Parathyroid hormone and parathyroid hormone-related peptide are two structurally similar peptide hormones that exert mainly identical effects on classical target cells of parathyroid hormone but remarkable differences in their effects on other target cells. This review focuses on their functional differences and the mechanisms underlying these effects.
Parathyroid hormone (PTH) is a peptide hormone that is secreted from the parathyroid glands. It is involved in systemic calcium homeostasis, with bone and kidney as main target organs. In most mammalian species investigated so far, PTH is an 84-amino acid-long peptide with only a few interspecies differences, specifically in the NH2-terminal part of the peptide (8). The classical target cells of PTH in bone and kidney, namely, chondrocytes, osteoblasts, and osteoclasts from bone tissue and tubule cells from kidney, respond commonly to PTH with an intracellular accumulation of adenosine 3',5'-cyclic monophosphate (cAMP). However, PTH also activates adenylate cyclase on other target cells, named nonclassical target cells for PTH. Examples are blood cells (erythrocytes and lymphocytes), liver cells, smooth muscle cells, pacemaker cells of the heart, and heart muscle cells. During the last two decades, this peptide hormone has become a subject of increasing interest. It shares its function with a group of structurally related peptides that are now named PTH-related peptide (PTHrP). PTHrP, however, differs from PTH in three important aspects: first, it is synthesized and released by various normal and malignant tissues; second, in comparison with PTH, it is elongated at its COOH-terminal end; and third, it acts in many tissues as a paracrine or autocrine factor rather than as a classical hormone. It is now clear that PTH and PTHrP, although structurally and sometimes functionally similar, exert different effects because of differences in targets and structure-function relationships. This review summarizes the present body of knowledge dealing with second messenger pathways and physiological responses of cells to PTH and PTHrP that are distinct from the classical actions previously identified on bone and kidney cells.
Parathyroid hormone: structure and signal transduction pathways
The identification of the primary structure of bovine PTH in 1970 was the starting point for the characterization of the structure-function relationship of PTH. Knowledge of the exact location of those parts of the PTH molecule that cause physiological responses is a prerequisite for the design of pharmacological antagonists that may be used in patients with excessive serum PTH levels. Because of the multiplicity of target cells of PTH, high serum levels of PTH may be pathogenetically related to a multitude of disorders, e.g., of bone, kidney, blood, or the cardiovascular system.
PTH consists of 84 amino acids. The NH2-terminal portion of the molecule exhibits the greatest homology between different species. The first 34 amino acids seem to be a full biological agonist on classical target cells. In this part of the molecule, PTH consists of two interacting α-helixes. As mentioned in the introduction, all classical target cells of PTH respond to PTH by an activation of adenylate cyclase. The responsible functional domain is located on the first two amino acids of the native molecule. Bone- or kidney-derived cells show an intracellular accumulation of cAMP when treated with PTH(1–84), PTH(1–34), and, to a lesser extent, PTH(2–34) but not after treatment with PTH fragments missing the first two amino acids. Elevation of intracellular cAMP levels leads subsequently to an activation of the cAMP-dependent protein kinase (PKA). Amino acids 3–6 play a small but still significant role in the activation of the PKA-dependent pathway. It was found in vivo that deletion of only the first two amino acids is insufficient to totally abolish PKA-dependent PTH effects (7). To decrease the potency of PTH to activate PKA, however, truncation of the first six rather than the first two amino acids is required. In conclusion, amino acids 1–6 are required for the full interaction of the PTH molecule with a domain of its receptor, which is coupled to activation of adenylate cyclase and, secondarily, of PKA. Interestingly, deletion of the first six amino acids in PTH results in functional antagonism.
From classical target cells of PTH, a PTH receptor was cloned (5). It is named the PTH/PTHrP receptor because it can bind the NH2-terminal portions of both PTH and PTHrP. This receptor is a classical G protein-linked receptor with seven transmembrane domains. The PTH/PTHrP receptor, like many other structurally related receptors, couples to more than one intracellular signaling pathway. By binding to the PTH/PTHrP receptor, PTH may activate either adenylate cyclase, and subsequently PKA, or the phospholipase C (PLC)/protein kinase C (PKC) pathway. There are two major differences for the PLC/PKC-dependent pathway compared to the well-known functional domain responsible for the cAMP-dependent effects of PTH. First, several target cells of PTH, which respond to PTH by the activation of adenylate cyclase, do not respond with activation of the PKC-dependent pathway, e.g., tubule cells derived from the distal part of the nephron. Second, there are two different functional domains within the PTH molecule that may activate PKC in its target cells. One of these functional domains is located between amino acids 28 and 34 (11). The smallest PTH peptide found to activate PKC covers amino acids 28–32. Site-directed mutagenesis showed that replacement of glutamine by alanine in position 29 destroys its function. Another PKC-activating domain was identified on some bone cells and is located in the NH2-terminal portion. When this is involved, PKC activation can be diminished by deletion of the first two amino acids. These observations suggest that this other PKC-activating domain is located near or is identical with the adenylate cyclase-activating domain. In none of the studies, however, were both PKC-activating domains found to be active on the same target cells. In summary, all classical target cells of PTH respond to PTH by an activation of adenylate cyclase and, subsequently, of PKA and PKC. The corresponding functional domains of PTH are located at the NH2-terminal portion (PKA, PKC) or cover amino acids 28–32 (PKC). On some bone cell preparations, however, activation of PKC was found only for those covering the NH2-terminal portion.
PTH: activities on classical target cells
On bone, PTH has a dualistic effect. On one hand, PTH exerts anabolic effects on bone cells, e.g., it stimulates the proliferation of osteoblasts or chondrocytes. On the other hand, PTH has catabolic effects on bone as well. It increases bone resorption by acting on osteoclasts. This might happen directly or indirectly, namely, via the release of cytokines from osteoblasts, which belong to the tumor necrosis factor (TNF) family.
PTH activates adenylate cyclase on chondrocytes. On most but not all chondrocyte preparations, PTH activates PKC as well. As a consequence of PKC activation, chondrocytes proliferate and their expression of the vitamin D3 receptor increases. On chondrocytes isolated from the resting zone, PTH activates alkaline phosphatase in a PKC-dependent manner, but on chondrocytes isolated from the growth zone, PTH activates alkaline phosphatase in a cAMP-dependent manner. In both cases, PTH(3–34) is devoid of intrinsic activity. This suggests primarily that in target cells of PTH both second messenger pathways, i.e., PKA and PKC activation, may couple to the same biochemical end point. Secondarily, it suggests that responsiveness to both PKC-activating domains described in the literature may be represented in chondrocytes, although probably at different states of maturation.
The osteoblasts represent the most thoroughly analyzed cell type of bone responding to PTH. In this cell type, PTH(1–34), often considered as a full biological agonist, activates adenylate cyclase and PKC. PKC activation was demonstrated for PTH(28–48), indicating an involvement of the midregional PKC-activating domain. One of the central questions in this field during the last decade has been whether PTH is a mitogen for osteoblasts. The results obtained using cell cultures are still controversial. In some preparations, PTH leads to proliferation in a cAMP-dependent manner. In other preparations, this response requires an activation of PKC. When both pathways are activated together, accumulation of cAMP can be attenuated by a PKC-dependent effect. Therefore, in cells without a PKC responsiveness to PTH, the cAMP-dependent proliferation effect is more pronounced. In exceptional cases, PKC activation by PTH inhibits the proliferation of osteoblasts independently from effects on adenylate cyclase activity. The main question is, therefore, Which cell culture model represents the in vivo situation? In vivo, PTH(1–34) or PTH(28–48) increases bone mass (9). The results suggest that both PKC-dependent and cAMP-dependent anabolic effects of PTH occur in vivo.
In contrast to the antagonistic effects of PKC activation on the cAMP-dependent cell proliferation under PTH, the PTH-dependent induction of c-fos and of the vitamin D3 receptor and the activation of the sodium-proton exchanger (NHE) type 3 are synergistically activated by both second messenger pathways. PKA- and PKCsignaling pathways are found to be antagonistic, again in respect to the activation of NHE-1 and L-type calcium channels on classical target cells.
The third target cell of PTH in the bone is the osteoclast, which is involved in bone resorption and therefore associated with the catabolic effects of PTH. Osteoclasts also respond to PTH by an activation of PKA- and PKC-dependent pathways (6). The NH2-terminal truncation of PTH, PTH(3–34), does not abolish the activation of PKC by PTH in osteoclasts. The PKC-activating domain of PTH acting on osteoclasts seems to be located at position 28–34. The main effect of PTH on osteoclasts is an increase in the release of protons. This acidifies the surrounding of the cells and supports bone resorption. This effect by PTH is synergistically activated by both intracellular signaling pathways. It is still not decided whether these direct effects of PTH on osteoclasts are responsible for the PTH-dependent activation of bone resorption. It has also been suggested that PTH acts mainly on osteoblasts, which then release cytokines acting on osteoclasts.
Kidney cells, namely, primary cultures of proximal tubule cells, respond to PTH by activation of either the adenylate cyclase pathway or the PKC pathway. The functional domains are the NH2-terminally located functional domain and the 28–34 domain. As an exception, opossum kidney cells, which represent a commonly used cell culture model for studying renal effects of PTH on kidney cells, respond to PTH by an activation of PKC through the NH2-terminally located functional domain. Primary cultures of rat kidney cells derived from different parts of the kidney have also been used. It was found that proximal tubule cells respond to PTH with an activation of both intracellular signal transduction pathways, whereas cells derived from the distal tubule respond only with activation of adenylate cyclase. Like osteoblasts, kidney cells show stronger activation of the adenylate cyclase pathway in cells with downregulated PKC.
In summary, PTH acts on all main cell types of bone and kidney. There, it activates the adenylate cyclase pathway via the classical NH2-terminally located functional domain and the PKC-dependent pathways via a midregionally located functional domain. It should be noted that in nearly all cell types thus far investigated PTH activates both second messenger pathways.
PTH: nonclassical target cells
Nonclassical target cells of PTH differ from the classical target cells described above in respect to the major second messenger pathways. They respond to PTH by an activation of PKC but not of adenylate cyclase. Examples are keratinocytes, pancreatic cells, murine T lymphocytes, and cardiomyocytes. The latter cells lose their adenylate cyclase responsiveness to PTH during maturation. Neonatal cardiomyocytes respond to PTH with an accumulation of cAMP, indicating adenylate cyclase activation. Adult cardiomyocytes, however, do not respond in the same way. They selectively respond to PTH by activation of the PLC/PKC pathway. The functional domain of PTH responsible for this effect is located within the 28–34 portion and seems to be identical to the functional domain identified with chondrocytes as target cells. The influence of PTH on the cardiovascular system is not limited to cardiomyocytes (see Ref. 12). A vasodilatory action of PTH is caused by an activation of adenylate cyclase of smooth muscle cells, which is an exception for the as yet characterized nonclassical target cells, because PTH exclusively activates adenylate cyclase in this vascular cell type. The responsible functional domain for this PTH effect on smooth muscle cells is the classical NH2-terminal portion of the molecule. A third cell population within the cardiovascular system that responds to PTH are pacemaker cells. Initially, it was thought from in vivo studies that the chronotropic effect of PTH is an indirect one, caused by its vasodilatory action. A direct activation of ionic channels on pacemaker cells, however, has meanwhile been demonstrated, indicating that PTH exerts direct chronotropic effects as well (4). PTH has several cardiovascular effects. It is a hypotensive drug because of the relaxation of smooth muscle cells. In intact circulation, a subsequent activation of the baroreflex may indirectly elicit a positive chronotropic response. The chronotropic effect in vivo is further enhanced by a direct action of PTH on pacemaker cells. There are no indications that PTH exerts direct inotropic effects on adult cardiac muscles. Using the PKC-activating domain, PTH also acts as a hypertrophic agonist on cardiomyocytes. The latter observation seems to have significant clinical relevance, because dialysis patients with extremely high serum PTH levels often develop left ventricular hypertrophy. This is reversible upon parathyroidectomy (10), suggesting that the growth-promoting effect of PTH, identified on cell culture models, has indeed a pathophysiological significance in vivo.
In summary, nonclassical target cells show an unusual coupling of the activated PTH receptor to intracellular signaling. Figure 1⇓ summarizes the actions of the functional domains of PTH in classical and nonclassical target cells.
PTHrP: structure and signal transduction pathways
In principle, all known target cells for PTH are also target cells for PTHrP because of the structural similarity in their NH2-terminal parts. Six of the first seven amino acids are identical between the two peptide hormones. Therefore, it is not surprising that PTHrP can mimic nearly all functions of PTH mediated by the NH2-terminally located adenylate cyclase-activating domain. The binding domain of PTH and PTHrP, identified between amino acids 18 and 34, does not show a great similarity in the primary structure, but the secondary structure is quite comparable. This seems to enable both peptide hormones to bind to the same receptor, namely, the PTH/PTHrP receptor. In addition, a second PTH receptor has been identified on classical target cells that binds PTH but not PTHrP, which has been named the type 2 receptor (1). Most natural target cells analyzed thus far express the classical PTH/PTHrP receptor. In exceptional cases they also express the type 2 receptor.
The PTHrP molecule contains further active parts with no homology to PTH. PTHrP from the rat consists of 141 amino acids and is therefore extended at its COOH-terminal part compared to PTH. In the human two additional variants are found: PTHrP(1–139) and PTHrP(1–173). All three forms are synthesized from a common gene and represent different splice variants. Whether these PTHrP forms differ in their physiological effects on target cells has yet to be elucidated.
Some functional domains of PTHrP are located in parts of the molecule with either no or only limited structural similarity to PTH. They couple, nevertheless, to the same second messenger pathways as PTH, namely, to PKA- and PKC-dependent intracellular signaling pathways. A first PTHrP-specific functional domain distinct from PTH couples to the PKA-dependent pathway and is located in the COOH-terminal direction from the well-known adenylate cyclase-activating domain of PTH. A second domain seems to be located between amino acids 37 and 107 and stimulates placental calcium transport. A third domain has been found in the region between amino acids 107 and 111. It couples to PKC-activated pathways (2). These functional domains are located in parts of the molecule with no similarity to PTH. This suggests the existence of PTHrP-specific receptors. Such receptors, however, have not yet been identified.
PTHrP: expression and actions
PTHrP is widely expressed in normal and malignant tissues. Initially, it was identified as a factor released from malignant tissues. It may act as a paracrine or autocrine factor. In most tissues in which PTHrP is expressed, target cells for PTHrP were found adjacent. They often respond to PTHrP by an activation of adenylate cyclase and subsequently PKA. Vascular smooth muscle cells represent an example. On these cells, the structure-function relationship for the adenylate cyclase-activating domain of PTHrP is identical to that of PTH: PTHrP activates adenylate cyclase, and PTHrP(3–34) and PTHrP(7–34) are devoid of intrinsic activity but antagonize the response to PTHrP not truncated at the NH2-terminal part. This indicates that the first two amino acids are responsible for the interaction between PTHrP and the receptor to activate adenylate cyclase. PTHrP also activates adenylate cyclase in opossum kidney cells, chondrocytes, osteoblasts, and neurons and thereby mimics the activity of the whole molecule in this aspect. In some other cell types, the cAMP-dependent effects of PTHrP could also be mimicked by NH2-terminal truncated PTHrP. Examples are human kidney cells, opossum kidney cells, osteoblasts, and cardiomyocytes. In these latter cell types the functional domain on PTHrP activating adenylate cyclase seems thus to be different from the NH2-terminally located domain of PTH. It is in line with this that PTHrP but not PTH activates adenylate cyclase on cardiomyocytes and osteoblasts from the cell line ROS 17/2.8.
In some exceptional target cells of PTHrP, e.g., pancreatic islets or carcinoma cells, it does not activate adenylate cyclase. There are numerous reports showing cAMP-independent PTHrP effects. An accumulation of intracellular calcium was observed in squamous carcinoma and pancreatic islet cells in the presence of PTHrP. This effect of PTHrP can be blocked by PTHrP(7–34) but occurs in the absence of PKA activation. On Walker 256 carcinoma cells, PTHrP acted as a mitogen in a cAMP-independent manner. This mitogenic effect of PTHrP requires PKC activation, suggesting that NH2-terminal peptides of PTHrP activate the PLC/PKC pathway. A candidate location of the PKC-activating domain is the peptide sequence 28–34, which represents the PKC-activating domain of PTH. The primary sequence between PTH and PTHrP in this part is different, but the secondary structure is similar. On osteoblasts, PTHrP(28–34) indeed activates PKC (3). In contrast to PTH, PTHrP does not activate PKC via this domain on cardiomyocytes. With the use of site-directed mutagenesis, it was shown that the difference of the amino acid in position 29, asparagine in the case of PTHrP and proline in the case of PTH, explains the different responses (13).
In addition, another PKC-activating domain, clearly distinct from that of PTH, was identified between amino acids 107 and 111. Initially, a COOH-terminal portion of PTHrP, PTHrP(107–139), was characterized as an inhibitor of osteoclastic bone resorption and named osteostatin. The functional domain responsible for this effect covers amino acids 107–111 and is linked to PKC activation. Because this part of PTHrP has no homology to PTH, this effect is unique for PTHrP. A specific receptor to which this part of the peptide hormone binds must be postulated but has yet to be identified. With one exception, an adenylate cyclase activation was not found for this fragment. PKC activation caused by PTHrP(107–111) has been shown on other classical target cells of PTH or PTHrP, e.g., osteoblasts, and nonclassical target cells such as cardiomyocytes.
It is an open question whether the identification of active functional domains on PTH and PTHrP molecules is now complete. There are a few reports in the literature indicating the existence of more domains than those identified to date. For example, a partial agonistic activity of PTH(18–48) has been reported on adenylate cyclase activity of periosteal cell culture systems of chick embryos, and PTH(52–84) increases the expression of collagen in proliferating chondrocytes. The effects of PTH and PTHrP are mediated by bioactive fragments produced in vivo with different half-times. This indicates that the identification of different functional domains may have physiological significance.
PTH versus PTHrP in the cardiovascular system
Although various target cells for PTH or PTHrP have been described within the cardiovascular system, the physiological role of PTH and PTHrP in cardiovascular biology has yet to be identified. Excess of PTH or PTHrP, however, is often accompanied by multiple forms of cardiovascular diseases, e.g., left ventricular hypertrophy, atherosclerosis, and hypertension.
The cardiovascular system contains three different target cells for PTH and PTHrP: smooth muscle cells, cardiomyocytes, and pacemaker cells. Two cell types express and release PTHrP in a nonconstitutive manner, namely, smooth muscle cells and endothelial cells. On the adult cardiomyocyte, PTHrP activates PKA, thereby improving cardiac contractility. This effect can be antagonized by PTH. PTH can activate the PKC-dependent pathway leading to myocardial hypertrophy. This response can be antagonized by PTHrP. In adult cardiomyocytes, PTHrP may also activate PKC, but via a functional domain not found on PTH. In smooth muscle cells, PTH and PTHrP synergistically activate the PKA pathway. This leads to vasorelaxation. A local overexpression of PTHrP in vessels has been observed under pathophysiological conditions, as in atherosclerosis and conditions characterized by high blood pressure. The regulation of PTHrP expression in cardiovascular cells, however, is only partly characterized. In smooth muscle cells, PTHrP expression increases under angiotensin II, which is a potent vasoconstrictor. These findings suggest that the expression of the vasorelaxant PTHrP is upregulated in response to a pathophysiological elevation of vasotonus. The regulation of PTHrP expression and release by endothelial cells has not yet been characterized. In the heart, microvascular endothelial cells express PTHrP but, unlike adjacent cardiomyocytes, they themselves do not respond to this hormone. In summary, PTHrP represents a paracrine factor of the cardiovascular system whose role in physiology and pathophysiology is still largely unknown. Figure 2⇓ summarizes the present knowledge on the response of either PTH or PTHrP in the cardiovascular system.
In conclusion, PTH and PTHrP show a great deal of variability in regard to target cells and intracellular signaling. Although PTHrP was initially characterized by its PTH-like activities, it must be considered as a peptide hormone on its own, with several newly discovered effects distinct from those exerted by PTH. The widespread expression of PTHrP in normal tissues and the extensive expression in malignant tissues indicate an important role of this peptide hormone. Local release of PTHrP occurs in several diseases, but the physiological and pathophysiological roles of PTHrP still remain to be clarified.
This work was supported by the Deutsche Forschungsgemeinschaft, Sonderforschungsbereich 547, project A1.
- © 1999 Int. Union Physiol. Sci./Am.Physiol. Soc.