Emerging Therapiesby Robert J. Majeska
(Issue 9 · posted May 30, 1997; archived June 13, 1997)
Lately, osteoporosis has become a hot topic in scientific as well as nonscientific circles. Rapid growth in our understanding of skeletal biology, coupled with increased awareness of health-related issues, has focused the attention of academic, government, and pharmaceutical researchers. This breadth of interest was evident at a conference on new approaches to osteoporosis therapy held April 28-29 in Boston and sponsored by International Business Communications. Speakers broached topics ranging from molecular and cellular studies to clinical trials. Presentations emphasized perspective and strategy over minute experimental detail - the latter were probably reserved for upcoming international meetings, e.g. the Endocrine Society and the American Society for Bone and Mineral Research (ASBMR). The meeting did not concentrate strictly on the cutting edges of osteoporosis research, but certainly showed where the blades are being pointed.
Osteoporosis and Bone Turnover
Osteoporosis results from cumulative tissue losses during normal bone turnover. In adults, turnover occurs mainly by remodeling, wherein a small region of bone is resorbed (literally bored out) by multinucleated osteoclasts, and soon after filled in with new bone by osteoblasts. Resorption and formation are tightly coupled in remodeling, but incremental differences between the two processes result in continual declines in bone mass after early adulthood. Sufficient losses eventually produce the clinical features of osteoporosis.
Attacking Bone Resorption
Osteoporosis could, in principle, be prevented by either inhibiting bone resorption or stimulating bone formation. Both avenues have been explored. Ronald Harning (Merck) described recent clinical studies with alendronate, a recently marketed bisphosphonate developed on the basis of its previously known ability to inhibit osteoclastic bone resorption. Maxine Gowen (SmithKline Beecham) outlined an alternate approach to developing antiresorptive drugs. Starting with a human osteoclastoma, a tumor producing large numbers of osteoclasts, a cDNA library was constructed and screened for presumptive osteoclast-specific genes. One such gene was eventually identified as cathepsin K, a cysteine protease potentially involved in breaking down the organic constituents of bone matrix. Cathepsin inhibitors were later found to inhibit bone resorption in vitro and in animal studies, offering promise for this line of development. Interestingly, substantiation of cathepsin K's importance came from data identifying this gene as the site of mutation in the genetic disease pycnodysostosis (Gelb et al., 1996), a condition characterized by impaired bone resorption.
Other targets potentially exploitable for antiresorptive therapeutics are the factors responsible for triggering osteoclast formation from its monocytic precursors. Mark Horowitz (Yale University) and Stavros Manolagas (University of Arkansas Center for Health Sciences) summarized studies on interleukin-6 (IL-6) and the related cytokines leukemia inhibitory factor (LIF), ciliary neurotrophic factor (CNTF) and oncostatin-M (OSM). All of these factors act through the same signal transduction pathway linked to the receptor/transducer protein gp130. IL-6 was previously implicated as a mediator of estrogen-deficiency-related bone loss in mice. IL-6 also stimulated osteoclast formation in vitro, and its expression by osteoblastic cells was inhibited by gonadal steroids. On the other hand, IL-6 is not essential for osteoclast activity, since this process occurs normally in genetically null IL-6 knockout mice. Still, these molecules are especially interesting since they participate in cross-talk between cells of the osteoblastic and osteoclastic lineage, which is necessary for resorption-formation coupling.
Stimulating Bone Formation
Inhibiting resorption will slow bone loss, but will not actively help to rebuild already lost bone or heal osteoporosis-associated fractures. As a result, much emphasis has been placed on understanding the osteoblast and how its function can be stimulated. Again, there are several steps potentially amenable to pharmacological intervention: (1) stimulation of osteoblast precursor proliferation, (2) stimulation of osteoblast differentiation, (3) enhancement of osteoblastic functions (i.e., matrix synthesis and mineralization), and (4) increasing osteoblast life span.
Much of our understanding of osteoblasts has come from cell culture, but the search for better models is never-ending. A nearly ideal situation would be to have clonal cell lines that are immortal, stable, and phenotypically "normal." Two approaches to this goal for osteoblastic cells were outlined by Emma Moore (ZymoGenetics) and Peter Bodine (Wyeth-Ayerst). Moore et al. established a series of bone cell lines from transgenic mice lacking the tumor suppressor gene p53. Because p53 is a tumor suppressor, mice tend to develop tumors as they age, but otherwise their tissues appear perfectly normal. On the other hand, p53-deficient cells are highly proliferative and easy to grow in culture but do not express a "transformed" phenotype of tumor-derived cells. Bodine et al. isolated normal human bone cells and infected them with a temperature-sensitive immortalizing virus. Infected cells are thus indefinitely proliferative at permissive temperatures but at non-permissive temperatures presumably revert to a non-immortalized phenotype. Several clonal lines were established and characterized from these immortalized human cells. In both the mouse and human cell systems, individual lines were shown to exhibit different properties thought to reflect distinct stages of osteoblastic differentiation or phenotypic "maturity." Interestingly, mouse cell lines characterized as immature or transitional showed the ability to differentiate into adipocytes as well as osteoblasts. On the further side of the maturation scheme, one of the human cell lines exhibited properties apparently consistent with osteocytes - the "progeny" of osteoblasts - which remain buried in bone matrix and are thought to be critical for the tissue to sense and respond to mechanical stimuli. Osteocytes have only recently been isolated and cultured from normal bone, and their properties are very poorly understood at the biochemical level. A line of "osteocyte-like" cells would be a very exciting development.
One of the causes of bone loss with age may be a diminution of osteoblast progenitor cells. Scott Bruder (Osiris Therapeutics) described studies in which cells were isolated from bone marrow stroma, a source of osteoprogenitors, and expanded in culture through multiple generations. These cells, which contain precursors (stem cells) capable of differentiation into osteoblasts, chondrocytes, muscle cells, and fibroblasts, could then regenerate tissues upon reintroduction to the body. This approach has been successful in animal models using local application (e.g., in bone defects) and early studies in which injection of marrow cells into ovariectomized rats increased bone formation indices suggest promise, although numerous questions remain.
Several groups have targeted known osteoblast-produced growth factors as possible means to stimulate bone formation. One group is that of the insulin-like growth factors (IGFs), and Clifford Rosen (University of Maine) discussed their possible roles in normal bone metabolism. A notoriously complicating aspect of IGF function, however, is the family of soluble binding proteins that interact with IGFs and modulate their activity. The full range of IGFBP activities and functional significance are not known, but David Rosen (Celtrix Pharmaceuticals) described the use of a complex of IGF-1 with IGFBP-3, the most abundant binding protein, as a bone anabolic agent. The IGF-1/IGFBP-3 complex (trademarked SomatoKine) had previously demonstrated efficacy in a published animal osteoporosis model study, and most of the current presentation dealt with toxicology and pharmacokinetics in other animals and humans. Chronic administration of a large protein complex may cause immune responses, but this drug seems to be targeted for patients with fractures, where it could potentially stimulate bone and muscle formation in a short-term scenario. No toxic effects were seen in the studies reported here.
The bone morphogenetic proteins (BMPs), members of the TGF-a superfamily, were originally identified as the component of demineralized bone matrix that induced cartilage and bone formation when implanted in soft tissues. The functions of these proteins were found to be much more extensive and include the control of tissue formation during embryonic development. Therapeutic uses were described for BMP-7/OP-1 by T. Kuber Sampath (Creative BioMolecules), for BMP-2 by Howard Seeherman (Genetics Institute). In addition to its effects on bone cells, OP-1 was shown to be important in the kidney. The kidney is a major site of OP-1 synthesis, and OP-1 knockout mice exhibit impaired kidney development. Prophylactic OP-1 treatment improved recovery from acute renal failure in an animal model. BMP-2 studies tested whether a single administration could promote bone formation at sites of high risk for osteoporotic fractures. BMP-2 induced bone formation following core decompression of sheep femoral head. This may be a particularly useful approach which emphasizes the local, rather than systemic role of morphogenetic factors.
In addition to growth factors, systemic hormones with long-known roles in bone metabolism are being reevaluated based on new knowledge of their molecular structures and mechanisms. Parathyroid hormone (PTH), the classic stimulator of osteoclastic resorption, has historically shown perplexing anabolic effects as well. These observations fueled attempts to develop hormone analogs with purely anabolic activities on bone. Structural studies of PTH (and the more recently discovered PTH-related peptide, PTHrP), plus new information on the structures of PTH/PTHrP receptors, have helped in the design of those analogs and an understanding of their likely mechanisms of action at the molecular level. Larry Suva (Brigham and Women's Hospital, Boston) described several molecular studies of PTH, PTHrP, and their receptors, while Brian Vickery (Roche) summarized several pharmacological studies with newly developed anabolic analogs of PTHrP.
Unlike PTH, estrogen has been a long-standing treatment to counteract bone losses in postmenopausal osteoporosis; however, its side effects, both real and perceived, have limited its usefulness. Recently several estrogen analogs, having favorable actions on bone without significantly affecting reproductive tissues, have been under development. Mas Sato (Eli Lilly) described developmental studies with one such analog, raloxifene, while David Thompson (Pfizer) summarized results on a novel analog currently under development at that company. Not unexpectedly, both drugs showed favorable results and eventual clinical promise; however, our understanding of estrogen action, much less its newly developed analogs, is still incomplete.
Genetics
One of the most intriguing questions about osteoporosis is whether it has an identifiable genetic component. This issue is important not only for predicting who might be at risk to develop osteoporosis, but also for identifying potential genes or gene products as therapeutic targets. The approach described by Geoffrey Joslin (Sequana Therapeutics) and Leah Rae Donahue (Jackson Laboratories) seeks to locate genes associated with high peak bone mass by linking them with known genetic polymorphisms in pedigree screenings. The Sequana studies involve both human and baboon populations. Human studies have the advantage of more polymorphic markers to screen, while the baboons have larger, better-defined populations for heritability analysis (e.g., numerous sibling and half-sibling pairs). The Jackson Lab studies involved analysis of two inbred mouse strains differing in peak bone mass and in most known polymorphic loci, but with similar weights, life spans, etc. In this work, cross-breeding of the two strains was used to develop pedigrees for mapping quantitative trait loci. These methods of genetic analysis are complicated, particularly for traits undoubtedly controlled by multiple genes. Questions were raised about the suitability of peak bone mass as a predictor of osteoporosis risk, and the ability to obtain reliable density measurements in mice; however, these techniques seem to be promising, especially at this early stage of development. Moreover, this genetic approach has the advantage of providing independent data to complement effectively the other approaches described here.
This meeting, while reflective of the diverse interests and approaches to research in skeletal biology and osteoporosis, could only provide a hint of the complexity of these subjects. Principles of Bone Biology, edited by J.P. Bilezikian, L.G. Raisz, and G.A. Rodan and published last year by Academic Press, offers 97 chapters and nearly 1,400 pages of up-to-date articles covering the entire field thoroughly . It is a heavy tome, however, perhaps unsuitable for the weak of muscle or skeleton.
Robert J. Majeska, Ph.D., is Assistant Professor of Orthopaedics at the Mount Sinai School of Medicine, New York City.
Endlinks
Osteovision - an interactive information center whose aim is to transmit information in the field of bone and calcium metabolism to scientists and physicians. Its use is free of charge.
National Osteoporosis Foundation - offers a plethora of information on such topics as: "Risk Factors," "Bone Health," "For Professionals," "Petition to Congress," "Cutting Edge Reports," and "Advocacy."
The Osteoporosis Center - contains useful information on risk, diagnosis, and current treatment.
The University of Toronto - provides extensive information about osteoporosis and its treatment.
Cal Scan -
a well-designed Web site of Demetech AB, a manufacturer of bone densitomers;
a rich source of information on osteoporosis diagnosis and treatment.