The role of cytokines in skeletal remodelling: possible consequences for renal osteodystrophy

The pathogenesis of renal bone disease, a common complication encountered in patients with renal insufficiency, has centred around perturbations in the parathyroid hormone–vitamin D axis [1]. Thus, excess parathyroid hormone (PTH) can give rise to the high bone turnover state of osteitis fibrosa, and conversely, relatively low levels of PTH are associated with the low bone turnover state of adynamic bone. However, there are many observations which imply that the protean manifestations of renal bone disease cannot be explained simply by abnormalities of PTH or vitamin D metabolites. For example, levels of PTH correlate relatively poorly with various parameters of bone histology, and there is clearly a wide scatter of the data points for any given level of PTH [2,3]. While low levels of calcitriol are clearly important in the pathogenesis of hyperparathyroidism, and the administration of vitamin D metabolites is a key component of therapy, a direct role of vitamin D in bone has recently come into question in that normalization of serum levels of calcium and phosphorus in vitamin D receptor (VDR) knockout mice results in complete normalization of bone histology and bone strength [4]. In recent times, there has been an explosion in our understanding of the complex biology of bone with the demonstration that a multitude of local and circulating growth factors and cytokines play major roles in bone cell biology. Abnormalities in several of these factors may be present during the course of chronic renal failure, and therefore should be considered for their contribution to the abnormalities of bone remodelling in uraemia. In order to maintain mechanical integrity, the skeleton is constantly undergoing remodelling; that is, old bone is being replaced by new bone. The normal remodelling cycle requires that the processes of bone resorption and bone formation take place in a coordinated fashion which in turn depends on the orderly development and activation of osteoclasts and osteoblasts respectively. Activation of the remodelling cycle by a factor such as PTH results in the stimulation of osteoblast and osteoclast differentiation from precursor cells in the bone marrow. Osteoclasts derive from the haematopoietic granulocyte‐macrophage colony forming unit whereas osteoblasts originate from the pluripotential mesenchymal stem cells. Bone remodelling is accomplished by bone resorption followed by new bone formation. Once osteoclasts have resorbed a cavity of bone, they detach from the bone surface and are replaced by cells of the osteoblast lineage which in turn initiate the process of bone formation. The osteoblasts fill the resorption cavity by laying down matrix that later becomes mineralized. When the cycle is completed, the amount of bone formed should equal the amount of bone resorbed. The complex series of events that take place during the different processes involved in bone remodelling are under the regulation of a multitude of systemic and local factors, and it is the integrated effects of these factors on cell differentiation, proliferation and activity that are responsible for the maintenance of a healthy skeleton. A detailed description of the different cytokine systems responsible for the regulation of bone remodelling is beyond the scope of this review; however, recent developments in the area of osteoclastogenesis are worth discussing in some detail. The terminology used in this rapidly developing field is somewhat confusing because of the several synonyms used to describe the same mediators. These terms are summarized in Table 1 and a consensus on terminology for this area is eagerly awaited. Terminology Terminology It had long been recognized that the development of osteoclasts in vitro required close interaction between osteoclast precursors and osteoblastic stromal cells, and that the latter produced an osteoclast differentiating factor (ODF) in response to a variety of stimuli which was essential for the formation of mature osteoclasts. This factor has recently been identified as RANKL (receptor activator of NF‐κB ligand), also known as OPGL, which is a membrane protein expressed on the osteoblastic stromal cell and belongs to the tumour necrosis factor family [5,6]. Osteoclast progenitors, in turn, express on their surface the receptor for RANKL which has been named RANK (receptor activator of NF‐κB) [7]. Several groups of investigators have demonstrated that the interaction between RANKL and RANK is essential for the development of the mature osteoclast (Figure 1) [5,8,9]. Thus, exposure of bone marrow cultures to RANKL in combination with M‐CSF stimulates osteoclast development [5]; similarly, osteoclasts can be generated from mouse spleen cells and monocytes as well as from cultured human monocytes in the presence of M‐CSF and RANKL [9]. In addition to stimulating osteoclastogenesis, it has been demonstrated that RANKL plays a key role in stimulating osteoclast activity in vivo and in vitro. Thus, injection of OPGL to experimental animals results in hypercalcaemia secondary to increased osteoclastic activity as suggested by the finding of large osteoclasts present in the bone of the animals examined [5]. Furthermore, mice lacking the gene for RANKL are characterized by the presence of severe osteopetrosis and the absence of mature osteoclasts [8]. Furthermore, antibodies to RANKL can inhibit the bone resorptive activity induced by known inducers of bone resorption such as calcitriol, PTH and PGE2 [10]. The actions of RANKL may be influenced by osteoprotegerin (OPG), also known as osteoclastogenesis inhibitory factor (OCIF) or TNF receptor‐like molecule 1 (TR1) [11–15]. This protein is a tumour necrosis factor receptor‐like molecule which is produced and secreted by osteoblastic stromal cells. OPG can interact with RANKL and prevent the binding of RANKL to its receptor on the osteoclast precursor, RANK (Figure 1). Therefore, the role of OPG as a decoy receptor for RANKL leads to failure of osteoblastic stromal cells to communicate with osteoclast precursors resulting in impaired osteoclastogenesis. Several pieces of evidence have confirmed the ability of OPG to counteract the actions of RANKL. Thus, in vitro, OPG has been shown to inhibit the increase in osteoclast development and activity induced by RANKL, and in experimental animals, OPG inhibits the increase in serum calcium caused by the administration of RANKL [5]. These findings support the notion that OPG acts as a decoy receptor for RANKL, and therefore, it should have a protective effect on bone by decreasing osteoclastogenesis and osteoclastic bone resorption. Indeed, in vitro studies have demonstrated that OPG inhibits osteoclast differentiation from precursor cells, a finding that is consistent with studies in experimental animals showing that the administration of OPG leads to increased bone mineral density, overexpression of OPG in transgenic animals results in osteopetrosis, and targeted disruption of the OPG gene results in osteoporosis [12,16,17]. Although the role of the RANK/RANKL/OPG system in disorders of bone remodelling in humans has not yet been well studied, recent findings indicate that this system is involved in the effects of oestrogen and glucocorticoids on bone, and strongly suggest that abnormalities in any of its components are likely to result in skeletal disorders [18,19]. The actions of many cytokines and hormones on the balance between activators and suppressors of osteoclast number and activity is depicted in Figure 2. Diagrammatic representation of the role of the RANK/RANKL/OPG system in osteoclast development. RANKL is expressed on the surface of the osteoblastic stromal cell and serves as a ligand for RANK, expressed on the osteoclast progenitor cell. The interaction between the osteoblastic stromal cell and the osteoclast progenitor triggers the development of the mature osteoclast. OPG, produced by the osteoblastic stromal cell, serves as a decoy receptor for RANKL and inhibits osteoclastogenesis by preventing the interaction between the osteoblastic stromal cell and the osteoclast progenitor. (Modified from Nakagawa et al. [7] with permission.) Diagrammatic representation of the influence of RANKL and OPG on osteoclast number and activity. RANKL will tip the balance towards increased osteoclast number and activity whereas increases in OPG will oppose this effect. The hormones and cytokines scattered around the ends of the balance beam will tip the balance in the direction indicated by the arrows. (Adapted from Hofbauer et al. [35,36] with permission.) There is accumulating evidence to suggest that the various cytokine systems involved in the regulation of the different stages of the bone remodelling cycle may be altered in patients with chronic renal failure, and therefore, are likely to contribute to the pathogenesis of the remodelling abnormalities of renal osteodystrophy [20,21]. For example, activation of the bone remodelling cycle may be influenced by the high levels of interleukin 1 (IL‐1) which have been reported in dialysis patients [22,23]. Similarly, high levels of IL‐1 receptor antagonist, which are expected to oppose the effects of IL‐1, have also been detected in these patients, and interestingly, Ferreira et al. reported an inverse relationship between the levels of IL‐1Ra and osteoblast surface [23,24]. Another important cytokine for the activation of the remodeling cycle is TNF‐α, which has also been found to circulate at high levels in uraemic patients [22]. Osteoclastogenesis and osteoclast function may be influenced by abnormalities in the major bone resorptive cytokines such as IL‐6 and the RANK/RANKL/OPG system. Elevated levels of IL‐6 have been reported in dialysis as well as in pre‐dialysis patients, and more recently, Montalban and his group found a correlation between IL‐6 levels and bone remodelling markers in patients with renal osteodystrophy [25,26]. High levels of soluble IL‐6 receptors (sIL‐6R), which modulate the actions of IL‐6 have been reported in haemodialysis patients, and studies by Ferreira and his group found an inverse relationship between the sIL‐6R : IL‐6 ratio and osteoclast surface in uraemic patients [23,27]. Langub and colleagues, using in situ hybridization, examined the expression of IL‐6 receptor mRNA in osteoclasts of uraemic patients and found that it paralleled bone resorbing activity [28]. Thus, although the exact contribution of abnormal IL‐6 metabolism to the pathogenesis of renal bone disease is not clear at the present time, it is likely to be of importance based on the role played by this cytokine in other states of abnormal bone resorption such as post‐menopausal osteoporosis. In view of the now established importance of the RANK/RANKL/OPG system in bone biology, it is necessary to consider how disturbances of this system may impact on renal bone disease. Although this cytokine system has not been well studied in uraemia, alterations in these patients are expected since the system is regulated by calciotropic hormones which are known to be abnormal in the uraemic setting. Preliminary studies have shown high circulating levels of OPG in uraemic patients, and are likely to have an impact on bone remodelling [29]. Alterations in factors known to regulate osteoblast growth, differentiation and activity have also been demonstrated in uraemic patients. For example, increased circulating levels of IGF‐I have been reported, and there is some evidence to suggest that bone formation rate correlates with the levels of IGF‐I in these patients [30]. Epidermal growth factor is also known to affect osteoblast development, and uraemia may affect the actions of EGF on bone by regulating the expression of EGF receptors [31]. Although abnomalities in TGF‐β have yet to be described in uraemia, the system is clearly under the influence of parathyroid hormone and vice‐versa thus introducing another potential site for derangements in bone remodelling in chronic renal failure [32,33]. Interestingly, a member of the TGF‐β family, known as osteogenic protein‐1 or bone morphogenetic protein 1 (OP‐1 or BMP‐1), which is crucial in osteoblast differentiation from mesenchymal stem cells, is normally expressed in the kidney [34]. Therefore, a deficiency of OP‐1 in uraemia as renal mass decreases could potentially lead to failure of osteoblast development and contribute to the low turnover bone disease (adynamic bone) which is part of the spectrum of skeletal abnormalities of renal osteodystrophy. In addition to the contributions of the well‐known calciotropic hormones, normal bone remodelling depends on the integrated effects of a variety of growth factors and cytokine systems. Renal osteodystrophy is a multifactorial disorder of bone remodelling commonly manifested in patients with chronic renal failure which has been traditionally attributed to disordered calciotropic hormone metabolism. However, there is accumulating evidence to suggest that abnormalities of bone acting cytokines and growth factors as well as their receptors and endogenous modulators are also present in uraemia, and therefore, should be considered as potential contributors to the pathogenesis of renal bone disease. Correspondence and offprint requests to: Esther A. González, MD, Division of Nephrology, Saint Louis University, 3635 Vista Avenue, St Louis, MO 63110, USA. This work was supported in part by grant DK‐51099 from the National Institutes of Health.