In rheumatoid arthritis (RA), periarticular destruction of cartilage and bone occurs in response to local invasion of the joint by an expanding tumor-like rheumatoid synovium (1, 2). Inflammatory cytokines produced by the invading synovium, such as tumor necrosis factor α (TNFα), interleukin-1β (IL-1β), and IL-6, are primary mediators of joint inflammation and articular destruction in RA (1, 2). Indeed, inhibition of cytokine action has proved to be an effective target for new methods of therapeutic modalities, as evidenced by the successful use of TNFα blocking agents in the prevention of joint inflammation in RA (1, 2).
Parathyroid hormone–related protein (PTHrP) is a peptide which, like TNFα, was first identified because of its biologic effects in the setting of malignancy. When produced in prodigious amounts by tumors, PTHrP, by virtue of its ability to bind to and activate the G protein–coupled PTH/PTHrP receptor in bone, is the humoral factor responsible for marked bone resorption and hypercalcemia in this setting (3). However, in the absence of malignancy, PTHrP is produced in many organs, where it acts locally, rather than systemically (3). For example, in response to inflammatory stimuli, PTHrP expression has been demonstrated to increase in the liver, where it activates the hepatic acute-phase response (4).
More recently, PTHrP has also been identified as a member of the cascade of cytokines produced in prodigious amounts by the rheumatoid synovium (5, 6). As with other cytokines, PTHrP is expressed by synoviocytes, the primary cell comprising the tumor-like rheumatoid synovium (5–9). TNFα and IL-β, 2 key inflammatory cytokines in RA, induce synoviocyte PTHrP expression, and PTHrP, in turn, induces synoviocyte secretion of IL-6 (5). IL-6, a cytokine that mediates joint destruction and inflammation in animal models of arthritis (10), is also a critical mediator of PTH/PTHrP peptide–induced bone resorption (11, 12). Moreover, continuous exposure of bone to PTH/PTHrP peptides is a potent stimulus for TRANCE/receptor activator of nuclear factor κB (RANK)–mediated osteoclastic bone resorption (13), a pathway previously demonstrated to mediate destruction of cartilage and bone in arthritis (14). In addition, studies demonstrating PTH/PTHrP inhibition of chondrocyte collagen synthesis and extracellular inorganic pyrophosphate secretion (15–17) have identified PTH/PTHrP peptide–mediated pathways that might also contribute to cartilage destruction. These facts led us to postulate that the local increase in synovial PTHrP expression in RA may be a critical mediator of bone and cartilage destruction and that, by virtue of its induction by TNFα and its ability to induce IL-6 secretion, PTHrP may also contribute to joint inflammation in this setting.
To test these hypotheses, we used a well-described animal model of RA, streptococcal cell wall (SCW)–induced arthritis in female Lewis rats (18). In this model, intraperitoneal administration of purified Streptococcus group A–derived peptidoglycan polysaccharide results in the preferential deposition of these complexes in joints, liver, and spleen (18–21). Within the joint, SCW initiates an inflammatory response that ultimately leads to synovial tissue proliferation and joint erosion by invading pannus, a histopathologic lesion very similar to that seen in RA (18–21). Moreover, the inflammatory cytokines and T cells that mediate joint destruction in RA are similarly responsible for joint destruction in SCW arthritis (20–23). Inflammatory cytokines, such as TNFα, are also involved in the granulomatous response that develops in the liver and spleen in response to SCW deposition (24, 25). Initial experiments were therefore undertaken to determine whether PTHrP, analogous to its inducible expression in the human rheumatoid synovium, was also expressed by the inflammatory synovium in SCW arthritis. After establishing the suitability of this model for testing the role of endogenous PTHrP production in joint destruction, treatment trials using a PTHrP 1–34 neutralizing antibody were then initiated to determine whether joint inflammation and/or joint destruction could be prevented by blocking the effects of PTHrP. As a secondary outcome, the effects of PTHrP antibody treatment on the inflammatory granulomatous response in liver and spleen were also evaluated.
- Top of page
- MATERIALS AND METHODS
Two novel biologic effects of PTHrP are demonstrated in this study: PTHrP mediation of joint destruction in inflammatory arthritis, and PTHrP regulation of granuloma formation and neutrophil function. The first effect of PTHrP appears to be analogous to the well-described ability of PTHrP to mediate osteolysis in metastatic bone disease, while the second effect of PTHrP represents an entirely new aspect of PTHrP bioactivity that would not be easily predicted from our current understanding of this multifunctional hormone. Indeed, because neutrophils are believed to contribute to bone loss in other types of inflammation and to mediate cartilage destruction in RA (40, 41), it is possible that normalization of circulating neutrophil counts by PTHrP blockade may also have contributed to the protective effect of this treatment on cartilage and bone during inflammation.
The joint protective effects of PTHrP blockade in SCW arthritis are similar to those reported by Guise et al (42) for an animal model of breast metastases wherein PTHrP neutralizing antibody treatment prevented osteolytic bone destruction in a setting of localized, but not systemic, increases in PTHrP expression by invading tumor metastases. Similarly, in the SCW model, because the inflammatory pannus that invades cartilage and bone is composed of PTHrP-positive cells, while systemic PTHrP levels are not elevated, it would appear that locally produced PTHrP within the tumor-like, invading pannus mediates osteolytic bone resorption as well as cartilage destruction. Because the TRANCE/RANK pathway has previously been demonstrated to mediate the osteolytic bone loss occurring both in malignancy and arthritis, as well as PTH/PTHrP-peptide–dependent bone loss (13, 14), these findings suggest that PTHrP induction of TRANCE and the subsequent induction of osteoclastogenesis may be an important mechanism of bone loss in both of these disease states.
In contrast to its effect in preventing destruction of cartilage and bone, PTHrP blockade had no effect on joint inflammation. As with any negative finding, it is impossible to rule out insufficient antibody dosing or delivery as the cause of this negative result. However, PTHrP antibody treatment did prevent other aspects of SCW-induced joint pathology. Moreover, other similarly designed studies have demonstrated the ability of systemically administered antibodies to distribute to joints and to prevent joint inflammation (37, 38, 43). Therefore, these results strongly suggest that endogenous PTHrP production does not mediate joint inflammation.
The ability of PTHrP to mediate joint destruction while not affecting joint inflammation is similar to the reported effect of IL-1 in arthritis, since IL-1 blockade in animal arthritis models also predominantly prevents joint destruction rather than inflammation (44). Because IL-1 induces PTHrP expression in synoviocytes (5), it is therefore possible that PTHrP mediates some of the joint-destructive effects of IL-1 in inflammatory arthritis. Alternatively, because PTHrP has previously been shown to synergize with TNF or IL-1 in destroying bone (45, 46), it is also possible that the partially protective effect of PTHrP antibody treatment, which prevented 30–60% of all measures of bone or cartilage destruction, is due to synergistic or additive effects of PTHrP with other inflammatory cytokines, such as IL-1β or TNFα.
The ability of PTHrP blockade to prevent SCW-induced granuloma formation in both liver and spleen was an unanticipated finding in these studies. Because the granulomatous response serves a protective function in walling off foreign bodies, stimulation of granuloma formation by endogenous PTHrP should be inherently beneficial. Conversely, inhibition of granuloma formation could lead to increased morbidity. However, in this noninfectious model, liver function was normal in SCW animals with granulomatous livers and remained unchanged with PTHrP antibody treatment (data not shown). Growing evidence suggests that PTHrP may be involved in the pathogenesis of multiple types of granulomatous disorders, because its increased expression has now been documented in sarcoidosis, schistosomiasis, giant cell granulomas, and idiopathic systemic granulomatous disease (47–50). However, to our knowledge, the results presented here are the first to identify a function for PTHrP in this pathologic process.
Because neutrophils are an important component of SCW granuloma formation (19), the evidence presented here would suggest that inhibition of neutrophil accumulation in the liver and spleen contributes to the protective effect of PTHrP blockade. Indeed, because perivascular epithelioid cell accumulation precedes neutrophil influx and SCW granuloma formation in the liver and spleen (19, 20), we hypothesized that PTHrP expression by the vasculature and/or epithelioid cells during early inflammation could induce the chemotaxis of PTH/PTHrP receptor–positive neutrophils to sites of SCW deposition in these organs. Consistent with this hypothesis, PTHrP 1–34, the peptide neutralized in our in vivo SCW studies, stimulated the chemotaxis of neutrophils in vitro. To our knowledge, neither PTH/PTHrP receptor expression nor PTHrP 1–34–stimulated chemotaxis has previously been reported in neutrophils. However, G protein–coupled receptors have been identified as critical mediators of neutrophil chemotaxis (51), and chemotaxic effects of PTH/PTHrP 1–34 peptides have been demonstrated in other cell types (52, 53). Indeed, Halstead et al (53) have reported a maximal response of osteoblasts to the chemotaxic effect of low-dose (10−10M) PTH 1–34 that is similar to the concentration-dependent effects demonstrated here for PTHrP 1–34 and neutrophils.
The apparent decrease in neutrophil influx into the livers and spleens of animals treated with PTHrP antibody may also be related to the ability of PTHrP blockade to prevent SCW-induced neutrophilia. While the exact mechanisms contributing to neutrophilia in SCW arthritis are not known, previous demonstrations of PTH 1–34 stimulation of granulocyte-macrophage colony-stimulating factor production are consistent with possible stimulatory effects of PTHrP 1–34 on neutrophil production (54, 55).
PTHrP is a multifunctional peptide whose role in inflammation is only beginning to be understood (3, 56). Previous studies conducted in our laboratory suggest that its function in this setting is dependent on the type of inflammatory stimulus and the site of PTHrP induction (56, 57). Thus, just as induction of PTHrP in the liver during endotoxemia can stimulate the hepatic acute-phase response (4), or induction of PTHrP in the cerebrovasculature during ischemia may prevent neuronal death by enhancing local blood flow (57), induction of this bone-resorbing peptide within the joint during inflammatory arthritis may contribute to periarticular joint destruction and previously unreported effects of PTHrP on neutrophil function may contribute to local granuloma formation.
In conclusion, the data presented here demonstrate that PTHrP-mediated osteolytic bone destruction is not only an important component of metastatic bone disease (42, 58), but that localized PTHrP expression can also significantly contribute to joint destruction in inflammatory arthritis. Moreover, because inhibition of PTHrP 1–34 activity blocked degradation of cartilage and bone during established arthritis in this animal model, PTHrP may also be a potential target for therapeutic interventions in patients with RA. At the same time, a cautionary note must also be sounded in light of the other striking finding in this study: the critical role of PTHrP in regulating granuloma formation in both the liver and the spleen. Just as reports have documented an increase in disseminated tuberculosis in RA patients treated with TNFα neutralizing agents (59), this same risk could also accompany the use of PTHrP blocking agents. At the same time, the identification of novel PTHrP actions in regulating the granulomatous response and/or neutrophil function during inflammation also provides fresh insights into the multifunctional effects of PTH/PTHrP peptides as they become available for clinical use.