Interleukin-1beta (IL-1beta) and tumor necrosis factor-alpha (TNF-alpha) synergistically induce and sustain arthritis. Two competing hypotheses of arthritis induction are 1) that TNF preferentially mediates inflammation, whereas IL-1 impels bone destruction, or 2) that either cytokine controls the entire process. In this study, these propositions were tested in two experiments by instilling IL-1beta or TNF-alpha into one knee of Lewis rats (n = 6/group) to incite arthritis, after which semiquantitative scores for inflammation, bone resorption, osteoclasts, and cartilage integrity were acquired. In the induction study, IL-1beta or TNF-alpha (3, 10, or 30 micro g) was given once to incite arthritis. After 2 days, IL-1beta induced significant, dose-dependent increases in inflammation (mild to marked), bone resorption (minimal to moderate), and osteoclasts (minimal to moderate). In contrast, TNF-alpha induced minimal to mild inflammation but had little impact on resorption or osteoclasts. Both IL-1 and TNF (>/=10 micro g) yielded mild cartilage degeneration. Most lesion scores in TNF-treated rats were significantly lower than those in animals given the same dose of IL-1beta. In the persistence study, rats were injected once with IL-1 or TNF (10 micro g) and maintained for 2, 3, or 7 days. IL-1beta significantly enhanced inflammation (all 3 days), bone resorption (days 2 and 3), osteoclasts (days 2 and 3), and cartilage matrix loss (days 2 and 3), whereas TNF-alpha augmented inflammation (days 2 and 3) and cartilage degeneration (day 2) but not bone resorption or osteoclasts. Thus, both IL-1beta and TNF-alpha can launch inflammation, but IL-1beta drives skeletal destruction.

Annals of the New York Academy of SciencesVolume 987, Issue 1 p. 295-298 IL-1β and TNF-α Produce Divergent Acute Inflammatory and Skeletal Lesions in the Knees of Lewis Rats GIUSEPPE CAMPAGNUOLO, GIUSEPPE CAMPAGNUOLO Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USA G.C. and B.B. are equal contributors.Search for more papers by this authorBRAD BOLON, BRAD BOLON Department of Pathology, Amgen Inc., Thousand Oaks, California 91320-1789, USA G.C. and B.B. are equal contributors.Search for more papers by this authorL.I. ZHU, L.I. ZHU Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USASearch for more papers by this authorDIANE DURYEA, DIANE DURYEA Department of Pathology, Amgen Inc., Thousand Oaks, California 91320-1789, USASearch for more papers by this authorULRICH FEIGE, ULRICH FEIGE Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USASearch for more papers by this authorDEBRA ZACK, Corresponding Author DEBRA ZACK Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USAAddress for correspondence: Dr. Debra Zack, Amgen, One Amgen Center Drive, M/S 29-M-B, Thousand Oaks, CA 91320-1799. Voice: 805-447-6684; fax: 805-480-1329. dzack@amgen.comSearch for more papers by this author GIUSEPPE CAMPAGNUOLO, GIUSEPPE CAMPAGNUOLO Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USA G.C. and B.B. are equal contributors.Search for more papers by this authorBRAD BOLON, BRAD BOLON Department of Pathology, Amgen Inc., Thousand Oaks, California 91320-1789, USA G.C. and B.B. are equal contributors.Search for more papers by this authorL.I. ZHU, L.I. ZHU Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USASearch for more papers by this authorDIANE DURYEA, DIANE DURYEA Department of Pathology, Amgen Inc., Thousand Oaks, California 91320-1789, USASearch for more papers by this authorULRICH FEIGE, ULRICH FEIGE Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USASearch for more papers by this authorDEBRA ZACK, Corresponding Author DEBRA ZACK Department of Inflammation, Amgen Inc., Thousand Oaks, California 91320-1789, USAAddress for correspondence: Dr. Debra Zack, Amgen, One Amgen Center Drive, M/S 29-M-B, Thousand Oaks, CA 91320-1799. Voice: 805-447-6684; fax: 805-480-1329. dzack@amgen.comSearch for more papers by this author First published: 24 January 2006 https://doi.org/10.1111/j.1749-6632.2003.tb06066.x Read the full textAboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Volume987, Issue1IMMUNE MECHANISMS AND DISEASEApril 2003Pages 295-298 RelatedInformation

Daily osteoprotegerin (OPG) injection for 7 or more days prevents bone loss for 3 weeks in rats with adjuvant-induced arthritis (AdA). The present experiments defined the duration of bone protection in AdA provided by a single OPG bolus. Male Lewis rats received OPG at the onset or peak of clinical disease, after which bone mineral density (BMD), erosions, and osteoclasts were evaluated. An OPG bolus (4 mg/kg subcutaneously) at onset eliminated osteoclasts, preserved BMD for 7 days, and prevented bone erosions for 4 days. In contrast, an OPG bolus (1, 3, 10, or 30 mg/kg intravenously) given at the peak of disease eradicated osteoclasts in a dose-dependent manner but had no impact on bone integrity due to extensive pre-existing bone loss. These data indicate that one OPG injection will inhibit joint erosions for several days, and confirm that bone-sparing therapy must be initiated early in disease to protect joint integrity.

Bone growth, maintenance, and repair during normal and pathologic skeletal remodeling are mediated by the reciprocal activities of osteoblasts and osteoclasts. Osteoblasts, which originate from mesenchymal stem cells in bone marrow, produce bone matrix and express receptors for most bone-regulating molecules (regardless of whether the predominant action is to drive bone formation or bone resorption) (1). Osteoclasts, which are derived from phagocytic precursors of the monocyte/macrophage lineage, degrade bone matrix. Normal skeletal remodeling is tightly controlled via cell-to-cell communication between osteoclasts and osteoblasts, with osteoblasts both supporting and regulating osteoclast activity (2). However, in pathologic states, "activated" cells (e.g., infiltrating leukocytes, synovial fibroblasts) contribute other molecules that shift the balance between osteoblastic and osteoclastic activities (3-5). The chief intercellular signaling pathway that dictates bone remodeling is mediated by 3 members of the tumor necrosis factor (TNF) and TNF receptor (TNFR) superfamily: receptor activator of nuclear factor κB (NF-κB) (RANK), a membrane-bound osteoclast receptor that initiates osteoclastic bone resorption after binding RANK ligand (RANKL); RANKL, the main osteoclast-stimulating factor (also known as osteoprotegerin ligand [OPGL]); and OPG, the key inhibitor of bone resorption because of its function as a soluble, nonsignaling decoy receptor for RANKL. Thus, the balance between OPG and RANKL in the bone microenvironment regulates bone resorption. The biology of the OPG/RANK/RANKL signaling system has been described in several recent reviews (6-10). The present analysis emphasizes the clinical opportunities that might derive from targeting components of this pathway. In particular, OPG supplementation is considered a potent and specific disease-modifying agent for preserving bone in patients with rheumatoid arthritis (RA) and/or other bone-destructive diseases. The interactions between OPG, RANK, and their common ligand OPGL/RANKL are shown in Figure 1. These functions have been elucidated both in vitro, using human and mouse bone marrow cultures, and in vivo, through the construction of genetically engineered mice or therapeutic interventions that introduce recombinant proteins. Details of these experiments are recapitulated briefly in the text below. In doing so, the term OPGL has been used instead of RANKL, to emphasize the therapeutic potential of OPG. Schematic representation of the osteoprotegerin (OPG)/OPG ligand (OPGL)/receptor activator of nuclear factor κB (RANK) signaling pathway in rheumatoid arthritis. OPGL, expressed by activated T cells and synovial fibroblasts as well as bone marrow stromal cells under the influence of many proresorptive stimuli, binds to its specific membrane-bound receptor RANK, thereby triggering a network of kinase cascades (see Figure 3) that promotes osteoclast differentiation, activation, and survival. OPG down-regulates osteoclastic resorption by acting as a soluble sink for OPGL. M-CSF = macrophage colony-stimulating factor; CFU-M = colony-forming unit–macrophage; mϕ = macrophage; LIF = leukemia inhibitory factor; PTH = parathyroid hormone; IL-11 = interleukin-11; PTHrP = PTH-related protein; PGE2 = prostaglandin E2; TNF = tumor necrosis factor. RANK mediates all known actions of OPGL. RANK is expressed by osteoclasts and their precursors (11) as well as by dendritic cells and mature T cells (12). Human RANK (616 amino acids [aa], MW 97 kd) is 70% homologous to the mouse form at the amino acid level (13). Manipulations that either delete cellular RANK or provide an excess of soluble RANK (to prevent binding of OPGL to the cellular receptor) down-regulate osteoclast production. For example, in bone marrow cultures osteoclastogenesis is blocked by adding antagonistic antibodies directed against the extracellular, ligand-binding portion of RANK (11). Injection of RANK-Fc (a soluble, nonsignaling fusion protein incorporating the OPGL-binding domain of RANK and the constant [Fc] region of human IgG1) into normal mice induces accumulation of cancellous bone adjacent to growth plates (11). In contrast, transgenic overexpression of RANK-Fc (11) or deletion of cellular RANK (RANK−/−) in genetically engineered mice (14) leads to marked osteopetrosis (Table1 and Figure 2). Osteoclast precursors of RANK−/− mice have an inherent defect in osteoclast development (14), as indicated by their inability to form osteoclasts when treated with OPGL and macrophage colony-stimulating factor (M-CSF), a proliferation and survival factor for osteoclast precursors (15). Interestingly, gene therapy that reintroduces RANK to spleen cells from RANK−/− mice restores their ability to form osteoclasts in vitro in response to OPGL and M-CSF (15). Furthermore, transplantation of the reconstituted cells into osteopetrotic RANK−/− mice results in formation of functional osteoclasts and significant reduction in the excess bone mass (14). Radiographs showing osteopetrosis and osteoporosis in genetically engineered mice. Mice with an engineered change in the OPG/OPGL/RANK signaling pathway exhibit osteopetrosis (indicative of a high OPG:OPGL ratio and osteoclast depletion) or osteoporosis (resulting from a low OPG:OPGL ratio and osteoclast expansion), shown here in the femur (angled bone crossing the center) and tibia of young adults. The entire skeleton is affected if the mutation is active throughout gestation and postnatal development, but effects are limited to regions adjacent to growth plates (arrowheads) if the molecule is given beginning in adulthood (RANK-Fc injection schedule: 5 mg/kg/day subcutaneously for 5 days). Radiographs were obtained using a cabinet x-ray system (model 43855A; Faxitron X-ray, Buffalo Grove, IL) set at 0.3 mA and 55 kV for 49 seconds. See Figure 1 for definitions. The transmembrane protein OPGL partners with M-CSF as the necessary and sufficient signals that regulate osteoclast differentiation, activation, and survival (15, 16). Synonyms for OPGL include ODF (osteoclast differentiation factor) (16) and TRANCE (TNF-related activation-induced cytokine) (17). Soluble OPGL (26 kd) is a homotrimer that is produced by proteolytic cleavage of the full-length membrane-bound protein (317 aa, MW 45 kd). Using mouse OPGL (316 aa), this conversion has been shown to be mediated by the metalloprotease TNFα-convertase enzyme (TACE) (18); to our knowledge, the interaction of TACE with human OPGL has not been reported. However, although TACE cleaves OPGL in the correct position, the specificity constant for OPGL is 1,000-fold lower than the constant for TNFα, suggesting that an as yet undiscovered metalloprotease with similar substrate specificity may represent the principal pathway for processing OPGL (18). Both the soluble and membrane-associated forms of OPGL are produced by CD4+ and CD8+ T cells, and both are potent RANK agonists (19, 20). Human OPGL is 87% homologous to the mouse variant at the amino acid level (15). OPGL is highly expressed in bone, intestine, and peripheral lymphoid organs (15). In adults, the expression of OPGL in normal bone is highest near remodeling zones such as the hypertrophic chondrocytes of growth plates, the primary spongiosa, the metaphysis, and the periosteal side of the diaphyseal collar (15). Incubation of cultured bone marrow cells with OPGL, either with or without M-CSF, results in the generation of myriad multinucleated cells that express the osteoclast marker protein tartrate-resistant acid phosphatase (15). OPGL results in rapid activation (within 3 hours) of mature osteoclasts, as shown by the swift increase in serum calcium levels after injection of recombinant protein (15). OPGL also promotes osteoclast survival by inhibiting apoptosis (21). In animal models, OPGL deficiency induces osteopetrosis as a consequence of essentially universal osteoclast extinction (Table 1 and Figure 2). OPGL null mutant (OPGL−/−) mice exhibit a phenotype that is nearly indistinguishable from that of RANK−/− animals, including profound osteopetrosis with both bone deformity and failed tooth eruption (22, 23). However, unlike RANK deficiency (14), ablation of OPGL does not generate an inherent defect in osteoclasts, because in vitro treatment of OPGL−/− osteoclast precursors with OPGL and M-CSF results in the production of osteoclasts (22). The glycoprotein OPG is a unique TNFR superfamily member because it lacks the usual hydrophobic membrane-spanning sequence (24, 25), thereby resulting in its secretion as a soluble decoy receptor for OPGL. Thus, OPG functions as an endogenous antagonist for the RANK signal transduction pathway. Synonyms for OPG include OCIF (osteoclastogenesis inhibitory factor) (24), FDCR-1 (follicular dendritic cell receptor) (26), and TR1 (TNFR superfamily member 1) (27). Human OPG (401 aa, MW 44 kd), which is increased in size by posttranslational N-glycosylation to 55 kd or 110 kd, respectively, for monomeric and dimeric forms, is ∼85% homologous to the rat form at the amino acid level (25). OPG is secreted predominantly as the disulfide-linked homodimer, although monomeric OPG is produced to some extent (25). Monomeric OPG is also formed by limited proteolysis in plasma (28). Both forms of OPG are cleaved during processing, giving rise to a mature protein of 380 aa (28). The OPGL-binding domains on OPG are located in 4 highly conserved, cysteine-rich, TNFR-like regions located near the N-terminus. The C-terminus contains a heparin-binding domain as well as 2 death domain homologous (DDH) regions (29). The functions of the DDH regions in OPG have not been well characterized, but, because of their extracellular location, they are considered unlikely to regulate cell death. OPG can exist in a membrane-bound form, likely as a result of binding to cell surface heparin (26). In humans, OPG is expressed in virtually all human tissues (including bone), except peripheral blood lymphocytes (24). In genetically engineered animals, excessive expression of OPG results in osteopetrosis secondary to the depletion of osteoclasts, and OPG deficiency is characterized by osteoporosis resulting from enhanced osteoclast numbers (Table 1 and Figure 2). Transgenic mice engineered to overexpress rat OPG are markedly osteopetrotic at birth (25, 30). This phenotype increases in severity with age but is not associated with bone deformities or defective tooth eruption (25, 30). The extent of osteopetrosis is highly correlated with circulating OPG concentrations; 100% of mice are affected if serum OPG levels exceed 10 ng/ml (normal ≤1–2) (25). Bones of OPG-transgenic mice, although lacking osteoclasts, have normal populations of macrophages, indicating that the defect in cells of the osteoclast lineage develops after differentiation of their precursors. As predicted, the skeletal phenotype of OPG knockout (OPG−/−) mice (marked progressive osteoporosis with reduced tensile strength and numerous pathologic fractures) (31) is the reverse of that seen in OPG-transgenic mice. These effects are the consequence of systemic increases in osteoclast numbers. Interestingly, heterozygous mice lacking 1 OPG copy (OPG+/−) also experience a significant decrease in bone mass over time (31), indicating that diminished OPG expression is sufficient to cause substantial bone loss. The osteoporotic phenotype of OPG−/− mice can be reversed by providing exogenous OPG, either by crossbreeding with transgenic mice that overexpress OPG or through injection of recombinant OPG (30). Thus, OPG supplementation can directly counteract a deficit of endogenous OPG. As with other TNFR family members, signal transduction by RANK is mediated through an intracellular kinase cascade that controls gene transcription (Figure 3). Signaling by TNFR proteins progresses by 2 pathways: those that contain cytoplasmic death domains and can directly induce apoptosis, and those (such as RANK) that do not. The cytoplasmic tail of RANK has binding sites for TNFR-associated factors (TRAF) 2, 5, and 6. Both TRAF2 and TRAF5 are involved in osteoclastogenesis, while TRAF6 controls both the formation and activation of osteoclasts (for review, see ref. 8). The RANK/TRAF complexes have been reported to mediate further amplification of OPGL signaling via the serine/threonine kinase/Akt pathway, the extracellular signal–regulated kinase pathway, the inhibitor of NF-κB kinase pathway, the c-Jun N-terminal kinase pathway, and the p38 kinase pathway (8, 32-34). Activation of these cascades by RANK is important for osteoclast differentiation and function. Schematic diagram of osteoclast differentiation, activation, and survival. Stimulation of cellular RANK (white rectangles) by OPGL enables TNF receptor–associated factors (TRAFs) to associate with RANK. Each RANK/TRAF pair triggers 1 or more intracellular signal transduction pathways, including the serine/threonine kinase/Akt, extracellular signal-regulated protein kinase (ERK), inhibitor of nuclear factor κB (NF-κB) kinase (IKK), c-Jun N-terminal kinase (JNK), and p38 mitogen-activated protein kinase (MAPK) systems. These cascades subsequently interact with 1 or more transcription factors (TFs) to initiate gene transcription. As shown in the right quadrant of this figure, binding of OPGL by OPG blocks RANK-mediated signal transduction, thereby resulting in osteoclast apoptosis. AP-1 = activator protein 1; ATF-2 = activating transcription factor 2 (see Figure 1 for other definitions). The significance of these pathways in RANK-mediated osteoclast activity is shown by the presence of osteopetrosis in mice with null mutations of TRAF6, the Src tyrosine kinase (a binding partner of TRAF6), and the RANK-activated transcription factors NF-κB and activator protein 1 (Table 1). Importantly, the cellular receptors for proinflammatory cytokines with osteoclastogenic activity, including interleukin-1 (IL-1) and TNFα, also communicate via the TRAF cascade (8). In fact, TNFα appears to couple the OPGL/RANK and TNFRI pathways to stimulate osteoclastogenesis (35). Thus, the signaling pathways involved in osteoclast differentiation and activation by OPGL and other osteoclast-modulating cytokines are intricately networked. OPGL promotes the differentiation, activation, and survival of osteoclasts. By binding OPGL, OPG inhibits differentiation of new osteoclasts while preventing the activation and persistence of previously formed cells. With respect to differentiation, OPG blocks osteoclastogenesis in mouse bone marrow (36) and human peripheral blood mononuclear cells (37) cultured with OPGL and M-CSF, and it essentially eliminates their production in severely inflamed joints of rats with adjuvant arthritis (38). Interestingly, OPG obstructs OPGL-induced fusion of osteoclast precursors (39) but does not reduce the number of precursors (25). Activation of mature osteoclasts is blocked by OPG in vitro (15, 40, 41). Injection of recombinant OPG into normal and hypercalcemic mice reduces circulating calcium levels (which depend on degradation of bone matrix by mature osteoclasts) within 2 hours (42, 43), a time course that can be explained only by direct inhibition of mature cells. Finally, OPG greatly reduces osteoclast survival in vitro by counteracting the capacity of OPGL to inhibit osteoclast apoptosis (21, 44). More mature osteoclasts are particularly sensitive to OPG (44). Significant osteoclast apoptosis requires the presence of OPG for from 1 (40, 45) to 3 days (21, 44). However, in mice and rats, a single OPG injection in vivo induces substantial osteoclast apoptosis within 6 hours (21), leading to a sizable decrease in osteoclast numbers within 12 to 48 hours (44, 46). Thus, OPG simultaneously thwarts osteoclastogenesis, foils osteoclast activation, and forces many cells to self-destruct. These 3 actions of OPG provide "triple-threat" potential for preventing pathologic destruction of bone. Despite its pleiotropic effects on osteoclasts, OPG induces few obvious effects outside the skeletal system. However, recent publications detailing phenotypes in nonosseus tissues show that molecules of this signaling pathway also have fundamental extraskeletal roles. Surprisingly, OPG appears to have several roles in vascular biology. OPG is an NF-κB–dependent endothelial cell survival factor (47) and thus may modulate capillary homeostasis. OPG also functions to maintain arterial wall integrity, as OPG knockout mice develop progressive calcification of the tunica media in large arteries (31); this arterial layer normally expresses OPG (25) but not OPGL or RANK (30). Calcified foci express OPGL and RANK but are not associated with atherosclerotic plaques. The lesions cannot be reversed by injecting OPG into young mice but can be reversed by crossbreeding OPG knockout mice with OPG-transgenic mice (30). A link between bone resorption and arterial wall mineralization under the control of OPG has been confirmed in warfarin-treated rats (48). Osteoporosis has been associated with vascular calcification in postmenopausal women (49), although mineralization in this patient population typically occurs in the tunica intima adjacent to atherosclerotic plaques. Thus, although a link between bone resorption and arterial wall mineralization, regulated by OPG, has been confirmed, the ability of OPG to block vascular calcification in humans remains to be determined. Another recent discovery is that OPGL is an essential mammary gland morphogen. Pregnant mice lacking OPGL or RANK fail to form lobulo-alveolar mammary glands (50). The defect, which results from diminished OPGL-mediated signaling on RANK-expressing mammary epithelium, can be reversed by therapy with recombinant OPGL. The implications in the human clinical setting (if any) are unknown. Finally, OPG—by its ability to sequester OPGL—has immunoregulatory activity. OPGL regulates B cell and T cell differentiation, directs T cell activation, augments cooperative interactions between T cells and dendritic cells, and controls dendritic cell survival (for review, see refs. 12 and51). Interestingly, after treatment with OPG, antigen-sensitized mice increase their production of specific IgM, IgG, and IgE antibodies; this effect is observed regardless of whether the response depends on T cell control (Stolina M: unpublished observations). OPG does not affect proliferation of cultured mouse B cells, nor does it affect the progression of cell-mediated disease in mouse models of contact hypersensitivity or granulomatous inflammation (Stolina M: unpublished observations). It has been reported that OPG can also bind to TNF-related apoptosis-inducing ligand (TRAIL), which acts on activated T cells and antigen-primed mature T cells, with OPG acting as a decoy receptor for TRAIL during immune responses and TRAIL modulating OPG-induced osteoclast inhibition (52). The implications, if any, of this immunoregulatory capacity with respect to OPG intervention in human patients with inflammatory disease have yet to be explored. However, arthritic rats treated with OPG exhibit neither an exacerbation nor a reduction in the severity of inflammation (38). This fact suggests that any impact on OPG therapy with respect to immunomodulation will be dwarfed by the benefit associated with bone preservation in severely inflamed joints. Osteoclast populations are regulated by the relative amounts of OPG and OPGL generated by other cells. For example, mature osteoblasts and bone marrow stromal cells have been suggested to produce more OPG than OPGL, and less mature osteoblasts secrete more OPGL than OPG (53); mature osteocytes are characteristic of normal adult bone, and immature cells are found in rapidly remodeling bone, whether for a physiologic purpose (e.g., bone growth during development) or in a pathologic state (e.g., RA). OPGL appears to be the final common mediator for many proresorptive ligands under both physiologic and pathologic conditions (Table 2). Thus, OPG, by serving as an OPGL sink, might control osteoclast net activity directed by the interaction of numerous hormones and cytokines. As shown in Table 2, many osteoclast-activating factors appear to act indirectly by inducing OPGL and/or by suppressing OPG (7, 41, 54). Under normal conditions, osteolysis is prevented by agents that augment OPG but lessen OPGL expression, including high calcium levels (24) and (likely) estrogen (55-58). In contrast, many molecules that support osteoclastic bone resorption increase OPGL but reduce OPG expression. Agents that elicit this response in cultured human osteoblasts or bone marrow stromal cells include glucocorticoids (59, 60), immunosuppressants (61), and hormones (62). Paradoxically, however, expression of both OPG and OPGL is increased by some cytokines (Table 2), including the proinflammatory agents IL-1β (63, 64), IL-11 (16, 65), IL-17 (66), TNFα (14, 63, 64, 67), and TNFβ (67) as well as transforming growth factor β (45, 68). OPG has been shown in vitro to block osteoclastogenesis and bone resorption mediated by OPGL-expressing cells isolated from joints of patients with RA. Resorption in this system is known to be well correlated to the OPG:OPGL ratio (69). Thus, control of osteoclast activity in RA and other osteolytic conditions resides not in absolute quantities of either OPG or OPGL but rather in their ratio. Numerous reports implicate altered OPG physiology as a factor in diseases associated with excessive bone resorption. Three such conditions for which a growing literature base exists are RA, osteoporosis, and skeletal metastases. A major consequence of RA is irreversible joint destruction leading to profound disability. Several lines of evidence indicate that an imbalance in the OPG/OPGL/RANK signaling pathway that controls osteoclast activity contributes significantly to the skeletal erosions that are characteristic of RA as well as animal models of arthritis (8, 69). First, osteoclast numbers are increased in patients with RA (70, 71). Primary control of osteoclast activity in this disease appears to be mediated by OPGL derived from activated T cells (20) and synovial fibroblasts (72); OPGL elaborated by osteoblasts and bone marrow stromal cells may also play a role (16, 73). In addition to the direct production of OPGL, activated T cells from patients with RA may activate osteoclasts indirectly through increased production of IL-1β (64), IL-17 (66), and/or TNFα (64), all of which are robust mediators of bone damage in inflamed joints. For these reasons, OPG therapy to inhibit OPGL-mediated effects on osteoclast activity represents a significant means of ameliorating bone destruction in inflammatory joint diseases. Age-related osteoporosis is common in postmenopausal women, possibly as a sequel to decreased OPG production resulting from estrogen deficiency (55). In support of this hypothesis, production of OPG messenger RNA by cultured human bone marrow stromal cells is decreased in older donors (74). Interestingly, the relationship between aging, menopause, and OPG levels may not be so simple. Healthy men and women exhibit higher serum OPG levels with age (75). Intriguingly, postmenopausal women with osteoporosis have significantly higher circulating OPG concentrations than do age-matched controls, possibly as a futile compensatory response to declining bone mass (75). However, other studies with human subjects suggest that OPG expression by osteoblasts is not substantially enhanced by estrogen (60), that serum OPGL levels do not differ significantly with age (76), and that expression of OPG and OPGL in bone of older individuals is more reflective of the serum parathyroid hormone level than the menopausal status (77). Thus, additional work will be required to examine OPG cycling (if any) associated with aging and/or estrogen loss. Osteoporosis also is a common consequence of chronic inflammation in RA patients, occurring at both periarticular (70) and distant (78) sites. A plausible mechanism for this condition is excess production of OPGL by the inflamed joint. Thus, in both age-related and inflammatory settings, the likely pathogenesis of osteoporosis is a relative excess of OPGL to OPG (i.e., a lowered OPG:OPGL ratio), which suggests that OPG supplementation to restore the ratio represents a logical means of inhibiting bone destruction. Neoplasia represents a third bone-destructive condition in which OPG, OPGL, and RANK have been implicated (for review, see ref. 79). OPGL clearly has a crucial role in many osteolytic human cancers. For example, neoplastic cells in giant cell tumors (80) and multiple myeloma (81) express OPGL as well as other osteolytic factors. The OPG:OPGL ratio is significantly lower in patients with these osteolytic tumors compared with those with nonlytic neoplasms (76, 80). However, patients with some lytic solid tumors, including advanced prostate cancer (82) and Hodgkin's disease (76), have higher serum OPG concentrations than do healthy individuals; the levels often are more elevated in individuals with metastatic rather than localized disease (82). An unconfirmed hypothesis for this paradoxical rise in OPG is that the increase represents an inadequate compensatory response to even more marked elevations in OPGL. The efficacy of recombinant OPG is readily apparent in normal animals as well as in many disease models (Table 3), which underscores the importance of the OPG:OPGL balance in both physiologic and pathologic bone resorption. Use of recombinant OPG to remove OPGL from the osteoclast microenvironment is a practical means of limiting skeletal destruction in many disease conditions. OPG prevents most of the bone loss that develops in immune-mediated arthritis. RA is characterized by osteolysis in conjunction with elevated OPGL expression in infiltrating T cells (20). Activated T cells of humans and mice secrete more OPGL and promote osteoclastogenesis when cultured with bone marrow cells; addition of OPG blocks this induction (20). Similarly, OPGL knockout mice with serum transfer-induced arthritis (a model that bypasses the need for T cell activation) develop less bone loss than do arthritic, wild-type littermates (83). In contrast, inhibition of the proinflammatory cytokines granulocyte–macrophage colony-stimulating factor, interferon γ, IL-1, IL-6, IL-17, and TNFα—all of which are found in arthritic joints—does not prevent T cell–induced osteoclastogenesis in vitro (20, 66). OPG (15 mg/kg given intravenously every third day for 5 weeks) is efficacious in preventing bone loss in TNFα-transgenic mice with TNF-driven chronic polyarthritis (84). In Lewis rats with adjuvant-induced arthritis, OPG treatment completely blocks the loss of bone mineral density and bone erosion (Figure 4) but does not inhibit inflammation (20, 38). The preservation of bone by OPG is associated with conservation of the articular cartilage matrix (20), an effect that appears to be the indirect result of preserving the adjacent subchondral bone (38) (Figure 4). The minimal OPG dose that provides essentially complete bone protection to inflamed joints in rats with adjuvant arthritis is 1 mg/kg/day (20, 38), and early intervention (near disease onset) is significantly more effective than delayed therapy in preventing skeletal dissolution (38). Comparable bone-sparing effects are observed for OPG in rats with collagen-induced arthritis (38, 85), although at a 17-fold lower dose (≥0.06 mg/kg/day) as compared with the adjuvant-induced condition (38). The appropriate dose for use in RA is yet to be determined. Radiographs and microscopic images showing that OPG prevents bone and cartilage destruction even in the presence of severe inflammation. Relative to normal controls, characteristic features of joints in the tibiotarsal region of rats with adjuvant-induced arthritis (7 days after the onset of paw swelling; i.e., at the peak of clinical disease) include significantly lower bone mineral density (BMD; assessed by dual-energy x-ray absorptiometry [DEXA] analysis in the region outlined in yellow), extensive leukocyte infiltration, marked bone erosion (including penetration of the growth plate [asterisks]) in the distal tibia), and degeneration of cartilage matrix (defined by loss of toluidine blue staining). OPG therapy (1 mg/kg/day for 7 days, beginning at disease onset) does not reduce inflammation but does prevent erosions and loss of BMD; cartilage is also protected if subchondral bone is intact (a), unless pannus extends into the joint cavity (b). Untreated arthritic rats cannot walk, but arthritic animals given OPG move without difficulty. Photomicrographs show histopathologic findings in the area within the green box shown in the DEXA view of the normal rat (and from comparable regions in the adjuvant arthritic and OPG-treated arthritic rats). In rats with adjuvant arthritis, a short course of OPG (7 consecutive days) prevents erosions in severely inflamed joints for almost 3 weeks (38), while a single OPG dose at disease onset effectively preserves articular architecture for only 4 days (46). Thus, a multidose loading regimen likely will be required to provide the best efficacy in active RA. Plausible explanations for this finding are that a single OPG treatment will remove only mature osteoclasts, while several days of dosing will also eliminate newly differentiated and recently activated cells, or that so much OPGL is produced in this condition that a single OPG treatment yields only incomplete OPGL inhibition. Taken together, the results of these studies indicate that OPG therapy should be initiated as early as possible to reduce the degree of bone and cartilage destruction in arthritic joints. OPG may exert a greater antierosive effect in combination with agents that inhibit IL-1 and TNFα. These potent proresorptive cytokines are produced at the cartilage–pannus junction in RA (86). In osteoclast biology, TNFα has been shown in vitro to induce osteoclast differentiation, and IL-1 is required for osteoclast function (87, 88). It has been suggested, however, that these actions—at least in the case of TNFα—may not occur in vivo (89) except under pathologic conditions. Both IL-1 and TNFα activate human vascular endothelial cells in vitro to yield a sustained upswing in OPGL expression (90). IL-1 also has been shown to enhance osteoclastogenesis in vitro in the presence of OPGL-producing activated T cells (91); a possible basis for this effect is the presence of IL-1 type I and IL-1 type II receptors on osteoclasts (92). Although their effects on osteoclasts are mainly mediated by OPGL (14), both IL-1β and TNFα have been reported to stimulate osteoclast differentiation by a pathway driven by M-CSF but independent of OPGL (87, 93). Thus, coadministration of a cytokine inhibitor with OPG might provide additive bone-sparing efficacy. However, IL-1 and TNFα do not induce osteoclasts in RANK knockout mice (14), signifying that the bone-sparing effect of IL-1 and TNF inhibitors is mediated indirectly by their antiinflammatory properties. An additive bone-protective effect also has been described following intermittent coadministration of OPG (15 mg/kg intravenously every third day for 10 treatments) and pamidronate to TNFα-transgenic mice with polyarthritis (84). However, our experience (38) suggests that OPG alone (≥1 mg/kg subcutaneously given daily for 7 days) would have provided comparable protection. A growing body of knowledge indicates that OPG could represent a potent new agent for treating human patients with RA. OPGL derived from activated T cells and synovial fibroblasts has been shown to instigate osteoclast formation by cells in the rheumatic joint (20, 69, 94). Although OPG clinical trials in RA have yet to be initiated, OPG has been used successfully to modify bone turnover in postmenopausal women. In this randomized, double-blind, placebo-controlled trial, a single subcutaneous OPG injection reduced serum levels of the bone resorption marker N-telopeptide within 12 hours, reaching a maximal reduction of 80% within 4 days at the highest dose (3.0 mg/kg) (Figure 5) (95). In fact, a single OPG dose ≥0.3 mg/kg effectively reduced the level of N-telopeptide for weeks (Figure 4) (95). Undesirable pharmacologic effects were not observed in any woman who received OPG. Graph showing reduced serum levels of N-telopeptide following injection of osteoprotegerin (OPG). A single subcutaneous OPG bolus on day 1 rapidly and profoundly reduced bone turnover in postmenopausal women for weeks, as measured by changes in the urinary N-telopeptide:creatinine ratio (from second morning void samples), expressed here as a percent change from baseline. Bars show the mean and SEM. Reproduced, with permission of the American Society for Bone and Mineral Research, from ref. 95. The restricted distribution of OPGL in adults (15) suggests that OPG therapy should have few or no undesirable pharmacologic effects in the rheumatology setting. In rodents, the high degree of OPGL expression near zones of bone remodeling explains the tendency of OPGL-inhibiting agents to induce accumulation of cancellous bone adjacent to growth plates (11). Although this finding has not been associated with clinical alterations in bone growth or strength, it suggests that OPG therapy in younger patients with arthritis should be undertaken with due care until bone growth is completed. The continued formation of osteophytes in arthritic joints despite daily OPG administration (38) indicates that the repair functions of osteoblasts are not impacted by OPG. Thus, fracture repair should not be affected by OPG therapy until the final step (osteoclastic remodeling to remove the bony callus). The apparent importance of OPGL as a mammary gland morphogen (50) suggests that OPG therapy might interfere with the ability of female RA patients to breast-feed. However, the availability of milk replacement products would significantly reduce the clinical importance of this event to both mother and child. We do not anticipate that OPG supplementation will adversely impact vascular health, because arterial calcification develops only if OPG is deficient (31). Finally, we believe that OPG therapy will not exacerbate the immune dysregulation that drives RA, because OPG does not affect the progression of inflammation in rodent models of adjuvant-induced and collagen-induced arthritis (38), contact hypersensitivity (Stolina M: unpublished observations), or granulomatous inflammation (Stolina M: unpublished observations). Instead, OPG may actually serve to down-regulate antigen-specific T cell (12) and antibody (Stolina M: unpublished observations) responses. Pharmacologic actions of OPG observed in animals and human patients suggest that this therapeutic protein rapidly terminates osteoclastic skeletal destruction in many pathologic conditions, including inflammatory, metabolic, and neoplastic diseases of bone. The impact of OPG on osteoclast numbers is mediated by 3 distinct mechanisms: inhibiting the function of mature activated osteoclasts, preventing the differentiation of precursor cells into new osteoclasts, and curtailing osteoclast survival. The primary consideration for using exogenous OPG to treat RA is that control of osteoclast net activity depends on the ratio of OPG to OPGL. Thus, OPG therapy will shift the balance toward preservation of bone. Prior animal arthritis studies (20, 38, 85) and a human osteoporosis trial (95) indicate that the bone-protective response will depend on both the OPG dose as well as the injection frequency. Data on the ability of OPG to preserve bone mineral density and skeletal integrity in human patients will be required to definitively establish its potential, including dose and schedule, as a bone-preserving therapy in clinical practice. The most important message is that OPG will almost completely halt bone destruction, even in the presence of severe inflammation, with maximal efficacy if treatment is started before bone damage has begun (38). The authors thank Dr. Colin Dunstan for critical discussion and Stephen Kaufman for technical assistance with dual-energy x-ray absorptiometry imaging.

Background Our laboratory has shown that bone destruction in the Lewis rat model of adjuvant arthritis (AdA) can be significantly reduced by daily administration of recombinant human osteoprotegerin (rhOPG) if treatment is initiated at the onset of clinical arthritis.1 Objectives The current study was performed to assess the ability of rhOPG to spare bone using an abbreviated dosing schedule. Methods AdA was induced in male Lewis rats (8 weeks old, n = 6/group) on Day 0; clinical arthritis (indicated by hind paw swelling) developed on Day 9. Arthritic animals received rhOPG (daily s.c. bolus of 4 mg/kg) by one of four schedules: no rhOPG (Group 1, control for AdA induction); daily rHOPG, Days 9 to 15 (Group 2, control for rhOPG efficacy); early intervention, Days 9 to 11 (Group 3); or delayed intervention, Days 13 to 15 (Group 4). Rats were necropsied on Day 16, and bone mineral density [BMD] of bones in the tibiotarsal (ankle) joint was measured by DEXA (dual X-ray absorptiometry). Next, sections of formalin-fixed, decalcified, paraffin-embedded hind paws were stained by an indirect immunohistochemical method to detect cathepsin K, an osteoclast marker. Histopathologic scores to assess lesion severity (given below as group means ± standard deviation) were acquired for inflammation, bone erosion, and osteoclast numbers using tiered, semi-quantitative grading scales. Results Joints of vehicle-treated arthritic animals (Group 1) exhibited measurable loss of BMD as well as extensive inflammation (4.0 ± 0), skeletal erosion (4.5 ± 0.5), and osteoclast production (4.0 ± 0). As noted previously,2 rhOPG therapy throughout the 7-day clinical course of AdA (Group 2) preserved BMD and almost completely inhibited erosions (1.2 ± 0.4) and osteoclastogenesis (0.8 ± 0.8) but did not modify the extent of inflammation (3.7 ± 0.5). Early initiation of rhOPG treatment (Group 3) yielded similar retention of BMD as well as reductions in erosion (1.2 ± 0.4) and osteoclast (1.7 ± 0.5) scores. However, in comparison to Groups 2 and 3, delayed rhOPG administration (Group 4) protected BMD and lowered the mean osteoclast score (2.0 ± 0.3) but was significantly less effective in preventing erosions (2.5 ± 1.4). Conclusion Our data indicate that early intervention with a bone-sparing agent such as rhOPG is needed to obtain the greatest clinical benefit with respect to protecting skeletal integrity in arthritic joints but also show that delayed rhOPG therapy can reduce the loss of joint BMD even in the presence of severe inflammation. References Kong, et al. Nature 1999;402:304 Feige, et al. Ann Rheum Dis. 2000;59(Suppl I):148

Interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α) play dominant roles in mediating the progression of many inflammatory joint diseases, including rheumatoid arthritis in humans, collagen-induced arthritis in mice and rats, and adjuvant arthritis in rats. Blockade of either cytokine partially controls these diseases. The present study investigated the value of combination anti-cytokine therapy in arthritis: the efficacy of IL-1 receptor antagonist (IL-1ra) and 30 kDa polyethylene glycol (PEG)-conjugated soluble TNF receptor type I (PEG sTNF-RI) given together was assessed in Lewis rats with adjuvant arthritis. Administration of either IL-1ra or PEG sTNF-RI partially alleviated joint inflammation, loss of bone mineral density, and loss of body weight. In contrast, combination of these anti-cytokine treatments exhibited a synergistic capacity to inhibit these changes, even when combining doses of IL-1ra and PEG sTNF-RI that did not affect lesion severity when used alone. Statistical analysis of these adjuvant arthritis data using the isobologram method proved that IL-1ra and PEG sTNF-RI were clearly synergistic in inhibiting inflammation, loss of bone mineral density, loss of body weight, and histopathologic parameters of inflammation and joint destruction. These results suggest that treating autoimmune arthritic diseases with combinations of anti-IL-1 and anti-TNF molecules will achieve superior efficacy compared to the use of a single class of anti-cytokine agent and may allow for dose reductions that could prove useful in minimizing potential side effects.