| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology, The University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania 15213
Address all correspondence and requests for reprints to: Andrew F. Stewart, M.D., Chief, Division of Endocrinology, BST E-1140, University of Pittsburgh School of Medicine, 200 Lothrop Street, Pittsburgh Pennsylvania 15213. E-mail: stewart{at}msx.dept-med.pitt.edu.
The parathyroid gland was discovered by Ivar Sandstrom in 1880. The destructive skeletal disease, osteitis fibrosa cystica (OFC), was described by von Recklinghausen in 1891. In 1903, Askanazy reported that parathyroid tumors were present at necropsy in patients dying with the presumed inflammatory, destructive skeletal disease, OFC. (For a review of the history of hyperparathyroidism, see Ref. 1 .) Surprising as it may seem in the current era, this led to the prevailing hypothesis (supported by the Viennese pathologist Jacob Erdheim), that the skeletal disease, OFC, was the primary pathophysiological event, and the parathyroid enlargement was a secondary consequence of, or response to, the skeletal disease.
Serum calcium measurements first became available in the 1920s, and they were rapidly noted to be elevated in patients with OFC. Thinking the iconoclastic idea that the parathyroid enlargement might be the cause of the hypercalcemia and the OFC, in Vienna in 1925 Mandl performed a successful parathyroidectomy on a patient named "Albert," with a return of the serum calcium to normal and improvement in the skeletal disease. In 1926 in the United States, thinking similarly iconoclastic thoughts, Joseph Aub, Oliver Cope, Fuller Albright, and their colleagues at the Massachusetts General Hospital performed the first American parathyroidectomy on Captain Charles Martell, a Boston sea captain with severe hypercalcemia and advanced and destructive OFC. A parathyroid adenoma was identified and removed (at the seventh operative attempt!), and the serum calcium returned to normal. These two cases confirmed what the iconoclasts dared to believe: the parathyroid glands secreted a "parathyroid hormone" that regulated calcium in an upward direction, and that the "parathyroid hormone" was a skeletal catabolic agent, with its oversecretion causing a destructive skeletal disease.
In 1932, J. B. Collip, thinking along similar lines, was able to collect the parathyroid glands of dogs, and then infuse an extract of these parathyroid glands into hypocalcemic thyroparathyroidectomized dogs. His studies showed that parathyroidectomy causes hypocalcemia, and infusion of PTH corrects this hypocalcemia. Collectively, these studies, together with those of Albright, Aub, Ellsworth, Cope, and others, laid the foundation for the prevailing understanding of PTH for the next six decades: PTH is the central regulator of bone and mineral metabolism, and its chronic excess is bad.
Surprisingly enough, and in seeming contradiction, Albright in 1929 and Selye in 1932 demonstrated that daily sc injection of PTH had anabolic effects on the rat and dog skeleton. These studies were widely ignored by the medical establishment because they did not fit prevailing dogma. After all, how could a hormone that was such a dramatic skeletal catabolic agent in Captain Martell and Albert, at the same time be a skeletal anabolic agent? This seemed impossible. We will return to these observations later.
In 1987, we and others (2) described the isolation of PTHrP. We first identified PTHrP in tumors from patients suffering from a lethal, rapidly fatal skeletal catabolic syndrome, humoral hypercalcemia of malignancy (HHM). PTHrP was thus a second calcemic, skeletal-destructive hormone. This was confirmed directly using skeletal biopsies from patients with HHM: they displayed severe reductions in bone mass associated with markedly reduced osteoblastic activity but strikingly increased osteoclastic activity (3). Thus, the second calcemic, skeletal catabolic hormone was identified.
We and others (4) showed that synthetic PTHrP was able to bind to and activate the PTH receptor with kinetics and affinities identical with those of PTH, observations later confirmed with the molecular cloning of the common PTH/PTHrP receptor, the PTH-1 receptor. Further exploration of the physiology of PTHrP demonstrated that it is secreted by essentially every tissue and cell type in the body, that the receptor is equally widely distributed, that PTHrP is expressed in these normal tissues at levels far lower than those observed in HHM-associated cancers, and that PTHrP does not normally appear in the circulation. PTHrP appears to stimulate cellular proliferation, survival, and differentiation, in cell type-specific ways. Said another way, PTHrP is a ubiquitously produced local paracrine and autocrine factor (5) [and in some instances an intracrine factor but that is another story! (6)], the role of which is to regulate cellular differentiation, proliferation, cell death, and epithelial calcium transport, both during development as well as in adult life (5). This physiological scenario is different from that of PTH in the following senses: PTH is produced not by many tissues, but by a single one, the parathyroid gland. PTH acts not as a local paracrine, autocrine factor, but rather acts exclusively in an systemic endocrine fashion. Thus, PTH is to PTHrP as insulin is to IGF-I. PTH and insulin are endocrine and metabolic hormones, whereas IGF-I and PTHrP are paracrine/autocrine factors that do not normally act systemically, but locally regulate cellular death, proliferation, and differentiation. Moreover, for both PTHrP and IGF-I, tumors may overproduce the paracrine factor, leading to an ectopic hormone or paraneoplastic syndrome, HHM, and tumor-induced hypoglycemia, respectively.
In 1996, to answer the question "what precisely is the physiological role of PTHrP?" the Karaplis group prepared mice in which both alleles of the PTHrP gene were disrupted. These PTHrP –/– mice died at or shortly after birth and displayed a dramatic form of chondrodysplasia as the most striking phenotype (7). These studies pointed to the growth plate chondrocyte as one of the key physiological sites of PTHrP action in embryonic life. Interestingly, and more germane to the current theme, the PTHrP +/– heterozytoges survived and grew normally but displayed a significant reduction in bone mass as adults (8). Was this the residual result of an embryonic or developmental deficiency of PTHrP? Or was it a residual manifestation of the embryonic chondrocyte abnormality in the growth plate?
In 2002, Miao, working with the same group of investigators (9, 10), generated mice that are homozygously null for the PTH gene. That is, they created an animal model of human congenital hypoparathyroidism. As one would anticipate from the work of Collip, Selye, and Albright and others described above, these mice displayed hypocalcemia and hyperphosphatemia. They also displayed a reduction in circulating 1,25(OH)2 vitamin D as occurs in humans with hypoparathyroidism. Perhaps surprisingly, as has also been described in humans with congenital hypoparathyroidism, these PTH –/– mice displayed increases in bone mass. This increase in bone mass was characterized histomorphometrically by increases in trabecular and cortical bone and was associated with reductions in both sides of the bone turnover equation: osteoblastic bone formation and osteoclastic bone resorption were both reduced. Obviously, however, because bone mass was increased, osteoclastic bone resorption must have been decreased more than osteoblastic bone formation. So what was maintaining bone formation in these PTH null mice?
In their study reported in this issue of Endocrinology (11), Miao and his collaborators have now created mice doubly deficient for both PTH and PTHrP. The mice are homozygously null for PTH, but only heterozygously deleted (haploinsufficient) for PTHrP (because the investigators wanted to examine adult bone mass, and PTHrP –/– mice are unable to survive postpartum). The adult PTH –/–, PTHrP +/– mice are full of surprises. First, their bone mass, both cortical and trabecular, is low. This is exactly the opposite of events in the PTH –/– mice describe above and suggests that PTHrP is required for developing and maintaining bone mass in the adult.
Second, compared with normal mice, PTHrP is overproduced within the skeleton in PTH –/– mice. That is, PTH deficiency leads to local skeletal overproduction of PTHrP, as determined using RT-PCR and Western blotting. Immunohistochemically, it appears that the source of the increase in skeletal PTHrP in PTH –/– mice may include at least osteoblasts and chondrocytes. This raises the question of how bone cells may know that PTH is absent, i.e. what is the signal to up-regulate the expression of PTHrP? The authors suggest that it may be reductions in circulating 1,25(OH)2 vitamin D that result from PTH deficiency. Other options might be the hypocalcemia, hyperphosphatemia, or absence of basal PTH stimulation in osteoblasts, chondrocytes, or their precursors.
Third, whereas the rates of bone resorption are low and comparable in both the PTH –/– and the PTH –/–, PTHrP +/– mice, the rates of bone formation in the PTH –/–, PTHrP +/– mice are even lower than the already depressed rates in the PTH –/– mice. The authors reasonably attribute this further reduction in bone formation in the doubly deficient mice compared with the single PTH –/– mice to the additional loss of the anabolic effects of PTHrP. Said another way, in PTH deficiency, bone mass increases because rate of bone resorption falls, but bone formation is sustained to some extent by the anabolic effects of PTHrP. When this is lost, low bone mass results.
A fourth surprise is that the phenotypes of these mice as adults is quite different from their phenotypes in fetal life. Mouse embryos that lack PTH display low bone mass at birth (in contrast to the adult PTH null mice that have high bone mass), presumably indicating that PTH is required for skeletal anabolism in utero (9, 10). In contrast, PTHrP –/– mice, to the extent it can be studied, appear to have normal bone mass at sites where bone has been formed in late embryonic life (9, 10). These earlier results in fetal mice, in concert with the results of the study in this current issue of Endocrinology (11), indicate a skeletal role reversal for PTH and PTHrP in embryonic vs. postpartum life: in embryonic life, PTH is the skeletal anabolic hormone, and the role of PTHrP is to drive calcium transport across the placenta from the maternal circulation to the fetal circulation. Postpartum, however, things change. In the new postpartum environment where the neonate no longer gets its calcium from a maternal placental source but from the intestine, the major goal of PTH becomes one of a systemic calcium-regulating hormone, acting via stimulation of osteoclastic bone resorption, stimulation of renal calcium conservation, and indirectly via 1,25(OH)2 vitamin D, stimulation of intestinal calcium absorption. Conversely, the skeletal anabolic role of PTH is replaced by PTHrP, which now acts not on the placenta, but acts as a local paracrine and/or autocrine factor, the role of which is to drive and maintain osteoblastic bone formation and to maintain skeletal mass.
This brings us back to studies in the 1920s and 1930s of Albright and Selye showing that the sc daily administration of PTH increased bone mass. Despite the neglect and disbelief these observations encountered, and the resultant passage of 80 yr, many recent studies have now shown that PTH is indeed anabolic when given as a daily sc injection. For example, Neer et al. (12) in 2001 convincingly showed that the daily administration of PTH to postmenopausal women not only increases bone mass, but also markedly reduces postmenopausal osteoporotic fractures. This has led to the approval by the FDA of PTH, once viewed as the premier skeletal catabolic agent, as the first authentic skeletal anabolic agent. The widely accepted explanation for this paradox is that whereas chronic intermittent (i.e. daily sc) administration of PTH activates both the bone formation pathway (recruiting osteoblasts from their stromal precursors and maintaining or prolonging the life span of existing osteoblasts), and the bone resorption pathway (leading to osteoclast recruitment from macrophages), the formation effects outweigh the resorption effects, and an increase in bone mass results. In contrast, continuous exposure of the skeleton to PTH activates both the formation pathway and the resorption pathway, but the resorptive effects predominate. Thus, the net effect of chronic or continuous exposure of the skeleton to PTH is a reduction in bone mass.
So this leads to a fifth surprise in the Miao manuscript (11): if chronic exposure to either PTH (as in hyperparathyroidism) or PTHrP (as occurs in HHM) leads to net activation of osteoclastic bone resorption, why is it that the PTH –/– mice, with continuous local overproduction of PTHrP within the skeleton, do not also display increases in osteoclast recruitment and activity? Is it because the PTHrP that is being made locally cannot access the macrophages that serve as osteoclast precursors? This seems unlikely because these cells are abundant within the bone marrow, and adjacent to the osteoblasts and stromal cells. Is it because they need to be primed in some way by PTH, and this does not happen in the PTH –/– mice? Is it because the local concentrations of PTHrP are not high enough to initiate the resorption program? This seems unlikely because the increases in PTHrP expression observed by Miao et al. (11) are not small. This observation will require additional studies to understand.
The 80 yr of skepticism regarding the anabolic skeletal effects of PTH have finally been undone. Many studies have now demonstrated PTH is the first in a new class of therapeutic agents for osteoporosis, the skeletal anabolic agents. PTHrP, like PTH, was initially discovered as an agent that causes dramatic skeletal bone resorption and hypercalcemia, and one that leads to striking and rapid bone loss (3). That is, it is the quintessential skeletal catabolic agent. Yet, paradoxically, the work of Miao et al. (11) clearly demonstrates that the normal physiologic role of PTHrP is to serve as an anabolic agent within the skeleton. Indeed, it appears to be absolutely required for the attainment and maintenance of skeletal mass in adult mice.
Is this true in humans? And is there any potential that PTHrP might also be a skeletal anabolic agent in humans, as is the case with PTH? We have recently reported that a single daily injection of PTHrP (1–36) to postmenopausal women with osteoporosis results in a striking (almost 5%) increase in spine bone mineral density in only 3 months (13). Compared with the antiresorptive agents currently available for treating osteoporosis, these changes are quantitatively large and temporally rapid. Additional studies using PTHrP (1–36) designed to extend, expand, and confirm this pilot study are underway. From a physiological standpoint, they strongly support and complement the notion of Miao et al. (11) that PTHrP is indeed a naturally occurring and physiologically requisite skeletal anabolic agent.
These are exciting times in understanding bone formation and identifying novel skeletal anabolic agents and mechanisms. The convergence of mouse genetics, gene linkage studies to discover new genes, signal transduction, intercellular signaling, and old-fashioned physiology with pharmaceutical development has created a completely new focus in skeletal biology over the past 3–5 yr: the identification of novel skeletal anabolic agents for the treatment of osteoporosis. One such agent, human PTH (1–34), is currently available commercially, and others are surely soon to follow. It is a beautiful example of the translational convergence of purely basic and purely clinical research streams to yield biologically fascinating new paradigms associated with medically and pharmaceutically important outcomes. More excitement, more questions, and more surprises can only be expected.
 |
Footnotes |
|---|
Disclosures: A.F.S. is the CEO of Osteotrophin LLC and is a consultant to Eli Lilly Inc.
Abbreviations: HHM, Humoral hypercalcemia of malignancy; OFC, osteitis fibrosa cystica.
Received April 20, 2004.
Accepted for publication April 27, 2004.
|
References |
|---|
 Top References |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. Seeman and G. J. Strewler Clinical and Basic Research Papers - February 2005 Selections IBMS BoneKEy, March 1, 2005; 2(3): 1 - 5. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
