patients at best.
The striking similarity of histopathologic changes in RA and in joints of occasional
patients with Lyme disease
TABLE 64-2 -- Possible Infectious Causes ofRheumatoid Arthritis
Infectious Agent Potential Pathogenic Mechanisms
Mycoplasma Direct synovial infection; superantigens
Parvovirus B19 Direct synovial infection
Retroviruses Direct synovial infection
Enteric bacteria Molecular mimicry (QKRAA)
Mycobacteria Molecular mimicry (proteoglycans, QKRAA)
Epstein-Barr virus Molecular mimicry (QKRAA)
Bacterial cell walls Macrophage activation
caused by the spirochete Borrelia burgdorferi leaves open the possibility that an as-yet-unappreciated or unknown organism is causative. Lyme disease is different from RA in its immunogenetic background. When HLA-DR4-positive patients are excluded from consideration, a secondary association with HLA-DR2 is noted to be an independent, dominant marker for Lyme disease. The general presumption is that a spirochetal antigen can be presented by DR2 much like a rheumatoid agent could bind to DR4 (and QKRAA). The role of class II molecules and cellular immunity in the pathogenesis of Lyme arthritis, however, is unclear since severe combined immunodeficiency (SCID) mice develop a progressive tenosynovitis after Borrelia infection despite the virtual absence of T cell function. [40] However, Lyme arthritis provides a clear example of an infectious arthritis caused by an organism that can be very difficult to detect or grow from joint tissue. It is not known if chronic Lyme arthritis requires continued antigen exposure or whether, as might occur in RA, the inciting organism can be eradicated without necessarily curing the synovitis.
Epstein-Barr Virus, dnaJ Proteins, and Molecular Mimicry
Epstein-Barr virus (EBV) has been indirectly implicated in the pathogenesis, if not the etiology, of RA. The EBV is a polyclonal activator of B lymphocytes that could result in the production of RF, and, as reiterated later in more detail, rheumatoid macrophages and T cells combine to generate a defect in suppression of EBV proliferation in human B cells. Rheumatoid patients appear to have higher levels of EBV shedding in throat washings, an increased number of virus-infected B cells in the circulating blood, higher levels of antibodies to the EBV antigens, and abnormal EBV-specific cytotoxic T cell responsiveness compared with controls. [41] Patients with RA have problems with the control and elimination of EBV-transformed lymphocytes [42] ; this has helped fuel speculation that a lymphocyte defect is the principal triggering event in this disease and that abnormal control of EBV is pathogenic.
Additional intriguing data implicating EBV in RA are derived from sequence homology between the susceptibility cassette in HLA-DR4 and -DR1 and the EBV
glycoprotein gp110. Gp110, like DRB*0401, contains the QKRAA motif, and patients with serologic evidence of a previous EBV infection have antibodies against this epitope. [43] An inference from these data can be that T cell recognition of EBV epitopes in some patients with HLA-DR4, -DR14, or -DR1 may cause an immune response directed at innocent bystander cells through "molecular mimicry," whereas in those with other class II MHC alleles, no cross-reactivity with EBV proteins would exist. This hypothesis could potentially account for disease perpetuation in the absence of active infection in patients with a specific MHC genotype, a scenario that is consistent with many observations in the chronic rheumatoid joint. However, the data are primarily circumstantial, and gp110 is only one of many xenoproteins that contain QKRAA. The E. coli dnaJ protein, which is a bacterial heat shock protein (HSP), also contains the cassette and might represent a potential link between gut bacteria and chronic arthritis. RA T cells, especially synovial fluid T cells, but not normal peripheral blood cells have increased proliferative responses to this protein, perhaps supporting the molecular mimicry link between any of a variety of QKRAA-containing proteins and arthritis. [44]
Parvovirus
A fraction of patients with early RA have serologic evidence of a recent infection with B19. [45] Anti-human parvovirus (IgG) levels later decline but are still present 8 months or more after the onset of symptoms. Despite these cases, it is important to point out that very few rheumatoid patients have evidence of such a coincident infection; in a total of 69 patients with RA, only 4 acquired the parvovirus infection near the time of onset of their RA. Using sensitive polymerase chain reaction methods to detect B19 genes in synovial tissue, however, 75 percent of RA synovium samples were positive compared with 17 percent of non-RA controls. [46] Many of the B19-positive RA patients did not have serum anti-B19 antibodies. More recently, immunohistochemical evidence of the B19 protein VP-1 was detected in 27 of 27 patients with RA but not in other forms of arthritis. [47] Sublining mononuclear cells, rather than the intimal lining, demonstrated the most intense staining. The majority of the patients also had detectable B19 DNA in their synovium. Furthermore, evidence of infective B19 virus was provided by coculture experiments with uninfected cell lines and RA synovial cells. In other studies, however, there was no evidence of the B19 genome in synovial fluid or synovial fluid cells, [48] and serologic evidence for B19 infection was present in only 3 percent of patients with early inflammatory polyarthritis. [49] How is one to interpret these disparate data? Unfortunately, a final conclusion cannot be drawn regarding the role of B19 in RA. This virus clearly causes a distinctive arthritis in children and might be responsible for a subset of patients with an RA-like disease.
Other Viruses
Lentiviruses are a subfamily of retroviruses that derive their name from the slow time course of the infections they cause in humans and animals. Pathologic changes in lentivirus infections are, for the most part, indirectly mediated by the immune and inflammatory responses of the host. Finding cells infected by virus is often extremely difficult. An epidemic deforming arthritis in goats and sheep is caused the Visna lentivirus. The pathogenesis of the disease appears to be infection of monocytes that subsequently migrate to the synovium; resultant cytokine production leads to the accumulation of lymphocytes and other cells. [50] Of interest, the cytokine profile of this disease is very similar to RA in that T cell products are low and macrophage
cytokines are abundant. [51] Hence, a "Trojan horse" mechanism can be invoked; the viral genome can be concealed within monocytes and transported without detection to other
sites. Restricted viral gene expression is the underlying mechanism in the persistence and spread of lentiviruses and the slow evolution of the disease that they produce.
Although similar retroviral infections have been suggested many times as the cause of RA, extensive searches for potential agents have not been fruitful. [52] This does not rule out the possibility that difficult-to-detect agents might be present, or even that endogenous retroviruses might play a role. In fact, endogenous retroviruses are abundant in inflamed and normal synovium, although certain transcripts are differentially expressed in RA cells. [53] Some indirect studies are suggestive of retroviral infection, such as the demonstration of zinc-finger transcription factors in cultured synoviocytes. [54] In addition, the pX domain of one human retrovirus, human T-lymphotropic virus-1 (HTLV-1), causes synovitis in transgenic mice, [55] and synoviocytes from patients infected with HTLV-1 express some features of a transformed phenotype, with increased proliferation and cytokine production. [56]
Because rubella virus and the rubella vaccine can cause synovitis in humans, there is interest in this virus as a triggering agent. In one series, 21 instances were reported in which live rubella virus was isolated from synovial fluid obtained from six patients with inflammatory oligoarthritis or polyarthritis over a period of 2 years in the absence of firm clinical evidence of rubella. [57] However, none of these patients had the classic polyarticular involvement seen so often in RA; most had an oligoarthritis involving large joints. As with B19 infection, it is possible that a subset of patients with chronic polyarthritis have disease due to direct infection with wild-type or attenuated rubella virus.
In summary, the hypothesis that one or more viral infections may serve as a triggering agent in the genetically susceptible host is both appealing and intellectually satisfying. Qualities of candidate viruses should be similar to those of the lentivirus, which has a restricted expression within cells and can remain hidden from defense mechanisms in the host. It is possible that just a small alteration in T cell reactivity or responsiveness induced by a viral infection, or generation of a neoantigen by insertion of a viral genome into the host itself, could be sufficient to trigger the disease. Alternatively, certain specific HLA-DR haplotypes associated with RA might not permit efficient presentation of pathogenic viral or retroviral antigens, and subsequent infection of synovial cells could lead to autonomous activation (as observed with HTLV-1-infected synoviocytes in humans or lentivirus-infected macrophages in goats) and an inability to mount an appropriate immune response that would clear infected cells. It is also possible that subsequent infection by a ubiquitous second virus (such as EBV) capable of generating polyclonal B cell stimulation would be sufficient to amplify the immune response and, in the immunogenetically appropriate host, enable it to become self-perpetuating through cross-reactivity of epitopes.
Autoimmunity
Type II Collagen Autoimmunity
The discoveries that immunization with type II collagen can cause arthritis in rats and mice and that the disease
TABLE 64-3 -- Potential Autoantigens in Rheumatoid Arthritis
HLA-DR (QKRAA)
Heat shock proteins
Immunoglobulins (IgG)
Cartilage antigens
Type II collagen
gp39
Cartilage link protein
Proteoglycans
can be passively transferred by IgG fractions containing anticollagen antibodies [58] or by transfer of lymphocytes from affected animals [59] have spawned extensive experiments that illustrate the antigenicity of collagen, the arthrotropic nature of the disease produced, and the dependence of experimental animals on immune response genes for reactivity. It is clear that functional T cells are necessary to initiate a collagen-induced arthritis that a major immunogenic and arthritogenic epitope on type II collagen resides in a restricted area of the type II collagen chains. [60] Furthermore, susceptibility to collagen-induced arthritis is directly linked to the class II MHC locus. [61]
Most data in humans are consistent with the hypothesis that RA is not caused by the development of antibodies to type II collagen but that the inflammatory response is amplified by the production of these antibodies (see Table 64-3 for a list of potential autoantigens in RA). Sera from patients with RA contain antibody titers to denatured bovine type II collagen that are significantly higher than those found in control sera [62] ; however, there is no difference in antibody titers to native collagen, suggesting that the denatured form generated after the breakdown of connective tissue might serve as the immunogen. It has been suggested that antibodies against collagen have pathogenic capability in RA, especially among the IgG3 subclass of anti-type II collagen antibodies. [63] It has also been demonstrated that the anti-collagen antibodies purified from the sera of patients with RA can activate complement, generating C5a when these antibodies became bound to cartilage. This has relevance to the observations that anticollagen antibodies can be eluted from rheumatoid articular cartilage. [64] In addition, isolated synovial tissue B lymphocytes actively secrete anti-type II collagen antibodies in almost all patients with seropositive RA, whereas articular cells from non-RA patients do not. [65]
Heat Shock Proteins
The HSPs are a family of mainly medium-sized (60 to 90 kD) proteins produced by cells of all species in response to stress. These proteins have conserved amino acid sequences; for example, certain HSPs of Myobacterium tuberculosis have a 65 percent sequence homology with HSPs of humans. The HSPs may facilitate intracellular folding and translocation of proteins as they protect from insults induced by heat, bacteria, and oxygen radicals. Immunity against HSPs might directly contribute to synovitis and joint destruction in the adjuvant arthritis model in rats in which T lymphocytes recognize an epitope of mycobacterial HSP65 (amino acids 180
through 188). Some of these cells also recognize cartilage proteoglycan epitopes, [66] perhaps explaining the targeting of the joint. Experimental arthritis can be transferred to naive animals by such reactive T cells, and protection is conferred on rats by preimmunization with mycobacterial HSPs.
Some patients with RA have elevated levels of antibodies to mycobacterial HSPs, especially in synovial fluid. [67] The majority of T cell clones isolated from RA synovial fluid with specificity to mycobacterial components have the gammadelta-T cell receptor (instead of the more common alphabeta form) and express neither CD4 nor CD8 surface antigens. Freshly isolated synovial fluid T cells from patients with RA demonstrate a brisk proliferative response to both the acetone-precipitable fraction of M. tuberculosis recombinant 65-kD HSP. [68] However, proliferation in response to other recall antigens, like tetanus toxoid, is not increased. Synovial fluid mononuclear cells activated by 60-kD mycobacterial HSP inhibit proteoglycan production by human cartilage explants. [69] This effect is dependent on the generation of cytokines like IL-1 and TNF-alpha by the activated cells. Human 60-kD HSP is expressed in the synovium, although the amount expressed per cell appears to be similar in osteoarthritis (OA), RA, and normal tissue. [70]
How could HSPs be related to the cause of RA? As noted earlier, the possibility exists of cross-reactive epitopes, and HSPs from bacteria can mimic numerous human proteins. At least one HSP, the E. coli dnaJ protein, contains the QKRAA motif; this raises additional questions about the relationship between HSPs and immunity directed against the HLA-DR molecule vis-a-vis an autologous mixed lymphocyte reaction. Within this paradigm, the identity of a specific etiologic agent could vary among patients; the critical step required to initiate RA would reside in the combination of an appropriate class II MHC haplotype and a pathogen that expresses cross-reactive HSPs.
Other Cartilage Autoantigens
Several cartilage components besides type II collagen have been implicated as potential autoantigens in RA. Among the most interesting is cartilage glycoprotein gp39. Several gp39 peptides are capable of binding to the HLA-DR*0401 molecule and stimulating proliferation of T cells from patients with RA. [71] BALB/c mice, which are often resistant to experimental arthritis, develop a polyarticular inflammatory arthritis after immunization with gp39 in complete Freund's adjuvant. Other examples
of potential cartilage autoantigens include proteoglycans, aggrecan, cartilage link protein, and other types of collagen.
Rheumatoid Factor: Cause or Effect?
The identification and characterization of RF as an autoantibody was the first direct evidence that autoimmunity might play a role in RA. For many years, immune complexes comprising RF and other immunoglobulins were thought to contribute mightily to RA as a primary pathogen (see Chapter 10) . One piece of indirect evidence suggesting that this is not necessarily the case, documented in 1957, was that RA and agammaglobulinemia caused by inactive B lymphocytes occurred simultaneously in a number of patients. [72] Thus, B cells are not essential in the development of RA. Of note, one of the key animal models used to study RA, collageninduced arthritis in mice, is dependent on B cells because B cell-deficient mice are not susceptible (hence raising issues about how one interprets this animal model). [73] Generally, current thinking suggests that cellular immune processes predominate, whereas RFs exacerbate rheumatoid synovitis through activation of complement and the formation of immune complexes in synovial fluid. This acute inflammatory response can ultimately recruit additional cells to the joint. Even so, one can learn a great deal about RA by examining its most characteristic and reproducible laboratory abnormality: the presence of RF in blood and synovial fluid.
Archived serum samples have shown that although some patients have RF present in their sera before the onset of any symptoms of arthritis, this does not usually occur. However, the report of cross-reactive idiotypes among monoclonal IgM proteins with anti-IgG activity, [74] and the possible implications of expression of this germline by many individuals, has continued to stimulate interest in the possibility that antibodies against IgG might contribute to the triggering event. Some patients who are seronegative but otherwise have a clinical diagnosis consistent with RA have "hidden" RFs in their 19-S or 7-S serum fractions, and these can be identified by antibody specific for the major RF cross-reactive idiotype. [75]
The RFs produced by RA B cells differ from those produced by B cells from healthy individuals or from patients with paraproteins. [76] The avidity of RF for the Fc portion of IgG is several orders greater in RA than in Waldenstrom's macroglobulinemia or in cryoglobulins. The germline-derived RFs are produced by immature CD5-positive B cells, and many paraproteins expressed by malignant B cells (like Waldenstrom's macroglobulinemia) are derived from the germline. In addition, some normal B cells in adult human tonsil tissue express and synthesize germline-encoded RFs, although they do not appear to secrete the protein. RFs produced by RA B cells are distinct in that these proteins are often not encoded by germline sequences. Instead, their sequence appears to be derived through rearrangements and somatic mutations that occur during life. RF analysis in synovial membrane cultures from patients with a variety of diseases have indicated that only cells from patients with seropositive RA synthesize RF spontaneously. [77] IgM RF represents about 7 percent of the total IgM produced by cells, and IgG RF represents 3 percent of IgG synthesized in those cultures.
The expression of any particular idiotype on RFs (or other immunoglobulins, for that matter) is under genetic control. This limited response is related to restriction of the number of relevant or expressible V genes available in the germline. [78] RF in RA primarily use the VH3 gene and a variety of VL genes, whereas natural antibodies
use VH1/VH4 and the Vk3 genes. The kappa light chain repertoire expressed in RF-producing cells isolated from a patient with chronic RA revealed enrichment for two specific Vkappa genes, known as Humkv325
and Humkv328, which are also frequently associated with RF factor paraproteins. [79] However, the kappa variable domains contained many somatic mutations and nongermlike encoded nucleotides. Based on the extent of substitutions, the selection and production of these specific RFs was likely due to antigenic drive rather than derived directly from the germline as is the case with many paraproteins. Additional RFs have been identified with characteristics similar to an antigen-driven response, although some examples of germline RFs have also been isolated from RA synovium. A crystal structure of one IgM RF bound to IgG showed a key contact residue of the RF with the Fc portion of IgG containing a somatic mutation, thereby supporting the notion that the mutations are related to affinity maturation. [80] Of interest, only a few amino acid contact points were identified, and it was suggested that this antibody might have arisen in response to another, as-yet-unidentified, antigen and the reactivity with IgG Fc was due to cross-reactivity. The binding site did not include any contacts with carbohydrates.
The ability to generate high-affinity RFs similar to those found in RA is tightly controlled. Transgenic mice engineered to produce RFs indicate that high-affinity RFs are deleted in the thymus, whereas the low-affinity RFs are not. [81] Therefore, mechanisms for inactivation of higher-affinity RF B cell clones must exist. Exposure of transgenic mice expressing a human IgM RF to soluble human IgG in the absence of T cell help causes antigen-specific B cell deletion in 2 to 3 days. [82] However, B cell activation and sustained RF secretion does occurs if T cell help is provided simultaneously. This suggests that the high-affinity RF production, as occurs in RA, is associated with T cell activation, whereas these RF-producing cells are deleted in normal individuals.
Spontaneous Autoimmunity in a Transgenic Mouse
One the most interesting new arthritis models demonstrated that antigen-specific immunity against a seemingly irrelevant nonarticular antigen can lead to destructive arthritis. [83] A transgenic mouse strain was developed by breeding nonobese diabetic mice (which develop autoimmune diabetes) with a strain that expressed a transgenic T cell receptor recognizing bovine pancreas ribonuclease. The cross-bred animals developed spontaneous polyarticular inflammatory arthritis after 1 month that required both CD4+ T cells and B cells. Although hypergammaglobulinemia was present, the mice lacked RFs. Hence, joint-specific disease can occur in a host with a preexistent autoimmune diathesis and systemic self-reactivity. The mechanism of disease relates to the fortuitous formation of antibodies to the ubiquitous enzyme glucose-6-phosphate isomerase. Because the model is now known to be merely an immune complex-mediated process, its relevance to RA is not certain.
SYNOVIAL ABNORMALITY
Cellular Infiltration into the Synovium
The primary site of immune activation in RA is the synovium. Infiltration of the synovium with mononuclear cells, especially T cells and macrophages, and synovial intimal lining hyperplasia are hallmarks of the disease (Fig. 64-2) . In this section, histopathologic and functional aspects of the inflamed joint are reviewed.
Synovial Intimal Lining
The synovial intimal lining is a loosely organized collection of cells that form an interface between the synovium and the synovial fluid space. Two major cell types are found in the lining: a macrophage-like cell known as a type A synoviocyte and a fibroblast-like cell called a type B synoviocyte. The former are derived from the bone marrow and express macrophage surface markers and abundant HLA-DR, whereas the latter express little if any class II MHC antigens, are devoid of macrophage
Figure 64-2 Histopathologic appearance of rheumatoid arthritis (RA) synovium. Intimal lining hyperplasia, angiogenesis, and a prominent mononuclear cell infiltrate are present. (Courtesy of Dr. P.-P. Tak.)
markers, and have a scant endoplasmic reticulum. The numbers of type A and B cells are relatively equal in normal synovium. There is an absolute increase in both cell types in RA, although the percentage increase in macrophage-like cells appears to be greater. In addition, the type A synoviocytes tend to accumulate in the more superficial regions of the intimal lining.
The synovial intimal lining cells are loosely associated with each other and lack tight junctions and a definite basement membrane. The increase in cell number can be quite substantial. In the normal joint, the lining is only 1 to 2 cell layers deep, whereas in RA it is usually 4 to 10 cells deep (and sometimes over 20). Although macrophages are terminally differentiated cells that presumably do not divide in the joint, it is presumed that the mesenchymally derived type B synoviocytes divide locally in response to the proliferative factors generated by the activated immune response. It is reasonable to assume that platelet-derived growth factor (PDGF),
TNF-alpha, and IL-1 produced by many different cells combine with products of arachidonic acid metabolism to generate proliferation of these cells presumed to be synovial fibroblasts.
One problem with this hypothesis is that studies attempting to quantify active cell division in the rheumatoid synovium rarely show mitotic figures, and thymidine uptake occurs in only a small percentage of synovial cells. [84] Using a monoclonal
antibody that recognizes dividing cells, an even lower rate of cell division ( 0.05 percent) was found. [85] More recently, however, a much higher percentage of cells that express the cell cycle-specific antigen proliferating cell nuclear antigen were identified in RA lining and the nuclear rearrangement apparatus (associated with cell division) was compared with that of OA. [86] This correlated with the expression of the protooncogene c- myc by lining cells, a gene that is intimately linked with fibroblast proliferation. [87] The reason for the differences with earlier studies is not clear, although it is possibly related to variations in the stage of the disease or the location within the joint. When all the data are considered, it is likely that type B cell proliferation is excessive in RA and contributes to lining hyperplasia.
With the use of enzymatically dissociated rheumatoid cells, adherent cells from rheumatoid synovial tissue can be divided into three types by the use of monoclonal antibodies. [88] As might be expected from studies of intact tissue, one type of cell is macrophage-like; these cells have DR antigens, Fc receptors, and monocyte lineage differentiation antigens and are capable of phagocytosis. They have a limited life span in vitro, rarely surviving more than a few weeks even in the presence of exogenous growth factors or colony-stimulating factors. A second distinctive cell population is nonphagocytic and has abundant DR antigens but lacks IgG Fc receptors, monocyte lineage antigens, or fibroblast-associated antigens. Many cells in this group have a dendritic morphology. A third type is defined by the presence of antigens expressed primarily on fibroblasts and by the absence of phagocytic capability, demonstrable DR antigens, or antigens of the monocytic lineage. When the enzymatically dispersed cells are cultured for several passages, it is this last cell type that ultimately survives and proliferates, resulting in a relatively homogeneous population of fibroblast-like cells that are presumed (but not proved) to be derived from the type B synoviocytes in the intimal lining.
A successful attempt to clone dissociated rheumatoid synovial cells and place them in long-term culture has supported this classification. [89] Each type of cell could be cloned, but the macrophage-like and fibroblast-like cells grew slowly, with a doubling time of 5 to 7 days. The dendritic cells had a doubling time of 1 to 2 days. Fibroblast-like cells generated a dendritic appearance when they were incubated with prostaglandin E2 (PGE2 ).
[90] After removal of the prostaglandin, the cells reverted to normal appearance, unlike the dendritic-like cells that maintained the stellate appearance through their slow doubling times. A study of cloned synovial cells indicated that the dendritic-like cells produced significant amounts of IL-1.
The fibroblast-like cells that can be grown from the dispersed cells can be passaged for several months in vitro. Their doubling time is rapid at first, perhaps due to the presence of cytokines produced by contaminating macrophages in the culture or due to a carryover effect from the synovial milieu. Over time, the rate of proliferation slows, and after 12 to 15 passages, the cells gradually become senescent and ultimately cease to grow. Synovial fibroblasts from RA have some characteristics
reminiscent of tumors or transformed cells, a notion that seems to fit well with the concept of RA as a locally invasive mesenchymal tumor (see later).
Synovial Lymphocyte Infiltration
In chronic RA, the synovium contains a collection of T lymphocytes that can lead to an organizational structure that often resembles a lymph node. The distribution of lymphocytes in the tissue varies from discrete lymphoid aggregates to diffuse sheets of mononuclear cells, with the most prominent location for T cells being the perivascular region. These collections consist of small, CD4-positive memory T cells (CD45RO positive) with scant cytoplasm. Peripheral to these foci is a transitional zone with a heterogeneous mixture of cells, including lymphocytes, occasional undifferentiated blast cells, plasma cells, and macrophages (see Fig. 64-2) . It is likely that much intercellular communication by mediators takes place here, and it has been shown that activated T cells bind to synovial cells in vitro, a process mediated by their lymphocyte function-associated antigen. [91] Immunoelectron microscopic study of the distribution of T cell subsets has shown that the majority of lymphocytes present in the lymphocyte-rich areas near blood vessels have CD4 surface proteins, indicating that they are helper-inducer cells. The transitional areas of cellular heterogeneity show a mixture of both CD4+ and CD8+ cells amidst the macrophages.
Overall, within the synovium, T cells predominate over B cells. T cells constitute about 50 percent or more of cells in most RA synovia, whereas only 5% or less of cells are B lymphocytes. The B cells are located primarily within reactive lymphoid centers, whereas plasma
cells and macrophages are often found outside these centers. This arrangement is consistent with T cell-dependent B lymphocyte activation; plasma cells, the main immunoglobulin producers, migrate away from the germinal centers after differentiation. CD4+ cells in RA synovium are intimately related to B lymphocytes and to HLA-DR-positive cells that resemble morphologically the interdigitating cells of lymph nodes.
In contrast to peripheral blood lymphocytes in RA, synovial lymphocytes bear a more activated phenotype, with fewer CD8+ cells and high expression of DR antigens. [92] Synovial lymphocytes also bear adhesion molecules of the very late activation antigen (VLA) and lymphocyte function antigen (LFA) superfamily of integrins, which may enable the inflammatory response to localize and persist within the synovium. [93] However, T cell activation and the induction of adhesion molecules probably do not occur within the joint; rather, the reason that T cells enter the synovium in the first place and then remain within the joint is their armamentarium of adhesion molecules. The cytokine milieu of the joint induces adhesion molecules like intercellular adhesion molecule-1 (ICAM-1), vascular adhesion molecule-1 (VCAM-1), and connecting segment-1 (CS-1) fibronectin on vascular endothelium, and these, in conjunction with chemokines and other chemoattractants, call the cells to the joint based precisely on this phenotype. Hence, it is this antigen-independent process that is responsible for the mononuclear cell infiltrate, not antigen-specific
local expansion. It is likely that T cells that might respond to a specific "RA pathogen" constitute a small percentage of cells, perhaps less than 1 percentage. Local proliferation and activation probably account for only a very small percentage of T cells in the synovium. The few T cells that proliferate are primarily CD8+ , not CD4+ . [94]
Other Cell Types
Despite the abundance of neutrophils in RA synovial effusions, only rare polymorphonuclear leukocytes (PMNs) have been described in the synovium. Natural killer (NK) cells have also been identified in RA synovium. [95] The NK cells contain large amounts of granzymes, which are serine proteases produced by activated cytotoxic NK cells. One potentially important immunoregulatory role of NK cells is that they can stimulate B cells to produce RFs.
Mast cells are present in the synovial membranes of patients with RA and may be localized in some patients at sites of cartilage erosion. [96] In one study, rheumatoid synovial membranes contained over 10 times as many mast cells in histologic sections than control synovial samples from patients undergoing surgery for meniscectomy. Patients with high numbers of mast cells had more intense clinical synovitis in the affected joints. Mast cells can be found in a majority of synovial fluid specimens from inflammatory synovitis, and there is measurable histamine content in these synovial fluids. Rheumatoid synovial fluids contained higher levels of histamine than corresponding plasma samples. [97] A detailed analysis of several indicators of proliferation and the enumeration of synovial mast cells has demonstrated strong positive correlations between the number of mast cells per cubic millimeter of synovial tissue and the degree of lymphocyte infiltration (the number of helper T lymphocytes and plasma cells). [98] Mast cells from RA synovium express significantly higher amounts of the C5a receptor compared with OA synovium. [99] In addition, C5a only released histamine from the RA cells.
What could be the contribution of mast cells to rheumatoid synovitis? They could be responding to cytokines that stimulate mast cell growth and chemotaxis. Extracts of mast cells can induce adherent rheumatoid synovial cells to increase production of PGE2 and collagenase. [100] Heparin, however, does have significant effects on connective tissue. In particular, it may modulate the effects of bone hormones on osseous cells and thereby alter the balance of bone synthesis toward degradation.
Synovial Histopathology in Early versus Late Rheumatoid Arthritis
Previous studies have suggested that the earliest phases of RA, that is, during the first few weeks of symptoms, exhibit distinct abnormality. Examination of synovial biopsies from early RA revealed a paucity of lymphocyte infiltration in the presence of endothelial cell injury, tissue edema, and neutrophil accumulation. [101] More recent studies, however, suggest that the histologic appearance of RA is similar regardless of the duration of clinical symptoms. [102] The extent of lymphoid aggregation, T cell infiltration, and synovial lining hyperplasia can resemble chronic disease even when symptoms have been present for a very short period of time. The cytokine patterns of these biopsies as determined by immunohistochemical analysis indicated similar levels of T cell (like interferon-gamma [IFN-gamma]) and non-T cell factors (like IL-1 and TNF-alpha). The tumor suppressor gene p53 is also expressed in early RA,
most likely due to the intensely genotoxic environment generated by hypoxia and oxygen radicals (see later).
Biopsies of asymptomatic joints from patients with early or late RA also have lymphocyte infiltration, cytokine production, and p53 expression. [103] Although IFN-gamma, IL-1, and TNF-alpha levels are increased compared with those of normal synovium, they are still modestly lower than those of clinically active joints. Of interest, studies in animal models of arthritis also demonstrate increased expression of proinflammatory transcription factors such as AP-1 and NF-kappaB well before clinically evident arthritis. [104] These studies suggest that patients with "early" RA as defined by the duration of symptoms might, in fact, already have chronic disease and that evaluation of truly early disease might require assessment of patients long before the onset of symptoms (if this is even possible).
Engraftment of Rheumatoid Arthritis Synovium into Severe Combined Immunodeficiency Mice
SCID mice, which lack a functional immune system, have been utilized to evaluate the biology of rheumatoid
synovium. Several variations of this model have been developed, and in each case explanted synovium is capable of engrafting successfully. A blood supply develops over a period of weeks, and many of the resident human cells remain in the synovium. When coimplanted with cartilage, synoviocytes from the explant attach to and invade cartilage matrix. CD4- and CD8-bearing lymphocytes in the synovium produce a variety of cytokines, including IFN-gamma. CD8+ T cells in the explant produce IL-16, which appears to be an endogenous suppressor of CD4+ T cell function. Despite the appeal of these models, it is important to recognize their limitations, especially when evaluating T cell function. Although the SCID mice lack an immune system, the explanted synovium contains functional human lymphocytes and antigen-presenting cells that, when implanted into mice, can respond to host of murine proteins. Hence, the cytokine profile of the explant could also reflect that of human T cells exposed to xenoantigens.
A second model utilizes enzymatically dispersed synovial tissue cells that have been co-implanted into SCID mice with cartilage explants and invade into the cartilage matrix, looking very much like destructive pannus. [105] Perhaps more important, this phenomenon still occurs even if pure populations of long-term cultured RA fibroblast-like synoviocytes are used (Fig. 64-3) . [106] Because these synoviocytes are devoid of T cells and macrophages, there can be no contribution from an immune response to murine antigens. The invading cells express VCAM-1, which could potentially facilitate adhesion to cartilage or chondrocytes, as well as proteases that digest the cartilage matrix. Control synoviocytes from OA patients and normal dermal fibroblasts do not invade the cartilage, indicating that the activity is unique to rheumatoid fibroblast-like synoviocytes. Using viral vectors to introduce cytokine genes into this explanted synoviocyte, one can evaluate their respective role in cartilage invasion. IL-1Ra, a natural antagonist to IL-1, had no effect on synoviocyte invasion but decreased perichondrocyte matrix loss. [107] In contrast, IL-10 decreased
invasion but did not alter matrix loss. Finally, overexpression of soluble TNF receptors had no consistent effect in this model. These studies suggest that excessive production of IL-1 and underexpression of IL-10 contribute to the invasive properties of RA synoviocytes.
IMMUNE RESPONSE
For several decades, it has generally been assumed that T lymphocytes, particularly the CD4+ helper-inducer cells, are a crucial component of the early rheumatoid response. Although these conclusions are based largely on circumstantial evidence, the data are, in some cases, compelling. This section is designed to explore some of the cell-mediated immune responses in RA that might establish and perpetuate rheumatoid synovitis. The histopathologic appearance of RA, with exuberant infiltration
Figure 64-3 Invasion of RA synoviocytes into cartilage explants in severe combined immunodeficiency syndrome (SCID) mice, RA fibroblast-like synoviocytes were co-implanted with normal human cartilage into the renal capsule of SCID mice. Note that the synoviocytes have attached to the cartilage and invade into the matrix. Several chondrocytes in lucanae are also present. (Courtesy of Dr. S. Gay.)
of the synovium with T lymphocytes, is often pointed to as evidence of a T cell-mediated disease because this is characteristic of antigen-specific responses. However, the synovium can only respond to inflammation in a limited number of ways; in fact, the histologic appearance of chronic arthritides that are clearly not mediated by T cells (e.g., chronic tophaceous gout) exhibits many of the same features as RA. Many animal models of arthritis, like collagen-induced arthritis and adjuvant arthritis, are clearly T cell-dependent. The relevance of these models to RA is not always clear, given that the clinical courses are highly compressed and the inciting agents are not thought to be specifically related to RA. Some animal models have also been developed that appear to be relatively independent of T cells, such as Lyme arthritis in SCID mice that lack T cell function or transgenic mice that overexpress TNF-alpha or oncogenes like c- fos.
More direct evidence for T cell activation is often cited from experimental therapeutics. Fistula of the thoracic duct that removes T lymphocytes from the body was shown in 1970 have some efficacy in RA, [108] and leukapheresis has been associated with brief improvement associated with the return toward normal in vitro anergy of peripheral blood mononuclear cells. [109] Total nodal irradiation is also a very effective means of suppressing systemic helper T cell function. In clinical
studies, this procedure did offer some benefit, although symptoms could return despite persistent suppression of delayed-type hypersensitivity. [110] Treatment with antibodies that deplete T cells, including anti-CD4, anti-CD5, and anti-CD52, has been reported at various times. [111] [112] [113] The responses, if any, were usually very modest and transient despite profound and sometimes prolonged lymphopenia and immunosuppression. Interpretation of such studies needs to take into account the fact that some anti-T cell therapies (e.g., anti-TcR antibody) can prevent arthritis in T cell-dependent animal models; however, if treatment is delayed until after the onset of clinical signs, the results are less impressive and the therapy can sometimes worsen the disease. The stage of disease might therefore be an important determinant of the clinical response after T cell depletion. However, even treatment of early RA with anti-CD4 antibody was ineffective. [114] In addition, the degree of peripheral lymphopenia does not necessarily correlate with depletion of the synovial T cell population. [115] Potent immunosuppressive drugs like cyclosporine have also shown efficacy in RA. Yet, this agent, which can suppress allograft rejection, results in clinically significant responses in only a minority of patients. [116] IL-2-diphtheria toxin fusion protein designed to kill IL-2 receptor-bearing T cells has not been effective in a controlled trial. [117] Reports of progressive destruction in RA patients despite acquired immunodeficiency syndrome suggests that non-T cell mechanisms are also important. [118]
T Lymphocyte Activation in the Synovium
In the subsynovial areas around the small capillaries and what will become high endothelial venules in synovium, tissue macrophages and dendritic-like cells are available to process and present antigen to T lymphocytes (Fig. 64-4) . Many different types of cells in chronically inflamed tissues, including rheumatoid synovium, bear HLA-DR antigens on their surfaces, although this does not always correlate with antigen-presenting function. Normal synovial lining cells can mediate T lymphocyte proliferation, but their activating ability appears to be lower than that of epidermal Langerhans' cells, which stand as barriers to the entrance of antigen through the skin. Rheumatoid synovial dendritic cells, however, are
Figure 64-4 T lymphoblast in rheumatoid synovial tissue surrounded by three macrophages (Mp). The arrows point to probable intercellular bridging. This may be the morphologic manifestation of presentation of antigen to the helper T cell by the antigen-presenting cells. (Courtesy of Dr. H. Ishikawa and Dr. M. Ziff.)
extremely efficient in allogeneic T cell activation. [119] The microheterogeneity of the rheumatoid synovial tissue, with different numbers and proportions of cell lineages in each area, suggests that antigens presented at each location might differ, with type II collagen presented to T cells one place, proteoglycans presented elsewhere, and responses to HSPs or viral antigens in yet-another region.
At many sites in the synovium, the histopathologic findings resemble those of a classic delayed-type hypersensitivity reaction. Large strongly HLA-DR-positive macrophages or dendritic-like cells form close contact with lymphocytes bearing CD4 markers. B cells also express class II MHC antigens and can present antigen and produce activating cytokines. When standard light microscopy and immunohistochemical methods are used, the majority of synovial T cells are of the small memory type that express activation antigens like HLA-DR, transferrin receptors, and LFA-1 on their surface. This should not necessarily be construed as a priori evidence that they were activated within the articular cavity, because this is precisely the phenotype of cells that is preferentially recruited from peripheral blood by the activated endothelium of RA synovium. In some areas, especially in the "transitional zones" between lymphocyte-enriched regions, ultrastructural studies show direct contact between antigen-presenting cells and blastlike cells, with long processes enveloping the latter. [120] Although the antigen remains undefined, this presumably results in helper cell activation. In addition to professional antigen-presenting cells, IFN-gamma-stimulated fibroblast-like synoviocytes can present antigen to T cells. [121]
For maximal T cell responsiveness, a second signal in addition to antigen stimulation is usually required. CD28 is one of these co-stimulatory molecules on T lymphocytes and is expressed by synovial T cells in RA. [122] Its ligands, CD80 and CD86 (also called B7-1 and B7-2), are also displayed on antigen-presenting cells in the joint, thereby providing an excellent environment for T cell activation. A population of CD28-negative T cells has been identified in rheumatoid blood and joint samples. [123]
These cells appear to respond briskly to autoantigens and may have escaped peripheral tolerance. CD40, another co-stimulatory molecule, and its ligand on T cells, CD40L, have also been identified in RA synovium. [124] CD40L is capable of synergizing with IL-1 for the production of cytokines like granulocyte-macrophage colony-stimulating factor (GM-CSF) by CD40-bearing synoviocytes. [125]
Peripheral Blood T Lymphocytes
The number of CD4+ helper T cells are mildly increased in the peripheral blood of patients with RA, with a concomitant decrease in cytotoxic (CD8) lymphocytes (and an increased CD4/CD8 ratio). The surface phenotype of circulating T cells in RA suggests activation in some studies, but not in others. For instance, an increased percentage of cells express HLA-DR and the adhesion protein VLA-4 (alpha4 beta1 -integrin). The latter is especially critical in that VLA-4 plays an important role in the recruitment of cell to the synovium through interactions with counter-receptors VCAM-1 and CS1 fibronectin on endothelial cells. Increased HLA-DR is present on gammadelta TcR-bearing lymphocytes as well as alphabeta cells. [126] Surface CD4 is lower on RA peripheral T cells compared with those of healthy controls, a phenomenon observed after lymphocyte activation. The decreased surface CD4 is not matched by an increase in circulating soluble CD4 protein. [127] Other markers of activation are not necessarily elevated on RA T cells, including IL-2 receptors, the co-stimulatory molecule B7, and VLA-1. [128] Elevated levels of soluble IL-2 receptors, which are shed by activated T cells, are found both in sera and synovial fluid of rheumatoid patients, and the levels appear to correlate with disease activity. [129] In summary, these studies suggest that peripheral blood T cells express phenotypic characteristics of partial activation. It is not clear whether this process occurs in the periphery or whether cells are activated in the synovium and reenter the circulation
via the synovial lymphatics.
In 1981, the first suggestions of an interesting immunoregulatory dysfunction in RA were noted. [130] It was observed that the outgrowth of EBV-infected B lymphocytes was inadequately suppressed by lymphoid cells from rheumatoid patients. This was related to a defect in suppressor T cell function. The deficient T cell response could be correlated somewhat with disease activity, but it was also noted that the abnormality was present in T cells of patients with inflammatory arthropathies other than RA. [131] A more specific defect was also apparent in the autologous mixed lymphocyte reaction in which T cells proliferate and produce cytokines in response to class II MHC antigens expressed on autologous antigen-presenting cells. [132] IFN-gamma production is significantly suppressed in RA cultures, and the abnormality is corrected by the addition of indomethacin. Additional studies suggest that IFN-gamma production is low because of a heightened sensitivity of RA cells to PGE. Peripheral blood T lymphocytes from rheumatoid patients also demonstrate defective IL-2 production. [133] Indomethacin can partially reverse this defect in the presence of mononuclear cells, again implicating prostaglandin sensitivity. Unlike lymphocytes from normal individuals, recombinant IL-2 has minimal effect on IFN-gamma production by peripheral blood lymphocytes from patients with active RA. [134]
Synovial Fluid Lymphocytes
The cell mix in synovial fluid differs from that of peripheral blood as well as that of synovial tissue. Therefore, analysis of synovial fluid cells is not necessarily an accurate reflection of the synovium. Even though synovial effusions contain an abundance of T cells, the CD4/CD8 ratio is actually reversed compared with that of blood or synovial tissue, with an excess of CD8+ suppressor cells relative to CD4+ lymphocytes. In addition, synovial tissue is nearly devoid of neutrophils, which often constitute 50 to 75 percent of synovial fluid cells. Hence, the synovial fluid does not contain a random distribution
of cells shed from synovial tissue, and data regarding TcR usage or the state of activation should be interpreted with this in mind (see later).
Synovial fluid, like peripheral blood, contains activated T cells based on the presence of surface HLA-DR antigens. [135] Other activation antigens that are not increased on peripheral blood cells, however, are increased on synovial fluid lymphocytes, including VLA-1. Surprisingly, IL-2 receptor expression is not increased. Of CD4+ cells in rheumatoid synovial fluid, most are memory cells and express CD45RO on their surface. [136] Despite the phenotypic appearance of activation, synovial fluid T cell function is rather deficient when compared with that of peripheral blood cells. For instance, synovial fluid lymphocyte proliferation in response to mitogens or most recall antigens, like tetanus toxoid, is significantly lower than paired blood T lymphocytes. Mycobacterial antigens and the 60-kD HSP appear to be exceptions, in which proliferation is greater in cells isolated from rheumatoid effusions. Cytokine production by synovial fluid T cells in vitro is also low. IFN-gamma release after stimulation by mitogens like phytohemagglutinin is defective. The amount of IL-2
produced by synovial fluid lymphocytes is significantly less than that produced by the corresponding blood cells. [137]
A possible mechanism that might contribute to defective T cell responses among synovial fluid mononuclear cells from rheumatoid patients is the presence of local inhibitors of cell activation. IL-1Ra and transforming growth factor-beta (TGF-beta) are possible T cell suppressants in the joint because both have been identified as components of synovial effusions that can suppress thymocyte proliferation. [138] [139] Nonspecific components of joint effusions like hyaluronic acid can be toxic to cells and can indirectly suppress T cell activation. The mechanism of diminished T cell activation could be related to abnormalities in TcR signaling. Articular T cells have diminished tyrosine phosphorylation of proteins after stimulation, especially the key signal transduction pathway p38 mitogen-activated protein (MAP) kinase. [140] Furthermore, tyrosine phosphorylation of the TcR zeta chain, an early event in TcR signaling, is also defective compared with that of peripheral blood T cells. Decreased levels of the zeta protein were also observed, suggesting that the TcR apparatus is abnormal in RA. In addition, the hyporesponsiveness of synovial fluid T cells correlates with a significant decrease in the levels of the intracellular redox-regulating agent glutathione. [141]
T Cell Oligoclonality
Another approach to assessing T cell activation and proliferation in peripheral blood is to perform detailed analysis of TcR expression. The vast majority of peripheral blood T cells express an alphabeta heterodimer. TcR specificity is determined by genetic rearrangement of variable (V), diversity, and joining segments of germline sequences that are then combined with constant regions. Antigen recognition and the specificity of T cell responses results from the many permutations of these segments and the occurrence of deletion or point mutations during rearrangement. In theory, the TcR repertoire in immune-mediated diseases should lead to the expansion of cells with similar alpha- and beta-chains. The repertoire can be determined using a number of techniques, each of which has advantages and disadvantages.
A number of studies have examined the T cell repertoire in peripheral blood in RA, although most appropriately focused on differences between blood and joint samples (reviewed in detail by Fox and Singer [142] ). Although some studies suggest a preponderance of one or another gene, no specific TcR genes are consistently overexpressed in RA peripheral blood compared with normal individuals, especially when one considers the lack of control for HLA haplotypes in the normal population. Studies of articular cells are probably the most relevant, because the joint is the site of the disease and antigen-driven paradigms of RA imply that in situ T cell activation and expansion are pivotal steps. Several dozen studies were published in the late 1990s purporting to find one or another gene overexpressed in the joint (either synovial fluid or tissue) compared with the blood. In some cases, a pattern appears to be emerging suggesting an increased number of T cells expressing Vbeta3, -14, and -17, especially in synovial tissue. These particular Vbeta genes are structurally related and are unusually susceptible to activation by superantigens. This supports the hypothesis that specific T cells can be activated by bacterial or mycoplasmal superantigens, leading to oligoclonal expansion and release of cytokines. The final answer to the question of T cell oligoclonality is still open, and other studies have either not found evidence for the restricted clonality of T cells in RA synovial fluid,
synovial tissue, and blood or identified expansion of different Vbeta or Valpha genes. Although it would be nice to believe that "oligoclonal expansion" results from exposure to a primary pathogenic antigen, secondary antigens like proteoglycans or type II collagen that do not initiate RA but could elicit a T cell response might be responsible. Therapeutic interventions designed to target a specific Vbeta gene (if one were convincingly shown) might not eliminate pathogenic clones but instead delete less important T cells involved in secondary responses. Furthermore, if RA is caused by an infectious pathogen that is partially controlled by expanded T cells in the joint, one runs the risk of exacerbating disease.
The best opportunity for detecting the appropriate pathogenic T cell clones in RA would be to study very early RA, because disease chronicity causes T cells to accumulate in an antigen-independent manner. This is best exemplified in the experimental allergic encephalitis model in rats, which is caused by autoimmunity against myelin basic protein. In the earliest phases, there is a strong bias in the central nervous system for a few antigen-specific pathogenic TcR genes. [143] As the disease progresses, this bias is rapidly overwhelmed by the continued influx of nonspecific cells recruited into the brain by chemokines and adhesion molecules, not by antigen. The same situation likely occurs in RA and other chronic inflammatory arthritides, and the study of TcRs in chronic disease might not give accurate information on the inciting populations.
T Cell Cytokines
Cytokines are hormone-like proteins that enable immune cells to communicate. They can either interact with cells after being released from cells in a soluble form or can be involved with direct cell-cell communication through membrane-bound factors (such as TNF-alpha). In addition to playing a critical role in normal immune responses, they play an integral role in the initiation and perpetuation of synovitis, according to a remarkable explosion of information in the last decade. The cytokine milieu in RA is not random, although early studies suggested an unrestricted abundance of cytokines. However, there is increasing evidence that factors produced by T lymphocytes are actually diminished in RA, whereas those generated by macrophages and by synovial fibroblasts are increased (see later and Table 64-4 ). [144]
Helper T cells have been divided into cytokine-specific subsets. Helper T cells 1 (Th1 cells) produce IFN-gamma and IL-2 but not IL-4, IL-5, or IL-10. In contrast, Th2 cells produce the opposite profile (IL-4+ IL-5+ IL-10+ IL-2- IFN-gamma- ). Some cytokines are produced by both subsets, including TNF-alpha, IL-3, and GM-CSF. A third subset with an unrestricted cytokine profile is called Th0, [145] whereas Th3 cells produce TGF-beta. Other phenotypes have since been reported, and, although it is clear that
TABLE 64-4 -- Level of Production of Synovial Cytokines in Rheumatoid Arthritis According to Cellular Source
Cellular Source Level of Production inRheumatoid Arthritis
Synovium
T cells
IL-2 -
IL-3 -
IL-4 -
IL-6 ±
IL-13 -
IL-17 +
Interferon-gamma ±
TNF-alpha -
TNF-beta -
GM-CSF -
Macrophages or fibroblasts
IL-1 +++
IL-1Ra +
IL-6 +++
IL-10 ++
IL-15 ++
TNF-alpha ++
M-CSF (CSF-1) +
GM-CSF ++
TGF-beta ++
Chemokines (IL-8, MCP-1, etc.) +++
Fibroblast growth factor ++
Abbreviations: -, Absent or very low concentrations; +, present; GM- CSF, granulocyte-macrophage colony-stimulating factor; IL, interleukin; IL-1Ra, interleukin 1 receptor antagonist; MCP-1, macrophage chemotactic protein-1; M-CSF, macrophage colony-stimulating factor; TNF, tumor necrosis factor.
Adapted from Firestein GS, Zvaifler NJ: How important are T cells in chronic rheumatoid synovitis? Arthritis Rheum 33:768, 1990.
Tissue macrophages or type A synoviocytes.
Tissue fibroblasts or type B synoviocytes.
Th0, Th1, and Th2 cells predominate, some T cells do not fit into well-defined categories (especially in humans). Th1 cells primarily mediate delayed-type
hypersensitivity in vivo, whereas Th2 cells are more prominent regulators of isotype switching and antibody production. Some cytokines produced by Th2 cells are immunosuppressive, because IL-4 and IL-10 downregulate Th1 cell differentiation and activation as well as delayed-type hypersensitivity.
Helper T Cell 1 Cytokines: Interferon-gamma and Interleukin 2
IFN-gamma is the most potent inducer of MHC class II antigen mononuclear cells. It concurrently activates synthetic and secretory activity in the monocyte-macrophages. After incubation with IFN-gamma, monocytes show morphologic, metabolic, and phenotypic changes consistent with activation to vigorous macrophages; they also begin to express class II MHC antigens and Fc receptors while simultaneously downregulating the expression of CD14. IFN-gamma induces adhesion molecules like VCAM-1 and ICAM-1 on the surface of endothelial cells and can help recruit inflammatory cell accumulation at sites of injury.
One of the most important functions of IFN-gamma is its capacity to alter the balance of extracellular matrix synthesis and degradation. IFN-gamma inhibits collagen synthesis both in vitro and in vivo and has been shown to decrease levels of type I and type III procollagen messenger RNAs (mRNAs) in rheumatoid synovial fibroblast-like cells. [146] IFN-gamma also inhibits metalloproteinase production by cultured fibroblast-like synoviocytes that have been stimulated with IL-1 or TNF-alpha. [147] [148] In the case of IL-1, IFN-gamma specifically decreases stromelysin gene expression and protein production. IFN-gamma also inhibits a variety of TNF-alpha-mediated activities of synoviocytes, including GM-CSF production, collagenase activity, and proliferation. This is not due to downregulation of TNF receptors on synoviocytes, since IFN-gamma paradoxically increases expression of the TNF receptors. [149]
Despite the putative intense T cell activation in the rheumatoid synovium, only very low concentrations of IFN-gamma have been detected, [150] far below the amounts needed to induce HLA-DR expression on monocytes. The relative lack of IFN-gamma in rheumatoid joints has since been confirmed, including studies employing reverse transcriptase-polymerase chain reaction (RT-PCR) to detect specific RNA transcripts. [151] The relative deficiency in IFN-gamma production is even more striking if one considers that peripheral blood monocytes from patients with RA also have defective HLA-DR and HLA-DQ induction by IFN-gamma compared with normal cells. [152] The difficulty detecting IFN-gamma in RA does not appear to be due to methodologic problems, because it is easily measured in other diseases known to be mediated by T cells, such as tuberculous pleuritis. IFN-gamma mRNA was detected in tuberculous pleura by in situ hybridization, [153] but RA synovial tissue was negative using similar techniques. Although immunohistochemical analysis demonstrates IFN-gamma in a small percentage of RA synovial T cells,
the percentage is far less than in chronically inflamed tonsils. [102] [154]
IL-2 is a T cell-derived cytokine that serves as a major autocrine or paracrine T cell
growth factor and was originally reported to be present in synovial fluid when biologic assays were used. [155] However, specific monoclonal antibodies that block the IL-2 receptor do not interfere with this activity. [156] More specific immunoassays showed that IL-2 was found in only a small percentage of RA synovial effusions and, when detected, was present in low concentrations. [157] An immunofluorescence study of RA synovium also demonstrated only trace amounts of immunoreactive protein in frozen sections of RA synovium. [158] Results of studies of IL-2 gene expression in synovial tissue are mixed, and some studies detect specific IL-2 mRNA whereas others do not. Moreover, the possible presence of IL-2 mRNA without protein production could suggest that the T cells are anergic.
TNF-alpha, GM-CSF, and IL-6 can be made by both Th1 and Th2 cells under some circumstances and are present in synovial fluid, but the primary sources of these cytokines in the rheumatoid joint are macrophages and fibroblasts rather than T cells.
Helper T Cell 2 Cytokines: Interleukin 4, Interleukin 10, and Interleukin 13
With the use of immunoassays, IL-4 and TNF-beta (lymphotoxin) were not detected in RA synovial fluid. In situ hybridization also showed little or no IL-4 in RA synovial tissue even though a small amount of IFN-gamma was detected using the same method. When extremely sensitive nested RT-PCR techniques were used on synovial biopsies, Th2 cytokines IL-4 and IL-13 were absent in RA, whereas both IFN-gamma and IL-12 (a cytokine that induces T cell maturation toward the Th1 phenotype) were present. IL-10, which has potent anti-inflammatory activities, is expressed in RA synovium. However, macrophages, not T cells, are the major producers of IL-10 in RA.
T Cell Subsets: Imbalance between Helper T Cells 1 and 2
As noted previously, Th1 cytokine mRNA, such as IFN-gamma, can be detected in RA synovial tissue with the use of extremely sensitive techniques. T cell clones from RA joints tend to exhibit a Th1 phenotype, and in vitro stimulation of synovial fluid mononuclear cells demonstrates a Th1 bias compared with peripheral blood. [159] [160] Some Th2 and Th0 clones have also been isolated. Another unique T cell type that produces both the Th1 cytokine IFN-gamma and the Th2 cytokine IL-10 has been isolated from RA synovium. In contrast to the Th1 factors, Th2 cytokines (especially IL-4 and IL-13) are absent from the joint, even using nested RT-PCR. Furthermore, IL-12 is produced by macrophages and NK cells in the joint. Therefore, the synovial milieu (low IL-4 and high IL-12) supports the maturation of T cells toward Th1. Surface display of certain chemokine receptors are characteristic of Th1 cells, with the former expressing CXCR3 and CCR5. [161] Almost all the T cells in RA synovial fluid have this phenotype, which is greatly enriched compared with peripheral blood. In toto, these data suggest that the T cells in RA are biased toward the Th1 lineage.
The cytokine most closely associated with resolution of Th1-mediated inflammatory diseases in animal models of arthritis is IL-10, and administration of this cytokine to mice with collagen-induced arthritis was effective. [162] In a second study, IL-4 and IL-10 were administered individually or in combination in the same model. [163] The cytokines had modest or no benefit when used separately, but together the effect was more impressive. Clinical improvement correlated with decreased synovial IL-1 and TNF-alpha and cartilage destruction. Anti-IL-10 antibody therapy in
collagen-induced arthritis accelerated disease. The complexity of cytokine networks in inflammatory arthritis was underscored by studies on the role of IL-12 in collagen-induced arthritis. In early arthritis, IL-12 administration increases the incidence of collagen-induced arthritis, whereas anti-IL-12 is beneficial. [164] However, in late disease, IL-12 administration suppresses arthritis and anti-IL-12 cause an exacerbation.
An alternative way to alter the Th1-Th2 balance in animal models is with a nondepleting anti-CD4 antibody. [165] Treatment of collagen-induced arthritis with such antibodies suppressed arthritis and prevented passive transfer with lymphocytes from treated to naive animals. IgG2a anti-type II collagen antibody production was suppressed by the antibody, implying a decline in endogenous IFN-gamma production, whereas in vitro IL-4 production was enhanced. These data suggest that the nondepleting antibody might act by increasing Th2 activity.
The relative lack of Th2 cytokines in RA might contribute to exacerbation of synovitis. One of the most important suppressive cytokines, IL-10, was originally characterized as a cytokine synthesis inhibitory factor based on its ability to block T cell and macrophage cytokine production. Abundant IL-10 protein is present in RA synovial fluid, and the gene is expressed by synovial tissue cells. [166] However, in vitro studies of cultured synovial cells suggest that not enough IL-10 is produced to suppress IFN-gamma production.
The relative absence of most Th2 cytokines in RA results in enhanced release of inflammatory mediators by RA synovial explants. Addition of exogenous IL-10 or IL-4 to cultures of synovial tissue cells or synovial tissue explants suppresses synthesis of proinflammatory cytokines like IL-6, IL-1, TNF-alpha and GM-CSF as well as metalloproteinases. [167] [168] The inhibitory action of IL-4 might be mediated by decreased c-jun and c-fos expression, which is required for efficient production of metalloproteinases and cytokines. [169] IL-4 inhibits IL-8 production by synovial fluid mononuclear cells. [170] IL-10 and IL-4 also increase the release of other anti-inflammatory cytokines, like IL-1Ra, by synovial cells. [171] This suggests that RA might be driven by a complex process involving macrophage and fibroblast cytokine networks in combination with defective T cell responses.
Activation of Synoviocytes by T Cells: Interleukin-17 and Cell-Cell Contact
Although T cell activation is surprisingly modest in rheumatoid synovium, there are alternative mechanisms that
permit these cells to participate in synovial cytokine networks and matrix destruction. For instance, the recently described T cell cytokine IL-17 is present in small, but functionally relevant, concentrations in RA synovial effusions. IL-17 mimics many of the activities of IL-1 and TNF-alpha with respect to fibroblast-like synoviocyte function, including collagenase and cytokine production. [172] More important, T cell-derived IL-17 in synovial tissue can synergize with IL-1 to activate synoviocytes.
A second mechanism by which T cells can activate macrophages and fibroblasts in RA is through direct cell-cell contact. Membranes prepared from activated T cells can directly stimulate macrophages and fibroblast-like synoviocytes to produce cytokines and metalloproteinases. [173] The membrane constituents that regulate this process vary, depending on the particular culture conditions, but include adhesion molecules like LFA-1 and membrane-bound TNF-alpha. Hence, a T cell displaying these proteins, even if the cell is no longer activated, can potentially contribute to macrophage and fibroblast activation in an antigen-independent fashion.
CYTOKINES: THE ROLE OF MACROPHAGE AND FIBROBLAST PRODUCTS IN DISEASE PERPETUATION
Although the production of T cell cytokines is low in RA, the same is not true for products of macrophages and fibroblasts. A reductionist view would be to state that virtually every macrophage and fibroblast mediator that has been sought in the RA synovium has been detected. Although this is obviously a simplification, it is not far from the truth. In this section, some of the major cytokines and effectors produced in the joint are enumerated, with an emphasis on the prevalence of macrophage and fibroblast products as dominant driving forces during the perpetuation phase of RA. Macrophages, in particular, are the most vigorous producers of intercellular mediators. These cells, which are present in small numbers in normal synovium, increase in number by migration from extrasynovial sites (e.g., the bone marrow) after inflammation begins. Their responses include secretion of more than 100 substances and cover a biologic array of activity from the induction of cell growth to cell death.
Proinflammatory Macrophage and Fibroblast Cytokines
Interleukin 1
IL-1 is a ubiquitous family of polypeptides with a wide range of biologic activity. Its actions make it a candidate for the major amplification factor and translator of the inflammatory response of RA into a destructive one. IL-1, including its properties and its actions in comparison with those of other active factors, is discussed in Chapter 20 . It has been demonstrated, for example, that recombinant IL-1beta, injected into rabbit knee joints induces the accumulation of PMNs and mononuclear leukocytes in the joint space and the loss of proteoglycan from articular cartilage. [174] The amount of IL-1 produced by peripheral blood monocytes is much greater when cells are isolated from rheumatoid patients who had a recent onset of disease or an exacerbation of disease than from patients with stable arthritis or from controls. [175] IL-1 activity sufficient to stimulate collagenase and PGE2 production from synovial lining fibroblasts has been shown to be generated by monocyte-macrophages isolated from synovial fluid of patients with RA. [176] High-affinity receptors for IL-1alpha and IL-1beta have been identified on cultured human rheumatoid synovial cells. [177] Even PMNs stimulated by phagocytosis or by other activating substances produce IL-1. Thus, the macrophages, synovial fibroblasts, PMNs, and endothelial cells can be induced to generate this powerful mediator.
In terms of cytokine mRNA production, the synovial macrophage is the most prolific cell in the joint, and nearly half of all CD11b-positive macrophages from the RA synovium contain significant amounts of IL-1beta mRNA. [178] Immunohistologic studies confirm this, with especially abundant IL-1 protein in synovial lining macrophages adjacent to type B synoviocytes and in sublining macrophages near
blood vessels. The IL-1 in the lining can subsequently activate type B synoviocytes to proliferate and secrete a variety of mediators, including metalloproteinases. IL-1 activity has been detected in culture supernatants of rheumatoid synovial biopsies, and in one study the amount produced correlated with joint destruction found on roentgenograms. [179] A broad range of substances are capable of inducing IL-1 production; for example, immunoglobulin Fc fragments and, to a lesser extent, immune complexes can generate IL-1 production by rheumatoid synovial monocyte-macrophages. Collagen fragments can induce IL-1 production, and it is intriguing that type IX collagen, which has been found only in articular cartilage and localized into intersections of collagen fibrils, is a potent inducer of IL-1 by human monocyte-macrophages. [180] Signals from T cells can also enhance IL-1 production during antigen presentation.
Within the rheumatoid joint, IL-1 also induces fibroblast proliferation, stimulates the biosynthesis of IL-6 and GM-CSF by synovial cells, and enhances collagenase production. [181] IL-1 increases glycosaminoglycan production in human synovial fibroblast cultures, [182] although the effect of IL-1 on the production of intact proteoglycan molecules by articular cartilage in some models seems to be decreased, indicating that the production of components of the glycosaminoglycan complex by IL-1 may be altered. Additionally, IL-1 stimulates human synovial cells to increase both cell-associated and extracellular plasminogen activator activity, and it has been demonstrated that the cytokine osteoclast-activating factor that is capable of stimulating bone resorption is identical to IL-1beta. [183] IL-1 induces a number of adhesion molecules on fibroblast-like synoviocytes and endothelial cells, including VCAM-1 and ICAM-1. Animal models of arthritis, including streptococcal cell wall arthritis, antigen-induced arthritis, and collagen-induced arthritis,
suggest that IL-1 is the key cytokine that regulates bone and cartilage destruction.
Tumor Necrosis Factor-alpha
TNF-alpha is a pleiotropic cytokine that has been implicated as a key proinflammatory cytokine in RA and has been detected in rheumatoid synovial fluid and serum. Notably, another member of the TNF family, TNF-beta (which is primarily a T cell-derived lymphokine), is not present in RA synovial effusions. Levels of TNF-alpha correlate with erythrocyte sedimentation rates and synovial fluid leukocyte counts. TNF-alpha is produced as a membrane-bound protein that is released from the cell surface after cleavage by a TNF convertase, a membrane metalloproteinase. IL-1 and TNF have many similar activities, including the ability to enhance cytokine production, adhesion molecule expression, proliferation, and metalloproteinase production by cultured synoviocytes. In general, IL-1 is more potent, especially with regard to collagenase and stromelysin production. In some systems, the effects of these two agents are synergistic.
TNF-alpha stimulates collagenase and PGE2 production by human synovial cells and dermal fibroblasts, [184] induces bone resorption, inhibits bone formation in vitro, and stimulates resorption of proteoglycan and inhibits its biosynthesis in explants of cartilage. [185] In situ hybridization and immunohistochemical studies suggest that
TNF-alpha, like IL-1, is primarily a product of synovial macrophages in RA. TNF blockade has demonstrated some efficacy in animal models of arthritis, like collageninduced arthritis in mice, [186] although the effects on bone and cartilage destruction are much less prominent (especially when compared with IL-1). Overexpression of TNF-alpha in transgenic mice leads to an aggressive and destructive synovitis. In fact, the arthritis also spontaneously occurs in transgenic mice that express only a membrane-bound form of TNF-alpha on T cells. [187] Clinical studies demonstrate the critical importance of this cytokine, since treatment with anti-TNF-alpha antibodies or soluble TNF-receptor-Fc fusion protein to block TNF function have striking anti-inflammatory activity in RA. [188]
Interleukin 6
IL-6 is an IL-1-inducible protein produced by T cells and monocytes and is also spontaneously produced by cultured fibroblast-like synoviocytes. [189] It can induce immunoglobulin synthesis in B cell lines, is involved in the differentiation of cytotoxic T lymphocytes, and is the major factor in the regulation of acute-phase response proteins by the liver. A striking correlation between serum IL-6 activity and serum levels of C-reactive protein, alpha1 -antitrypsin, fibrinogen, and haptoglobin was found in patients with RA. [190] Very high levels of IL-6 are present in RA synovial fluid, and synovial cells in culture from diverse inflammatory arthropathies produce IL-6. [191] In situ hybridization of frozen sections of synovial tissue also show IL-6 mRNA in the intimal lining, and immunoperoxidase studies show IL-6 protein in the lining and sublining regions. [192] Although many synovial macrophages express the IL-6 gene, the majority of IL-6 appears to be produced by type B synoviocytes. T cells might contribute to synovial IL-6 production, but the amount is small.
Interleukin 15
IL-15 is a novel cytokine that exhibits IL-2-like activity (and, in fact, can bind to subunits of the IL-2 receptor) and can serve as an alternative mechanism for inducing T cell proliferation. It is primarily produced by macrophages and might provide a key link between T cell activation and macrophages in an IL-2-deficient environment. Even more important, IL-15 can serve as a mediator that permits antigen-independent regulation of TNF-alpha production by T cells. [193] Synovial macrophages produce IL-15, which in turn induce local T cells to stimulate macrophage TNF-alpha production through a poorly defined mechanism. This provides a potential critical link between T cells and the cytokine networks that does not require an antigen-specific responses. Soluble IL-15 receptors, which bind to and inactivate IL-15, decrease joint inflammation in murine collagen-induced arthritis. [194] Treated animals also had decreased IFN-gamma production and proliferative responses to type II collagen, suggesting that the clinical efficacy also involved suppression of T cell activation. IL-15 is has been detected in RA synovial macrophages cells using immunohistochemical techniques. [195]
Granulocyte-Macrophage Colony-Stimulating Factor
GM-CSF has the ability to support differentiation of bone marrow precursor cells to mature granulocytes and macrophages. As with other major colony-stimulating factors (e.g., macrophage-CSF, granulocyte-CSF, and IL-3), GM-CSF also participates in normal immune responses. It is potent macrophage activator,
including the induction of HLA-DR expression, tumoricidal activity, IL-1 secretion, intracellular parasite killing, and priming for enhanced release of TNF-alpha and PGE2 . Neutrophil function is also regulated by GM-CSF, which enhances antibody-dependent cytotoxicity, phagocytosis, chemotaxis, and the production of oxygen radicals.
GM-CSF is present in RA synovial fluid and is produced by RA synovial tissue cells. [196] The major source in the synovium appears to be macrophages, although IL-1- or TNF-alpha-stimulated fibroblast-like synoviocytes also express the GM-CSF gene. [197] In situ hybridization studies show little or no GM-CSF mRNA in synovial T cells. Its ability to induce HLA-DR gene expression on macrophages might be of particular importance in RA: GM-CSF, not IFN-gamma, is the major DR-inducing cytokine in RA synovial fluid and in supernatants of cultured synovial tissue cells.
Chemokines
Chemokines are a family of related chemoattractant peptides that, with the assistance of adhesion molecules, summon cells into inflammatory sites. Chemokines,
TABLE 64-5 -- Classification of Chemokines
C-X-C Subfamily C-C Subfamily
Chromosome 4 (q12-21) 17 (q11-32)
Chemokine IL-8 Monocyte chemoattractant protein 1, 2, 3, and 4
GRO alpha, beta, and gamma Monocyte inhibitory protein 1alpha and 1beta
Platelet factor 4 RANTES
Epithelial neutrophil-activating peptide-78
Interferon-inducible protein-10
Granulocyte chemoattractant protein-2
Platelet basic protein and derivatives
Abbreviations: IL-8, Interleukin 8; RANTES, "regulated or activation, normally T-cell expressed and secreted."
which generally are relatively small proteins (8 to 10 kD), are divided into two major families, known as C-C or C-X-C based on the position of characteristic cysteine residues ( Table 64-5 ). In the former, two conserved cysteines are adjacent to one another, whereas in the C-X-C family, the cysteines are separated by a
nonconserved amino acid. The families are also encoded on different chromosomes, with C-C chemokines found on chromosome 17 and C-X-C factors on chromosome 4. Each individual factor has the ability to attract specific lineages of cells after interacting with specific cell surface receptors. A host of chemokines have been identified in the rheumatoid joint. IL-8, a C-X-C chemokine that was originally characterized as a potent chemoattractant for neutrophils (although it now known to also attract monocytes and other cells), along with immune complexes and other chemotactic peptides like C5a, contributes to the large influx of PMNs into the joint. Immunohistochemical analysis of synovial tissue demonstrates IL-8 protein in sublining perivascular macrophages as well as in scattered lining cells. [198] Cultured synovial tissue macrophages constitutively produce IL-8, and fibroblast-like synoviocytes express the gene if they are stimulated with IL-1 or TNF-alpha. Although proinflammatory cytokines IL-1 and TNF-alpha are capable of inducing the expression of a large number of chemokines by cultured synoviocytes, IL-8 is accounts for the majority of neutrophil-attracting activity. The addition of anti-IL-8 neutralizing antibodies eliminates about 40 percent of the neutrophil chemoattractant activity in synovial fluid. IL-8 has a number of other activities: it activates neutrophils through G-protein-coupled receptors and is a potent angiogenesis factor.
IL-8 is certainly not the only chemokine present in the RA synovium. Macrophage-inhibitory protein-1alpha, macrophage-inhibitory protein-1beta macrophage chemoattractant protein-1 (MCP-1), and RANTES ("regulated on activation, normally T cell expressed and secreted") (all members of the C-C subfamily) are produced by RA synovium. [199] Epithelial neutrophil-activating peptide-78 (ENA-78), which is a C-X-C chemokine, is also abundant. [200] ENA-78 accounts for about 40 percent of the chemotactic activity for neutrophils in RA synovial fluid. In each case, the source of the chemokine appears to be synovial macrophages or cytokine-stimulated type B synoviocytes. The regulation of each chemokine appears to be distinct in fibroblast-like synoviocytes. For instance, IL-8 production is inhibited by IFN-gamma and enhanced by IL-4, whereas the opposite is true for RANTES. [201] The concentrations of chemokines are higher in RA synovial effusions compared with samples from noninflammatory arthritides like OA. Although the chemokines can also be detected in the blood, the levels are considerably lower than in the joint, thereby providing a gradient that signals cells to migrate into the synovium.
Transforming Growth Factor-beta
TGF-beta is widely distributed in different tissues and produced by many cells, including T cells, monocytes, and platelets. Although TGF-beta alone has a modest effect on the expression of genes for collagenase and collagenase inhibitor, in the presence of other growth factors, it suppresses the production of collagenase and superinduces the expression of tissue inhibitor of metalloproteinases (TIMP). It accelerates the healing of incisional wounds and induces both fibrosis and angiogenesis in experimental animal models. Substantial amounts of TGF-beta are present in synovial fluid (although it is mainly present in an inactive latent form), and the mRNA can be detected in RA synovial tissue. [202] Although typically thought of as an immunosuppressive cytokine with wound-healing properties, the role of TGF-beta in RA is quite complex as demonstrated by the conflicting results of its administration in various animal models. When it is injected directly into the knees of animals, cellular infiltration and massive synovial lining hyperplasia develop. [203] In streptococcal cell wall arthritis, parenteral administration of the protein ameliorates
the disease. [204] Intra-articular administration of anti-TGF-beta antibody decreases arthritis in the injected joint but not in the contralateral joint in the same model. [205] Systemic gene therapy with the TGF-beta gene also suppresses streptococcal cell wall arthritis in rats. [206]
Platelet-Derived Growth Factor
PDGF is a potent growth factor comprising two chains (A and B) that are about 60 percent homologous. The A and B forms can combine to form either heterodimers or homodimers. The B chain predominates in macrophages, whereas the A chain is preferentially produced by endothelial cells. PDGF is both chemoattractant and mitogenic for fibroblasts and induces collagenase expression. PDGF is the most potent stimulator of longterm
growth of synovial cells in culture. [207] PDGF is overexpressed in vascular endothelial cells in other synovial sublining cells in rheumatoid compared with healthy synovium. [208] The PDGF receptor also is expressed in the same regions of RA synovium, suggesting the presence of an autocrine or paracrine system.
Fibroblast Growth Factors
Fibroblast growth factors (FGFs) are a family of peptide growth factors with pleiotropic activities. In rheumatoid patients, it is likely that heparin-binding growth factor, the precursor of acidic fibroblast growth factor, is a major mitogen for many cell types and stimulates angiogenesis. The interaction between FGF and proteoglycans is required for biologic activity. [209] FGF induces capillary endothelial cells to invade a three-dimensional collagen matrix, organizing themselves to form characteristic tubules that resemble blood capillaries. FGF is present in RA synovial fluid, and the genes are expressed by synovial cells. [210] Synovial fibroblasts express FGF receptors and proliferate after exposure to the growth factor.
Suppressive Cytokines and Cytokine Antagonists
The proinflammatory cytokine network described previously is balanced by a variety of suppressive factors that attempt to reestablish homeostasis. Underproduction of these suppressive cytokines could potentially contribute to the perpetuation of the RA. As described earlier, relative deficiency of IFN-gamma or IL-4 might contribute to unopposed activation of synoviocytes by TNF-alpha or other cytokines. However, in addition to these, there are many other cytokine antagonists or natural immunosuppressives that represent potential therapeutic targets for the treatment of inflammatory diseases.
Interleukin 1 Receptor Antagonist
IL-1Ra is a naturally occurring IL-1 inhibitor that binds directly to types I and II IL-1 receptors and competes with IL-1 for the ligand-binding site. Interaction of IL-1Ra with the IL-1 receptors does not result in signal transduction, and, in contrast to the
internalization of the complex after the binding of IL-1alpha or beta, the receptor-ligand complex is not internalized after it binds to the IL-1 receptor. Even though IL-1Ra has high affinity for the IL-1 receptor, it is a relatively weak inhibitor because IL-1 can activate cells even if only a small percentage of IL-1 receptors are occupied. Because of this, a large excess of the inhibitor is needed to saturate the receptor and thereby block IL-1-mediated stimulation (usually 10- to 100-fold excess of IL-1Ra). Recombinant IL-1Ra inhibits a variety of IL-1-mediated events in cultured cells derived from the joint, including the induction of metalloproteinase and prostaglandin production by chondrocytes and synoviocytes. It can block synovitis in rabbits induced by direct intra-articular injection of recombinant IL-1 but not antigen-induced arthritis. [211] [212] Two major structural variants of IL-1Ra have been described: (1) secretory IL-1Ra, which is synthesized with a signal peptide that allows it to be transported out of cells; and (2) intracellular IL-1Ra, which lacks a leader peptide due to alternative splicing of mRNA and therefore remains intracellular. Secretory IL-1Ra is a major product of mononuclear phagocytes, particularly mature tissue macrophages, and intracellular IL-1Ra is the dominant form in cultured fibroblast-like synoviocytes as well as keratinocytes and epithelial cells.
High concentrations of the IL-1Ra (up to 50 ng/ml) are present in rheumatoid synovial effusions; much is produced by neutrophils and macrophages. [213] Immunohistochemical studies of rheumatoid synovium reveal abundant IL-1Ra protein especially in perivascular mononuclear cells and the synovial intimal lining. [214] The IL-1Ra protein and mRNA can be detected in synovial macrophages and, to a lesser extent, in type B synoviocytes (Fig. 64-5 ). The presence of IL-1Ra in synovium is not specific to RA, since OA synovial tissue also contains IL-1Ra, albeit in lesser amounts; normal synovium contains little, if any, IL-1Ra protein. Despite the presence of significant amounts of IL-1Ra in synovial tissue, its importance as an IL-1 antagonist can only be
Figure 64-5 Localization of interleukin/receptor antagonist (IL-1Ra) messenger RNA in RA synovial tissue by in situ hybridization. The specific RNA transcript was detected in perivascular cells, especially macrophages. A, bright field view. B, Same area using a dark field filter. Silver grains in the dark field view show the location of IL-1Ra-positive cells.
evaluated in the context of the IL-1/IL-1Ra ratio. Although it is impossible to measure the amounts of IL-1 and IL-1Ra directly in the synovial microenvironment, studies of synovial cell culture supernatants show that the amount of IL-1Ra is insufficient to antagonize synovial IL-1. [215] There are two reasons for the relatively low levels of IL-1Ra released by synovial cells: (1) fibroblast-like synoviocytes, which account for much of the immunoreactive IL-1Ra in the synovial intimal lining, selectively produce
the intracellular form of IL-1Ra; and (2) synovial macrophages appear to have defective IL-1Ra production compared with macrophages isolated from other sites, secreting only about 1 percent as much as alveolar macrophages or monocyte-derived macrophages. Its possible use in therapy is detailed in Chapter 62 .
Interleukin 10
IL-10 is a major immunosuppressive cytokine that was originally characterized as a cytokine synthesis inhibitory factor based on its ability to block T cell cytokine production. Its immunosuppressive actions might be important in pregnancy to suppress an immune response directed against paternal MHC antigens, and it might regulate susceptibility to some parasitic infections. As noted previously, IL-10 protein is present in RA synovial fluid, and the gene is expressed by synovial tissue cells. Synovial macrophages are the major source of IL-10 in RA.
Transforming Growth Factor-beta
In addition to its tissue repair activities described earlier, TGF-beta also has immunosuppressive actions that might be important. For instance, the ability of RA synovial fluid to inhibit IL-1-mediated thymocyte production is neutralized by anti-TGF-beta antibody. TGF-beta also downregulates IL-1 receptor expression on chondrocytes.
Soluble Cytokine Receptors
Soluble cytokine receptors and binding proteins can absorb free cytokines and prevent them from engaging functional receptors on cells. Although these obviously could inhibit cytokine action, it should be kept in mind that they also could act as carrier proteins that protect cytokines from proteolytic degradation or deliver them directly to cells, or both. For instance, type II receptor is present in RA synovial fluid, along with lesser amounts of the type I receptor. [216] These soluble receptors can bind to IL-1 or IL-1Ra in synovial effusions; depending on the relative affinities for the agonist or antagonist, the soluble IL-1 receptors can have anti-inflammatory or proinflammatory activity.
Soluble TNF receptors have also been detected in RA synovial fluid. [217] Both the p55 and p75 receptors are present, sometimes in very high concentrations (>50 ng/ml). This is considerably higher than the concentration of TNF-alpha in blood or synovial fluid and probably explains why biologically active TNF is difficult to detect in RA synovial fluid despite the presence of immunoreactive protein. Synovial membrane mononuclear cells have increased surface expression and mRNA levels for both TNF receptors compared with OA synovial tissue cells or peripheral blood cells. [218] Cultured fibroblast-like synoviocytes express TNF receptors and constitutively shed them into culture supernatants. It is not clear how TNF-alpha functions in RA in the presence of such an enormous excess of soluble receptor.
Perpetuation of Synovitis by Macrophage-Fibroblast Cytokine Networks
To incorporate information on the cytokine profile into current concepts of RA, a variety of alternative models have been proposed. A central theme of these
paradigms is that the chronic inflammatory process might achieve a certain degree of autonomy that permits inflammation to persist after a T cell response has been downregulated. This could occur if the inflammation is sustained by factors produced by neighboring macrophages and synovial fibroblasts in the joint lining in paracrine or autocrine networks. Several cytokines that have been identified in the synovium or synovial fluid can participate in this system and might explain lining cell hyperplasia, HLA-DR and adhesion molecule induction, and synovial angiogenesis. The list of potential candidates in this highly redundant system is very long. For example, one can assume that at least two, IL-1 and TNF-alpha, play particularly central roles. Both are produced by synovial macrophages and stimulate synovial fibroblast proliferation and secretion of IL-6, GM-CSF, and chemokines as well as effector molecules like metalloproteinases and prostaglandins. GM-CSF, which is produced by both synovial macrophages and IL-1beta- or TNF-alpha-stimulated synovial fibroblasts, can in turn, induce IL-1 secretion to form a positive feedback loop. GM-CSF, especially in combination with TNF-alpha, also increases HLA-DR expression on macrophages. Macrophage and fibroblast cytokines could also indirectly contribute to the evidence for local T cell and B cell activation, including RF production.
This model for the perpetuation of RA clearly does not eliminate the likelihood that synovitis is initiated by a specific arthritogenic antigen. In fact, unless RA is truly caused by transformed cells, it requires an external stimulus to initiate the process, with periodic restimulation possibly required. T cell-mediated responses, directed against either an inciting antigen or a secondary target like type II collagen or proteoglycans can occur along with this macrophage-fibroblast cytokine network and might enhance the local inflammatory response. Although the factors released by macrophages and fibroblasts are reasonably well defined, the precise function of synovial T cells remains unknown. The role of T cells and antigen-specific stimulation might even change during various phases of the disease. In early RA, antigen-specific T cell activation might be most important. As the disease progresses, the cytokine networks become established and antigen-independent processes might assume a central position. In addition, nonspecific T cell functions, such its role in IL-15-mediated induction
of TNF-alpha, can also help sustain the cytokine networks.
SIGNAL TRANSDUCTION AND TRANSCRIPTION FACTORS
By diverse stimuli, intracellular signal transduction systems transduce extracellular signals initiated from the cell surface to the nucleus, where they are subsequently integrated at the level of transcription factor activity. The transcription factors bind to specific DNA sites in promoter regions and regulate the expression of the appropriate genes. The remarkable diversity of signal transduction pathways as well as transcription factors provides a selective mechanism for orchestrating activation and repression for appropriate arrays of genes in response to an extracellular stress. Many of the inflammatory responses observed in RA synovium, including the activation of cytokine and adhesion molecule genes, can be traced to specific transcription factors and signal transduction pathways. Although an extensive description of these mechanisms is beyond the scope of this chapter, it is reviewed in detail elsewhere. [219]
Nuclear Factor-Kappa B
Nuclear factor-kappa B (NF-kappaB) is a ubiquitous transcription factor that plays a key role in the expression of many genes central to RA, including IL-1beta in monocytes, as well as ICAM-1, TNF-alpha, and IL-6 in rheumatoid synoviocytes. [220] NF-kappaB normally resides as a hetero- or homodimer in the cytoplasm in an inactive form associated with an inhibitory protein called IkappaBeta that regulates the DNA binding and subcellular localization of NF-kappaB proteins by masking a nuclear localization signal. Extracellular stimuli initiate a signaling cascade leading to activation of two IkappaB kinases (IKKs), which phosphorylate IkappaB at specific NH2 -terminal serine residues. [221] Phosphorylated IkappaB is then selectively ubiquitinated and degraded by the 26S proteasome. This process permits NF-kappaB to migrate to the cell nucleus, where it binds its target genes to initiate transcription.
NF-kappaB is abundant in rheumatoid synovium, and immunohistochemical analysis demonstrates p50 and p65 NF-kappaB proteins in the nuclei of cells in the synovial intimal lining. [222] [223] Although the proteins can also be detected in OA synovium, NF-kappaB activation is much greater in RA because of phosphorylation and degradation of IkappaB in RA intimal lining cells (Fig. 64-6) . Nuclear translocation of NF-kappaB in cultured fibroblast-like synoviocytes occurs rapidly after stimulation by IL-1 or TNF-alpha through the activation of IKK. [224] Both IKK-1 and IKK-2 (also called IKKalpha and IKKbeta) are constitutively expressed by synoviocytes, and IKK functional activity increases 10- to 20-fold within 10 minutes of cytokine exposure. IKK-2 appears to be the primary path by which NF-kappaB is activated in synoviocytes after cytokine stimulation.
There is some evidence that NF-kappaB function can be abnormal in RA synovial cells. Constitutive IL-6 production is high in some RA synoviocyte clones, and this is associated with increased NF-kappaB activation involving p50 and p65 subunits. [225]
Deletion analysis of the IL-6 promoter shows that the phenomenon is regulated by two positive elements located at -159 to -142 and -77 to -59 base pairs. [226] After IL-1 treatment of the synoviocytes, NF-kappaB p50-p65 heterodimer and p65-p65 homodimer bind to the -77 to -59 site.
Activator Protein-1
Like NF-kB, activator protein-1 (AP-1) regulates many genes implicated in RA, including TNF-alpha and many metalloproteinases. AP-1 activity can be induced by extracellular signals including cytokines, growth factors, tumor promoters, and the Ras oncoprotein. AP-1 includes members of the Jun and Fos families of transcription factors, which are characterized by leucine zipper DNA-binding domains. AP-1 proteins bind to DNA and activate transcription as Jun homodimers, Jun-Jun heterodimers, or Jun-Fos heterodimers. Multiple Jun and Fos family members (c-Jun, JunB, JunD, c-Fos, FosB, Fra-1, Fra-2) are expressed in different cell types that mediate the transcription of both unique and overlapping genes.
AP-1 proteins and mRNA, including c- jun and c- fos, are expressed in RA synovium, especially in the nuclei of cells in the intimal lining layer. [227] c-Jun and c-Fos proteins are also expressed in the sublining inflammatory infiltrate, albeit to a lesser degree. Localization of AP-1 to the intimal lining correlates with the site where various protease and cytokine genes are overexpressed in RA. AP-1 proteins are usually not detected in normal synovium, although modest amounts have also been detected in OA synovium. Electromobility shift assays (EMSAs) demonstrate very high levels of AP-1 activity in nuclear extracts from RA synovium compared with OA tissue. [104] [228]
Cytokines like IL-1 and TNF-alpha probably contribute to the activation of AP-1 in RA synovium. These factors are potent inducers of AP-1 nuclear binding in cultured fibroblast-like synoviocytes. This is accompanied by increased c- jun and c- fos mRNA and enhanced collagenase gene transcription. [229] The specific Jun family genes that constitute AP-1 in synoviocytes have a clear effect on function. For instance, c-Jun increases the production of proinflammatory mediators, whereas JunD suppresses cytokine and metalloproteinase production. [230]
MAP kinases, especially Jun N-terminal kinase (JNK), are a cascade of signal transduction enzymes that regulate c-Jun phosphorylation as well as cytokine production. Three of these MAP kinases pathways, extracellular regulated kinase, p38, and JNK, are rapidly activated by phosphorylation in RA synoviocytes after brief exposure to IL-1. [231] JNK is especially important in AP-1 regulation because it is the primary kinase family that phosphorylates c-Jun. JNK2, which is one of the three JNK (JNK 1 through 3) isoforms, is constitutively expressed by RA synoviocytes and appears to be the dominant JNK. Phosphorylated JNK can be detected by Western blot analysis in extracts of rheumatoid synovium, suggesting local activation of the JNK pathway. Fas-mediated apoptosis
Figure 64-6 Nuclear factor-kappa B (NF-kappaB) activation in RA synovium. Electromobility shift assays were performed on extracts of RA and osteoarthritis (OA) synovium. NF-kappaB activity was significantly higher in RA synovial tissue extracts compared with those of OA. This is consistent with increased expression of NF-kappaB-driven genes in RA synovium, such as proinflammatory cytokines and vascular adhesion molecules. Mutant probe is shown on the left as a negative control, and C is a positive control. (From Han Z, Boyle DL, Manning AM, Firestein GS: AP-1 and NF-kappaB regulation in rheumatoid arthritis and murine collagen-induced arthritis. Autoimmunity 28:197, 1998.)
in cultured synoviocytes appears to utilize the JNK- AP-1 pathway and involves CPP32 but not IL-1-converting enzyme. [232] [233]
Activator Protein-1, Nuclear Factor-Kappa B, and Mitogen-Activated Protein Kinases in Animal Models of Arthritis
The expression and regulation of AP-1, NF-kappaB, and some MAP kinases have been evaluated in animal models of arthritis. For instance, NF-kappaB expression is increased in the adjuvant arthritis model in rats as early as 3 days after immunization even though arthritis does not occur until about 1 week later. [234] Inhibitors of IkappaB phosphorylation suppress clinical arthritis in this model. [235] Decoy oligonucleotides that bind intracellular AP-1 inhibit joint inflammation and destruction in murine collageninduced arthritis. [236] In another study, inhibition of local NF-kappaB activation using an adenoviral vector encoding a dominant negative IkappaB gene increased synovial apoptosis and suppressed inflammation in the rat streptococcal cell wall arthritis model. [237] The p38 inhibitors have demonstrated efficacy in murine collagen-induced arthritis and rat adjuvant arthritis, possibly by decreasing the production of IL-1 and TNF-alpha. [238] Treated animals also had improved bone mineral density and decreased histologic evidence of joint inflammation.
The time course of transcription factor activation and matrix metalloproteinase (MMP) expression in inflammatory arthritis has been evaluated in murine collagen-induced arthritis using EMSA on joint extracts. [104] Notably, synovial AP-1 and NF-kappaB activation increase long before the onset of arthritis: NF-kappaB activation first occurs on day 10 and peaks in late disease, whereas AP-1 activation increases by day 20. In contrast, clinical synovitis is delayed until day 25 to 30. Synovial collagenase and stromelysin gene expression do not increase until day 35 to 45. These studies are consistent with clinical studies in RA suggesting that even asymptomatic joints in RA exhibit synovial inflammation, p53 expression, and cytokine production.
Signal Transducers and Activators of Transcription
The signal transducers and activators of transcription (STATs) are a family of latent cytoplasmic transcription factors that are activated in response to cytokine stimulation of cells. STAT proteins contain domains that promote the docking to the appropriate tyrosine-phosphorylated cytokine receptor. STATs have been implicated
in the expression of many proinflammatory genes. Active STAT3 has been detected in cells from inflamed joints, [239] and synovial fluid from RA patients can activate STAT3 but not STAT1 in monocytes. [240] The induction of STAT
translocation in the joint is independent of IFN-gamma and appears to be regulated primarily by IL-6.
LIFE AND DEATH IN THE RHEUMATOID SYNOVIUM: OXIDATIVE DAMAGE, APOPTOSIS, AND TRANSFORMATION
Studies defining the life cycle of cells have opened a new door to understanding the pathogenesis of neoplastic and inflammatory diseases. One could posit that the increased cellularity is not necessarily due to enhanced proliferation but could also result from defective elimination of cells through apoptosis, or programmed cell death.
Reactive Oxygen and Nitrogen
Oxidative stress in the joints of RA patients results from a confluence of several stimuli, including increased pressure in the synovial cavity, reduced capillary density, vascular changes, an increased metabolic rate of synovial tissue, and locally activated leukocytes. The generation of reactive oxygen species can also be facilitated by a repetitive ischemia reperfusion injury in the joint. [241] Tissue injury releases iron and copper ions and heme proteins that are catalytic for free radical reactions. [242] Electron transport chains are also disrupted in the mitochondria and endoplasmic reticulum, leading to leakage of electrons to form superoxide.
Evidence for increased production of reactive oxygen species in RA patients includes elevated levels of lipid peroxidation products, degradation of hyaluronic acid by free radicals, decreased levels of ascorbic acid in serum and synovial fluid, and increased breath pentane excretion. Moreover, the levels of thioredoxin, which is a marker of oxidative stress, are significantly higher in synovial fluid from RA patients compared with that from patients with other forms of arthritis. [243] Peripheral blood lymphocyte DNA from RA patients contains significantly increased levels of the mutagenic 8-oxohydrodeoxyguanosine, [244] which is a product of oxidative damage to DNA, pointing to the genotoxic effects of oxidative stress.
The production of nitric oxide (NO) is also upregulated in rheumatoid synovial tissue. [245] Low levels of NO are constitutively produced by endothelial or neuronal synthases, and this is increased substantially by inducible NO synthase after stimulation by cytokines or bacterial products. The nitrite levels in synovial fluid are elevated in RA patients, indicating local NO production. [246] In addition, the urinary nitrate-to-creatinine ratio is increased and inducible NO synthase is present in the synovium.
Apoptosis
Programmed cells death, or apoptosis, is a process by which cells can be eliminated in the midst of living tissue. This stereotypic response provides a mechanism for tissue development, remodeling, or cell deletion without instigating an inflammatory response. Apoptosis is a normal process that is tightly regulated and can be initiated
by withdrawal of hormones and growth factors and is evident in the elimination of autoreactive cells like thymocytes in the thymus gland and the loss of cells in response to DNA damage.
The accumulation of cells in RA is typically considered as a process involving in situ cell proliferation or recruitment of cells from the blood stream. However, it is equally tenable that increased cell numbers could collect in the synovium due to insufficient cell deletion as a result of unexpectedly low levels of apoptosis. There is some evidence that T cell apoptosis in RA synovial effusions is defective. [247] Using flow cytometry to assess the percentage of live and apoptotic cells, substantial cell death was detected among the RA synovial fluid neutrophil population, but little or no apoptosis of T cells was seen. This contrasted with crystal-induced arthropathy, in which abundant T cell death occurred. Evaluation of the RA T cells showed high Fas expression, high Bax, and low Bcl-2, which is a phenotype typically associated with increased susceptibility to apoptosis. This contrasts with synovial tissue cells in which high Bcl-2 expression is found in lymphoid aggregates. Resistance of RA synovial fluid lymphocytes to apoptosis can be maintained by coculturing RA or gout synovial fluid T cells with fibroblast-like synoviocytes or IL-15. The specific adhesion molecules involved are not known, although the RGD motif (arginine-glycine-asparagine) could block the protective effects of synoviocytes. This suggests a role for integrins like alphav beta3 .
The role of Fas (CD95), which is a potent membrane cell death receptor on many cell types, in synovial T cell homeostasis is uncertain. Higher amounts of Fas are found on synovial fluid T cells in RA, although the number of Fas+ cells in the peripheral blood of RA patients is greater than in healthy controls. An anti-Fas antibody, which cross-links Fas on cell surfaces, rapidly causes apoptosis in synovial fluid B and T lymphocytes in RA, although peripheral blood T cells are resistant. The presence of the natural ligand to Fas, Fas ligand is controversial, with some groups [248] detecting Fas ligand whereas others [249] do not.
Studies of apoptosis in RA synovial tissue have relied on a number of techniques that label damaged DNA. Using the most stringent methods, only a small number of apoptotic nuclei have been detected in both the intimal lining and sublining. [250] This has been confirmed using gel electrophoresis to identify fragmented DNA. With the use of morphologic criteria, only rare cells exhibit the typical findings of programmed cell death. Surprisingly, less specific techniques that detect any DNA damage caused by other genotoxic stimuli show that 30 to 50 percent of cells in the intimal lining have evidence for nuclear fragmentation. [251] Hence, there is an unexpected discrepancy between the cytologic evidence of DNA damage and the rarity of typical morphologic changes of apoptosis, even when synovial membranes are studied by electron microscopy. [252] Despite the dearth of apoptotic cells in the lining, Bcl-2 expression
(which inhibits apoptosis) is low in this region. Furthermore, p53 expression is increased in the synovial lining and sublining. As discussed later, functional p53 should induce cell cycle arrest and either DNA repair or apoptosis. Notably, abnormal p53 function would permit the persistence of DNA damage without
significant apoptosis.
Because of these surprising findings, the regulation of apoptosis was evaluated in cultured fibroblast-like synoviocytes in RA. Fas is constitutively expressed by cultured synoviocytes, and programmed cell death is initiated in a minority of cells when it is cross-linked by anti-Fas antibody. [251] Although there is some disagreement in the literature, by and large most investigators find that RA and OA synoviocytes are equally susceptible to anti-Fas-mediated death and that a relatively small percentage of cells are killed by this process. This process can also be initiated by oxidative stress, such as hydrogen peroxide, or by exposure to NO. Oxidation-induced apoptosis is closely regulated by the p53 tumor suppressor gene. When p53 activity is blocked, apoptosis is enhanced. TGF-beta decreases Fas antigen expression and upregulates Bcl-2 in synoviocytes. [253] Synoviocytes stimulated with TGF-beta become markedly resistant to anti-Fas-mediated apoptosis. Fas-mediated synoviocyte death appears to utilize the JNK signal transduction pathway and involves activation of the transcription factor AP-1. [254]
The potential relevance of Fas-induced death has been demonstrated in murine collagen-induced arthritis, in which high levels of Fas and low levels of Fas ligand are expressed on synovial cells. [255] Mice received an intra-articular injection with an adenoviral vector encoding Fas ligand, which led to a decrease in synovial inflammation. DNA-labeling studies showed that the construct increased synovial apoptosis. Therefore, the induction of programmed cell death in the joint represents a possible therapeutic target for RA.
Synoviocyte Transformation
Although RA fibroblast-like synoviocytes often appear and behave like normal fibroblasts, they also exhibit certain characteristics that suggest partial transformation. Certainly, the dichotomy between the extent of DNA fragmentation in the synovial lining and the lack of effective apoptosis is consistent with this notion. Several properties of cell transformation have been evaluated in RA and have led to the conclusion that synoviocytes have been permanently altered by their environment. For instance, adherence to plastic or extracellular matrix is generally required for normal fibroblasts to proliferate and survive in culture. Transformed cells, on the other hand, can grow in suspension or in semisolid medium without contact with a solid surface. Although fibroblast-like synoviocytes typically grow and thrive under conditions that permit adherence, they can also proliferate in an anchorage-independent manner. [256]
A second unusual feature of cultured RA synoviocytes suggestive of partial transformation is the loss of contact inhibition. Normal cultured fibroblasts proliferate in culture only until they reach confluence. Malignant or transformed fibroblasts lose this control mechanism and proliferate even after contact. Under some conditions, however, cultured fibroblast-like synoviocytes also escape contact inhibition. This is especially true if the cells are cultured in the presence of PDGF, which is a key fibroblast-like synoviocyte growth factor. Unrestricted cell growth likely occurs in vivo as well, and studies examining X-linked genes demonstrated clonality in the synoviocyte population from RA but not OA synovium. [257] This is especially true of cells derived from the invading pannus, which is the most aggressive region of the synovium.
RA synoviocytes also express an array of transcription factors that regulate DNA synthesis similar to those observed in tumor cells. One of the most important is c-Myc, which is a critical signal that initiates cell proliferation and is expressed by both cultured fibroblast-like synoviocytes and intimal lining cells in RA. [258] Although late-passage OA and RA synoviocytes are often indistinguishable, some data suggest that the latter spontaneously secrete greater amounts of cytokines and growth factors that can subsequently regulate expression of c-myc and other oncogenes. [259] Elevated production of growth factors like TGF-beta or FGF could increase proliferation and invasiveness through autocrine stimulation.
The most convincing data indicating permanent changes in RA synoviocyte function were demonstrated in the SCID mouse model utilizing cultured fibroblast-like synoviocytes that have been co-implanted with cartilage into the renal capsule. RA synoviocytes, but not OA or normal synoviocytes or normal skin fibroblasts, invade the cartilage explants. These data provide strong evidence that fibroblast-like synoviocytes, are irreversibly altered in RA and that an autonomous process allows them to remain activated even after removal from the articular inflammatory milieu. The cells continue to migrate and invade without additional exogenous stimulation.
Tumor Suppressor Genes and Mutations: A Mechanism for Autonomous Synoviocyte Activation
The p53 tumor suppressor is a key regulator of DNA repair and cell replication. Although not an oncogene itself, p53 is under the transcriptional control of oncogenes like c-myc and provides the critical signals to arrest cell growth or induce apoptosis, or both. The suppressor p53 has several domains that serve distinct functions. A transactivation region stimulates transcription of a number of genes, including p21waf , the ribosomal gene cluster, and GADD45. In contrast, a transrepression region decreases expression of proteins like RB1 and PCNA. Many of the suppressed genes regulate cell proliferation, and down regulation arrests the cell cycle at the G1 phase. This is especially prominent after cells are transformed by oncogenes like ras or sustain DNA damage, thereby leading to apoptosis or providing sufficient time for DNA repair.
The p53 protein is overexpressed in the rheumatoid synovium. In long-standing disease marked by marked joint destruction, immunostaining localizes the protein to sublining mononuclear cells as well as the intimal lining. [260] Immunoreactive p53 was also detected in some OA joint tissues, but the degree of staining was significantly less than in RA. Of interest, p53 protein can also be detected in RA synovium from patients with very early RA as well as in asymptomatic rheumatoid joints. [261] However, its expression is much lower in other inflammatory arthropathies like reactive arthritis. RA fibroblast-like synoviocytes also constitutively express significantly greater amounts of p53 compared with skin fibroblasts or OA synoviocytes.
In seeking a unified explanation for these observations, the possibility that somatic mutations in the p53 gene might explain, in part, the high p53 expression, the
"transformed" phenotype of synoviocytes, and inadequate apoptosis in rheumatoid synovial tissue was considered. [262] cDNAs from cultured fibroblast-like synoviocytes and synovial tissue from RA patients were subcloned and sequenced, revealing mutations in the synovium of most RA patients requiring joint replacement surgery. Mutations were not found in the skin or blood from the same patients, indicating that they were somatic rather than germline. Transition mutations, which are characteristic of oxidative deamination, were present in more than 80 percent of the patients. Although the existence and function of p53 mutations in RA remain controversial, the presence of p53 mutations in synovium and cultured fibroblast-like synoviocytes from patients with long-standing, erosive RA has been confirmed. [263]
Almost all the mutations identified in the rheumatoid samples were similar to those previously identified in neoplastic diseases, suggesting that they have functional significance. At least some of them were shown to exhibit dominant negative characteristics, thereby suppressing the function of the wild-type allele. [264] Hence, synoviocytes containing these dominant negative mutations lack detectable p53 function. The effect of losing p53 function in fibroblast-like synoviocytes was investigated by transducing fibroblast-like synoviocytes expressing wild-type p53 with the human papilloma virus 18 E6 gene, which inactivates p53. [265] The loss of p53 function led to enhanced fibroblast-like synoviocyte proliferation, anchorage-independent growth, and invasiveness into cartilage extracts as well as impaired apoptosis. Thus, cells with p53 mutations might have a selection advantage, because these mutations produce resistance to p53-dependent apoptosis. In RA, this process could provide a mechanism for the creation of partially transformed synoviocyte islands in the synovium marked by increased invasive potential.
Mutations of p53 could have an impact on cytokine and metalloproteinase production in rheumatoid synovium. For instance, expression of IL-6, inducible NO synthase, and collagenase are repressed by wild-type p53. A dominant negative p53 mutation could therefore inhibit increased expression of these genes and contribute to the perpetuation of synovitis as well as increase joint destruction. This hypothesis suggests that mutations in key genes like p53 are a result of long-standing inflammation that subsequently changes the character of the disease. At present it is unclear whether p53 mutations will also be observed in synovial cells other than fibroblast-like synoviocytes or whether mutation in other key genes will be identified.
Other Mutations in Rheumatoid Arthritis
Although p53 mutations are the best characterized mutations in RA, changes in other genes have also been reported. For instance, synovial T cells in RA have an increased incidence of mutations in the HPRT gene. [266] Although not functionally significant, these act as a marker for oxidative damage that occurs in the synovial milieu. Some of these abnormal lymphocytes can also be detected in the peripheral blood, suggesting that articular T cells can migrate out of the joint to other parts of the body. Abnormalities of the ras gene are perhaps more interesting. The ras family of oncogenes include three genes ( H-ras, K-ras, and N-ras) that encode guanine-binding proteins. Ras proteins are localized on the inner leaflet of the plasma membrane and are involved in many signal transduction pathways, including MAP kinases. Certain ras point mutations in codons 12, 13, and 61 result in a constitutively active protein associated with cell transformation. The H-ras gene is expressed in the synovium of patients with a variety of arthritides. [267] Mutations in
the H-ras gene have been identified in both RA and OA synovium. [268] Surprisingly, the incidence of these mutations is actually greater in OA. One of the mutations (Gly to Asp at codon 13) is known to be activating and occurs in a small percentage of patients.
BLOOD VESSELS IN ARTHRITIS: ADHESION MOLECULES AND ANGIOGENESIS
Blood vessels used to be thought of as passive conduits through which red blood cells and leukocytes circulated while en route to an inflammatory site. This is now known to be far from the truth: The microvasculature plays an extremely active role in such processes, not only as the means of selecting which cells should enter the tissue but also as a determinant of tissue growth and nutrition through the proliferation of new capillaries.
Angiogensis in Rheumatoid Arthritis: Feeding the Starving Synovium
From the vantage of the proliferation of new blood vessels in the synovium, synovitis in RA resembles both tumor growth and wound healing. The importance of luxurious new capillary growth early in the development of synovitis was emphasized by Kulka and coworkers many years ago (Fig. 64-7) . [269] Decades later, Folkman and colleagues demonstrated the first soluble factors responsible for inducing an endothelial cell with
Figure 64-7 Human rheumatoid synovial membrane (4-mm thickness) stained with rabbit anti-human type IV collagen. This gives precise definition in blood vessels, the only structures in synovium that contain type IV collagen. Virtually all of these blood vessels have formed in response to angiogenic stimuli after the rheumatoid process had been initiated. (Courtesy of Drs. S. and R. Gay.)
the capability to proliferate and develop new capillaries. The importance of new blood vessel formation in inflammatory arthritis has been elegantly demonstrated in the collagen-induced arthritis model. The disease was markedly attenuated in animals pretreated with an angiostatic compound similar to fumagillin, which is derived from Aspergillus.[270] This compound is cytotoxic to proliferating but not resting endothelial cells. In addition, there was regression of established arthritis if treatment was initiated well into the course of the disease. Hence, angiogenesis is essential for the establishment and progression of inflammatory arthritis, because of the need for blood vessels either to recruit leukocytes or to provide nutrients and oxygen to starved tissue.
Taxol, which can induce endothelial cell apoptosis, was also effective in an animal model of arthritis. [271] Vascular corrosion casts of arthritic rat synovium revealed marked expansion of the blood vessel volume and an extensive interconnecting network. The vasculature reverted to the normal synovial morphology in the treated animals as arthritis diminished. An alternative approach to angiogenesis blockade in arthritis was a cyclic RGD peptide that blocks alphav beta3 integrin. [272] This integrin is intimately involved in blood vessel growth, especially in wound healing and neoplasms. As with RA synovium, alphav beta3 is expressed by proliferating blood vessels in inflamed rabbit synovium. When the cyclic peptide was used, modest clinical efficacy was observed in both preventative and therapeutic dosing regimens. This was associated with decreased synovial inflammation and increased endothelial cell apoptosis. Bone and cartilage protection were more impressive than the effect on standard measures of inflammation, suggesting that synovial expansion supported by angiogenesis is more relevant to joint destruction than recruitment of inflammatory cells.
The absolute number of blood vessels is increased in RA synovium, with a rich network of sublining capillaries and postcapillary venules in histologic sections stained with endothelium-specific antibodies. However, the mass of tissue appears to outstrip angiogenesis in RA as determined by the number of blood vessels per unit area. [273] This could result in local tissue ischemia, a situation that has been amply documented in vivo. Synovial fluid oxygen tensions can be remarkably low, lactate measurements are frequently high, and a pH as low as 6.8 has been found. [274] The mean rheumatoid synovial fluid P O2 in samples from rheumatoid knees was 27 mm Hg, [275] and in a subsequent study, a P O2 of less than 15 mm Hg was measured in a quarter of fluids examined. [276] Another cause of diminished blood flow may be the increased positive pressure exerted by synovial effusions within the joint, a process that could effectively obliterate capillary flow and exacerbate the hypoxia in these tissues while producing ischemia-reperfusion injury in the joint. [277] Physiologic determinations by Simkin and colleagues have supported this [278] ; clearance values generated by the kinetics of iodine-123 removal from synovial fluid have shown that small solute clearance from rheumatoid synovial effusions is less than in normal individuals or in patients with other rheumatic diseases. Patients with the lowest synovial iodine clearance have the lowest synovial fluid pH, the lowest synovial fluid glucose-to-serum glucose ratios, the lowest synovial fluid temperatures, the highest synovial fluid lactate levels, and the highest numbers of synovial fluid neutrophils. Thus, the most seriously affected rheumatoid joints may be both hypoperfused and ischemic. Diminished blood flow relative to need may be decreased further by high intrasynovial pressure from effusions, as well as the existence of abnormal microvascular structures. [279] Altered vascular flow may not be the only cause of hypoxia in joints. It has been estimated that the oxygen consumption of the rheumatoid synovium (per gram of tissue) is 20 times normal. [280]
Hypoxic drive is a potent stimulus for angiogenesis. One of the mechanisms by which this occurs is through the production of angiogenic factors like vascular endothelial growth factor (VEGF). [281] VEGF, which is also known as vascular permeability factor, is a specific endothelial cell mitogen that also appears to possess some chemotactic activity. It is present in high concentrations in synovial fluid, and immunoreactive VEGF is readily detected in the synovium in and about blood
vessels. [282]
VEGF is also able to stimulate the expression of collagenase, which can degrade the extracellular matrix to make room for advancing pannus. [283] The VEGF receptor is present in the same area. VEGF is also highly expressed in the synovial intimal lining and is produced by cultured fibroblast-like synoviocytes that have been exposed to hypoxia and IL-1. [284] In addition to the hypoxia-driven stimulus for blood vessel growth, the inflammatory cytokine milieu of the joint also encourages angiogenesis. Several cytokines known to be produced in the joint, including IL-8, FGF, and TNF-alpha, are known to be angiogenic factors. Additional angiogenesis factors, such as soluble E-selectin and soluble VCAM, are produced in the rheumatoid joint and contribute to vascular proliferation. [285] Some factors that inhibit capillary proliferation, such as platelet factor-4 and thrombospondin, are also produced by the joint. [286]
Blood vessels also control the rate that various proteins leave the circulation and enter the joint. It has been known for many years that there is an inverse relationship between the molecular weight of proteins and their concentrations in minimally inflamed synovial fluid; the high-molecular-weight serum proteins gain access more easily to synovial fluid in inflamed joints, and the relatively high concentration of IgG in RA synovial fluid is good evidence for local (synovial) synthesis of IgG. [287] "Protein traffic" in human synovial effusion has been measured by determining the clearance of albumin and other proteins from synovial fluid. This gives a useful measure of afferent synovial lymph flow. An increased "permeance" of proteins in rheumatoid patients was found to be more than seven times greater than that suggested by synovial fluid-to-serum ratios and underscores the severity of the microvascular lesion in rheumatoid synovitis. [288]
Adhesion Molecule Regulation
The formation of new capillaries is only one aspect of blood vessel involvement in the rheumatoid process. Endothelial cells are also activated by cytokines to express adhesion proteins that bind to counterreceptors on mononuclear cells and neutrophils from the circulation and facilitate their transfer from the circulation into the subsynovial tissue (see Chapter 17) . When activated, some endothelial cells in postcapillary venules take on a tall, plump appearance, and these cells in the aggregate are referred to as high endothelial venules. These are a major site of blood cell adhesion and subsequent migration into the tissue, a process mediated by adhesion proteins. There are two primary categories of vascular adhesion molecules. The selectins (E-, L-, and P-selectin) are a family of adhesion molecules whose primary ligands are carbohydrates, especially sialyl Lewisx and related oligosaccharides. E-selectin is transiently expressed on endothelial cells that have been activated by cytokines like IL-1 or TNF-alpha, whereas L-selectin is expressed on most leukocytes. Because L-selectin is glycosylated, it can serve as a counterreceptor for E-selectin on vascular endothelium. L-selectin is shed from neutrophils within minutes after they are activated by cytokines like IL-8 or platelet-activating factor. P-selectin (originally characterized in platelets) is expressed in endothelial cell cytoplasmic granules and can be rapidly mobilized to the cell
surface after activation by thrombin, histamine, or C5a. The integrin family is complex and is more fully described in Chapter 17 . These adhesion proteins are heterodimers comprising alpha- and beta-chains. The counterreceptors depend on the specific combination of these chains and are frequently proteins in the immunoglobulin supergene family (e.g., the combination of ICAM-1 and alphaM beta2
) or extracellular matrix proteins (e.g., the combination of fibronectin and alpha5 beta1 or vitronectin and alphav beta3 ).
The currently accepted paradigm for leukocyte recruitment into inflammatory sites is as follows: selectins initiate the first interactions between the circulating cell and the endothelium. The resultant low-affinity binding causes cells to slow down and roll along the blood vessel wall. In the second stage, endothelial cells release specific activating factors like IL-8 that cause the cells to stick firmly and spread out. This process is mediated by the integrins, especially the beta2 family on leukocytes and the ICAM counterreceptors on endothelial cells. VLA-4, also known as alpha4 beta1 , on lymphocytes and monocytes, can also mediate immobilization via interactions with VCAM-1 or CS1 fibronectin on endothelium. Finally, cells migrate through endothelial pores into the tissue by an integrin-dependent process.
As one might expect, adhesion molecule expression is increased in the RA synovium ( Table 64-6 ). This is almost certainly due to exposure of the vasculature to the rich cytokine milieu, especially IL-1 and TNF-alpha. IFN-gamma and IL-4 are also known to increase adhesion molecule expression, but their role is uncertain due to the relatively low production in RA. High levels of ICAM-1 are expressed in the rheumatoid synovium. [289] Immunohistochemical techniques have localized ICAM-1 to sublining macrophages, macrophage-like synovial lining cells, and fibroblasts in greater amounts than normal tissue. Significant amounts are also present on the majority of vascular endothelial cells, although the ICAM-1 levels are quantitatively similar to those of vessels in normal endothelium. Cultured fibroblast-like synoviocytes constitutively express ICAM-1. TNF-alpha, IL-1, and IFN-gamma dramatically increase synoviocyte ICAM-1 expression. Maintenance of ICAM-1 expression requires continuous cytokine exposure, and ICAM-1 levels decrease to baseline within a couple of days if the cytokine is removed. The function of ICAM-1 on nonendothelial cells like synoviocytes is not known. It might serve as a counterreceptor to leukocyte integrins and act as a barrier to cells trying to migrate through the tissue to the synovial fluid space. Alternatively, integrin counterreceptors are co-stimulatory molecules that participate in the activation of T lymphocytes through alphaL beta2 (LFA-1).
Adhesion of alpha4 beta1 -expressing cells to cytokine-activated endothelial cells can be mediated by VCAM-1. Under some culture conditions, VLA-4 mediates CD18-independent monocyte transendothelial migration. [290] VLA-4, which is predominately expressed on lymphocytes, monocytes, and eosinophils, but not on neutrophils, serves as a receptor for both the six- and seven-
TABLE 64-6 -- Major Adhesion Molecule Interactions with Rheumatoid Arthritis Synovial Endothelium
Endothelial Cell Adhesion Molecule
Leukocyte Counterreceptors
Leukocytes Expressing Counterreceptor
ICAM family beta2 -integrins Neutrophils, lymphocytes, monocytes
VCAM-1 alpha4 beta1 (VLA-4); alpha4 beta7
Lymphocytes, monocytes
CS1 fibronectin alpha4 B1 (VLA-4) Lymphocytes, monocytes
E-selectin, P-selectin L-selectin; sLex Neutrophils, lymphocytes, monocytes
PECAM-1 alphav beta3 Lymphocytes, monocytes
Hyaluronate CD44 Neutrophils, lymphocytes, monocytes
CS1, Connecting segment; ICAM-1, intracellular adhesion molecule-1; PECAM-1, platelet-endothelial cell adhesion molecule; VCAM-1, vascular cellular adhesion molecule-1; VLA-4, very late activation antigen.
domain forms of VCAM-1 and a 25-amino acid sequence in an alternatively spliced region of fibronectin (FN) known as CS1. The binding sites for CS1 and VCAM-1 on VLA-4 are either very close to each other or physically overlap.
A role for VLA-4 in arthritis has been suggested by a number of experimental observations. In adjuvant arthritis in rats, anti-alpha4 antibody decreased lymphocyte accumulation in the joint but not lymph nodes, suggesting that VLA-4 is more important in recruitment to inflamed sites than to noninflamed sites. [291] In streptococcal cell wall arthritis, intravenous injection of CS1 peptide decreased the severity of acute and chronic arthritis. [292] T lymphocytes isolated from the synovial fluid and synovial membrane of RA patients exhibit increased VLA-4-mediated adherence to both CS1 and VCAM-1 relative to autologous peripheral blood lymphocytes. [293] These studies also suggest that leukocytes expressing functionally activated VLA-4 are selectively recruited to inflammatory sites in RA.
Moderate amounts of VCAM-1 are expressed in RA synovial blood vessels. Surprisingly, the intimal lining is the location of the most intense staining with anti-VCAM-1 antibodies on histologic sections of synovium. Even normal synovial tissue expresses VCAM-1 in the lining, albeit less than in RA tissue. Cultured fibroblast-like synoviocytes constitutively express small amounts of VCAM-1, and the level can be increased by IL-1, TNF-alpha, IFN-gamma, and IL-4. VCAM-1 on synoviocytes is functionally active and can support T cell binding. VCAM-1 also contributes to T cell adhesion to high endothelial venules in frozen sections of RA synovium. [294]
The expression and functional significance of CS1-containing forms of FN generated by alternative splicing has also been studied in RA. [295] Unlike most molecular forms of FN typically found in the extracellular matrix, CS1 expression is restricted to inflamed RA vascular endothelium and the synovial intimal lining. In situ hybridization
studies confirm that the CS1 FN gene is expressed in the synovial endothelial cells. [296] Normal synovial tissue contains little, if any, CS1 fibronectin. Ultrastructural studies show that CS1-expressing fibronectin molecules decorate the lumen of RA endothelial cells but not the abluminal side of the endothelium. RA synovial endothelium binds activated T lymphocytes; this can be blocked by anti-alpha4 antibody and synthetic CS1 peptide but not anti-VCAM-1 antibody, suggesting that the CS1-VLA-4 interaction is critical to lymphocyte homing to the joint. Immunoelectron microscopy also showed that CS1 fibronectin is expressed on the surface of fibroblast-like synoviocytes in the synovial intimal lining. Binding studies confirmed this in vitro, because a portion of VLA-4-mediated T cell adhesion to cultured synoviocytes is blocked by CS1 peptides.
The integrin alpha4 beta7 , which can also bind to VCAM-1, is analogous to LPAM-1, a specific adhesion molecule involved in lymphocyte homing to Peyer's patches. Nearly all intraepithelial and 40 percent of lamina propria lymphocytes express alpha4 beta7 ; this molecule is rarely identified in other lymphoid tissues. The expression of alpha4 beta7 on peripheral blood lymphocytes from patients with RA is similar to that of controls (7.3 percent), whereas 25.4 percent of synovial fluid lymphocytes express this adhesion molecule. [297] Moreover, 62 percent of the alpha4 beta7 -positive cells in synovial fluid lymphocytes are of the CD8 subtype. Why this occurs when other lymphoid tissues do not express alpha4 beta7 is undefined, and it is not known whether cells bearing this activation marker will be found within the synovial tissues. It does suggest, however, another potential linkage between the gastrointestinal epithelium and the joint, which has been a source of much discussion in several inflammatory joint diseases.
E-selectin expression has also been described in rheumatoid synovium, although the levels are much less than those of the integrins. [298] This might be due, in part, to the kinetics of E-selectin expression on endothelial cells. The protein is not found on resting endothelial cells and peaks after about 3 hours of cytokine stimulation. However, even in the continued presence of cytokine, E-selectin expression then declines to near basal levels after about 6 hours. This might explain the relatively low amounts of this protein in chronically inflamed tissue. In one study, E-selectin expression was decreased in synovial biopsies after patients were treated with injectable gold and corticosteroids. [299]
The therapeutic potential for anti-adhesion therapy has been studied in the SCID mouse model. Labeled human peripheral mononuclear cells were injected into engrafted mice, and migration into the tissue was examined. [300] If the mice were treated with TNF-alpha, trafficking into synovium was significantly increased. This was accompanied by increased ICAM-1 expression, and anti-ICAM-1 antibody blocked leukocyte migration into the explant. Tonsil mononuclear cells also migrated into the RA synovial grafts in SCID mice. [301] In this case, leukocyte accumulation was blocked by anti-LFA-1 antibody
(a counterreceptor for ICAM-1) but not by antibody to alphaE beta7 integrin (CD103).
CARTILAGE DESTRUCTION
Cartilage Destruction and the Pannus-Cartilage Junction
In RA, the cartilage is often covered by a layer of tissue comprising fibroblast-like cells, which might represent the progenitor of the aggressive mature pannus. In the established lesion, numerous areas are seen in which "aggressive cell clusters" of mesenchymal cells, both macrophage-like and fibroblast-like, appear to have a leading edge of penetration into cartilage matrix far from blood vessels of lymphocytes. [302] However, some areas show relatively acellular pannus tissue, suggesting that there is little if any enzyme action in these areas. [303] Whereas other sections show microfoci of one particular cell type, including microabscesses of PMNs, mast cells, or dendritic or "stellate" cells. Macrophage and fibroblast foci were three to five times more common than those containing mast cells or PMNs. [304] Rarely, proliferating small blood vessels surrounded by cellular infiltrates penetrate deeply into the cartilage.
Multinucleate giant cells are particularly common at the erosive front when penetration is from the subchondral side of cartilage. Both osteoclasts and large giant cells named chondroclasts have been observed degrading bone and mineralized cartilage, as well as areas of unmineralized (hyaline) articular cartilage. [305] These multinucleate cells stain brightly for acid phosphatase. It should be remembered that the experimental production of multinucleate cells in cultures of synovial fibroblasts is associated with an enormous increase in production in these cultures of collagenase and, presumably, other MMPs. [306] It may be that bradykinin released by the proliferating cells has a significant effect on bone resorption at the pannus-cartilage junctions; it appears to stimulate osteoclast-mediated bone degradation by a process that is dependent on endogenous prostaglandin formation. [307]
There may be consistent variations in the histopathologic appearance of the cartilage-pannus junction between different joints. Invasive pannus is more commonly found in sections from involved metatarsophalangeal joints compared with hip and knee joints in which a layer of resting fibroblasts appeared to separate pannus from cartilage. [308] This could explain the fact that erosions are seen more often around small joints such as metatarsophalangeal joints, whereas joint space narrowing without erosions is more common in knees. A more primitive cell type might play a role in RA, especially at the cartilage-pannus junction. This less differentiated cell type could, then, be responsible for the aggressive degradation of the extracellular matrix and help drive the destructive phase of the disease. Fibroblast-like synoviocytes from the intimal lining exhibit some characteristics of transformed cells but are not necessarily alone in their ability to degrade articular structures. Other cells in the joint, especially from the pannus that erodes directly into cartilage, could also be responsible for cartilage and bone erosions. Since the cartilage-pannus junction is essentially devoid of lymphocytes, this cell would likely
be of mesenchymal origin and derived from synoviocytes or chondrocytes. Such cell types have been isolated directly from the cartilage-pannus junction where it burrows into cartilage.
These primitive cells consistently express phenotypic and functional features of both synoviocytes and chondrocytes and have been referred to as pannocytes because of their site of origin. [309] They exhibit a distinctive rhomboid morphology and can grow in culture for a prolonged time without becoming senescent. In addition, VCAM-1 surface expression is constitutive and very high compared with that of synoviocytes or chondrocytes. The VCAM-1 is functional and can support the adhesion of T cells. Pannocytes exhibit some features of chondrocytes in that both express inducible lymphocyte antigen, which is a surface receptor in the TNF-nerve growth factor family, and inducible NO synthase. Neither of these genes are expressed by fibroblast-like synoviocytes. Interestingly, pannocytes, like synoviocytes, do not produce NO synthase even though they contain NO synthase mRNA. Pannocytes are more fibroblast-like in that they produce type I but not type II collagen and, like synoviocytes, contain intracellular vimentin. Others have suggested that pannus-derived cells can express the type II collagen gene or produce proteoglycans that stain positively with safranin O. [310] Taken together, these data still do not allow one to determine whether pannocytes are dedifferentiated chondrocytes or fibroblast-like synoviocytes, or whether they represent a truly separate cell lineage. Given that they are fully capable of producing collagenase and are isolated directly from eroding cartilage, these cells likely have great relevance to RA.
Cartilage is destroyed in RA by both enzymatic and mechanical processes. The enzymes induced by factors such as IL-1, TNF-alpha, phagocytosis of debris by synovial cells, and both free and bound iron cause the joint destruction. Early in synovitis, proteoglycans are depleted from the tissue, most likely due to the catabolic effect of cytokines like IL-1 on chondrocytes, and this leads to mechanical weakening of cartilage. As proteoglycans are depleted from cartilage (Fig. 64-8) , it loses the ability to rebound from a deforming load and thereby becomes susceptible to mechanical fragmentation and fibrillation and eventually a loss of functional integrity concurrent with its complete dissolution by collagenase and stromelysin. [311] There is also increasing evidence that the metalloproteinases responsible for this process are also derived from the chondrocytes themselves. Both stromelysin and collagenase mRNA levels are increased in RA cartilage, as measured by Northern blot analysis, and in situ hybridization studies confirm the presence of the specific RNA transcripts within chondrocytes. [312] Hence, the cartilage is under attack from a multitude of sources: not only is it being bathed in protease-rich synovial fluid and under extrinsic attack from the invasive pannus, but the chondrocytes themselves contribute to destruction from within.
Although many studies have focused on the pannus-cartilage
Figure 64-8 Human articular cartilage from active RA removed at joint arthroplasty and stained for metachromasia. The only metachromatic stain surrounds a few chrondrocytes that, presumably, are actively making proteoglycan only to have it broken down by proteinases derived from synovial fluid, chondrocytes, or synovial tissue. The form of this depleted cartilage is normal; however, its functional capacity to rebound from a deforming load is seriously impaired.
junction, enzyme release from PMNs in synovial fluid could also have an effect on the loss of cartilage. Consistent with the findings of immune complexes in superficial layers of cartilage, electron microscopic examinations of articular cartilage in RA have revealed amorphous-appearing material and evidence of breakdown of collagen and proteoglycan consistent with superficial diffuse activity of joint fluid enzymes. [313] However, in a rabbit model of arthritis in which IL-1 was injected directly into the joint, the degree of cartilage damage as measured by proteoglycan levels in synovial fluid correlated best with the stromelysin concentrations in synovial effusions (presumably derived from synoviocytes). Neutrophil depletion of animals did not interfere with subsequent destruction of extracellular matrix, suggesting that synovium-derived metalloproteinases were more important.
By and large, most animal studies indicate that IL-1 is a key regulator of matrix degradation in arthritis. [314] [315] This has been true across a broad range of arthritis models, including zymosan-induced arthritis, collagen-induced arthritis, antigen-induced arthritis, and streptococcal cell wall-induced arthritis. Although TNF-alpha blockade has clear anti-inflammatory effects, chondroprotection is less prominent. This does not mean that TNF-alpha cannot mediate joint destruction; the TNF transgenic mice that develop polyarticular arthritis have severe erosions and deformities. TNF inhibitors appear to have chondroprotective effects in human RA.
The rate-limiting step in cartilage loss is the degradation of collagen, because proteoglycans are degraded very soon after inflammation begins. Metalloproteinases, released into the extracellular space and active at neutral pH, are probably responsible for most of the effective proteolyses of articular cartilage proteins, but other classes of enzymes may play a role in joint destruction. Enzymes such as cathepsins B, D, G, K, L, and H may play a role within and outside cells in degrading noncollagenous matrix proteins. Serine proteinases (e.g., elastase and plasmin) and aggrecanase are doubtless involved.
Matrix Metalloproteinases--Key Mediators of Joint Destruction
The metalloproteinases are a family of enzymes that participate in extracellular matrix degradation and remodeling (see Chapter 4) . Although the substrate specificities of individual members of the family differ, the MMPs have several structural and functional similarities. For instance, metalloproteinases are usually secreted as inactive proenzymes. Their proteolytic activity requires limited cleavage or denaturation to reveal a zinc cation at the core that is normally chelated to a cysteine residue in the latent form. Metalloproteinase activation can be mediated by
other proteases, including trypsin, plasmin, or tryptase. They share other structural features, including a catalytic domain and, in some cases, a fibronectin-like region. The substrates for metalloproteinases are varied, but quite specific for individual members of the family. Collagenase degrades native collagen types I, II, III, VII, and X, whereas gelatinase is able to degrade denatured collagen. Stromelysin has broader specificity and can degrade proteoglycans in addition to proteins. It also cleaves procollagenase to the active form, thereby serving as a positive feedback signal for matrix destruction. Many different families of proteinases are found in the joint ( Table 64-7 ), but the metalloproteinases are thought to play a pivotal role in joint destruction.
TABLE 64-7 -- Key Proteases and Inhibitors in Rheumatoid Arthritis Synovium
Protease Inhibitor
Metalloproteinases Collagenase-1 Collagenase-3 Stromelysin-1 92-kD gelatinase Aggrecanase
TIMP familyalpha2 -macroglobulin
Serine proteases Trypsin Chymotrypsin Plasmin Tryptase
SERPINsalpha2 -macroglobulin
Cathepsins Cathepsin B Cathepsin L Cathepsin K
Cystatinsalpha2 -macroglobulin
Abbreviations: SERPIN, Serine protease inhibitor; TIMP, tissue inhibitor of metalloproteinases.
The cytokine milieu has the capacity to induce the biosynthesis of metalloproteinases by synovial cells and alter the balance between extracellular matrix production and degradation. PDGF directly induces proteinase production by mesenchymal cells, as well as being both a chemoattractant and a mitogen for these cells. It is likely that PDGF is released into rheumatoid tissues in large quantities as platelet activation in the presence of angiogenesis and clot formation occurs. IL-1 and TNF-alpha also directly induce metalloproteinase gene expression by many cells, including fibroblast-like synoviocytes. This process is mediated by both a change in collagenase gene transcription and mRNA stabilization. [316] Increased cyclic AMP levels, such as by activation of A2b adenosine receptors, in synoviocytes can suppress collagenase expression by decreasing mRNA half-life. [317] These two cytokines are additive or synergistic when used in combination. In models in vitro, it has been shown that although culture medium from rheumatoid synovium stimulates cartilage degradation, this can be inhibited by an antibody against IL-1; these data
implicate rheumatoid synovium as a source of IL-1 that activates chondrocytes to produce proteases. [318] IL-6 does not induce metalloproteinase production by synovial cells but instead increases the production of TIMP-1, a naturally occurring inhibitor of metalloproteinases. [319]
In contrast to cytokines that induce the production of enzymes that degrade connective tissue, there are several that inhibit the biosynthesis of proteolytic enzymes. One of these, TGF-beta, inhibits collagenase synthesis in vitro and enhances the production of TIMP by fibroblasts and chondrocytes. [320] TGF-beta also increases collagen production, apparently shifting the balance to matrix production and preservation.
Substances other than cytokines are capable of inducing synovial cells to produce metalloproteinases in vitro, and it is probable that many have a role in vivo as well. It is not known for many whether their effects are mediated through cytokines such as PDGF. Proteinases, soluble iron, collagens, crystals of monosodium urate monohydrate, and various calcium crystals are found in joints at one time or another and stimulate collagenase biosynthesis. The crystals, and perhaps other substances in this group, stimulate metalloproteinase production by triggering IL-1 (and possibly other cytokine) production. Some proteases, such as the 92-kD gelatinase, are made constitutively by early-passage RA synoviocytes. [321] These factors operate, it is presumed, by activating receptors that, in turn, enhance the expression of
transacting factors that blind to the cis element in the 5 flanking region of the metalloproteinase genes. One of these, AP-1, is constitutively expressed by fibroblast-like synoviocytes and is activated by IL-1, thereby enhancing the expression of metalloproteinase mRNA. Glucocorticoid-mediated inhibition of collagenase gene expression is due to interference with the Fos-Jun complex by the glucocorticoid receptor. [322]
Collagenase and Stromelysin
Collagenases and stromelysins have, between them, the capacity to degrade all the important structural proteins in the extracellular tissues within joints. Collagenase-1 was first found in culture medium of explants of rheumatoid synovium in 1967. [323] The rheumatoid synovial collagenase (MMP-1) is a metalloproteinase with maximal activity in a range between pH 7 and 8; in pure form, the collagenase has very little activity against substrates other than collagen, including denatured collagen (i.e., gelatin). It cleaves through the triple-helical collagen molecule at a single glycine-isoleucine bond approximately three quaters of the distance from the NH2 terminus of the 300-nm protein. This enzyme has the capability to degrade only the interstitial helical collagens (e.g., types I, II, III, and X collagens); it has little or no activity against types IV, V, and IX and other nonhelical collagens. The process of degradation is slow; kinetics for the human enzyme indicated that only 25 molecules of collagen are cleaved per molecule of enzyme per hour at 37°C.
The collagenase-1 gene is expressed by RA synovial tissue as well as by cartilage. In situ hybridization studies show that the primary location of collagenase gene expression is the intimal lining, especially in fibroblast-like cells. [324] Collagenase protein cannot be readily detected in frozen sections by immunostaining unless synovial explants are first cultured briefly in the presence of monensin, which interferes with protein transport, [325] indicating that collagenase protein is not stored
intracellularly but is secreted very rapidly after synthesis. Increased metalloproteinase gene expression is an early feature of RA and occurs during the first few weeks or months of disease. [326] This underscores the need for early therapy if one wants to prevent joint destruction.
Two additional members of the collagenase family have been identified. Neutrophil collagenase, or MMP-8, is constitutively stored in neutrophil granules and is released into the milieu after degranulation. The relative importance of this enzyme in inflammatory arthritis is uncertain, and neutrophil depletion does not prevent cartilage damage in most animal models. A more important enzyme, collagenase-3 (MMP-13), is highly expressed in chondrocytes in both OA and RA cartilage. In addition, collagenase-3 is also produced by RA synovium in amounts significantly greater than in OA. [327] Collagenase-3 might ultimately be considered the most relevant MMP because its specific activity is much higher than that of collagenase-1. Of note, rodents lack the collagenase-1 gene whereas the collagenase-3 gene is preserved. This is especially important to note when evaluating effects of MMP inhibitors in animal models (rabbits, however, do possess a functional collagenase-1 gene). Like collagenase-1, collagenase-3 has an AP-1-binding site in the promoter that is an important regulator of MMP-13 gene transcription. [328]
Stromelysin (MMP-3) is a metalloproteinase of similar molecular weight to that of collagenase, having the same pH range for activity. It has no activity against interstitial collagens but effectively degrades type IV collagen, fibronectin, laminin, proteoglycan core protein, and type IX collagen. Stromelysin removes the NH2 -terminal propeptides from type I procollagen, producing products similar to those produced by procollagen N-proteinase, the enzyme believed responsible in
vivo for performing this function. Stromelysin is integrally involved in the activation of procollagenase. Like collagenase gene expression, stromelysin gene expression is almost exclusively in the intimal lining (Fig. 64-9) . [329]
Most data indicate that there is a cascade of activation of the MMPs. Prostromelysin is activated, and the presence of stromelysin is essential for the subsequent activation of procollagenase. Prostromelysin from human synovial cells can be activated by other proteases, including cysteine proteases (the cathepsins), trypsin, chymotrypsin, plasma kallikrein, plasmin, and mast cell tryptase. Despite the putative importance of this enzyme in matrix destruction, stromelysin knockout DBA/1 mice are susceptible to collagen-induced arthritis and develop as much joint destruction as the mice with intact stromelysin. [330]
Inhibitors of Metalloproteinase Activity
In serum, alpha2 -macroglobulin (alpha2 M) accounts for more than 95 percent of collagenase inhibitory capacity in serum (see Chapter 4) . The mechanism of inhibition by alpha2 M involves hydrolysis by the proteinase of a susceptible region in one of the four polypeptide chains of alpha2 M (sometimes called the "bait") with subsequent trapping of the proteins within the interstices of the alpha2 M. Ultimately,
the protease is covalently linked to a portion of the alpha2 M molecule. The serine protease inhibitors are also abundant in synovial effusions and plasma and can serve a dual purpose of directly blocking serine protease function and indirectly decreasing metalloproteinase activity by preventing serine proteases from activating metalloproteinase proenzymes. One serine protease inhibitor, alpha1 -antitrypsin, has been well characterized in
Figure 64-9 Localization of stromelysin, tissue inhibitor of metalloproteinases-1, and actin mRNA in RA synovial tissue by in situ hybridization. All three genes are mainly expressed in the synovial intimal lining, presumably by cytokine-stimulated type B synoviocytes. Bright field and dark field views are shown. (Courtesy of D. Boyle.)
synovial fluid and is frequently in an inactivated state after oxidation by reactive oxygen species. [331]
The first inhibitor of mammalian collagenase to be isolated and purified from human tissues was a protein of 25 kD produced by cells in explants of human tendons. This inhibitor, TIMP, blocked trypsin-activated rheumatoid synovial collagenase as well as the collagenase obtained from PMNs. [332] Additional members of the TIMP family have been cloned and characterized; each has distinctive patterns of affinity for each metalloproteinase. [333] The TIMP proteins block proteinase activity by binding directly to metalloproteinases in a 1:1 molar ratio. TIMP generally binds only to the active enzyme; there are some exceptions, such as TIMP-2, which can interact with type IV procollagenase. The inhibitors bind to metalloproteinases with extremely high avidity, and although the interaction does not result in new covalent bonds, it is essentially irreversible. TIMP protein is present in RA synovial fluid in excess. It is, in fact, very difficult to detect free active collagenase or stromelysin, and they are usually complexed with TIMP. [334] The majority of metalloproteinase is not complexed and is in the proenzyme form. Immunohistochemical studies have localized the inhibitor in hyperplastic synovial lining cells in rheumatoid synovium, but not in the cells of normal synovium. TIMP gene expression is not significantly altered by IL-1 or TNF-alpha but is increased by IL-6.
TIMP mRNA, that of like metalloproteinases, is primarily localized to the synovial intimal lining. When double-label in situ hybridization was used, three phenotypes of cells could be detected in the intimal lining and among cultured synoviocytes: TIMP negative-stromelysin positive, TIMP positive-stromelysin negative, and TIMP positive-stromelysin positive. [335] This underscores the differential regulation of TIMP and the metalloproteinases and shows that some cells simultaneously seek matrix destruction and protection.
Given the important role of metalloproteinases in tissue destruction, the relative balance between metalloproteinases and TIMPs ultimately determines the fate of the extracellular matrix. Presumably, RA, with its more destructive potential, would have a ratio that favors degradation whereas OA might have a lower MMP:TIMP ratio.
This has been examined directly using quantitative in situ hybridization to study synovial mRNA levels, and it appears that the metalloproteinase-to-TIMP ratio is, indeed, higher in RA. The levels of TIMP gene expression are very similar in the two disease and are likely maximal. The cause of the higher ratio in RA is, therefore, the increased amount of metalloproteinase mRNA. The differences between OA and RA are not dramatic, and it may be that rather subtle changes in the balance can have profound effects over years (or decades) of disease. Perhaps more important is the question of whether drug treatment can alter this balance. Intra-articular corticosteroid injections markedly decreased synovial collagenase, stromelysin, and TIMP gene expression. In contrast, chronic low-dose methotrexate therapy specifically decreased collagenase mRNA (by about two thirds) but not stromelysin or TIMP-1 mRNA. [336] The specificity for collagenase, but not stromelysin
gene expression, is a bit puzzling, but other studies have shown discoordinate expression of these two genes. [337] The explanation might lie in the hypothesis that methotrexate works, in part, through increased endogenous production of adenosine. [338] The selective decrease in collagenase gene expression suggests that a decreased collagenase-to-TIMP ratio is the mechanism of decreased bone destruction observed in a some patients treated with methotrexate. In another study, tenidap, but not a conventional nonsteroidal anti-inflammatary drug, significantly decreased stromelysin gene expression. [339]
Cysteine Proteases--The Cathepsins
Cathepsins are an extensive family of cysteine proteases that have broad proteolytic activity, including activity on types II, IX, and XI collagen and proteoglycans. [340] Like MMPs, the cathepsins can be regulated by cytokines and by proto-oncogenes like ras. IL-1 and TNF-alpha induce cathepsin L expression in cultured fibroblast-like synoviocytes. [341] In situ hybridization studies demonstrate expression of cathepsin B and L in RA synovium, especially at sites of erosion. A novel cysteine protease called cathepsin K has been implicated in bone resorption by osteoclasts. It is unique among the cathepsins in its ability to degrade native type I collagen. [342] Cathepsin K is expressed in RA synovial tissue by both macrophages and fibroblasts. [343] A potential role of cathepsins as mediators of bone destruction in arthritis was confirmed in studies in which a cysteine protease inhibitor significantly decreased joint damage in the rat adjuvant arthritis model. [344]
B CELL ACTIVATION AND RHEUMATOID FACTOR
Activated B lymphocytes are present in peripheral blood as well as in the rheumatoid synovium. The process of B cell activation and the mediators controlling it are detailed elsewhere (Chapter 9) , but it is useful to review this in the context of RA. As with other activation systems in this disease (as well as in normal physiology), cytokines play a major role in antibody production and isotype switching. The B cell subset that is enriched in autoantibody production is characterized by a surface determinant CD5. [345] In humans, IL-2 plays a leading role in inducing all immunoglobulin isotypes; although other cells can enhance this response, none affect lymphocyte responses of activation, proliferation, or differentiation in the absence of IL-2. IL-4 has growth-promoting effects on human B cells and induces class II MHC expression on B cells. IL-4 directly enhances both T cell proliferation and IL-2 production and is responsible for isotype switching to IgE antibodies.
B cell maturation is a complex process and might involve the nurselike cells that are analogous to nurse cells in the thymus that support thymocytes. Such cells have been isolated from RA synovium and have the capacity to rescue B cells from spontaneous apoptosis and facilitate immunoglobulin production. [346] Some evidence suggests that subsequent activation is, in fact, antigen driven. Many RA patients with normal circulating numbers of lymphocytes show an abnormal kappa-to-lambda chain analysis compared with controls, implying oligoclonal B cell proliferation. [347] It is not known whether this reflects expansion of the restricted number of clones capable of producing RF or whether an inciting antigen is something other than IgG and related specifically to RA. However, shared mutations containing an identical sequence throughout the variable domain of immunoglobulins have been identified in RA synovial tissue. [348] Preferential utilization of a limited number of VH and DH gene segments and marked preference for a DH reading frame encoding predominantly hydrophilic residues have also been observed. Analysis of expressed heavy chain variable domains supports the notion that the B cell response in RA synovium is oligoclonal.
Despite the evidence for antigen-mediated responses as a regulator of RFs, the driving force behind RF production has not been fully elucidated (see Chapter 10) . Enhanced helper T cell function has been correlated with the spontaneous production of RF, although only for the IgM isotype. [349] The NK cells and the cytokine profile of the joint (especially IL-6) can also support nonspecific B cell activation. In addition, terminally differentiated plasma cells that spontaneously secrete RF are present in RA synovial fluid, and RF production markedly increased when the cells were cultured in the presence of fibroblast-like synoviocytes and IL-10. [350] These CD20- CD38+ cells were also demonstrated in other forms of arthritis, but they did not release significant amounts of RF.
Although no data clearly implicate RF as a principal causative agent in RA, its role in
the amplification and perpetuation of the process is well supported:
1. Although some patients with virtually no circulating IgG develop RA, patients with a positive test result for RF in blood have more severe clinical disease and complications than do seronegative patients.
2. Polyclonal IgM RF is able to fix and activate complement by the classic pathway.
3. IgG RF produced in large quantity in rheumatoid synovial tissue can form large complexes of itself through self-association, because these molecules have a much higher frequency of double-valent Fc-binding regions than do most normal IgG molecules. It appears that these large complexes fix complement and can bind to IgM RF.
4. Immune complexes containing RF have been localized within synovial tissues by immunofluorescent techniques.
5. Increased levels of IgG RF have been associated with a high frequency of subcutaneous nodules, vasculitis, an elevated erythrocyte sedimentation rate, decreased complement levels, and an increased number of involved joints.
6. In experiments performed in patients with RA, a marked inflammatory response was elicited when RF from the patient was injected into a joint, but not when normal IgG was given. [351] RF becomes involved in pathogenesis when it forms immune complexes sufficiently large to activate complement or be phagocytized by macrophages or PMNs, or both.
Several hypotheses have been advanced to explain how IgG could become immunogenic. First, new determinants on IgG might be exposed after polymerization among molecules to form aggregates or as IgG complexes with specific non-IgG antigens. [352] Second, structural anomalies in the IgG of rheumatoid patients may render it immunogenic, such as a possible defect in the hinge region of rheumatoid IgG that could increase the binding affinity to membrane Fc receptors on B lymphocytes. Alternatively, depletion of suppressor T lymphocytes might allow B lymphocytes to produce autoantibodies against certain determinants on IgG. Finally, autoantigenic reactivity of IgG could be related to demonstrated changes in the relative extent of galactosylation. A deficiency of the galactosylation enzyme machinery may increase the relative risk of developing RA.
In addition to RFs of the mu and gamma isotypes in RA, epsilon isotypes also have been demonstrated in certain patients. Many patients with sera containing IgE immune complexes have extra-articular manifestations. IgE RF might complex with aggregated (self-associating) IgG in synovial tissue, and the IgG-IgE complexes could then activate mast cells and basophils in the synovium. This is of particular interest in light of reports that there are numerous mast cells in rheumatoid synovium and that these may release factors capable of stimulating collagenase production by synovial cells.
Considering the restricted number of idiotypes of RF, it is interesting that the four major classes of immunoglobulins (IgM, IgG, IgA, and IgE) are all produced in RA. Whereas only about three quarters of patients with RA are seropositive by standard tests for RF, over 90 percent are positive for IgM RF using enzyme-linked
immunosorbent assays. [353] This study is significant because it lowers the numbers of "seronegative" RA patients in this and presumably other populations. In the same group, two thirds were positive for IgA, IgE, and IgG RF. Disease activity correlated with IgM RF and IgA RF, as did levels of circulating immune complexes. Extra-articular features correlated positively with levels of IgA and IgE RF.
SYNOVIAL FLUID
Although it would be of great interest to examine synovial tissue from each patient with RA to compare histologic changes with those from previous specimens and perhaps assay for T cell subsets, RF production, and cytokine production by synovial cells, this is impractical (although miniarthroscopy might someday permit one to "stage" RA in this manner to determine the prognosis and appropriate therapy). As stressed earlier, blood is far removed from the site of disease and the focus of inflammatory activity. Despite some limitations, synovial fluid is a reasonable compromise; by examination of its characteristics one can gain a good appreciation for the extent of inflammation, and by using synovial fluid investigators can learn much about events within the synovium itself. Chapter 42 describes techniques for the analysis of synovial fluid. Components of the inflammatory and proliferative response that can be dissected by examination of synovial fluid from patients are discussed in the following sections.
Polymorphonuclear Leukocytes
The number of PMNs remains one of the most accurate indices of inflammation within a particular joint. These are cells that truly amplify inflammatory responses, and although it is unlikely that they destroy much cartilage directly, their role in amplification, and, thereby, perpetuation of the inflammation within joints, is probably significant. The joint space serves as a depository for PMNs; they enter the synovial fluid by direct passage from postcapillary venules in the synovium. Neutrophils adhere to activate synovial microvasculature because of the action of selectins and the beta2 integrins. After adherence, however, agents such as IL-8 produced by endothelium and fibroblasts may facilitate egress through the capillaries into the chemoattractant gradients of the synovium. Once they arrive in the joint space, they have no way to leave. Thus, considering the survival time of PMNs in synovial fluid, it has been estimated that the breakdown with an average (30 ml) rheumatoid effusion containing 25,000 PMNs per mm3 may well exceed 1 billion cells each day (Fig. 64-10) . The ultimate fate of many of these cells is apoptosis.
The physiology of granulocytes is discussed in detail in Chapter 14 . As described there, the strong attraction of chemotactic agents within the synovial fluid in RA is responsible for the large number of cells found there (occasionally up to 100,000/mm3 ). Few PMNs are seen in the pannus itself and subsynovial tissue; once in the synovium they move rapidly to the synovial fluid, drawn by the activated component of cleavage of the fifth component of complement (C5a), leukotriene B4 (LTB4 ), platelet-activating factor, and chemokines. The reason that neutrophils can move into the articular cavity without resistance whereas mononuclear cells collect in the sublining is not clear. It might be related to the fact that neutrophils express very little VLA-4 compared with mononuclear cells and therefore are unimpeded by VCAM-1 and CS1 fibronectin in the intimal lining. In the synovial fluid, PMNs come in
contact with immune complexes and particulate material (i.e., fibrin, cell membranes, cartilage fragments). Phagocytosis occurs, particularly to particles coated with IgG, and the PMNs are activated. Through a complex set of changes in the membrane potential of the cell involving calcium flux and both phospholipid metabolism and cyclic nucleotide activation, the neutrophil begins to degranulate, to generate products of oxygen metabolism, to metabolize arachidonic acid, and to develop the capacity for aggregation. In addition, PMNs from synovial fluid in RA release de novo synthesized proteins, including fibronectin, neutral proteinases, and IL-1. Neutrophils also secrete IL-1Ra as a major product. [354] Although the
Figure 64-10 Normal and RA peripheral blood polymorphonuclear neutrophils (PMNs). A, Scanning electron micrograph (SEM) of a PMN from peripheral blood of a normal 73-year-old man (×30,000). It is characteristically apolar and spherical with surfaces completely covered by plasma membrane elaborated into irregular ridges or small ruffles. B, Transmission electron micrograph (TEM) of PMN from a normal man. The apolar cell has few phagocytic vacuoles but many undischarged electron-dense granules (×19,800). C, SEM of a PMN from peripheral blood of a 78-year-old woman with RA. This striking polarized appearance is much more common in rheumatoid patients than in healthy individuals and suggests that the cells have been activated (×19,200). D, TEM of a PMN from peripheral blood of a 59-year-old woman with RA. This cell has many phagocytic vacuoles but relatively few undischarged electron-dense granules (×17,500). ( A- D from McCarthy DA, Holburn CM, Pell BK, et al: Scanning electron microscopy of rheumatoid arthritis peripheral blood polymorphonuclear leucocytes. Ann Rheum Dis 45:899, 1986.)
amount of IL-1Ra that each neutrophil produces is low compared with that produced by macrophages, the sheer number of PMNs allows them to produce massive amounts in synovial effusions.
Immune Complexes
The significance of immunoglobulin complexes circulating in blood and in synovial fluid was appreciated several decades ago. [355] However, it was not until more reliable assays for immune complexes were available that broad studies correlating disease activity and immune complexes could be generated (see Chapter 12) .
Studies of immune complexes in synovial fluid have generated data more relevant to the pathophysiology of the disease than in the blood of the same patients, because the disease process is initiated and perpetuated in the synovium. Findings in blood may reflect only what "spills out" from the synovial fluid and synovial tissue. Levels of IgM-containing circulating immune complexes are elevated in both RA and systemic lupus
erythematosus, although levels of IgG immune complexes are not. [356] In all assays for circulating immune complexes, the possibility of in vitro formation of IgM and IgG complexes giving false-positive tests must be considered. Assays such as the C1q-binding assay overestimate the concentration of immune complexes. False-positive results are also found in the Raji cell test, which is one reason why it is now rarely used in clinical situations. In studies designed to identify the components of immune complexes in the circulation of rheumatoid patients, most data have found no specific antigen other than IgG complexed with RF. Using more sensitive techniques, it has been found that circulating immune complexes in RA are composed of as many as 20 polypeptides, including albumin, immunoglobulin, complement, and acute-phase reactants.
Most relevant to the pathogenesis of joint destruction in RA has been the identification of immunoglobulins and complement in articular collagenous tissues from RA patients. Over 90 percent of cartilage and meniscal samples from rheumatoid patients have evidence of these components in the avascular connective tissue. Electron microscopic morphology of immunoglobulin aggregates show that there are pathologic changes in the matrix of cartilage in the microenvironment of the aggregates themselves. [357] Immune complexes were absent under areas in cartilage invaded actively by synovial pannus, suggesting that phagocytic cells in the invasive synovium had perhaps ingested the immune complexes, [358] lending credence to the possibility that immune complexes deposited in the avascular superficial layers of cartilage in the joint may serve as chemoattractants for the pannus and be an explanation for the centripetal orientation of the rheumatoid lesion. Immune complexes have been extracted from cartilage of RA and OA patients. Rheumatoid cartilage contained 37 times more IgM and 14 times more IgG than did healthy cartilage extracts. IgM RF was found in 13 of 16 RA cartilage extracts and in none of 11 OA or 6 healthy control extracts. In addition, more than 60 percent of the RA cartilage extracts were positive for native and denatured collagen type II antibody, as were OA specimens.
These observations help support the hypothesis that the presence of cartilage itself, and perhaps these complexes, contributes to the chronicity and persistence of rheumatoid inflammation. Orthopedic surgeons have noted many times that joints from which all cartilage is removed do not participate in general flares of rheumatoid disease after surgery. "Burnt out" RA may mean that with the combination of loss of motion and loss of cartilage in a joint, there is nothing to sustain continued inflammation.
Arachidonate Metabolites
Accompanying activation of PMNs is the increased mobilization of membrane phospholipids in these cells to arachidonic acid and its subsequent oxidation by cyclooxygenases (COX) to prostaglandins and thromboxanes, or by lipoxygenases to leukotrienes. Although the stable prostaglandins, especially PGE2 , do produce vasodilation, cause increased vascular permeability, and are involved centrally in fever production, there is increasing evidence that they have significant antiinflammatory activities as well. For example, stable prostaglandin can retard the
development of adjuvant arthritis, [359] and the drug misoprostol, a prostaglandin analogue, may have significant anti-inflammatory or immunomodulatory effects. Physiologic concentrations of PGE2 inhibit IFN-gamma production by T cells, HLA-DR expression by macrophages, and T cell proliferation.
Production of prostaglandins in the rheumatoid synovium is dependent on two distinct cyclooxygenase enzymes, known as COX-1 and COX-2 (see Chapter 15) . The former is constitutively expressed and is responsible for the normal endogenous production of prostaglandins in the joint as well as in other tissues. COX-1 inhibition in the gastrointestinal tract leads to a precipitous drop in prostanoid synthesis and increases susceptibility to ulceration. The role of COX-1 in the joint homeostasis is not certain. COX-2, on the other hand, is an inducible enzyme that appears to be responsible for increased prostaglandin synthesis in inflamed tissue. Cytokines like IL-1 and TNF-alpha induce COX-2 gene expression in cultured synoviocytes and macrophages. COX-2 mRNA and immunoreactive protein are increased in RA synovium. [360] Most nonsteroidal anti-inflammatory drugs, including indomethacin and ibuprofen, inhibit both COX-1 and COX-2. However, it appears that most of the anti-inflammatory activity (and analgesia) results from inhibition of the latter. For instance, selective COX-2 inhibitors like celecoxib are as effective as nonselective inhibitors in animal models of arthritis. [361] Clinical studies using similar compounds in patients with OA or RA also indicate that COX-2 blockade is sufficient for clinical benefit.
LTB4 has also received considerable attention as a proinflammatory product of neutrophil activation. It is chemotactic for neutrophils, eosinophils, and macrophages; it promotes neutrophil aggregation; it enhances neutrophil adherence to endothelium; and it enhances NK cell cytotoxic activity. Metabolism of arachidonic acid into the pathway mediated by lipoxygenase is directly proportional to the breakdown of these cells in the joint cavity; cell breakdown facilitates the access of arachidonic acid to lipoxygenase from the cytosol. It is of interest that peripheral blood PMNs from rheumatoid patients have an enhanced capacity for the production of LTB4 compared with similar cells from control groups. [362] No mechanism for this has been elucidated, and it is not known whether synovial fluid leukocytes have the same enhanced capacity for LTB4 release. In murine collagen-induced arthritis, a specific LTB4 antagonist significantly decreased paw swelling and joint destruction, suggesting a pivotal role for this potent chemoattractant. [363]
Complement
The components and pathways involved in complement activation are described in Chapter 13 . As in studies of
lymphocyte function and in measures of inflammation, synovial fluid serves as a better index of complement metabolism in RA than does peripheral blood. The liver is the major source of complement synthesis in humans, and passive transfer of serum proteins into effusions can account for some of the complement proteins found there. The synovial tissue also actively produces complement proteins. [364]
Macrophages and fibroblasts produce complement proteins under the influence of cytokines. IFN-gamma induces C2, whereas IL-1 and TNF-alpha increase C3 production. [365] In situ hybridization shows that C2 is expressed in the synovial intimal lining, whereas C3 appears to be produced by synovial sublining macrophages. Northern blot analysis of synovial tissue shows that all complement genes from the classic pathway are expressed in RA synovium as well as in healthy synovium. [366] Despite the local production of complement components, the activities of C4, C2, and C3 and total hemolytic complement in rheumatoid (seropositive) synovial effusions are lower than in synovial fluids from patients with other joint diseases. [367] A low synovial fluid C3 level might be modestly predictive of more erosive disease.
Using a sensitive solid-phase radioimmunoassay to quantify the activation of the classic pathway of complement by RF, IgM RF appears to be a much more important determinant of complement activation than IgG RF in both sera and synovial fluids. [368] Combined with other data showing that there is accelerated catabolism of C4 in RA, and that the presence of C4 fragments in the plasma of rheumatoid patients correlates with titers of IgM RF, the weight of evidence indicates a role in vivo for IgM RF in complement activation. [369]
The biologically active products of complement activation are probably the most important consequence of intra-articular complement consumption. Like proteinases from PMNs, these inflammatory components may build up in synovial fluid during acute inflammation. The potential for interaction between PMNs and the complement system is substantial. Neutrophil lysosomal lysates contain enzymatic activity capable of generating chemotactic activity (probably C5a) from fresh serum. [370] C5a, in addition to being a principal chemotactic factor in inflammatory effusions, is capable of mediating lysosomal release from human PMNs. This sets up one of many amplification loops in inflammatory synovial fluid. The role of complement has been confirmed in an antigen-induced arthritis model in rats, intra-articular treatment with a soluble complement receptor (sCR1) inhibited joint swelling. [371]
SUMMARY
Understanding the etiology and pathogenesis of RA remains a complex problem that continues to defy solution despite decades of effort. We have certainly come long way from the simple notion that immune complexes and complement fixation account for rheumatoid synovitis. There is currently an appreciation that both T cell-dependent and -independent processes might contribute to disease initiation and perpetuation. Moreover, it might be important to appreciate differences in disease pathogenesis at various stages of the process. These hypotheses have unveiled many novel therapeutic targets and interventions that might lead to significant clinical benefit. Such was the case with the TNF-inhibitory proteins that have joined the pharmacopoeia for the treatment of RA; initial observations that defined the cytokine profile in arthritis and that delineated the biology of macrophage cytokines led to this breakthrough. Similarly, it is possible that understanding of apoptotic pathways, abnormalities in tumor suppressor genes, the function of the susceptibility cassette, or the mechanisms of Th1-Th2 balance might have unexpected rewards. The most important path to follow, however, is one marked by an open mind to new ideas that might ultimately lead to better treatments or (perhaps) cures.
REFERENCES
1. Stastny P: Association of the B-cell alloantigen DRw4 with rheumatoid arthritis. N Engl J Med 298:869, 1978.
2. Gregersen PK, Shen M, Song QL, et al: Molecular diversity of HLA-DR4 haplotypes. Proc Natl Acad Sci U S A 83:2642, 1986.
3. Nepom GT, Byers P, Seyfried C, et al: HLA genes associated with rheumatoid arthritis: Identification of susceptibility alleles using specific oligonucleotide probes. Arthritis Rheum 32:15, 1989.
4. Weyand CM, Hicok KC, Conn DL, Goronzy JJ: The influence of HLA-DRB1 genes on disease severity in rheumatoid arthritis. Ann Intern Med 117:801, 1992.
5. Boki KA, Drosis AA, Tzioufas GA, et al: Examination of HLA-DR4 as a severity marker for rheumatoid arthritis in Greek patients. Ann Rheum Dis 52:517, 1993.
6. Hameed K, Bowman S, Kondeatis E, et al: The association of HLA-DRB genes and the shared epitope with rheumatoid arthritis in Pakistan. Br J Rheumatol 36:1184, 1997.
7. Wakitani S, Imoto K, Murata N, et al: The homozygote of HLA-DRB1*0901, not its heterozygote, is associated with rheumatoid arthritis in Japanese. Scand J Rheumatol 27:381, 1998.
8. Calin A, Elswood J, Klouda PT: Destructive arthritis, rheumatoid factor, and HLA-DR4: Susceptibility versus severity, a case control study. Arthritis Rheum 32:1221, 1989.
9. Thomson W, Pepper L, Payton A, et al: Absence of an association between HLA-DRB1*04 and rheumatoid arthritis in newly diagnosed cases from the community. Ann Rheum Dis 52:539, 1993.
10. Brown JH, Jardetzk TS, Gorga JC, et al: Three dimensional structure of the human class II histocompatability antigen HLA-DR1. Nature 364:33, 1993.
11. Kirschmann DA, Duffin KL, Smith CE, et al: Naturally processed peptides from rheumatoid arthritis and non-associated HLA-DR alleles. J Immunol 155:5655, 1995.
12. Walser-Kuntz DR, Weyand CM, Weaver AJ, et al: Mechanisms underlying the formation of the T cell receptor repertoire in rheumatoid arthritis. Immunity 2:597, 1995.
13. Auger I, Escola JM, Gorvel JP, Roudier J: HLA-DR4 and HLA-DR10 motifs that carry susceptibility to rheumatoid arthritis bind 70 kD heat shock proteins. Nature 2:306, 1996.
14. McCusker CT, Reid B, Green D, et al: HLA-D region antigens in patients with rheumatoid arthritis. Arthritis Rheum 34:192, 1991.
15. Zanelli E, Gonzalez-Gay MA, David CS: Could HLA-DRB1 be the protective locus in rheumatoid arthritis? Immunol Today 16:274, 1995.
16. Doherty DG, Vaughan RW, Donalson PT, Mowat AP: HLA DQA, DQB, and DRB genotyping by oligonucleotide analysis: Distribution of alleles and haplotypes in British caucasoids. Hum Immunol 34:53, 1992.
17. Perdriger A, Chales G, Semana G, et al: Role of HLA-DR-DR and DR-DQ associations in the expression of extraarticular manifestations and rheumatoid factor in rheumatoid arthritis. J Rheumatol 24:1272, 1997.
18. Zanelli E, Krco CJ, Baisch JM, et al: Immune response of HLA-DQ8 transgenic mice to peptides from the third hypervariable region of HLA-DRB1 correlates with predisposition to rheumatoid arthritis. Proc Natl Acad Sci U S A 93:1814, 1996.
19. Moxley G: DNA polymorphism of immunoglobulin kappa confers risk of rheumatoid arthritis. Arthritis Rheum 32:634, 1989.
20. Perdriger A, Chales G: Influence of non HLA factors in rheumatoid arthritis: Role of enzyme abnormalities in joint destruction. Rev Rhum Engl Ed 64:523, 1997.
21. Tsuchiya N, Endo T, Shiota M, et al: Distribution of glycosylation abnormality among serum IgG subclasses from patients with rheumatoid arthritis. Clin Immunol Immunopathol 70:47, 1994.
22. Young A, Sumar N, Bodman K, et al: Agalactosyl IgG: An aid to differential diagnosis in early synovitis. Arthritis Rheum 34:1425, 1991.
23. Rademacher TW, Williams P, Dwek RA: Agalactosyl glycoforms of IgG autoantibodies are pathogenic. Proc Natl Acad Sci U S A 91:6123, 1994.
24. Jacob CO, McDevitt HO: Tumour necrosis factor-alpha in murine autoimmune "lupus" nephritis. Nature 331:356, 1988.
25. Jacob CO, Fronek Z, Lewis GD, et al: Heritable major histocompatibility complex class II-associated differences in production of tumor necrosis factor alpha: Relevance to genetic predisposition to systemic lupus erythematosus. Proc Natl Acad Sci U S A 87:1233, 1990.
26. Vinasco J, Beraun Y, Nieto A, et al: Polymorphism at the TNF loci in rheumatoid arthritis. Tissue Antigens 49:74, 1997.
27. Hajeer A, John S, Ollier WE, et al: Tumor necrosis factor microsatellite haplotypes are different in male and female patients with RA. Rheumatology 24:217, 1997.
28. McDowell TL, Symons JA, Ploski R, et al: A genetic association between juvenile rheumatoid arthritis and a novel interleukin-1alpha polymorphism. Arthritis Rheum 38:221, 1995.
29. Tarlow JK, Blakemore AI, Lennard A, et al: Polymorphism in human IL-1 receptor antagonist gene intron 2 is caused by variable numbers of an 86-bp tandem repeat. Hum Genet 91:403, 1993.
30. Persellin RH: The effect of pregnancy on rheumatoid arthritis. Bull Rheum Dis 27:922, 1977.
31. Quinn C, Mulpeter K, Casey EB, Feighery CF: Changes in levels of IgM RF and alpha 2 PAG
correlate with increased disease activity in rheumatoid arthritis during the puerperium. Scand J Rheumatol 22:273, 1993.
32. Lin H, Mosmann TR, Guilbert L, et al: Synthesis of T helper 2-type cytokines at the maternal-fetal interface. J Immunol 151:4562, 1993.
33. Buzas EI, Hollo K, Rubliczky L, et al: Effect of pregnancy on proteoglycan-induced progressive polyarthritis in BALB/c mice: Remission of disease activity. Clin Exp Immunol 94:252, 1993.
34. Nelson JL, Hughes KA, Smith AG, et al: Maternal-fetal disparity in HLA class II alloantigens and the pregnancy-induced amelioration of rheumatoid arthritis. N Engl J Med 329:466, 1993.
35. Van der Heijden IM, Wilbrink B, Tchetverikov I, et al: Presence of bacterial DNA and bacterial peptidoglycans in the joints of patients with rheumatoid arthritis and other arthritides. Arthritis Rheum 41:S162, 1998.
36. Wilbrink B, van der Heijden IM, Schouls LM, et al: Detection of bacterial DNA in joint samples from patients with undifferentiated arthritis and reactive arthritis, using polymerase chain reaction with universal 16S ribosomal RNA primers. Arthritis Rheum 41:535, 1998.
37. Wilkinson NZ, Kingsley GH, Sieper J, et al: Lack of correlation between the detection of Chlamydia trachomatis DNA in synovial fluid from patients with a range of rheumatic diseases and the presence of an antichlamydial immune response. Arthritis Rheum 41:845, 1998.
38. Cole BC, Griffiths MM: Triggering and exacerbation of autoimmune arthritis by the Mycoplasma arthritidis superantigen MAM. Arthritis Rheum 36:994, 1993.
39. Hoffman RW, O'Sullivan FX, Schafermeyer KR, et al: Mycoplasma infection and rheumatoid arthritis: Analysis of their relationship using immunoblotting and an ultrasensitive polymerase chain reaction detection method. Arthritis Rheum 40:1219, 1997.
40. Barthold SW, Sidman CL, Smith AL: Lyme borreliosis in genetically resistant and susceptible mice with severe combined immunodeficiency Am J Trop Med Hyg 47:605, 1992.
41. Depper JM, Zvaifler NJ: Epstein-Barr virus: Its relationship to the pathogenesis of rheumatoid arthritis. Arthritis Rheum 24:755, 1981.
42. Slaughter L, Carson DA, Jensen FC, et al: In vitro effects of Epstein-Barr virus on peripheral blood mononuclear cells from patients with rheumatoid arthritis and normal subjects. J Exp Med 148:1429, 1978.
43. Roudier J, Petersen J, Rhodes GH, et al: Susceptibility to rheumatoid arthritis maps to a T-cell epitope shared by the HLA-Dw4 DR beta-1 chain and the Epstein-Barr virus glycoprotein gp 110. Proc Natl Acad Sci U S A 86:5104, 1989.
44. Albani S, Ravelli A, Mass M, et al: Immune responses to the Escherichia coli dnaJ heat shock protein in juvenile rheumatoid arthritis and their correlation with disease activity. J Pediatr 124:561, 1994.
45. Cohen BJ, Buckley MM, Clewley JP, et al: Human parvovirus infection in early rheumatoid and inflammatory arthritis. Ann Rheum Dis 45:832, 1986.
46. Saal JG, Steidle M, Einsele H, et al: Persistence of B19 parvovirus in synovial membranes of patients with rheumatoid arthritis. Rheumatol Int 12:147, 1992.
47. Takahashi Y, Murai C, Shibata S, et al: Human parvovirus B19 as a causative agent for rheumatoid arthritis. Proc Natl Acad Sci U S A 95:8227, 1998.
48. Nikkari S, Luukkainen R, Mottonen T, et al: Does parvovirus B19 have a role in rheumatoid arthritis? Ann Rheum Dis 53:106, 1994.
49. Harrison B, Silman A, Barrett E, Symmons D: Low frequency of recent parvovirus infection in a population-based cohort of patients with early inflammatory polyarthritis. Ann Rheum Dis 57:375, 1998.
50. Narayan O, Sheffer D, Clements JE, Tennekoon G: Restricted replication of lentiviruses. Visna viruses induce a unique interferon during interaction between lymphocytes and infected macrophages. J Exp Med 162:1954, 1985.
51. Lechner F, Vogt HR, Seow HF, et al: Expression of cytokine mRNA in lentivirus-induced arthritis. Am J Pathol 151:1053, 1997.
52. Di Giovine FS, Bailly S, Bootman J, et al: Absence of lentiviral and human T cell leukemia viral sequences in patients with rheumatoid arthritis. Arthritis Rheum 37:349, 1994.
53. Nakagawa K, Brusic V, McColl G, Harrison LC: Direct evidence for the expression of multiple endogenous retroviruses in the synovial compartment in rheumatoid arthritis. Arthritis Rheum 40:627, 1997.
54. Aicher WK, Heer AH, Trabandt A, et al: Overexpression of zinc-finger transcription factor Z-225/Egr-1 in synoviocytes from rheumatoid arthritis patients. J Immunol 152:5940, 1994.
55. Yamamoto H, Sekiguchi T, Yamamoto I: Histopathological observation of joint lesions of extremities in mice transferred genome. Exp Toxicol Pathol 45:233, 1993.
56. Nakajima T, Aono H, Hasunuma T, et al: Overgrowth of human synovial cells driven by the human T cell leukemia virus type I tax gene. J Clin Invest 92:186, 1993.
57. Grahame R, Armstrong R, Simmons N, et al: Chronic arthritis associated with the presence of intrasynovial rubella virus. Ann Rheum Dis 42:2, 1983.
58. Stuart JM, Cremer MA, Townes AS, Kang AH: Type II collagen-induced arthritis in rats: Transfer with serum. J Exp Med 155:1, 1982.
59. Trentham DE, Dynesius RA, David JR: Passive transfer by cells of type II collagen-induced arthritis in rats. J Clin Invest 62:359, 1978.
60. Terato K, Hasty KA, Cremer MA, et al: Collagen-induced arthritis in mice: Localization of an arthritogenic determinant to a fragment of the type II collagen molecule. J Exp Med 162:637, 1985.
61. Wooley PH, Luthra HS, Stuart JM, David CS: Type II collagen-induced arthritis in mice. I. Major histocompatibility complex (I region) linkage and antibody correlates. J Exp Med 154:688, 1981.
62. Rowley M, Tai B, Mackay IR, Cunningham T, Phillips B: Collagen antibodies in rheumatoid arthritis. Significance of antibodies to denatured collagen and their association with HLA-DR4. Arthritis Rheum 29:174, 1986.
63. Watson WC, Cremer MA, Wooley PH, Townes AS: Assessment of the potential pathogenicity of
type II collagen autoantibodies in patients with rheumatoid arthritis. Arthritis Rheum 29:1316, 1986.
64. Jasin HE: Autoantibody specificities of immune complexes sequestered in articular cartilage of patients with rheumatoid arthritis and osteoarthritis. Arthritis Rheum 28:241, 1985.
65. Tarkowski A, Klareskog L, Carlsten H, et al: Secretion of antibodies to types I and II collagen by synovial tissue cells in patients with rheumatoid arthritis. Arthritis Rheum 32:1087, 1989.
66. Van Eden W, Holoshitz J, Nevo Z, et al: Arthritis induced by a T-lymphocyte clone that responds to Mycobacterium tuberculosis and to cartilage proteoglycans. Proc Natl Acad Sci U S A 82:5117, 1985.
67. Oda A, Miyata M, Kodama E, et al: Antibodies to 65Kd heat-shock protein were elevated in rheumatoid arthritis. Clin Rheumatol 13:261, 1994.
68. Gaston JSH, Life PF, Bailey LC, Bacon PA: In vitro responses to a 65-kilodalton mycobacterial protein by synovial T cells from inflammatory arthritis patients. J Immunol 143:2494, 1989.
69. Wilbrink B, Holewijn M, Bijlsma JW, et al: Suppression of human cartilage proteoglycan synthesis by rheumatoid synovial fluid mononuclear cells activated with mycobacterial 60-kd heat-shock protein. Arthritis Rheum 36:514, 1993.
70. Sharif M, Worrall JG, Singh B, et al: The development of monoclonal antibodies to the human mitochondrial 60-kd heat-shock protein, and their use in studying the expression of the protein in rheumatoid arthritis. Arthritis Rheum 35:1427, 1992.
71. Verheijden GF, Rijnders AW, Bos E, et al: Human cartilage glycoprotein-39 as a candidate autoantigen in rheumatoid arthritis. Arthritis Rheum 40:1115, 1997.
72. Good RA, Rotstein J, Mazzitello WF: The simultaneous occurrence of rheumatoid arthritis and agammaglobulinemia. J Lab Clin Med 49:343, 1957.
73. Svensson L, Jirholt J, Holmdahl R, Jansson L: B cell-deficient mice do not develop type II collagen-induced arthritis. Clin Exp Immunol 111:521, 1998.
74. Kunkel HG, Agnello V, Jaslin FG, et al: Cross-idiotypic specificity among monoclonal IgM proteins with anti-IgG activity. J Exp Med 137:331, 1973.
75. Bonagura VR, Wedgwood JF, Agostino N, et al: Seronegative rheumatoid arthritis, rheumatoid factor cross reactive idiotype expression, and hidden rheumatoid factors. Ann Rheum Dis 48:488, 1989.
76. Carson DA, Chen PP, Kipps TJ: New roles for rheumatoid factor. J Clin Invest 87:379, 1991.
77. Wernick RM, Lipsky PE, Marban-Arcos E, et al: IgG and IgM rheumatoid factor synthesis in rheumatoid synovial membrane cell cultures. Arthritis Rheum 28:742, 1985.
78. Bouvet JP, Xin WJ, Pillot J: Restricted heterogeneity of polyclonal rheumatoid factor. Arthritis Rheum 30:998, 1987.
79. Lee SK, Bridges SL Jr, Koopman WJ, Schroeder HW Jr: The immunoglobulin kappa light chain repertoire expressed in the synovium of a patient with rheumatoid arthritis. Arthritis Rheum 35:905, 1992.
80. Sutton BJ, Corper AL, Sohi MK, et al: The structure of a human rheumatoid factor bound to IgG Fc. Adv Exp Med Biol 435:41, 1998.
81. Wang H, Shlomchik MJ: High affinity rheumatoid factor transgenic B cells are eliminated in normal mice. J Immunol 159:1125, 1997.
82. Tighe H, Warnatz K, Brinson D, Corr M, et al: Peripheral deletion of rheumatoid factor B cells after abortive activation by IgG. Proc Natl Acad Sci U S A 94:646, 1997.
83. Kouskoff V, Korganow AS, Duchatelle V, et al: Organ-specific disease provoked by systemic autoimmunity. Cell 87:811, 1996.
84. Nykanen P, Helve T, Kankaanpaa U, Larsen A: Characterisation of the DNA-synthesizing cells in rheumatoid synovial tissue. Scand J Rheumatol 7:118, 1978.
85. Revell PA, Mapp PI, Lalor PA, Hall PA: Proliferative activity of cells in the synovium as demonstrated by a monoclonal antibody, Ki67. Rheumatol Int 7:183, 1987.
86. Qu Z, Garcia CH, O'Rourke LM, et al: Local proliferation of fibroblast-like synoviocytes contributes to synovial hyperplasia. Results of proliferating cell nuclear antigen/cyclin, c-myc, and nucleolar organizer region staining. Arthritis Rheum 37:212, 1994.
87. Ritchlin C, Dwyer E, Bucala R, Winchester R: Sustained and distinctive patterns of gene activation in synovial fibroblasts and whole synovial tissue obtained from inflammatory synovitis. Scand J Immunol 40:292, 1994.
88. Burmester GR, Dimitriu-Bona A, Waters SJ, Winchester RJ: Identification of three major synovial lining cell populations by monoclonal antibodies directed to Ia antigens associated with monocytes/macrophages and fibroblasts. Scand J Immunol 17:69, 1983.
89. Goto M, Sasano M, Yamanaka H, et al: Spontaneous production of an interleukin 1-like factor by cloned rheumatoid synovial cells in long-term culture. J Clin Invest 80:786, 1987.
90. Baker DG, Dayer JM, Roelke M, et al: Rheumatoid synovial cell morphologic changes induced by a mononuclear cell factor in culture. Arthritis Rheum 26:8, 1983.
91. Haynes BF, Grover BJ, Whichard LP, et al: Synovial microenvironment-T cell interactions: Human T cells bind to fibroblast-like synovial cells in vitro. Arthritis Rheum 31:947:1988.
92. Nakao H, Eguchi K, Kawakami A, et al: Phenotypic characterization of lymphocytes infiltrating synovial tissue from patients with rheumatoid arthritis: Analysis of lymphocytes isolated from minced synovial tissue by dual immunofluorescent staining. J Rheumatol 17:142, 1990.
93. Postigo AA, Garcia-Vicuna R, Diaz-Gonzalez F, et al: Increased binding of synovial T lymphocytes from rheumatoid arthritis to endothelial-leukocyte adhesion molecule-1 (ELAM-1) and vascular cell adhesion molecule-1 (VCAM-1). J Clin Invest 89:1445, 1992.
94. Nykanen P, Bergroth V, Raunio P, et al: Phenotypic characterization of 3H-thymidine incorporating cells in rheumatoid arthritis synovial membrane. Rheumatol Int 6:269, 1986.
95. Tak PP, Kummer JA, Hack CE, et al: Granzyme-positive cytotoxic cells are specifically increased in early rheumatoid synovial tissue. Arthritis Rheum 37:1735, 1994.
96. Bromley M, Fisher WD, Woolley DE: Mast cells at sites of cartilage erosion in the rheumatoid joint. Ann Rheum Dis 43:76, 1984.
97. Frewin DB, Cleland LG, Johnsson JR, Robertson PW: Histamine levels in human synovial fluid. J Rheumatol 13:13, 1986.
98. Malone DG, Wilder RL, Saavedra-Delgado AM, Metcalfe DD: Mast cell numbers in rheumatoid
synovial tissues. Arthritis Rheum 30:130, 1987.
99. Kiener HP, Baghestanian M, Dominkus M, et al: Expression of the C5a receptor (CD88) on synovial mast cells in patients with rheumatoid arthritis. Arthritis Rheum 41:233, 1998.
100. Yoffe JR, Taylor DJ, Woolley DE: Mast cell products stimulate collagenase and prostaglandin E production by cultures of adherent rheumatoid synovial cells. Biochim Biophys Res Commun 122:270, 1984.
101. Schumacher HR, Kitridou RC: Synovitis of recent onset. A clinicopathological study during the first month of disease. Arthritis Rheum 15:465, 1972.
102. Smeets TJ, Dolhain RJEM, Miltenburg AM, et al: Poor expression of T cell-derived cytokines and activation and proliferation markers in early rheumatoid synovial tissue. Clin Immunol Immunopathol 88:84, 1998.
103. Kraan MC, Versendaal H, Jonker M, et al: Asymptomatic synovitis precedes clinically manifest arthritis. Arthritis Rheum 41:1481, 1998.
104. Han Z, Boyle DL, Manning AM, Firestein GS: AP-1 and NF-kB regulation in rheumatoid arthritis and murine collagen-induced arthritis. Autoimmunity 28:197, 1998.
105. Geiler T, Kriegsmann J, Keyszer GM, et al: A new model for rheumatoid arthritis generated by engraftment of rheumatoid synovial tissue and normal human cartilage into SCID mice. Arthritis Rheum 37:1664, 1994.
106. Muller-Ladner L, Kriegsmann J, Franklin BN, et al: Synovial fibroblasts of patients with rheumatoid arthritis attach to and invade normal human cartilage when engrafted into SCID mice. Am J Pathol 149:1607, 1996.
107. Muller-Ladner U, Roberts CR, Franklin BN, et al: Human IL-1Ra gene transfer into human synovial fibroblasts is chondroprotective. J Immunol 158:3492, 1997.
108. Wegelius O, Laine V, Lindstrom B, Klockars M: Fistula of the thoracic duct as immunosuppressive treatment in rheumatoid arthritis. Acta Med Scand 187:539, 1970.
109. Wahl SM, Wilder RL, Katona IM, et al: Leukapheresis in rheumatoid arthritis. Association of clinical improvement with reversal of anergy. Arthritis Rheum 26:1076, 1983.
110. Zvaifler NJ: Fractionated total lymphoid irradiation: A promising new treatment for rheumatoid arthritis? Yes, no, maybe. Arthritis Rheum 30:109, 1987.
111. Watts RA, Isaacs JD, Hale G, et al: CAMPATH-1H in inflammatory arthritis. Clin Exp Rheumatol 11 (Suppl 8):S165, 1993.
112. Strand V, Lipsky PE, Cannon GW, et al: Effects of administration of an anti-CD5 plus immunoconjugate in rheumatoid arthritis. Results of two phase II studies. The CD5 Plus Rheumatoid Arthritis Investigators Group. Arthritis Rheum 36:620, 1993.
113. Moreland LW, Pratt PW, Bucy RP, et al: Treatment of refractory rheumatoid arthritis with a chimeric anti-CD4 monoclonal antibody. Long-term followup of CD4+ T cell counts. Arthritis Rheum
37:834, 1994.
114. Peter A, van der Lubbe B, Dijkmans AC, et al: Lack of clinical effect of CD4 monoclonal antibody therapy in early rheumatoid arthritis; a placebo controlled trial. Arthritis Rheum 37:S294, 1994.
115. Ruderman EM, Weinblatt ME, Thurmond LM, et al: Synovial tissue response to treatment with CAMPATH-1H. Arthritis Rheum 38:254, 1995.
116. Wells G, Tugwell P: Cyclosporin A in rheumatoid arthritis: Overview of efficacy. Br J Rheumatol 32(Suppl 1):51, 1993.
117. Sewell KL, Parker KC, Woodworth TG, et al: DAB486IL-2 fusion toxin in refractory rheumatoid arthritis. Arthritis Rheum 36:1223, 1993.
118. Muller-Ladner U, Kriegsmann J, Gay RE, et al: Progressive joint destruction in a human immunodeficiency virus-infected patient with rheumatoid arthritis. Arthritis Rheum 38:1328, 1995.
119. Poulter LW, Janossy G: The involvement of dendritic cells in chronic inflammatory disease. Scand J Immunol 21:401, 1985.
120. Kurosaka M, Ziff M: Immunoelectron microscopic study of the distribution of T cell subsets in rheumatoid synovium. J Exp Med 158:1191, 1983.
121. Tsai C, Diaz LA Jr, Singer NG, et al: Responsiveness of human T lymphocytes to bacterial superantigens presented by cultured rheumatoid arthritis synoviocytes. Arthritis Rheum 39:125, 1996.
122. Liu MF, Kohsaka H, Sakurai H, et al: The presence of costimulatory molecules CD86 and CD28 in rheumatoid arthritis synovium. Arthritis Rheum 39:110, 1996.
123. Schmidt D, Goronzy JJ, Weyand CM: CD4+ CD7- CD28- T cells are expanded in rheumatoid arthritis and are characterized by autoreactivity. J Clin Invest 97:2027, 1996.
124. MacDonald KP, Nishioka Y, Lipsky PE, Thomas R: Functional CD40 ligand is expressed by T cells in rheumatoid arthritis. J Clin Invest 100:2404, 1997.
125. Rissoan MC, Van Kooten C, Chomarat P, et al: The functional CD40 antigen of fibroblasts may contribute to the proliferation of rheumatoid synovium. Clin Exp Immunol 106:481, 1996.
126. Lamour A, Jouen-Beades F, Lees O, et al: Analysis of T cell receptors in rheumatoid arthritis: The increased expression of HLA-DR antigen on circulating gamma delta+ T cells is correlated with disease activity. Clin Exp Immunol 89:217, 1992.
127. Dooley MA, Cush JJ, Lipsky PE, et al: The effects of nonsteroidal antiinflammatory drug therapy in early rheumatoid arthritis on serum levels of soluble interleukin 2 receptor, CD4, and CD8. J Rheumatol 20:1857, 1993.
128. Hemler ME, Glass D, Coblyn JS, Jacobson JG: Very late activation antigens on rheumatoid synovial fluid T lymphocytes. Association with stages of T cell activation. J Clin Invest 78:696, 1986.
129. Keystone EC, Snow KM, Bombardier C, et al: Elevated soluble interleukin-2 receptor levels in the sera and synovial fluids of patients with rheumatoid arthritis. Arthritis Rheum 31:844, 1988.
130. Tosato G, Steinberg AD, Blaese RM: Defective EBV-specific suppressor T-cell function in rheumatoid arthritis. N Engl J Med 305:1238, 1981.
131. Gaston JSH, Rickinson AB, Yao QY, Epstein MA: The abnormal cytotoxic T cell response to Epstein-Barr virus in rheumatoid arthritis is correlated with disease activity and occurs in other arthropathies. Ann Rheum Dis 45:932, 1986.
132. Hasler F, Bluestein HG, Zvaifler NJ, Epstein LB: Analysis of the defects responsible for the impaired regulation of EBV-induced B cell proliferation by rheumatoid arthritis lymphocytes. II. Role of monocytes and the increased sensitivity of rheumatoid arthritis lymphocytes to prostaglandin E. J Immunol 131:768, 1983.
133. Combe B, Pope RM, Fischbach M, et al: Interleukin 2 in rheumatoid arthritis: Production of and response to interleukin 2 in rheumatoid synovial fluid, synovial tissue, and peripheral blood. Clin Exp Immunol 59:520, 1985.
134. Hasler F, Dayer JM: Diminished IL-2-induced gamma-interferon production by unstimulated peripheral-blood lymphocytes in rheumatoid arthritis. Br J Rheumatol 27:15, 1988.
135. Fox RI, Fong S, Sabharwal N, et al: Synovial fluid lymphocytes differ from peripheral blood lymphocytes in patients with rheumatoid arthritis. J Immunol 128:351, 1982.
136. Lasky HP, Bauer K, Pope RM: Increased helper inducer and decreased suppressor inducer phenotypes in the rheumatoid joint. Arthritis Rheum 31:52, 1988.
137. Nouri AME, Panayi GS: Cytokines and the chronic inflammation of rheumatic disease. III. Deficient interleukin-2 production in rheumatoid arthritis is not due to suppressor mechanisms. J Rheumatol 14:902, 1987.
138. Wahl SM, Allen JB, Wong HL, et al: Antagonistic and agonistic effects of transforming growth factor-beta and IL-1 in rheumatoid synovium. J Immunol 145:2514, 1990.
139. Firestein GS, Berger AE, Tracey DE, et al: IL-1 receptor antagonist protein production and gene expression in rheumatoid arthritis and osteoarthritis synovium. J Immunol 149:1054, 1992.
140. Maurice MM, Lankester AC, Bezemer AC, et al: Defective TCR-mediated signaling in synovial T cells in rheumatoid arthritis. J Immunol 159:2973, 1997.
141. Maurice MM, Nakamura H, van der Voort EA, et al: Evidence for the role of an altered redox state in hyporesponsiveness of synovial T cells in rheumatoid arthritis. Immunology 158:1458, 1997.
142. Fox DA, Singer NG: T cell receptor rearrangements in arthritis. In Miossec P, van den Berg WB, Firestein GS (eds): T cells in Arthritis. Basel, Birkhauser, 1998, p 19.
143. Karin N, Szafer F, Mitchell D, et al: Selective and nonselective stages in homing of T lymphocytes to the central nervous system during experimental allergic encephalomyelitis. J Immunol 150:4116, 1993.
144. Firestein GS, Zvaifler NJ: How important are T cells in chronic rheumatoid synovitis? Arthritis Rheum 33:768, 1990.
145. Firestein GS, Roeder WD, Laxer JA, et al: A new murine CD4+ T cell subset with an unrestricted cytokine profile. J Immunol 143:518, 1989.
146. Stephenson ML, Krane SM, Amento EP, et al: Immune interferon inhibits collagen synthesis by rheumatoid synovial cells associated with decreased levels of the procollagen mRNAs. FEBS Lett 180:43, 1985.
147. Alvaro-Gracia JM, Zvaifler NJ, Firestein GS: Cytokines in chronic inflammatory arthritis. V. Mutual antagonism between interferon-gamma and tumor necrosis factor-alpha on HLA-DR expression,
proliferation, collagenase production, and granulocyte macrophage colony-stimulating factor production by rheumatoid arthritis synoviocytes. J Clin Invest 86:1790, 1990.
148. Unemori EN, Bair MJ, Bauer EA, Amento EP: Stromelysin expression regulates collagenase activation in human fibroblasts. Dissociable control of two metalloproteinases by interferon-gamma. J Biol Chem 266:23, 477, 1991.
149. Alvaro-Gracia JM, Yu C, Zvaifler NJ, Firestein GS: Mutual antagonism between interferon-gamma and tumor necrosis factor-alpha on fibroblast-like synoviocytes: Paradoxical induction of IFN-gamma and TNF-alpha receptor expression. J Clin Immunol 13:212, 1993.
150. Firestein GS, Zvaifler NJ: Peripheral blood and synovial fluid monocyte activation in inflammatory arthritis. II. Low levels of synovial fluid and synovial tissue interferon suggest that gamma-interferon is not the primary macrophage activating factor. Arthritis Rheum 30:864, 1987.
151. Chen E, Keystone EC, Fish EN: Restricted cytokine expression in rheumatoid arthritis. Arthritis Rheum 36:901, 1993.
152. Bergroth V, Zvaifler NJ, Firestein GS: Cytokines in chronic inflammatory arthritis. III. Rheumatoid arthritis monocytes are not unusually sensitive to gamma-interferon, but have defective gamma-interferon-mediated HLA-DQ and HLA-DR induction. Arthritis Rheum 32:1074, 1989.
153. Barnes PF, Fong SJ, Brennan PJ, et al: Local production of tumor necrosis factor and IFN-gamma in tuberculous pleuritis. J Immunol 145:149, 1990.
154. Smeets TJ, Dolhain RJ, Breedveld FC, Tak PP: Analysis of the cellular infiltrates and expression of cytokines in synovial tissue from patients with rheumatoid arthritis and reactive arthritis. J Pathol 186:75, 1998.
155. Ruschen S, Lemm G, Warnatz H: Interleukin-2 secretion by synovial fluid lymphocytes in rheumatoid arthritis. Br J Rheumatol 27:350, 1988.
156. Firestein GS, Xu WD, Townsend K, et al: Cytokines in chronic inflammatory arthritis. I. Failure to detect T cell lymphokines (interleukin 2 and interleukin 3) and presence of macrophage colony-stimulating factor (CSF-1) and a novel mast cell growth factor in rheumatoid synovitis. J Exp Med 168:1573, 1988.
157. Miossec P, Navillat M, Dupuy d'Angeac A, et al: Low levels of interleukin-4 and high levels of transforming growth factor beta in rheumatoid arthritis. Arthritis Rheum 145:2514, 1990.
158. Husby G, Williams RC Jr: Immunohistochemical studies of interleukin-2 and interferon-gamma in rheumatoid arthritis. Arthritis Rheum 28:174, 1985.
159. Dolhain RJ, van der Heiden AN, ter Haar NT, et al: Shift toward T lymphocytes with a T helper 1 cytokine-secretion profile in the joints of patients with rheumatoid arthritis. Arthritis Rheum 39:1961, 1996.
160. Kusaba M, Honda J, Fukuda T, Oizumi K: Analysis of type 1 and type 2 T cells in synovial fluid and peripheral blood of patients with rheumatoid arthritis. J Rheumatol 25:1466, 1998.
161. Qin S, Rottman JB, Myers P, et al: The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory reactions. J Clin Invest 101:746, 1998.
162. Walmsley M, Katsikis PD, Abney E, et al: Interleukin-10 inhibition of the progression of established collagen-induced arthritis. Arthritis Rheum 39:495, 1996.
163. Joosten LA, Lubberts E, Durez P, et al: Role of interleukin-4 and interleukin-10 in murine collagen-induced arthritis. Protective effect of interleukin-4 and interleukin-10 treatment on cartilage destruction. Arthritis Rheum 40:249, 1997.
164. Joosten LAB, Lubberts E, Helsen MMA, van den Berg WB: Dual role of IL-12 in early and late stages of murine collagen type II arthritis. J Immunol 159:4094, 1997.
165. Chu CQ, Londei M: Induction of Th2 cytokines and control of collagen-induced arthritis by nondepleting anti-CD4. Abs J Immunol 157:2685, 1996.
166. Katsikis KD, CQ Chu, FM Brennan, et al: Immunoregulatory role of interleukin 10 in rheumatoid arthritis. J Exp Med 179:1517, 1994.
167. Chomarat P, Banchereau J, Miossec P: Differential effects of interleukins 10 and 4 on the production of interleukin-6 by blood and synovium monocytes in rheumatoid arthritis. Arthritis Rheum 38:1046, 1995.
168. Lacraz S, Nicod LP, Chicheportiche R, et al: IL-10 inhibits metalloproteinase and stimulates TIMP-1 production in human mononuclear phagocytes. J Clin Invest 96:2304, 1995.
169. Doketer WH, Esselink MT, Halie MR, Vellenga E: Interleukin-4 inhibits the lipopolysaccharide-induced expression of c-jun and c-fos messenger RNA and activator protein-1 binding activity in human monocytes. Blood 81:337, 1993.
170. Deleuran B, Iversen L, Kristensen M, et al: Interleukin-8 secretion and 15-lipoxygenase activity in rheumatoid arthritis: In vitro anti-inflammatory effects by interleukin-4 and interleukin-10, but not by interleukin-1 receptor antagonist protein. Br J Rheumatol 33:520, 1994.
171. Chomarat P, Vannier E, Dechanet J, et al: Balance of IL-1 receptor antagonist/IL-1 beta in rheumatoid synovium and its regulation by IL-4 and IL-10. J Immunol 154:1432, 1995.
172. Chabaud M, Fossiez F, Taupin JL, Miossec P: Enhancing effect of IL-17 on IL-1-induced IL-6 and leukemia inhibitory factor production by rheumatoid arthritis synoviocytes and its regulation by Th2 cytokines. J Immunol 161:409, 1998.
173. Burger D, Rezzonico R, Li JM, et al: Imbalance between interstitial collagenase and tissue inhibitor of metalloproteinases 1 in synoviocytes and fibroblasts upon direct contact with stimulated T lymphocytes: Involvement of membrane-associated cytokines. Arthritis Rheum 41:1748, 1998.
174. Pettipher ER, Higgs GA, Henderson B: Interleukin 1 induces leukocyte infiltration and cartilage proteoglycan degradation in the synovial joint. Proc Natl Acad Sci U S A 83:8749, 1986.
175. Shore A, Jaglal S, Keystone EC: Enhanced interleukin 1 generation by monocytes in vitro is temporally linked to an early event in the onset or exacerbation of rheumatoid arthritis. Clin Exp Immunol 65:293, 1986.
176. Poubelle P, Damon M, Blotman F, Dayer J-M: Production of mononuclear cell factor by mononuclear phagocytes from rheumatoid synovial fluid. J Rheumatol 12:412, 1985.
177. Chin J, Rupp E, Cameron PM, et al: Identification of a high-affinity receptor for interleukin 1a and interleukin 1b on cultured human rheumatoid synovial cells. J Clin Invest 82:420, 1988.
178. Firestein GS, Alvaro-Gracia JM, Maki R: Quantitative analysis of cytokine gene expression in
rheumatoid arthritis. J Immunol 144:3347, 1990.
179. Miyasaka N, Sato K, Goto M, et al: Augmented interleukin-1 production and HLA-DR expression in the synovium of rheumatoid arthritis patients: Possible involvement in joint destruction. Arthritis Rheum 31:480, 1988.
180. Dayer JM, Ricard-Blum S, Kaufman MT, Herbage D: Type IX collagen is a potent inducer of
PGE2 and interleukin 1 production by human monocyte macrophages. FEBS Lett 198:208, 1986.
181. Postlethwaite AE, Lachman LB, Kang AH: Induction of fibroblast proliferation by interleukin-1 derived from human monocytic leukemia cells. Arthritis Rheum 27:995, 1984.
182. Yaron I, Meyer FA, Dayer JM, Yaron M: Human recombinant interleukin-1beta stimulates glycosaminoglycan production in human synovial fibroblast cultures. Arthritis Rheum 30:424, 1987.
183. Dewhirst FE, Stashenko PP, Mole JE, Tsurumachi T: Purification and particle sequence of human osteoclast-activating factor: Identity with interleukin 1beta. J Immunol 135:2562, 1985.
184. Dayer JM, Beutler B, Cerami A: Cachectin/tumor necrosis factor stimulates collagenase and prostaglandin E2 production by human synovial cells and dermal fibroblasts. J Exp Med 162:2163,
1985.
185. Saklatvala J: Tumour necrosis factor alpha stimulates resorption and inhibits synthesis of proteoglycan in cartilage. Nature 322:547, 1986.
186. Williams RO, Mason LJ, Feldmann M, Maini RN: Synergy between anti-CD4 and anti-tumor necrosis factor in the amelioration of established collagen-induced arthritis. Proc Natl Acad Sci U S A 91:2762, 1994.
187. Georgopoulos S, Plows D, Kollias G: Transmembrane TNF is sufficient to induce localized tissue toxicity and chronic inflammatory arthritis in transgenic mice. J Inflamm 46:86.1996.
188. Elliott MJ, Maini RN, Feldmann M, et al: TNF-alpha blockade in rheumatoid arthritis: Results of multicenter placebo-controlled, double-blind trial of chimeric anti-TNF-alpha monoclonal antibody (cA2) in refractory disease. Arthritis Rheum 37:S283, 1994.
189. Guerne PA, Zuraw BL, Vaughan JH, et al: Synovium as a source of interleukin 6 in vitro. Contribution to local and systemic manifestations of arthritis. J Clin Invest 83:585, 1989.
190. Houssiau FA, Devogelaer J-P, van Damme J, et al: Interleukin-6 in synovial fluid and serum of patients with rheumatoid arthritis and other inflammatory arthritides. Arthritis Rheum 31:784, 1988.
191. Guerne PA, Zuraw BL, Vaughan JH, et al: Synovium as a source of interleukin 6 in vitro: Contribution to local and systemic manifestations of arthritis. J Clin Invest 83:585, 1989.
192. Field M, Chu C, Feldman M, Maini RN: Interleukin-6 localisation in the synovial membrane in rheumatoid arthritis. Rheumatol Int 11:45, 1991.
193. McInnes IB, Leung BP, Sturrock RD, et al: Interleukin-15 mediates T cell-dependent regulation of tumor necrosis factor-alpha production in rheumatoid arthritis. Nat Med 3:189, 1997.
194. Ruchatz H, Leung BP, Wei XQ, et al: Soluble IL-15 receptor alpha-chain administration prevents murine collagen-induced arthritis: A role for IL-15 in development of antigen-induced immunopathology. J Immunol 160:5654, 1998.
195. Thurkow EW, van der Heijden IM, Breedveld FC, et al: Increased expression of IL-15 in the
synovium of patients with rheumatoid arthritis compared with patients with Yersinia-induced arthritis and osteoarthritis. J Pathol 181:444, 1997.
196. Xu WD, Firestein GS, Taetle R, et al: Cytokines in chronic inflammatory arthritis. II. Granulocyte-macrophage colony-stimulating factor in rheumatoid synovial effusions. J Clin Invest 83:876, 1989.
197. Alvaro-Gracia JM, Zvaifle NJ, Brown CB, et al: Cytokines in chronic inflammatory arthritis. VI. Analysis of the synovial cells involved in granulocyte-macrophage colony-stimulating factor production and gene expression in rheumatoid arthritis and its regulation by IL-1 and tumor necrosis factor-alpha. J Immunol 146:3365, 1991.
198. Koch AE, Kunkel SL, Burrows JC, et al: Synovial tissue macrophage as a source of the chemotactic cytokine IL-8. J Immunol 147:2187, 1991.
199. Koch AE, Kunkel SL, Harlow LA, et al: Enhanced production of monocyte chemoattractant protein-1 in rheumatoid arthritis. J Clin Invest 90:772, 1992.
200. Koch AE, Kunkel SL, Harlow LA, et al: Epithelial neutrophil activating peptide-78: A novel chemotactic cytokine for neutrophils in arthritis. J Clin Invest 94:1012, 1994.
201. Rathanaswami P, Hachicha M, Sadick M, et al: Expression of the cytokine RANTES in human rheumatoid synovial fibroblasts. Differential regulation of RANTES and interleukin-8 genes by inflammatory cytokines. J Biol Chem 268:5834, 1993.
202. Fava R, Olsen N, Keski-Oja J, et al: Active and latent forms of transforming growth factor beta activity in synovial effusions. J Exp Med 169:291, 1989.
203. Allen JB, Manthey CL, Hand AR, et al: Rapid onset synovial inflammation and hyperplasia induced by transforming growth factor beta. J Exp Med 171:231, 1990.
204. Brandes ME, Allen JB, Ogawa Y, Wahl SM: Transforming growth factor beta 1 suppresses acute and chronic arthritis in experimental animals. J Clin Invest 87:1108, 1991.
205. Wahl SM, Allen JB, Costa GL, et al: Reversal of acute and chronic synovial inflammation by anti-transforming growth factor beta. J Exp Med 177:225, 1993.
206. Song XY, Gu M, Jin WW, et al: Plasmid DNA encoding transforming growth factor-beta1 suppresses chronic disease in a streptococcal cell wall-induced arthritis model. J Clin Invest 101:2615, 1998.
207. Remmers EF, Lafyatis R, Kumkumian GK, et al: Cytokines and growth regulation of synoviocytes from patients with rheumatoid arthritis and rats with streptococcal cell wall arthritis. Growth Factors 2:179, 1990.
208. Remmers EF, Sano H, Lafyatis, et al: Production of platelet derived growth factor B chain (PDGF-B/c-sis) mRNA and immunoreactive PDGF B-like polypeptide by rheumatoid synovium: Coexpression with heparin binding acidic fibroblast growth factor-1. J Rheumatol 18:7, 1991.
209. Yayon A, Klagsbrun M, Esko JD, et al: Cell surface, heparin-like molecules are required for binding of basic fibroblast growth factor to its high affinity receptor. Cell 64:841, 1991.
210. Melnyk VO, Shipley GD, Sternfeld MD, et al: Synoviocytes synthesize, bind, and respond to basic fibroblast growth factor. Arthritis Rheum 33:493, 1990.
211. Henderson B, Thompson RC, Hardingham T, Lewthwaite J: Inhibition of interleukin-1-induced synovitis and articular cartilage proteoglycan loss in the rabbit knee by recombinant human interleukin-1 receptor antagonist. Cytokine 3:246, 1991.
212. Lewthwaite J, Blake SM, Hardingham TE, et al: The effect of recombinant human interleukin 1 receptor antagonist on the induction phase of antigen induced arthritis in the rabbit. J Rheumatol 21:467, 1994.
213. Malyak M, Swaney RE, Arend WP: Levels of synovial fluid interleukin-1 receptor antagonist in rheumatoid arthritis and other arthropathies. Potential contribution from synovial fluid neutrophils. Arthritis Rheum 36:781, 1993.
214. Firestein GS, Berger AE, Tracey DE, et al: IL-1 receptor antagonist protein production and gene expression in rheumatoid arthritis and osteoarthritis synovium. J Immunol 149:1054, 1992.
215. Firestein GS, Boyle DL, Yu C, et al: Synovial interleukin-1 receptor antagonist and interleukin-1 balance in rheumatoid arthritis. Arthritis Rheum 37:644, 1994.
216. Arend WP, Malyak M, Smith MF Jr, et al: Binding of IL-1 alpha, IL-1 beta, and IL-1 receptor antagonist by soluble IL-1 receptors and levels of soluble IL-1 receptors in synovial fluids. J Immunol 153:4766, 1994.
217. Cope AP, Aderka D, Doherty M, et al: Increased levels of soluble tumor necrosis factor receptors in the sera and synovial fluid of patients with rheumatic diseases. Arthritis Rheum 35:1160, 1992.
218. Brennan FM, Gibbons DL, Mitchell T, et al: Enhanced expression of tumor necrosis factor receptor mRNA and protein in mononuclear cells isolated from rheumatoid arthritis synovial joints. Eur J Immunol 22:1907, 1992.
219. Firestein GS, Manning AM: Signal transduction and transcription factors in rheumatic diseases Arthritis Rheum 42:609, 1999.
220. Fujisawa K, Aono H, Hasunuma T, et al: Activation of transcription factor NF-kB in human synovial cells in response to tumor necrosis factor alpha. Arthritis Rheum 39:197, 1996.
221. Mercurio F, Zhu H, Murray BW, et al: IKK-1 and IKK-2: Cytokine-activated IkB kinases essential for NF-kB activation. Science 278:860, 1997.
222. Marok R, Winyard PG, Coumbe A, et al: Activation of the transcription factor nuclear factor-kappaB in human inflamed synovial tissue. Arthritis Rheum 39:583, 1996.
223. Sioud M, Mellbye O, Forre O: Analysis of the NF-kappa B p65 subunit, Fas antigen, Fas ligand and Bcl-2-related proteins in the synovium of RA and polyarticular JRA. Clin Exp Rheumatol 16:125, 1998.
224. Bennett BL, Aupperle KR, Manning AM, Firestein GS: Activation of IKB kinase in fibroblast-like synoviocytes. Arthritis Rheum 41:S350, 1998.
225. Miyazawa K, Mori A, Yamamoto K, Okudaira H: Constitutive transcription of the human interleukin-6 gene by rheumatoid synoviocytes: Spontaneous activation of NF-kappaB and CBF1. Am J Pathol 152:793, 1998.
226. Miyazawa K, Mori A, Yamamoto K, Okudaira H: Transcriptional roles of CCAAT/enhancer binding protein-beta, nuclear factor-kappaB, and C-promoter binding factor 1 in interleukin (IL)-1beta-induced
IL-6 synthesis by human rheumatoid fibroblast-like synoviocytes. J Biol Chem 273:7620, 1998.
227. Kinne RW, Boehm S, Iftner T, et al: Synovial fibroblast-like cells strongly express jun-B and C-fos proto-oncogenes in rheumatoid-and osteoarthritis. Scand J Rheumatol Suppl 101:121, 1995.
228. Asahara H, Fujisawa K, Kobata T, et al: Direct evidence of high DNA binding activity of transcription factor AP-1 in rheumatoid arthritis synovium. Arthritis Rheum 40:912, 1997.
229. Boyle DL, Han Z, Rutter J, et al: Post-transcriptional regulation of collagenase gene expression in synoviocytes by adenosine receptor stimulation. Arthritis Rheum 40:1772, 1997.
230. Wakisaka S, Suzuki N, Saito N, et al: Possible correction of abnormal rheumatoid arthritis synovial cell function by jun D transfection in vitro. Arthritis Rheum 41:470, 1998.
231. Han Z, Boyle DL, Bennett B, et al: JunN-amino-terminal kinase in rheumatoid arthritis synoviocytes. J Pharm Exp Ther 294:124, 1999.
232. Okamoto K, Fujisawa K, Hasunuma T, et al: Selective activation of the JNK/AP-1 pathway in Fas-mediated apoptosis of rheumatoid arthritis synoviocytes. Arthritis Rheum 40:919, 1997.
233. Sekine C, Yagita H, Kobata T, et al: Fas-mediated stimulation induces IL-8 secretion by rheumatoid arthritis synoviocytes independently of CPP32-mediated apoptosis. Biochem Biophys Res Comm 228:14, 1996.
234. Tsao PW, Suzuki T, Totsuka R, et al: The effect of dexamethasone on the expression of activated NF-kappa B in adjuvant arthritis. Clin Immunol Immunopathol 83:173, 1997,
235. Pierce JW, Schoenleber R, Jesmok G, et al: Novel inhibitors of cytokine-induced IkappaBalpha phosphorylation and endothelial cell adhesion molecule expression show anti-inflammatory effects in vivo. J Biol Chem 272:21,096, 1997.
236. Shiozawa S, Shimizu K, Tanaka K, Hino K: Studies on the contribution of c-fos/AP-1 to arthritic joint destruction. J Clin Invest 99:1210, 1997.
237. Miagkov AV, Kovalenko DV, Brown CE, et al: NF-kappaB activation provides the potential link between inflammation and hyperplasia in the arthritic joint. Proc Natl Acad Sci U S A 95:13859, 1998.
238. Badger AM, Bradbeer JN, Votta B, et al: Pharmacological profile of SB 203580, a selective inhibitor of cytokine suppressive binding protein/p38 kinase, in animal models of arthritis, bone resorption, endotoxin shock and immune function. J Pharmacol Exp Ther 279:1453, 1996.
964
239. Wang F, Sengupta TK, Zhong Z, Ivashkiv LB: Regulation of the balance of cytokine production and the signal transducer and activator of transcription (STAT) transcription factor activity by cytokines and inflammatory synovial fluids. J Exp Med 182:1825, 1995.
240. Sengupta TK, Chen A, Zhong Z, et al: Activation of monocyte effector genes and STAT family transcription factors by inflammatory synovial fluid is independent of interferon. J Exp Med 181:1015, 1995.
241. Mapp PI, Grootveld MC, Blake DR: Hypoxia, oxidative stress and rheumatoid arthritis. Br Med Bull 51:419, 1995.
242. Halliwell B: Oxygen radicals, nitric oxide and human inflammatory joint disease. Ann Rheum Dis 54:505, 1995.
243. Maurice MM, Nakamura H, van der Voort EAM, et al: Increased expression of thioredoxin in rheumatoid arthritis. Arthritis Rheum 41:S319, 1998.
244. Bashir S, Harris G, Denman MA, et al: Oxidative DNA damage and cellular sensitivity to oxidative stress in human autoimmune diseases. Ann Rheum Dis 52:659, 1993.
245. Sakurai H, Kohsaka H, Liu MF, et al: Nitric oxide production and inducible nitric oxide synthase expression in inflammatory arthritides. J Clin Invest 96:2357, 1995.
246. Farrell AJ, Blake DR, Palmer RM, Moncada S: Increased concentrations of nitrite in synovial fluid and serum samples suggest increased nitric oxide synthesis in rheumatic diseases. Ann Rheum Dis 51:1219, 1992.
247. Salmon M, Scheel-Toellner D, Huissoon AP, et al: Inhibition of T cell apoptosis in the rheumatoid synovium. Clin Invest 99:439, 1997.
248. Asahara H, Hasumuna T, Kobata T, et al: Expression of Fas antigen and Fas ligand in the rheumatoid synovial tissue. Clin Immunol Immunopathol 81:27, 1996.
249. Cantwell MJ, Hua T, Zvaifler NJ, Kipps TJ: Deficient Fas ligand expression by synovial lymphocytes from patients with rheumatoid arthritis. Arthritis Rheum 40:1644, 1997.
250. Nakajima T, Aono H, Hasunuma T, et al: Apoptosis and functional Fas antigen in rheumatoid arthritis synoviocytes. Arthritis Rheum 38:485, 1995.
251. Firestein, GS, Yeo M, Zvaifler NJ: Apoptosis in rheumatoid arthritis synovium. J Clin Invest 96:1631, 1995.
252. Matsumoto S, Muller-Ladner U, Gay RE, et al: Ultrastructural demonstration of apoptosis, Fas and Bcl-2 expression of rheumatoid synovial fibroblasts. Rheumatology 23:1345, 1996.
253. Kawakami A, Eguchi K, Matsuoka N, et al: Inhibition of Fas antigen-mediated apoptosis of rheumatoid synovial cells in vitro by transforming growth factor beta 1. Arthritis Rheum 39:1267, 1996.
254. Okamoto K, Fujisawa K, Hasunuma T, et al: Selective activation of the JNK/AP-1 pathway in Fas-mediated apoptosis of rheumatoid arthritis synoviocytes. Arthritis Rheum 40:919, 1997.
255. Zhang H, Yang Y, Horton JL, et al: Amelioration of collagen-induced arthritis by CD95 (Apo-1/Fas)-ligand gene transfer. J Clin Invest 100:1951, 1997.
256. Lafyatis R, Remmers EF, Roberts AB, et al: Anchorage-independent growth of synoviocytes from arthritis and normal joints: Stimulation by exogenous platelet-derived growth factor and inhibition by transforming growth factor-beta and retinoids. J Clin Invest 83:1267, 1989.
257. Imamura F, Aono H, Hasunuma T, et al: Monoclonal expansion of synoviocytes in rheumatoid arthritis. Arthritis Rheum 41:1979, 1998.
258. Gay S, Gay RE: Cellular basis and oncogene expression of rheumatoid joint destruction. Rheumatol Int 9:105, 1989.
259. Bucala R, Ritchlin C, Winchester R, Cerami A: Constitutive production of inflammatory and
mitogenic cytokines by rheumatoid synovial fibroblasts. J Exp Med 173:569, 1991.
260. Firestein GS, Nguyen K, Aupperle K, et al: Apoptosis in rheumatoid arthritis: p53 overexpression in rheumatoid arthritis synovium. Am J Pathol 149:2143, 1996.
261. Tak PP, Smeets TJM, Boyle DL, et al: p53 overexpression in synovial tissue from patients with early and chronic rheumatoid arthritis. Arthritis Rheum 42:948, 1999.
262. Firestein GS, Echeverri F, Yeo M, et al: Somatic mutations in the p53 tumor suppressor gene in rheumatoid arthritis synovium. Proc Natl Acad Sci U S A 94:10895, 1997.
263. Reme T, Travaglio A, Gueydon E, et al: Mutations of the p53 tumour suppressor gene in erosive rheumatoid synovial tissue. Clin Exp Immunol 111:353, 1998.
264. Han Z, Boyle DL, Shi Y, et al: Dominant negative p53 mutations in rheumatoid arthritis. Arthritis Rheum 42:1088, 1999.
265. Aupperle KR, Boyle DL, Hendrix M, et al: Regulation of synoviocyte proliferation, apoptosis, and invasiveness by the p53 tumor suppressor gene. Am J Pathol 152:1091, 1998.
266. Cannons JL, Karsh J, Birnboim HC, Goldstein R: HPRT- mutant T cells in the peripheral blood and synovial tissue of patients with rheumatoid arthritis. Arthritis Rheum 41:1772, 1998.
267. Roivainen A, Soderstrom KO, Pirila L, et al: Oncoprotein expression in human synovial tissue: An immunohistochemical study of different types of arthritis. Br J Rheumatol 35:933, 1996.
268. Roivainen A, Jalava J, Pirila L, et al: H-ras oncogene point mutations in arthritic synovium. Arthritis Rheum 40:1636, 1997.
269. Kulka JP, Blocking D, Ropes MW, Bauer W: Early joint lesions of rheumatoid arthritis. Arch Pathol 59:129, 1955.
270. Peacock DJ, Banquerigo ML, Brahn E: Angiogenesis inhibition suppresses collagen arthritis. J Exp Med 175:1135, 1992.
271. Arsenault AL, Lhotak S, Hunter WL, et al: Taxol involution of collagen-induced arthritis: Ultrastructural correlation with the inhibition of synovitis and neovascularization. Clin Immunol Immunopathol 86:280, 1998.
272. Storgard CM, Stupack DG, Jonczyk A, et al: Decreased angiogenesis and arthritis in rabbits treated with an alphav beta3 antagonist. J Clin Invest 103:47, 1998.
273. Stevens CR, Blake DR, Merry P, et al: A comparative study by morphometry of the microvasculature in normal and rheumatoid synovium. Arthritis Rheum 34:1508, 1991.
274. Falchuk H, Goetzl J, Kulka P: Respiratory gases of synovial fluids. Am J Med 49:223, 1970.
275. Lund-Olesen K: Oxygen tension in synovial fluids. Arthritis Rheum 13:769, 1970.
276. Treuhaft PS, McCarty DJ: Synovial fluid pH, lactate, oxygen and carbon dioxide partial pressures in various joint diseases. Arthritis Rheum 14:475, 1971.
277. Jayson MIV, Dixon A St J: Intra-articular pressure in rheumatoid arthritis of the knee. Ann Rheum Dis 29:261, 1970.
278. Wallis WJ, Simkin PA, Nelp WB: Low synovial clearance of iodide provides evidence of
hypoperfusion in chronic rheumatoid synovitis. Arthritis Rheum 28:1096, 1985.
279. Kulka JP: Vascular derangement in rheumatoid arthritis. In Hill AGS (ed): Modern Trends in Rheumatology. London, Butterworths, 1961, p 49.
280. Dingle JTM, Page-Thomas DP: In vitro studies in human synovial membrane. A metabolic comparison of normal and rheumatoid disease. Br J Exp Pathol 37:318, 1956.
281. Shweiki D, Itin A, Soffer D, Keshet E: Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature 359:843, 1992.
282. Fava RA, Olsen NJ, Spencer-Green G, et al: Vascular permeability factor/endothelial growth factor (VPF/VEGF): Accumulation and expression in human synovial fluids and rheumatoid synovial tissue. J Exp Med 180:341, 1994.
283. Unemori EN, Ferrara N, Bauer EA, Amento EP: Vascular endothelial growth factor induces interstitial collagenase expression in human endothelial cells. J Cell Physiol 153:557, 1992.
284. Jackson JR, Minton JA, Ho ML, et al: Expression of vascular endothelial growth factor in synovial fibroblasts is induced by hypoxia and interleukin 1beta. Rheumatol 24:1253, 1997.
285. Koch AE, Halloran MM, Haskell CJ, et al: Angiogenesis mediated by soluble forms of E-selectin and vascular cell adhesion molecule-1. Nature 376:517, 1995.
286. Koch AE, Friedman J, Burrows JC, et al: Localization of the angiogenesis inhibitor thrombospondin in human synovial tissues. Pathobiology 61:1, 1993.
287. Kushner I, Somerville JA: Permeability of human synovial membrane to plasma proteins: Relationship to molecular size and inflammation. Arthritis Rheum 14:560, 1971.
288. Wallis WJ, Simkin PA, Nelp WB: Protein traffic in human synovial effusions. Arthritis Rheum 30:57, 1987.
289. Hale LP, Martin ME, McCollum DE, et al: Immunohistologic analysis of the distribution of cell adhesion molecules within the inflammatory synovial microenvironment. Arthritis Rheum 32:22, 1989.
290. Chuluyan HE, Issekutz AC: VLA-4 integrin can mediate CD11/CD18-independent transendothelial migration of human monocytes. J Clin Invest 92:2768, 1993.
291. Issekutz TB, Issekutz AC: T lymphocyte migration to arthritic joints and dermal inflammation in the rat: Differing migration patterns and the involvement of VLA-4. Clin Immunol Immunopathol 61:436, 1991.
292. Wahl SM, Allen JB, Hines KL, et al: Synthetic fibronectin peptides suppress arthritis in rats by interrupting leukocyte adhesion and recruitment. J Clin Invest 94:655, 1994.
293. Laffon A, Garcia-Vicuna R, Humbria A, et al: Upregulated expression and function of VLA-4 fibronectin receptors on human activated T cells in rheumatoid arthritis. J Clin Invest 88:546, 1992.
294. Van Dinther-Janssen AC, Pals ST, Scheper RJ, Meijer CJ: Role of the CS1 adhesion motif of fibronectin in T cell adhesion to synovial membrane and peripheral lymph node endothelium. Ann
Rheum Dis 52:672, 1993.
295. Elices MJ, Tsai V, Strahl D, et al: Expression and functional significance of alternatively spliced CS1 fibronectin in rheumatoid arthritis microvasculature. J Clin Invest 93:405, 1994.
296. Muller-Ladner U, Elices M, Kriegsmann JB, et al: Expression of the alternatively spliced CS-1 fibronectin isoform and its counter-receptor VLA-4 in rheumatoid arthritis synovium. J Rheumatol 24:1873, 1997.
297. Jorgensen C, Travaglio-Encinoza A, Bologna C: Human mucosal lymphocyte marker expression in synovial fluid lymphocytes of patients with rheumatoid arthritis. J Rheumatol 21:1602, 1994.
298. Koch AE, Burrows JC, Haines GK, et al: Immunolocalization of endothelial and leukocyte adhesion molecules in human rheumatoid and osteoarthritic synovial tissues. Lab Invest 64:313, 1991.
299. Corkill MM, Kirkham BW, Haskard DO, et al: Gold treatment of rheumatoid arthritis decreases synovial expression of the endothelial leukocyte adhesion receptor ELAM-1. J Rheumatol 18:1453, 1991.
300. Jorgensen C, Couret I, Canovas F, et al: Mononuclear cell retention in rheumatoid synovial tissue engrafted in severe combined immunodeficient (SCID) mice is up-regulated by tumour necrosis factor-alpha (TNF-alpha) and mediated through intercellular adhesion molecule-1 (ICAM-1). Clin Exp Immunol 106:20, 1996.
301. Jorgensen C, Couret I, Canovas F, et al: In vivo migration of tonsil lymphocytes in rheumatoid synovial tissue engrafted in SCID mice: Involvement of LFA-1. Autoimmunology 24:179, 1996.
302. Annefeld M: The potential aggressiveness of synovial tissue in rheumatoid arthritis. J Pathol 139:399, 1983.
303. Shiozawa S, Shiozawa K, Fujita T: Morphologic observations in the early phase of the cartilage-pannus junction. Arthritis Rheum 26:472, 1983.
304. Bromley M, Woolley DE: Histopathology of the rheumatoid lesion. Arthritis Rheum 27:857, 1984.
305. Bromley M, Woolley DE: Chondroclasts and osteoclasts at subchondral sites of erosion in the rheumatoid joint. Arthritis Rheum 27:968, 1984.
306. Brinckerhoff CE, Harris ED Jr: Collagenase production by cultures containing multinucleated cells derived from synovial fibroblasts. Arthritis Rheum 21:745, 1978.
307. Lerner UH, Jones IL, Gustafson GT: Bradykinin, a new potential mediator of inflammation-induced bone resorption. Arthritis Rheum 30:530, 1987.
308. Allard SA, Muirden KD, Maini RN: Correlation of histopathological features of pannus with patterns of damage in different joints in rheumatoid arthritis. Ann Rheum Dis 50:278, 1991.
309. Zvaifler NJ, Tsai V, von Kempis J, et al: Pannocytes: Distinctive cells found in articular cartilage erosions from the joints of patients with rheumatoid arthritis. Am J Pathol 150:1125, 1997.
310. Xue C, Takahashi M, Aono H, et al: Characterization of alternative fibroblast-like cells sharing the properties of fibroblasts and chondrocytes in RA pannus lesions. Arthritis Rheum 38:S344, 1995.
311. Harris ED Jr, Parker HG, Radin EL, Krane SM: Effects of proteolytic enzymes on structural and mechanical properties of cartilage. Arthritis Rheum 15:497, 1972.
312. Wolfe GC, MacNaul KL, Buechel FF, et al: Differential in vivo expression of collagenase messenger RNA in synovium and cartilage. Quantitative comparison with stromelysin messenger RNA levels in human rheumatoid arthritis and osteoarthritis patients and in two animal models of acute inflammatory arthritis. Arthritis Rheum 36:1540, 1993.
313. Cooke TD, Hurd ER, Jasin HE, et al: Identification of immunoglobulins and complement in rheumatoid articular collagenous tissues. Arthritis Rheum 18:541, 1975.
314. Kuiper S, Joosten LA, Bendele AM, et al: Different roles of tumour necrosis factor alpha and interleukin 1 in murine streptococcal cell wall arthritis. Cytokine 10:690, 1998.
315. Joosten LA, Helsen MM, van de Loo FA, van den Berg BW: Anticytokine treatment of established type II collage-induced arthritis in DBA/1 mice. A comparative study using anti-TNF-alpha, anti-IL-1alpha/beta, and IL-1Ra. Arthritis Rheum 39:397, 1996.
316. Rutter JL, Benbow U, Coon CI, Brinckerhoff CE: Cell-type specific regulation of human interstitial collagenase-1 gene expression by interleukin-1 beta (IL-1 beta) in human fibroblasts and BC-8701 breast cancer cells. J Cell Biochem 66:322, 1997.
317. Boyle DL, Han Z, Rutter J, et al: Post-transcriptional regulation of collagenase gene expression in synoviocytes by adenosine receptor stimulation. Arthritis Rheum 40:1772, 1997.
318. Yodlowski ML, Hubbard JR, Kispert J, et al: Antibody to interleukin 1 inhibits the cartilage degradative and thymocyte proliferative actions of rheumatoid synovial culture medium. J Rheumatol 17:1600, 1990.
319. Lotz M, Guerne PA: Interleukin-6 induces the synthesis of tissue inhibitor of metalloproteinases-1/erythroid potentiating activity (TIMP-1/EPA). J Biol Chem 266:2017, 1991.
320. Gunther M, Haubeck HD, van de Leur E, et al: Transforming growth factor beta 1 regulates tissue inhibitor of metalloproteinases-1 expression in differentiated human articular chondrocytes. Arthritis Rheum 37:395, 1994.
321. Unemori EN, Hibbs MS, Amento EP: Constitutive expression of a 92-kD gelatinase (type V collagenase) by rheumatoid synovial fibroblasts and its induction in normal human fibroblasts by inflammatory cytokines. J Clin Invest 88:1656, 1991.
322. Yang-Yen HF, Chambard JC, Sun VL, et al: Transcriptional interference between c-Jun and the glucocorticoid receptor: Mutual inhibition of DNA binding due to direct protein-protein interaction. Cell 62:1205, 1990.
323. Evanson JM, Jeffrey JJ, Krane SM: Human collagenase: Identification and characterization of an enzyme from rheumatoid synovium in culture. Science 158:499, 1967.
324. Firestein GS, Paine MM, Littman BH: Gene expression (collagenase, tissue inhibitor of metalloproteinases, complement, and HLA-DR) in rheumatoid arthritis and osteoarthritis synovium: Quantitative analysis and effect of intraarticular corticosteroids. Arthritis Rheum 34:1094, 1991.
325. Okada Y, Gonoji Y, Nakanishi I, et al: Immunohistochemical demonstration of collagenase and tissue inhibitor of metalloproteinases (TIMP) in synovial lining cells of rheumatoid synovium. Virchows Archiv B Cell Pathol 59:305, 1990.
326. Zvaifler NJ, Boyle D, Firestein GS: Early synovitis--synoviocytes and mononuclear cells. Semin Arthritis Rheum 23 (Suppl 2):11, 1994.
327. Lindy O, Konttinen YT, Sorsa T, et al: Matrix metalloproteinase 13 (collagenase 3) in human
rheumatoid synovium. Arthritis Rheum 40:1391, 1997.
328. Pendas AM, Balbin M, Llano E, et al: Structural analysis and promoter characterization of the human collagenase-3 gene (MMP13). Genomics 40:222, 1997.
329. Gravallese EM, Darling JM, Ladd AL, et al: In situ hybridization studies of stromelysin and collagenase messenger RNA expression in rheumatoid synovium. Arthritis Rheum 34:1076, 1991.
330. Mudgett JS, Hutchinson NI, Chartrain NA, et al: Susceptibility of stromelysin 1-deficient mice to collagen-induced arthritis and cartilage destruction. Arthritis Rheum 41:110, 1998.
331. Abbink JJ, Kamp AM, Nuijens JH, et al: Proteolytic inactivation of alpha 1-antitrypsin and alpha 1-antichymotrypsin by neutrophils in arthritic joints. Arthritis Rheum 36:168, 1993.
332. Vater CA, Mainardi CL, Harris ED Jr: An inhibitor of mammalian collagenases from cultures in vitro of human tendon. J Biol Chem 254:3045, 1979.
333. Stetler-Stevenson WG, Krutszch HC, Liotta LA: Tissue inhibitor of metalloproteinase (TIMP-2). A new member of the metalloproteinase inhibitor family. J Biol Chem 264:17,374, 1989.
334. Clark IM, Powell LK, Ramsey S, et al: The measurement of collagenase, tissue inhibitor of metalloproteinases (TIMP), and collagenase-TIMP complex in synovial fluids from patients with osteoarthritis and rheumatoid arthritis. Arthritis Rheum 36:372, 1993.
335. Firestein GS, Paine MM: Stromelysin and tissue inhibitor of metalloproteinases gene expression in rheumatoid arthritis synovium. Am J Pathol 140:1309, 1992.
336. Firestein GS, Paine MM, Boyle DL: Mechanisms of methotrexate action in rheumatoid arthritis. Selective decrease in synovial collagenase gene expression. Arthritis Rheum 37:193, 1994.
337. MacNaul KL, Chartrain N, Lark M, et al: Discoordinate expression of stromelysin, collagenase, and tissue inhibitor of metalloproteinases-1 in rheumatoid human synovial fibroblasts. Synergistic effects of interleukin-1 and tumor necrosis factor-alpha on stromelysin expression. J Biol Chem 265:17,238, 1990.
338. Morabito L, Montesinos MC, Schreibman DM, et al: Methotrexate and sulfasalazine promote
adenosine release by a mechanism that requires ecto-5 -nucleotidase-mediated conversion of adenine nucleotides. J Clin Invest 101:295, 1998.
339. Littman B, Schumacher R Jr, Boyle DL, et al: Tenidap decreases stromelysin gene expression in rheumatoid arthritis synovium: A synovial biopsy study. J Clin Rheumatol 3:194, 1997.
340. Muller-Ladner U, Gay RE, Gay S: Cysteine proteinases in arthritis and inflammation. Perspect Drug Discovery Design 6:87, 1996.
341. Lemaire R, Huet G, Zerimech F, et al: Selective induction of the secretion of cathepsins B and L by cytokines in synovial fibroblast-like cells. Br J Rheumatol 36:735, 1997.
342. Garnero P, Borel O, Byrjalsen I, et al: The collagenolytic activity of cathepsin K is unique among mammalian proteinases. J Biol Chem 273:32,347, 1998.
343. Hummel KM, Franz JK, Petrow P, et al: Cysteine proteinase cathepsin K mRNA is expressed in synovium of patients with rheumatoid arthritis and is detected at sites of synovial bone destruction. J Rheumatol 25:1887, 1998.
344. Esser RE, Angelo RA, Murphey MD, et al: Cysteine proteinase inhibitors decrease articular cartilage and bone destruction in chronic inflammatory arthritis. Arthritis Rheum 37:236, 1994.
345. Burastero SE, Casali P, Wilder RL, Notkins AL: Monoreactive high affinity and polyreactive low affinity rheumatoid factors are produced by CD5+ B cells from patients with rheumatoid arthritis. J Exp Med 168:1979, 1988.
346. Shimaoka Y, Attrep JF, Hirano T, et al: Nurse-like cells from bone marrow and synovium of patients with rheumatoid arthritis promote survival and enhance function of human B cells. J Clin Invest 102:606, 1998.
347. Fox DA, Smith BR: Evidence for oligoclonal B cell expansion in the peripheral blood of patients with rheumatoid arthritis. Ann Rheum Dis 45:991, 1986.
348. Clausen BE, Bridges SL Jr, Lavelle JC, et al: Clonally-related immunoglobulin VH domains and nonrandom use of DH gene segments in rheumatoid arthritis synovium. Mol Med 4:240, 1998.
349. Patel V, Panayi GS: Enhanced T helper cell function for the spontaneous production of IgM rheumatoid factor in vitro in rheumatoid arthritis. Clin Exp Immunol 57:584, 1984.
350. Reparon-Schuijt CC, van Esch WJ, van Kooten C, et al: Functional analysis of rheumatoid factor-producing B cells from the synovial fluid of rheumatoid arthritis patients. Arthritis Rheum 41:2211, 1998.
351. Rawson AJ, Hollander JL, Quismorio FP, Abelson NM: Experimental arthritis in man and rabbit dependent upon serum anti-immunoglobulin factors. Ann N Y Acad Sci 168:188, 1969.
352. Henney GS, Stanworth DR, Gell PGH: Demonstration of the exposure of new antigenic determinants following antigen-antibody combination. Nature 205:1079, 1965.
353. Paquet-Gioud M, Auvinet M, Raffin T, et al: IgG rheumatoid factor (RF), IgA RF, IgE RF, and IgG RF detected by ELISA in rheumatoid arthritis. Ann Rheum Dis 46:65, 1987.
354. McColl SR, Paquin R, Menard C, Beaulieu AD: Human neutrophils produce high levels of the interleukin 1 receptor antagonist in response to granulocyte/macrophage colony-stimulating factor and tumor necrosis factor alpha. J Exp Med 176:593, 1992.
355. Winchester RJ, Agnello V, Kunkel HG: Gamma globulin complexes in synovial fluids of patients with rheumatoid arthritis. Clin Exp Immunol 6:689, 1970.
356. Panush RS, Katz P, Longley S, Yonker RA: Detection and quantitation of circulating immune complexes in arterial blood of patients with rheumatic disease. Clin Immunol Pathol 36:217, 1985.
357. Ohno O, Cooke TD: Electron microscopic morphology of immunoglobulin aggregates and their interactions in rheumatoid articular collagenous tissues. Arthritis Rheum 21:516, 1978.
358. Shiozawa S, Jasin HE, Ziff M: Absence of immunoglobulins in rheumatoid cartilage-pannus junctions. Arthritis Rheum 23:816, 1980.
359. Zurier RB, Quagliata F: Effect of prostaglandin E1 on adjuvant arthritis. Nature 234:304, 1971.
360. Siegle I, Klein T, Backman JT, et al: Expression of cyclooxygenase 1 and cyclooxygenase 2 in human synovial tissue: Differential elevation of cyclooxygenase 2 in inflammatory joint diseases. Arthritis Rheum 41:122, 1998.
361. Anderson GD, Hauser SD, McGarity KL, et al: Selective inhibition of cyclooxygenase (COX)-2 reverses inflammation and expression of COX-2 and interleukin 6 in rat adjuvant arthritis. J Clin Invest 97:2672, 1996.
362. Elmgreen J, Haagen N, Ahnfelt-Ronne I: Enhanced capacity for release of leucotriene B4 by
neutrophils in rheumatoid arthritis. Ann Rheum Dis 46:501, 1987.
363. Griffiths RJ, Pettipher ER, Koch K, et al: Leukotriene B4 plays a critical role in the progression of collagen-induced arthritis. Proc Natl Acad Sci U S A 92:517, 1995.
364. Ruddy S, Colten HR: Rheumatoid arthritis. Biosynthesis of complement proteins by synovial tissues. N Engl J Med 290:1284, 1974.
365. Perlmutter DH, Goldberger G, Dinarello CA, et al: Regulation of class III major histocompatibility complex gene products by interleukin-1. Science 232:850, 1986.
366. Gulati P, Guc D, Lemercier C, et al: Expression of the components and regulatory proteins of the classical pathway of complement in normal and diseased synovium. Rheumatol Int 14:13, 1994.
367. Pekin TJ Jr, Zvaifler NJ: Hemolytic complement in synovial fluid. J Clin Invest 43:1372, 1964.
368. Sabharwal UK, Vaughan JH, Fong S, et al: Activation of the classical pathway of complement by rheumatoid factors. Assessment by radioimmunoassay for C4. Arthritis Rheum 25:161, 1982.
369. Elmgren J, Hansen TM: Subnormal sensitivity of neutrophils to complement split-product C5a in rheumatoid arthritis: Relation to complement catabolism and disease extent. Ann Rheum Dis 44:514, 1985.
370. Goldstein IM, Weissmann G: Generation of C5-derived lysosomal enzyme-releasing activity (C5a) by lysates of leukocyte lysosomes. J Immunol 113:1583, 1974.
371. Goodfellow RM, Williams AS, Levin JL, et al: Local therapy with soluble complement receptor 1 (sCR1) suppresses inflammation in rat mono-articular arthritis. Clin Exp Immunol 110:45, 1997.
Related Documents
