Chemistry

The first extracts from bovine parathyroid glands were described in 1925, and the content of biologically active PTH was assessed by their hypercalcemic and phosphaturic properties.^2,3 However, it was not until 1959, when Aurbach³³ and Rasmussen and Craig³⁴ developed improved extraction procedures, that it became possible to isolate and purify sufficient quantities to determine the primary structure of bovine, porcine, and human PTH through the protein sequence determination methods.^35–40 Two groups independently determined the sequences of human and bovine hormones.^35–40 Shown in Fig. 1-3A are the sequences of the bovine, porcine, and human hormones determined by one group.^36,37,39,40 Discordant sequences for the human PTH polypeptide, and in one position for the bovine hormone, published by the other group,^35,38 are not shown in Fig. 1-3A, since nucleotide sequence analysis of genomic and complementary DNA confirmed the amino acid sequences of the first group (the only exception was residue 76 in human PTH, which was determined to be glutamine instead of glutamic acid).^36,37,40 Based on these amino acid sequences, the PTH(1–34) fragments of the different species were synthesized, and their biological activities were compared in vitro and in vivo with those of highly purified intact PTH from the same species (Table 1-1). Molecular cloning techniques then led to the deduction of the amino acid sequences of rat, chicken, and dog PTHs,^41–44 followed more recently by the identification of PTH molecules in other mammals and fish, as shown in Fig. 1-3A and discussed later. The synthetic peptides used in parathyroid hormone research today are based largely on the (1-34) regions of the mammalian hormone sequences shown in Fig. 1-3A.^45,46

Extensive sequence homology is present in the mammalian PTH species (see Fig. 1-3A and D); these molecules consist of a single-chain polypeptide with 84 amino acids and a molecular weight of approximately 9400 daltons (that of human PTH[1–84] is 9425 D). The N-terminal region of PTH, which is necessary and sufficient for the regulation of mineral ion homeostasis, shows high sequence conservation among all the vertebrate species (see Fig. 1-3A). The middle portions of the different molecules exhibit the most structural variation, which could suggest that this region of PTH is only of limited functional importance. The nonmammalian PTH homologues of chicken^42,43 and fish species Danio rerio (zebrafish)⁴⁷ and Takifugu ruberipes (puffer fish)^48,49 diverge considerably from the mammalian hormones C-terminal of amino acid residue His32. Interestingly, both fish species have two distinct genes encoding two separate PTH molecules, called PTH1 and PTH2.^47,48 The zebrafish peptides are considerably shorter than mammalian PTH (67 and 68 residues), while fugu PTH1 is predicted to be 81 residues in length and fugu PTH2 is predicted to be 63 residues.⁴⁸

After the original work establishing that the first 34 amino acids of mammalian PTH were sufficient to produce a fully active synthetic peptide,^45,46 much work has centered on defining the minimum pharmacophore essential for biological activity. In a following section, we describe how sites of ligand interaction with the PTH/PTHrP receptor were defined by performing assays with products of various combinations of shortened and modified PTH ligands and mutagenized receptors. Furthermore, a fusion protein consisting of ligand substituted for most of the receptor amino-terminal extracellular domain was generated, and this helped define a much smaller minimum chain length of PTH peptide needed for biological activity.⁵⁰ As also discussed later in the section on hormone/receptor interactions, it has been determined that substitutions of non-naturally occurring amino acids (e.g., α-aminoisobutyric acid at positions 1 and 3 in the primary ligand structure) favoring formation of an α-helix, even in short peptides, such as PTH(1–14), produce peptides that are highly potent when tested in vitro using cell-based assays and have highly stabilized helical structure in solution^51–55(Fig. 1-4).

The in vitro activity of the native PTH(1-14), which is quite weak, is improved about 100,000-fold by the modifications indicated in Fig. 1-4. Shorter PTH peptides have also been shown to be active in vivo, since some cause hypercalcemia and are anabolic on bone, although their potency is much less than that of PTH(1-34) due to a more rapid clearance.⁵⁶ Replacement of valine-2 in these peptides with bulky amino acids, such as tryptophan or parabenzoyl-l-phenylalanine (Bpa), results in competitive antagonist peptides defective for AC/cAMP and PLC/IP₃/Ca²⁺ signaling, thus confirming the critical role that this conserved valine plays in receptor activation.⁵⁷ Other longer-length PTH or PTHrP analogs having residue-1 (serine or alanine) replaced by glycine,⁵⁸ Bpa,⁵⁹ or tryptophan⁶⁰ exhibit signal-selective properties in that they efficiently stimulate the cAMP cellular pathway but not the inositol tri-phosphate/intracellular Ca²⁺ pathway.

The three-dimensional structures of intact PTH(1-84) and the N-terminal biologically active fragments of both PTH and PTHrP have been analyzed by various solution-based methods, including nuclear magnetic resonance (NMR) spectroscopy. Interpretation of these results in terms of biological mode of action is not straightforward, because the ligand interacts with a membrane-embedded receptor, and the biophysical properties of this environment, as well as the receptor-induced conformational changes that occur, are mostly unknown. However, the bioactive (1-34) portions of the ligands are generally found to contain helical structure within their N-terminal and especially C-terminal portions, with some flexibility in the midregion.61–64 An x-ray crystallographic study of PTH(1-34) revealed α-helical structure extending nearly the full length of the peptide.⁶⁵ A persisting question is thus whether the ligand bound to the receptor adopts a linear and extended⁶² or “U-shaped” folded structure, the latter suggested by tertiary interactions seen in some solution-phase biophysical studies.^61,63,64 Of particular interest has been the question of whether common structural features would be discerned for PTH and PTHrP that could explain their use of apparently overlapping binding sites on the common PTH/PTHrP receptor; to some extent, the helical propensities of at least the C-terminal domains of the peptides are consistent with this possibility.^62,66,67

Therefore, consistent with their rather limited homology in primary structure (see Fig. 1-3), convincing evidence has not yet been provided for the conclusion that the N-terminal fragments of PTH and PTHrP display a similar secondary structure in solution. Because of their generally demonstrated similar potencies at the PTH/PTHrP receptor, it seemed likely that both ligands would adopt very similar conformations when part of the active hormone-receptor complex. However, recent data (discussed later) suggest that each ligand selectively binds to or induces a distinct receptor confirmation. Ideally, each hormone should be co-crystallized with the PTH/PTHrP receptor to permit analysis by x-ray diffraction of those intermolecular interactions that are characteristic of the biologically active hormone-receptor complexes. G protein–coupled receptors such as the PTH/PTHrP receptor have multiple membrane-embedded domains and are likely to have complex three-dimensional structures. Interaction with either PTH or PTHrP appears to involve several distinct receptor domains (see later discussion) that may undergo significant conformational changes after ligand binding has occurred, which makes it even more challenging to conduct x-ray or multidimensional NMR analyses. Recent advances, however, have been made it possible to co-crystallize the extracellular portion of the PTH/PTHrP receptor with the carboxyl end of the PTH (1-34) peptide⁶⁸ and to crystallize the membrane-spanning portion of a distantly related GPCR, the β₂-adrenergic receptor.⁶⁹

Evolution

To maintain extracellular calcium and phosphate concentrations within narrow limits, the intricate regulatory system outlined, in which PTH plays the most important role, developed in the terrestrial animals. In mammals, PTH is produced almost exclusively by the parathyroid glands (only small amounts of its messenger RNA [mRNA] have been detected elsewhere^70,71). During evolution, these glands first appear as discrete organs in amphibians, that is, with the migration of vertebrates from an aquatic to a terrestrial existence, and their appearance most likely represents an evolutionary adaptation to an environment that is, by comparison to seawater, low in calcium.^17,18,72 Parathyroid glands have not been identified in fish or invertebrate species, but earlier immunologic and RNA hybridization data from fish provided evidence for expression of PTH proteins in several tissues, including pituitary,^17,18,73 plasma, brain, kidney, spinal cord, ultimobranchial gland, as well as in the ventral neural tube and mineralizing jaw during development.^48,74 Now, with the rapid advances in characterization of complete genomes of multiple species, we have definitive proof of the earlier evolutionary origin of both PTH and PTHrP (see Fig. 1-3). Gene analyses of several teleost fish species, including the zebrafish, Danio rerio, and the puffer fishes Takifugu rubripes and Tetraodon fluviatilis, reveal the duplication of the PTH gene in each case.^48,75 Both the PTH1 and PTH2 peptides derived from the zebrafish activate the PTH/PTHrP receptors from different species,^49,75 and indeed, a fugu PTH(1-34) peptide has been shown to induce bone anabolic effects in osteopenic ovariectomized rats.⁷⁶

In addition to PTH, the teleost fish also express PTHrP, again encoded by duplicate genes.75,77–80 Furthermore, PTHrP immunoreactivity has been detected in the cartilaginous sharks and rays⁸¹ and in a more primitive agnathan, the lamprey.⁸² The teleost PTHrPs contain some amino acid residues characteristic of mammalian PTH; for example, fish PTHrP contains Met at position 8, Trp at position 23, and Leu at position 28, which are amino acid residues found in mammalian PTH. However, there is only one amino acid residue, Gln(Q)25, in fish PTH that is found in some mammalian PTHrP species but not in mammalian PTH (see Fig. 1-3). This pattern suggests that the fish proteins may be phylogenetically closer to a common PTH/PTHrP precursor than are the mammalian proteins (see later). Indeed, in addition to duplicate copies of PTH and PTHrP genes, the puffer fish genome contains a fifth gene that encodes a protein containing amino acid residues characteristic of both PTH and PTHrP. This gene, called PTH-L, is phylogenetically an intermediary to PTH and PTHrP and may thus represent first definitive evidence for an ancestral gene from which the two divergent ligand forms evolved (see Fig. 1-3C and D).^48,75

The PTH Gene and Its mRNA

The human PTH gene consists of three exons located on chromosome 11p15.83–86 The first exon is 85 nucleotides in length and is noncoding (Fig. 1-5). Exon 2 (90 bp) encodes most amino acids of the prepropeptide sequence, whereas the third exon (612 bp) encodes the remainder of the propeptide sequence and all amino acids of the mature peptide, and it constitutes the 3′ noncoding region.⁸⁷ Several frequent intragenic polymorphisms (TaqI and PstI⁸⁸; BstBI⁸⁹; DraIII⁹⁰; XmnI⁹¹), and a tetranucleotide repeat ([AAAT]n;⁹²) have been identified in the human PTH gene, and some were shown to be informative in genetic linkage studies.^93–95 Two mRNAs that are 822 and 793 bp in length are derived in the human gene from the two transcriptional start sites, which follow two different functional TATA boxes that are separated by 29 bp.⁸⁷ Two closely spaced TATA boxes and two distinct transcripts are also derived from the bovine PTH gene, while rat and chicken PTH genes give rise to only one transcript; as a consequence of a long 3′ noncoding region, the transcript from the chicken PTH gene is unusually long and comprises 2.3 kb.^25,96 The genes encoding zebrafish PTH1 and PTH2 have a similar overall organization as the mammalian PTH genes.⁴⁷

PTH Biosynthesis and Intraglandular Processing

During the synthesis of the preproPTH molecule, the signal sequence, which comprises the 25-amino-acid-containing “pre” sequence, is cleaved off after entry of the nascent peptide chain into the intracisternal space bounded by the endoplasmic reticulum. A heterozygous mutation in this leader sequence, which changes a cysteine to an arginine at position −8 and thus impairs processing of preproPTH to proPTH, has been identified as the most plausible molecular cause of an autosomal dominant familial form of hypoparathyroidism.^97,98 The mutant hormone was found to be trapped intracellularly, predominantly in the endoplasmic reticulum (ER), leading to a marked up-regulation of ER stress-responsive proteins (BiP and PERK) and the proapoptotic transcription factor, CHOP, indicating that apoptosis-mediated parathyroid cell death is the likely cause of the observed hypoparathyroidism.⁹⁹

Subsequent to the removal of the pre-sequence, the pro-peptide is transported to the trans-Golgi network, where the pro-sequence (amino acid residues −6 through −1) is removed.¹⁰⁰ This latter process may involve furin (paired basic amino acid cleaving enzyme) and/or proprotein convertase-7 (PC-7), which are both expressed in parathyroid tissue; their expression levels do not appear to be regulated by either calcium or 1,25(OH)₂D₃.^101,102 After removal of the basic pro-sequence, the mature polypeptide, PTH(1-84), is packaged into secretory granules. Two proteases, cathepsins B and H, are subsequently involved in the intraglandular generation of carboxyl-terminal PTH fragments from the intact hormone; no amino-terminal PTH fragments appear to be released from the gland.^103–105 Since small or intermediate-size carboxyl-terminal fragments of PTH are unlikely to be involved in the regulation of calcium homeostasis, the intraglandular degradation of intact PTH is thought to represent an inactivating pathway, at least with regard to the regulation of mineral ion homeostasis. Consistent with this conclusion, hypercalcemia results in a substantial decrease in PTH secretion and, furthermore, favors the secretion of carboxyl-terminal PTH fragments, including a previously undetected large molecular species that are truncated at the amino-terminus (see following section).^105–108 However, recent studies have shown that some amino-terminally truncated PTH fragments, such as PTH(7-84), have hypocalcemic properties in vivo and can furthermore reduce the formation of osteoclasts in vitro.¹⁰⁹

The pool of stored, intracellular PTH is small, and the parathyroid cell must therefore have mechanisms to increase hormone synthesis and release in response to sustained hypocalcemia. One such adaptive mechanism is to reduce the intracellular degradation of the hormone, thereby increasing the net amount of intact, biologically active PTH that is available for secretion. During hypocalcemia, the bulk of the hormone that is released from the parathyroid cell is intact PTH(1-84).103–105,107,108 As the level of Ca²⁺_o increases, a greater fraction of intracellular PTH is degraded, and with overt hypercalcemia, most of the secreted immunoreactive PTH consists of biologically inactive C-terminal fragments.^10,25,26

Regulation of PTH Gene Expression

Another adaptive mechanism of the parathyroid cell to sustained reductions in Ca²⁺_o is to increase cellular levels of PTH mRNA, a response that takes several hours. A reduction in Ca²⁺_o increases, whereas an elevation in Ca²⁺_o reduces the cellular levels of PTH mRNA by affecting both its stability and the transcriptional rate of its gene.11,26,110,111 Available data suggest that phosphate ions also regulate, directly or indirectly, PTH gene expression. In the rat, hypophosphatemia and hyperphosphatemia, respectively, lower and raise the levels of mRNA for PTH through a mechanism that is independent of changes in Ca²⁺_o or 1,25(OH)₂D₃. An elevated extracellular phosphate concentration could thus contribute importantly to the secondary hyperparathyroidism frequently encountered in patients with end-stage renal failure, who often have chronically elevated serum phosphate concentrations.

Metabolites of vitamin D, principally 1,25(OH)₂D₃, also play an important role in the long-term regulation of parathyroid function and may act at several levels: by affecting the secretion of PTH and regulation of its gene, by regulating transcriptional activity of the genes encoding the calcium-sensing receptor (CaSR; see later) and the vitamin D receptor (VDR), as well as by regulating parathyroid cellular proliferation.11,26,110,112 1,25(OH)₂D₃ is by far the most important vitamin D metabolite that modulates parathyroid function. It acts through a nuclear receptor, the VDR, often in concert with other such receptors (i.e., those for retinoic acid or glucocorticoids), on DNA sequences upstream from the PTH gene (see Chapter 3).^113,114 1,25(OH)₂D₃-induced upregulation of VDR and CaSR expression in the parathyroid could potentiate its inhibitory action(s) on PTH synthesis and secretion.^11,26,110 Noncalcemic or less calcemic analogs of 1,25(OH)₂D₃ inhibit PTH secretion while producing relatively little stimulation of intestinal calcium absorption and bone resorption^115–117 and may thus be attractive candidates for treating the hyperparathyroidism of chronic renal insufficiency.

Adjustment of the rate of parathyroid cellular proliferation is the third adaptive mechanism contributing to changes in the overall secretory activity of the parathyroid gland. Under normal conditions, parathyroid cells have little or no proliferative activity. The parathyroid glands, however, can enlarge greatly during states of chronic hypocalcemia, particularly in the setting of renal failure, probably because of a combination of hypocalcemia, hyperphosphatemia, and low levels of 1,25(OH)₂D₃ in the latter condition.

Regulation of Parathyroid Hormone Secretion

A large number of factors modulate PTH secretion in vitro,11,26,118 but most of these factors are not thought to control hormonal secretion in vivo in a biologically relevant manner. Therefore, we focus in this section principally on factors that are the most physiologically meaningful regulators of PTH secretion—that is, the extracellular ionized calcium concentration itself (Ca²⁺_o), 1,25(OH)₂D₃, and the level of extracellular phosphate ions. Of these three, Ca²⁺_o is most important in the minute-to-minute control of PTH secretion. Indeed, the actions of 1,25(OH)₂D₃ and phosphate ions on the secretion of PTH probably result at least in part from their effects on hormonal biosynthesis rather than secretion per se.^11,26,118 Ca²⁺_o also modulates several other aspects of parathyroid function that indirectly affect PTH secretion, including PTH gene expression, the hormone’s intracellular degradation, and parathyroid cellular proliferation, as described previously. Recent data have shown that novel factors playing key roles in phosphate homeostasis, especially FGF-23 and α-klotho (a coreceptor for FGF receptors), also modulate parathyroid function, inhibiting^119,120 and enhancing¹²¹ parathyroid function, respectively. Our rapidly improving understanding of how these factors participate in phosphate homeostasis is described in detail in Chapter 6.

Metabolism of Parathyroid Hormone

Studies performed over more than 3 decades by several laboratories have focused on the heterogeneity of circulating forms of PTH, which was first identified by Berson and Yalow in 1968 (Fig. 1-8).¹⁷⁰ From these investigations, it is now apparent that in addition to the full-length polypeptide PTH(1-84), which is the biologically active hormone, much of the circulating hormone lacks an intact N-terminus, and most of these fragments are thus devoid of biological activity, at least with regard to the PTH/PTHrP receptor-mediated regulation of mineral ion homeostasis.²⁵ C-terminal PTH fragments are produced in and released from the parathyroid gland, but they are also derived from circulating intact hormone by efficient, high-capacity degradative systems in the liver and kidney and most likely at other peripheral sites (Fig. 1-9). However, some PTH fragments, such as PTH(7-84), appear to be generated within the gland and to have some biological activity.^171–173

Direct measurement of arterial and venous differences in parathyroid effluent blood (with vigorous conditions to prevent any ex vivo cleavage of hormone after sample collection) were performed in cattle and confirmed that smaller C-terminal fragments and intact hormone, but not N-terminal fragments, are secreted into the circulation. The relative concentration of these C-terminal fragments released from the gland increases under conditions of systemic hypercalcemia, when overall secretion rates of intact hormone are lower.^105,174 The C-terminal PTH fragments are similar to those generated by peripheral metabolism (see later) but were not chemically characterized.

The peripheral metabolism of PTH has been analyzed by injecting intact hormone into the circulation of test animals. Such experiments have not been performed in human subjects, but it is assumed that the similar metabolism of PTH in rats, dogs, and cows is reflective of its metabolism in humans.¹⁷⁵ Clearance of intact PTH from plasma was found to be very rapid (half-life, 2 to 4 minutes),^176,177 the major sites of clearance being the liver and kidney. Clearance by the liver predominates over clearance by the kidney; the two organs together account for virtually all clearance of intact hormone. Hepatic clearance of intact hormone has been estimated to be 40% to 75% and renal clearance 20% to 30%.²⁵

In summary, current evidence indicates that intact PTH and multiple C-terminal fragments (which are derived from glandular and peripheral cleavage and may not be identical) are the principal circulating hormonal forms (however, recent data suggest the presence of a previously unrecognized large N-terminally truncated form of the hormone [discussed later]). Biologically active N-terminal fragments of PTH, if found in the circulation at all, are likely to circulate only at extremely low concentrations (<10-13 to 10-14 mol/L). More recent in vivo evidence regarding the renal clearance and metabolism of intact PTH (as distinct from C-terminal fragments) indicates a peritubular uptake process, rather than glomerular filtration followed by uptake from the tubular lumen and subsequent cleavage.¹⁷⁸ Furthermore, other studies indicate that megalin, a multifunctional endocytic receptor expressed in the proximal renal tubules, can mediate the reuptake and subsequent degradation of the fraction of PTH that is subject to glomerular filtration.¹⁷⁹ Megalin-mediated uptake depends on an intact N-terminus of PTH; C-terminal fragments that are eliminated by glomerular filtration are not recognized by megalin. The potential significance, quantitatively and biologically, of glomerular filtration and megalin-mediated uptake of intact PTH therefore remains uncertain. However, megalin-ablated mice excrete fourfold more N-terminal PTH in the urine than do wild-type animals.¹⁷⁹

Because of the lack of evidence of circulating forms of biologically active, N-terminal fragments, it was feasible to introduce immunometric assays that use two different antibodies: an immobilized capturing antibody directed against the C-terminal portion of PTH(1-84) and a radiolabeled or enzyme-labeled detection antibody directed against an epitope within the N-terminal portion of the intact molecule^180,181 (see Fig. 1-9).

Although these two site assays have been clinically useful and provided a great improvement over earlier assays,¹⁸⁰ recent studies have demonstrated the presence of circulating PTH fragments that are different in character and composition from any of the hormone fragments discussed previously.^108,128,182 The studies that led to the detection of these PTH species were at least partly stimulated by the clinical observations that two-site immunometric assays to measure intact PTH frequently gave high levels of nonsuppressible PTH in patients with end-stage renal disease, yet the patients had clinical and/or histologic evidence of a dynamic bone disease^{108,183–186} (see Chapter 14).

High-performance liquid chromatography analysis of blood samples from such uremic patients and from patients with other forms of hyperparathyroidism revealed two distinct immunoreactive peaks in column effluent that were detectable (although with differing sensitivity) by different commercial assays. One peak corresponded to intact PTH(1-84), whereas the other peak migrated close to the position occupied chromatographically by synthetic PTH(7-84).^108,128 This finding suggested that the epitope of the detection antibody in these assays did not require the presence of the first six or more amino acids of PTH(1-84).^128,129 The results were interpreted as being consistent with the view that the molecular entity or entities detected besides PTH(1-84) are N-terminally truncated forms of the intact molecule that are similar, but not necessarily identical, to synthetic PTH(7-84).¹²⁸ In fact, more detailed chromatographic studies performed with one of these “intact” PTH assays showed significant variation in the ratio of the N-terminally truncated PTH to PTH(1-84). Individuals with normal renal function showed a lower ratio than did patients with uremia; furthermore, the percentage of immunoreactivity representing N-terminally truncated forms of PTH (PTH[7-84]) rose in both groups when hypercalcemia was present.¹⁰⁸

The newer immunoradiometric assays that have N-terminal epitopes at the extreme amino-terminus of PTH(1-84) show significantly lower PTH concentrations in patients with chronic renal failure during stimulation and suppression of glandular activity with alterations in calcium, a result consistent with the conclusion that only full-length and therefore biologically active PTH(1-84) is detected by these assay systems.173,182,187

The findings outlined previously were expected to have considerable significance for the management of patients with parathyroid dysfunction, especially in the presence of renal failure. Treatment of patients with end-stage renal disease with large amounts of vitamin D and/or calcium (or calcimimetics) has been associated with adynamic bone disease, which can be deleterious, particularly in growing children.188–190 Particularly during hypercalcemia, N-terminally truncated forms of PTH become a significant, if not the dominant, PTH species.^191,192 Measurement of “intact” PTH by earlier assays¹⁸⁰ therefore overestimated the concentrations of biologically active PTH and could result in the overtreatment of uremic patients with vitamin D-analogs and/or calcium. The resulting suppression of the parathyroid gland, even without evoking an inhibitory effect of a fragment similar to PTH(7-84), could be acting together to excessively reduce PTH-dependent bone turnover. When secretion of PTH is suppressed, as in hypercalcemia, and the fractional concentration of the large N-truncated PTH increases, however, inhibition of the actions of native PTH on calcium or bone could occur. Recent experiments with synthetic PTH(7–84) support the conclusion that such actions occur in vivo and in vitro. For example, PTH(7-84) was shown to have hypocalcemic properties in vivo,^171,173 and it reduces, presumably via receptors specific for C-terminal PTH fragments, bone resorption that may be partly due to impaired osteoclast differentiation.^109,171 It is not yet established clinically whether the newer radioimmunoassays that do not detect large but amino-terminally truncated PTH molecules will have a decisive advantage in diagnosis and management of primary and secondary hyperparathyroidism¹⁸⁷ (see Chapters 7 and 14).

By contrast to the extensive metabolism of PTH, recent evidence suggests that intact FGF23, which is most likely the most important phosphate-regulating hormone,²³ does not appear to undergo significant metabolism.¹⁹³ In fact, immunoreactive FGF23, which can be extraordinarily elevated in patients with end-stage renal disease, appears to be largely intact and biologically active. This makes it unlikely that FGF23, unlike PTH, is metabolized in peripheral organs.

Actions of Parathyroid Hormone On Kidney

In addition to regulating renal calcium and phosphate transport, PTH modifies the tubular handling of magnesium, sodium, potassium, bicarbonate, and water and stimulates renal gluconeogenesis. Specific regions of the nephron that are involved in each of the PTH-dependent actions have been defined through in vivo micropuncture analyses and through in vitro studies with nephron segments that have been dissected from the remainder of the kidney. Furthermore, cell lines from specific regions of the kidney have also been used. However, interpretation of these results has to take into account the complexity of the organization of the kidney as an organ (for example, the complex anatomic relationship involving different renal tubular segments spanning both the cortex and medulla and the effects of countercurrent distribution that modify solute and water transport), particularly since certain in vivo features of renal tubules are not readily imitated in vitro through the use of isolated tubules or cells.

Renal Calcium Reabsorption

Most of the calcium in the glomerular filtrate is reabsorbed. The bulk of this reabsorption (65%) occurs via passive, paracellular mechanisms, both in the proximal tubules and, to a lesser extent, in the thick ascending limb of Henle’s loop and the distal convoluted tubule.194–198 As noted by Diaz and colleagues,²⁶ the calcium-sensing receptor plays an important PTH-independent role in the adjustment of renal calcium reabsorption in the cortical thick ascending limb. The physiologically important stimulation of renal calcium reabsorption by PTH occurs almost entirely in the distal nephron. In the cortical thick ascending limb, PTH increases the magnitude of the lumen positive potential that drives the passive paracellular reabsorption of calcium and magnesium. In the distal convoluted tubule, in contrast, PTH promotes increased transcellular Ca²⁺ reabsorption by coordinate up-regulation of molecular components of this transcellular pathway.¹⁹⁹ PTH enhances uptake of Ca²⁺ into the tubular cells via the (luminal) plasma membrane channel, TRPV5,^200,201 as well as its active extrusion against a steep electrochemical gradient at the basolateral membrane. This latter process of extrusion involves two types of transporters. One is the plasma membrane calcium pump (Ca²⁺, Mg²⁺-ATPase, PMCA) and the second is a Na⁺/Ca²⁺ exchanger (NCX1), which in turn is indirectly regulated by NA⁺/K⁻-ATPase(s), which maintains the transcellular Na⁺ gradient. Studies performed with membrane vesicles from the distal region of the kidney show increased activity of the Na⁺/Ca²⁺ exchanger in response to PTH.²⁰² PTH also stimulates the translocation of preformed Ca²⁺ channels sequestered within the interior of certain distal tubular cells, presumably TRPV5, apical surface (Fig. 1-10),²⁰³ which translocate to the apical (i.e., luminal) surface of the renal tubular epithelial cell and mediate increased cellular Ca²⁺ uptake. Because PTH simultaneously enhances the activity of Na⁺/Ca²⁺ exchangers in the basolateral (antiluminal) membrane, the overall process promotes an increase in transcellular Ca²⁺ uptake from the lumen to blood, that is, from the apical to the basolateral membrane.

Regulation Of 1α- and 24-Hydroxylase Activity

PTH is a major inducer of the activity of proximal tubular 1α-hydroxylase, a microsomal cytochrome P-450 enzyme that synthesizes biologically active 1,25(OH)₂D₃ from the substrate 25(OH)D₃.204–206 This effect of PTH on synthesis of the renal enzyme shows longer lag times than its effect on renal Ca²⁺ transport and is mediated, at least in part, by the protein kinase A (PKA) signaling pathway of the PTH/PTHrP receptor.^207,208 Although 1α-hydroxylase activity has been detected in several nonrenal tissues, its mRNA was found in abundant concentrations only in the kidney, thus confirming that this organ is the most important site for the generation of 1,25(OH)₂D₃. PTH-dependent synthesis of new 1α-hydroxylase protein requires several hours and is blocked by 1,25(OH)₂D₃ and actinomycin D, a blocker of protein synthesis.^208–210 PTH administration increases the mRNA encoding the 1α-hydroxylase.²⁰⁸ Hypophosphatemia is, similar to PTH, a major inducer of 1α-hydroxylase, whereas hypercalcemia, as would be generated by sustained increases in circulating levels of PTH or PTHrP, suppresses synthesis of the enzyme, thus limiting overall 1,25(OH)₂D₃ synthesis in a homeostatic manner. The molecules 25(OH)D₃ and 1,25(OH)₂D₃ can also be hydroxylated by the 24-hydroxylase, but the resulting metabolites, 24,25(OH)₂D₃ and 1,24,25(OH)₃D₃, appear to have no major role in the regulation of mineral ion homeostasis (see Chapter 3). However, PTH has an inhibitory effect on the 24-hydroxylase, thus reducing the inactivation of 1,25(OH)₂D₃; in contrast, 1,25(OH)₂D₃ stimulates the synthesis of 24-hydroxylase, thereby inducing its own metabolism.^211,212

Renal Phosphate Transport

PTH is of major importance for maintaining normal blood calcium levels, but it is not the principal regulator of the serum phosphate concentration, which appears to depend on specific phosphaturic factors, particularly FGF-23 (see Chapter 6).^19,23,213 However, when PTH causes an increase in bone resorption (as might occur with prolonged dietary calcium deprivation), calcium and phosphate increase simultaneously in the blood. Although calcium is needed, phosphate is best excreted, which is mainly accomplished by a PTH-stimulated increase in renal phosphate clearance.

PTH acts directly on proximal tubular cells, where it regulates expression of NPT2a and NPT2c in the brush border membrane. Whereas NPT2a is expressed in segments S1 through S3 of the proximal tubule, NPT2c is expressed only in the S1 segment. NPT2a protein undergoes internalization in response to PTH,²¹⁴ followed by lysosomal degradation²¹⁵; in contrast, recent evidence suggests that NPT2c can be recycled and reinserted into the brush border membrane.²¹⁶ In the proximal renal tubules, PTH furthermore enhances the production of biologically active 1,25(OH)₂ vitamin D.^217–220 All these actions of PTH are mediated through the PTH/PTHrP receptor (PTH1R), which is expressed at the basolateral membrane (BLM) and at much higher levels at the apical brush border membrane (BBM) of the proximal renal tubules.^{217,221–223}

The PTH-dependent renal actions probably involve cAMP/PKA-dependent and Ca²⁺/IP₃/PKC-dependnent signaling events at the BLM, and both pathways appear to contribute to the reduction in proximal tubular phosphate reabsorption.217,218,220 In contrast to these dual signaling properties of the PTH1R at the BLM, there is considerable evidence to suggest that the inhibition of phosphate uptake mediated through the PTH1R at the BBM involves predominantly, if not exclusively, a pertussis-toxin-sensitive, PKC-dependent pathway.^{217,224–227} PTH thus activates both major signaling pathways via PTH/PTHrP receptors located at either the BLM or the BBM, and it induces phosphaturia by reducing expression of the type II sodium-phosphate cotransporters, NPT2a and NPT2c.

Activation of the cAMP/PKA pathway down-stream of the PTH1R is undoubtedly involved in the PTH-dependent regulation of NPT2a expression.226–228 In fact, patients affected by pseudohypoparathyroidism type Ia (PHP-Ia), a disease caused by inactivating mutations in GNAS, the gene encoding the alpha subunit of the stimulatory G protein (G_sα), develop hyperphosphatemia due to a lack of functional G_sα in the proximal tubules and consequently show impaired urinary cAMP and phosphate excretion in response to PTH (see Chapter 6). However, the phosphaturic response stimulated by PTH is not totally absent, since PHP-Ia patients have a small but delayed increase in urinary phosphate excretion after challenge with PTH, which suggests that signaling molecules other than cAMP could also be involved in promoting phosphate excretion.^229–231

Besides regulating NPT2a expression, PTH also affects expression of the second kidney-specific sodium-dependent phosphate cotransporter, namely NPT2c.²²⁷ The mRNA encoding NTP2c is about 10-fold less abundant than that encoding NPT2a, which initially suggested that NTP2c plays only a minor biological role limited to some period during postnatal development.²³² It was then found, however, that homozygous and compound heterozygous mutations in NPT2c cause hereditary hypophosphatemic rickets with hypercalciuria (HHRH), an autosomal-recessive disorder.^233–236 This disease linkage makes it certain that NPT2c serves important functions in biology and is not redundant with NPT2a.^218–220 Nevertheless, it is so far unknown how the two transporters differ in their functional roles. Both proteins are internalized at the BBM in response to treatment with PTH, but the time courses of these agonist-dependent processes differ, raising the possibility that different scaffolding proteins regulated by PTH underlie the differential regulation of NPT2a and NPT2c. Furthermore, activation of different signaling pathways downstream of the PTH1R—namely, the cAMP-dependent activation of PKA or the Ca²⁺/IP₃-dependent or Ca²⁺/IP₃-independent activation of PKC—may have different roles in the regulation of NPT2a and NPT2c expression (Fig. 1-11).

These differences may involve “signalsomes,” that is, specialized intracellular domains in which proteins involved in a specific biochemical process or pathway are brought into close proximity by binding to “scaffold” proteins, thereby limiting their diffusion and altering receptor-mediated downstream signaling pathways.²³⁷ In renal proximal tubules, the sodium/hydrogen exchanger regulatory factors 1 and 2 (NHERF1/2) are prominent scaffold proteins that contain two PDZ (psd-95, discs large, ZO-1)-binding domains and a C-terminal ERM (ezrin, radixin, moesin)-binding motif (see review²³⁸). The PTH1R interacts with NHERF proteins through a C-terminal PDZ-binding motif ^239,240 (see Fig. 1-11), which leads in some cell lines to NHERF-dependent suppression of PTH-stimulated cAMP accumulation.²³⁹ In opossum kidney cells (OK), NHERF1 enhances PTH1R signaling through intracellular calcium and PLC.²⁴⁰ In fact, anchoring the receptor to the BBM of proximal tubules appears to promote signaling through the IP₃/PKC pathway, illustrating how domain-specific factors can influence receptor-dependent signaling.

Similar to the PTH1R, NPT2a also contains a C-terminal motif that binds to PDZ domains of NHERF1,²⁴¹ and NHERF1-null mice display phosphate wasting due to a reduction in NPT2a expression at the BBM of proximal tubules. Thus, this transporter requires a NHERF1-assembled scaffold for proper membrane expression.²⁴² Furthermore, NHERF1 has been shown to be an essential factor for PTH-mediated regulation of NPT2a expression in OK cells and in primary proximal tubular cell cultures isolated from NHERF1-null mice. Combined, these findings demonstrate that NHERF1 not only establishes NPT2a surface expression but also aids in the formation of a regulatory complex necessary for PTH-elicited phosphaturia. In vitro, NHERF1-assembled complexes consisting of PTH1R and NPT2a are present in apical domains of OK cells and in genetically modified LLC-PK1 cells developed for studying phosphate transport.²⁴³

The PTH fragment peptide, PTH(3-34), has been frequently used in efforts to discern the signaling events at the BBM from those at the BLM which regulate NPT2a and NPT2c; some data suggest that this analog may selectively activate the PKC signaling pathway244–246 and thus should act only at the BBM. There are, however, no data demonstrating that PTH(3-34) can directly activate the PLC-dependent formation of IP₃, and in fact substitutions or deletions at positions 1 and 2 of PTH result in markedly and/or selectively impaired activation of the PLC/IP₃/Ca²⁺ signaling response.^58–60,247 PTH(3-34) has been shown to activate PKC via a non-PLC/IP₃-dependent mechanism, perhaps involving another phospholipase such as PLA₂.²⁴⁶ In any case, several investigators have shown that PTH(3-34), as well as other PTH and PTHrP fragments truncated at the amino-terminus, can inhibit the uptake of phosphate by opossum kidney cells (which do not express significant amounts of NPT2c) in the apparent absence of significant cAMP accumulation.^248–252 Furthermore, PTH(3-34) was shown to promote, at least partially, urinary phosphate excretion in animals.²⁵³ This raises the possibility that PTH fragments such as PTH(7-84) and PTH(4-84), which are found in the circulation of patients with primary or secondary hyperparathyroidism, could induce phosphaturic effects.^{172,191,192,254,255} PTH(7-84), however, does not appear to have phosphaturic activity when tested in vivo, or in some in vitro systems. In addition, PTH(3-34) can activate at least partially the cAMP-dependent PKA in cells.^256,257 This makes it plausible that the capacities of amino-terminally truncated PTH or PTHrP analogs to promote the urinary excretion of phosphate arise at least to some extent from partial activation of the cAMP/PKA pathway at the BLM.

As mentioned, PTH treatment leads to the reduction of NPT2a and NPT2c in the kidney, but effects of the hormone on the two cotransporters appear to differ, as shown in thyro-parathyroidectomized rats.218–220,227 NPT2a is expressed in the segments S1-S3 of the proximal renal tubules, whereas NPT2c is present only in S1.^{218–220,227} NPT2a is the more abundantly expressed cotransporter, and it handles (in the mouse) 70% to 80% of renal phosphate reabsorption, with NPT2c accounting for the remaining 20% to 30%. In response to PTH, NPT2a disappears rapidly from the BBM and translocates to lysosomes where it undergoes degradation, whereas NPT2c disappears from the BBM surface at a much slower rate and does not seem to undergo lysosomal degradation but rather may recycle back to the BBM surface.²¹⁶ The mechanisms underlying these diverse response profiles are uncertain at present, but the generation of engineered LLC-PK1 cells that show PTH-dependent inhibition of phosphate transport via reductions in either NPT2a or NPT2c expression²⁴³ will likely lead to important new insights into the molecular and cellular processes involved.

Actions of Parathyroid Hormone On Bone

PTH affects a wide variety of the highly specialized bone cells, including osteoblasts, osteoclast/stromal cells, and osteocytes, the latter being the most numerous cell type in bone. Some of these effects reflect direct actions of PTH; others are indirect and mediated in an autocrine/paracrine manner through factors released by cells (osteoblasts) expressing PTH/PTHrP receptors that regulate the activity of yet other cells (osteoclasts) that lack these receptors.258–260

PTH action on bone cells can be considered from at least five distinct perspectives. First is its major physiologic role—the maintenance of calcium homeostasis. As noted in Fig. 1-2, this action can be analyzed most clearly in experimental animals through parathyroidectomy, with resultant hypocalcemia. Renal losses of calcium are significant contributors to the severe, sometimes fatal hypocalcemia, but the loss of PTH action on bone cells is the predominant cause. The traditional explanation is that the hypocalcemia is due principally to the loss of PTH stimulation of osteoclastic bone resorption, but other mechanisms may be involved (see later). A second perspective is the action of PTH when administered pharmacologically, resulting in elevation of serum calcium, immediately preceded by a short-lived fall in calcium. It is still unclear which specific cellular actions explain the physiologic actions known to follow PTH administration to animals in vivo. Earlier studies had shown that the administration of PTH leads within minutes to a transient lowering of blood calcium caused by uptake of the mineral by bone cells,²⁶¹ which is rapidly followed by increased mobilization of calcium from the mineral phase into the bloodstream.^261,262 Although there continues to be uncertainties regarding the cellular/anatomic basis of these rapid responses to PTH, considerable progress has been made in elucidating the mechanisms through which bone cells respond to PTH and to other autocrine/paracrine factors such as cytokines (see Chapters 4 and 12). A major pathway for calcium release from bone involves osteoclasts; these cells undergo multiple cellular changes involving the activation of cellular transporters and pumps, as well as the secretion of enzymes such as cathepsin K and collagenases, but other mechanisms of calcium release are proposed.^261,262

A third perspective regarding PTH action on bone cells can be appreciated through results seen when PTH is used as a therapy for osteoporosis. Chronic administration of PTH, if given intermittently, stimulates bone formation, an action that involves a complex set of cellular responses in bone, affecting principally stromal cell/osteoblast proliferation, differentiation, and cellular actions.263–267 It is also clear that actions on osteocytes are involved in PTH actions on bone^267,268; together, PTH effects on the two cell types, osteoblasts and osteocytes, stimulate bone matrix formation and bone mineral deposition, as well as bone mineral mobilization (see Chapters 4 and 5).

Fourthly, pathophysiologic actions of PTH are seen in hyperparathyroidism with chronic excess levels of PTH. In this situation, both bone formation and bone resorption are stimulated, the latter effect being predominant (see Chapters 4 and 7).

Finally, one must consider parathyroid hormone action on bone (and that of the closely related PTHrP) from the important role played, along with multiple other hormones and cytokines, in the complex system biology of embryonic bone formation and bone formation and remodeling in the adult. Numerous other hormones and autocrine/paracrine-acting cytokines interact with PTH and PTHrP in this still incompletely understood complex cellular biology of bone.

The PTHrP Gene In Comparison To The PTH Gene

The human PTHrP gene is located on chromosome 12p12.1-11.2,²⁹⁸ which has a region analogous to that containing the human PTH gene on chromosome 11p15.^83–85 Both the PTH and the PTHrP genes have a similar organization, including equivalent positions of the boundaries between some of the coding exons and the adjacent introns^{11,87,298,307} (Fig. 1-12).

Like the PTH gene, the PTHrP gene contains a single exon that encodes most amino acid residues of the prepropeptide sequence, and both genes have an exon that encodes the remainder of the propeptide sequence, that is, two basic residues (Lys and Arg) that are required for endoproteolytic cleavage of the mature peptide, and either all or most of the amino acids of the secreted peptides. The similarities in their protein sequence and in the structure of their genes, as well as the overlap in some of their functional properties, confirmed that both peptides are derived from a common ancestor.4,6,293 By now, both mammalian genes have diverged considerably; for example, in contrast to the less complex PTH gene, which gives rise to a single gene product, the PTHrP gene uses at least three different promoters and alternative splice patterns that lead to the synthesis of several different mRNA species encoding peptides with different C-terminal ends.^{6,293,298,307} A polymorphic dinucleotide repeat sequence that has been used for genetic linkage studies is located downstream of exon 4 (which is exon 6 according to a different nomenclature) of the human PTHrP gene³⁰⁸ (see Fig. 1-12); thus far, no human disorder has been discovered that is caused by mutations in the PTHrP gene.

When compared with each other, chicken PTHrP and the known mammalian species of PTHrP show strong amino acid sequence homology within the first 111 residues; the degree of amino acid sequence conservation of PTHrP is considerably higher than that of the known PTH species (see Fig. 1-3). The amino acid sequence homology between both peptide families is restricted to the N-terminal portion, where 6 of the first 12 amino acid residues are conserved in all PTH and PTHrP species (see Fig. 1-3); the middle and C-terminal regions of both peptides share no recognizable similarity.

PTHrP is likely to undergo extensive posttranslational processing, resulting in peptide fragments, some of which may have biological properties distinct from those of PTHrP(1-34) on calcium and bone metabolism. The N-terminal PTHrP fragment functionally homologous to PTH is derived through cleavage at amino acid residue Arg37.^309,310 PTHrP(1-36) interacts efficiently with the PTH1R,^7–9,311 and it has an in vivo efficacy similar to that of PTH(1-34).^312–314 Longer, glycosylated fragments of N-terminal PTHrP were also described; however, these forms appear to be generated predominantly by skin-derived cells, and their biological role, if different from that of PTHrP(1-34) or PTHrP(1-36), remains to be established.³¹⁵

In addition to the biologically active N-terminal PTHrP fragment, different C-terminal fragments are generated and accumulate in patients with end-stage renal disease.³¹⁶ A PTHrP fragment consisting of amino acids 107 to 139 and a shorter peptide, PTHrP(107-111), may be relevant to the control of bone metabolism because both fragments were shown to inhibit osteoclastic bone resorption^317,318 and to stimulate osteoblast activity and proliferation.³¹⁹ Although some investigators, using modified experimental conditions, were unable to confirm the in vitro findings with osteoclasts,³²⁰ more recent data indicate that PTHrP(107-139) reduces the number of osteoclasts and inhibits bone resorption in vivo.³²¹ Taken together, these findings suggest that C-terminal PTHrP fragments may have a role in regulating bone resorption and/or formation (Table 1-2).

Cleavage at amino acid residue Arg37 also generates PTHrP(38-94) amide, a PTHrP fragment that could be of considerable importance in maintaining fetal calcium homeostasis.^309,322 This peptide is found in the circulation³²³ and appears to interact with a distinct receptor that signals through changes in intracellular free calcium and is likely to be an important regulator of transplacental calcium transfer.^{309,322,324–326} Studies with PTHrP(38-94) amide and PTHrP(67-86) amide have shown that these fragments increase the blood calcium concentrations of parathyroidectomized fetal lambs and PTHrP-ablated murine fetuses, respectively^309,325,326 (see later). These results confirmed earlier data that had provided the first evidence for an important role of PTHrP in the regulation of fetal calcium homeostasis.³²⁷

Functions of PTHrP

The first actions of PTHrP to be defined were the PTH-like actions associated with the humoral hypercalcemia of malignancy.4,6,293 In this pathologic circumstance, PTHrP acts like a hormone, that is, it is secreted from the tumor into the bloodstream and then acts on bone and kidney to raise calcium levels (see Chapter 7). Whether PTHrP circulates at high enough levels in normal adults to contribute at all as a hormone to normal calcium homeostasis is an unanswered question; the levels are certainly low, and patients with congenital or acquired hypoparathyroidism are hypocalcemic despite the presence of PTHrP. Although incomplete and somewhat conflicting, growing evidence suggests, however, that PTHrP may act as a hormone in two special circumstances: during fetal life and during lactation as outlined in the next sections.

Three Parathyroid Hormone Receptor Subtypes

Subsequent to cloning of the PTH/PTHrP receptor, or the PTH1R, different efforts led to the identification of two novel receptors closely related to the initially isolated receptor. One of these receptors, called the PTH2 receptor (PTH2R), was obtained from a human brain cDNA library³⁷⁷ and found to have reactivity towards PTH but not towards PTHrP.^377–380 The PTH2R from rat, however, responded poorly to both PTH and PTHrP.^381,382 A search for the true ligand for this PTH2 receptor led to the isolation of a hypothalamic peptide of 39 amino acids (TIP39) called tubular infundibular peptide which efficiently binds to and activates both rat and human PTH2Rs.³⁸³ TIP39 shows weak amino acid sequence homology to the N-terminal 34 amino acids of PTH and PTHrP, as well as some overlap in secondary structure (see Fig. 1-3).³⁸⁴ Intact TIP39 binds relatively weakly to the PTH1R, but deletion of the first seven residues results in a high-affinity PTH1R antagonist.^385–387 The peptide TIP39 is produced at the hypothalamus and within the testes,^383,388 and together with the PTH2R appears to play an autocrine/paracrine role in germ cell development, inasmuch as mice lacking the gene for TIP39 are sterile due to a failed formation of spermatatids.³⁸⁹

The second novel PTH receptor variant isolated, the PTH3 receptor (PTH3R), has so far been found in fish³⁹⁰ and birds (see NCBI chicken genome website) but not in mammals (Fig. 1-15). Functional and phylogenetic analyses indicate that the PTH3R is more closely related to the PTH1R than to the PTH2R.³⁹⁰ In addition to the three identified PTH receptor subtypes, there are functional and physicochemical data suggesting that there may be additional receptors that interact with portions of either PTH or PTHrP which are C-terminal of the principal bioactive region defined by the (1–34) segment (see following and Table 1-2). As of yet, however, no gene or cDNA encoding such a novel receptor has been identified.

The PTH/PTHrP Receptor Type-1—Evolution, Gene Structure, and Protein Topology

The PTH1R belongs to a distinct subgroup of G protein–coupled receptors called the class B, or secretin family GPCRs. The first cDNAs encoding the PTH/PTHrP receptor were isolated through expression cloning techniques from two different model cell lines: OK cells, an opossum kidney–derived proximal tubule-like cell, and ROS 17/2.8 cells, a rat osteosarcoma-derived osteoblast-like cell.^7,8 Subsequently, cDNAs encoding the human,^9,391,392 mouse,³⁹³ rat,³⁹⁴ chicken,¹³ porcine,³⁹⁵ dog,³⁹⁶ frog,³⁹⁷ and fish PTH/PTHrP receptors³⁹⁸ were isolated through hybridization techniques from different tissue sources, i.e., kidney, osteoblast-like cells, brain, embryonic stem cells, and/or whole embryos. Northern blot and in situ studies^304,399,400 and data provided through available public EST (expressed sequence tag) databases confirmed that the PTH/PTHrP receptor is expressed in a wide variety of fetal and adult tissues. The tetraploid African clawed frog, Xenopus laevis, expresses two nonallelic isoforms of the PTH/PTHrP receptor,³⁹⁷ whereas all other investigated species have only one copy of the receptor gene per haploid genome.

The possible existence of PTH/PTHrP receptor subtypes with distinct, organ-specific characteristics had been suggested by distinct ligand binding properties401–403 or by the activation of distinct second messenger pathways in different clonal cell lines.^251,404,405 However, the cloning of identical PTH/PTHrP receptors from human kidney, brain, and bone-derived cells^9,391,392 indicated that the previously observed pharmacologic differences could likely be explained by species-based variations in receptor primary sequence, rather than to tissue-based differences in expression of various receptor subtypes.

The gene encoding the human PTH/PTHrP receptor is located on chromosome 3p (within the region 3p21.1-3p24.2). Its intron/exon structure406–408 is largely preserved in the genes encoding the rat and mouse receptor homologs.^409,410 The gene spans at least 20 kb of genomic DNA and consists of 14 coding exons and at least three noncoding exons. The size of the coding exons in the human gene ranges from 42 bp (exon M7) to more than 400 bp (exon T), and the introns vary in length from 81 bp (intron between exons M6/7 and M7) to more than 10,000 bp (intron between exons S and E1). Two promoters for the PTH/PTHrP receptor, P1 and P2, have been described in rodents.^409–412 The activity of P1 (also referred to as U3) is mainly restricted to the adult kidney, while that of P2 (also referred to as U1) is detected in several fetal and adult tissues, including cartilage and bone. In humans, a third promoter (P3, also referred to as S) appears to control PTH/PTHrP receptor expression in some tissues, including kidney and bone.^408,413,414

The expressed human PTH/PTHrP receptor protein is 593 amino acids in length, including the 22-amino-acid N-terminal signal peptide sequence that is removed in the mature receptor. The predicted topology is defined by a relatively large amino-terminal extracellular domain (∼160 amino acids in the human PTH1R), a juxtamembrane, or J, region containing seven membrane spanning helices and interconnecting loops, and a carboxy-terminal tail of about 110 amino acids (see Fig. 1-14). In mammals, the PTH1R N-domain is encoded by five exons: S, E1, E2, E3, and G. Exon S and G encode the signal peptide and a glycosylated segment,⁴¹⁵ respectively. Exons E1, E3, and G encode segments that include a number of conserved residues that are likely required for proper folding of the protein, surface expression, and/or ligand binding. These residues include six cysteine residues that form a disulfide bond network that stabilizes the N-domain structure.^68,416,417 Exon E2 encodes a non-conserved 41-amino-acid segment that is not essential for receptor function, since it can be targeted for deletion or large insertions, such as with green fluorescent protein (GFP), without a major loss in ligand binding affinity or surface expression.^416–418 Indeed, an equivalent of exon E2 is absent in the PTH1Rs from all nonmammalian species, as well as in other members of this GPCR subfamily (mammalian and nonmammalian), including the PTH2R.^390,397,419 The E2 segment of the mammalian PTH1R gene might thus reflect an evolutionarily recent genetic insertion event³⁹⁰ or perhaps an atavistic genomic remnant. In any event, the E2 segment appears to form a nonfunctional loop that is conformationally flexible; a recent x-ray crystallographic analysis of the PTH1R N-terminal domain structure found this segment to be disordered in the crystal and hence unresolved.⁶⁸

The initial cloning of the PTH1 receptor, cDN,^7,8 along with the cDNAs for the receptors for secretin⁴²⁰ and calcitonin,⁴²¹ led to the realization that these receptors form a distinct GPCR subfamily, now called the class B or secretin family receptors.^419,422,423 These family B receptors share a similar overall structural topology, which is defined by the large N-terminal extracellular domain containing the conserved disulfide bond network, as well as a number of other conserved amino acids dispersed throughout the receptor protein.⁴¹⁹ Protein alignment analyses reveal virtually no amino acid sequence homology between the family B GPCRs and the receptors comprising the four other major GPCR subgroups, including the well studied β₂-adrenergic receptor and rhodopsin, representing the family A GPCRs.⁴²² The structural similarity between these groups of receptors is therefore limited to the common use of a heptahelical domain organization for the membrane-spanning J region.^422,424 As an evolutionary protein class, the secretin receptor family appears to have arisen during the divergence of the metazoans. Representatives of the receptor family are found in the genomes of a variety of invertebrate species, including the insect Drosophila melanogaster, and the nematode Caenorhabditis elegans, but not in yeast or bacteria.⁴¹⁹ A plausible PTH/PTHrP receptor homolog is present in the genome of the tunicate sea squirt Ciona intestinalis (see Fig. 1-15) pointing to a much earlier evolutionary origin of the PTH receptor than previously known.^419,425

The three-dimensional molecular structures of the receptors in the secretin family are still only poorly defined because none has been adequately isolated or crystallized in an intact form. For several members of these receptors, however, including the PTH1R,⁶⁸ the CRFR1,⁴²⁶ the CRFR2,⁴²⁷ and the PACR1,⁴²⁸ high-resolution three-dimensional structures of the isolated N-terminal domain in complex with a cognate peptide ligand have recently been obtained, either by solution-phase NMR^427,428 or x-ray crystallography.^68,426 These studies reveal a similar overall fold composed of three layers of secondary structure: an upper layer formed by an N-terminal α-helix, a middle layer of β sheet, and a lower layer formed by a β strand and short C-terminal α-helix (Fig 1-16). The overall fold is stabilized by the three conserved disulfide bonds and an extensive array of hydrophobic and electrostatic packing interactions.⁶⁸ The ligand, represented by the PTH(15-34) domain, binds as an amphipathic α-helix within a central groove in the structure, as described further later. The membrane-spanning helical regions of the class B receptors may be predicted to at least generally follow the heptahelical domain topology used by the class A receptors, which is now revealed by three-dimensional crystal structure of the β₂-adrenergic receptor.⁶⁹

Mechanisms of Ligand Binding and Activation at The PTH1R

The ligand-binding mechanisms used by the PTH receptor have been extensively analyzed by a combination of approaches, including functional methods that are based on receptor site–directed mutagenesis and ligand analog design strategies50,51,429–434 and biophysical methods that are based on the use of photoreactive cross-linking analogs of the ligand that permit direct identification of sites of intermolecular proximity by protein fragmentation and mapping analysis. The cross-linking analogs used in these studies have been labeled at various positions with a photoreactive benzophenone group, incorporated either directly into the peptide chain, as parabenzoyl-l-phenylalanine (BPA),^434–437 or attached to the epsilon amino function of lysine-13 (BpLys).⁴³⁸ The combined functional and cross-linking data suggest that the ligand, as represented by the bioactive PTH(1-34) peptide, binds to the receptor via a two-step mechanism that involves two principal and somewhat autonomous components of the overall interaction.^439–442 The C-terminal portion of the ligand representing the principal binding domain^443,444 first contacts the N-terminal domain of the receptor to establish initial docking interactions,⁴¹⁷ and subsequently the N-terminal portion of the ligand interacts with the J domain portion of the receptor to produce the holo-enzyme bimolecular complex that can couple to and activate G proteins (Fig. 1-17). The overall interaction is certain to be more complex than indicated here, but this general scheme of interaction is now believed to extend to most of the class B family of peptide hormone–binding G protein–coupled receptors.^{440,441,445,446}

Specific contacts to the receptor’s N-domain have now been elucidated by the recent crystal structure of the PTH1R N-domain in complex with PTH(15-34) (see Fig. 1-16).⁶⁸ The PTH(15-34) domain binds as an amphiphilic α-helix, with the hydrophobic face of the helix formed by Trp23, Leu24, and Leu28, making extensive contacts with a hydrophobic surface formed on the floor of the central groove that runs through the N-domain structure. The guanidinium side chain of highly conserved Arg20 of the ligand makes additional contacts with at least five receptor residues that form an intricate pocket at the end of the groove. Binding studies performed using both the intact receptor and the isolated N-domain confirm the importance of the identified contacts to the development of ligand-binding affinity.⁶⁸

The ligand interactions that occur with the receptor’s J domain are known with far less certainty, but these appear to principally involve ligand residues extending from position 1 to about position 19.436,441,442 This portion of the ligand contains the activation pharmacophore, which has minimally been defined by peptide functional studies as the PTH(1-9) segment.⁵⁰ A contact site for conserved valine-2 in the ligand, which plays a critical role in receptor activation,^57,447 has been mapped by cross-linking methods to Met425 at the extracellular end of the transmembrane domain (TM) 6.^448,449 This interaction site is supported by functional data,^448,449 however, other sites in the receptor (e.g., Ser370 and Ile371) have also been implicated by mutagenesis approaches to be likely contact sites for Val2.⁴⁵⁰ Other cross-linking studies identified proximities between residue 19 in the ligand and Lys240 at the extracellular end of TM2,⁴⁵¹ and between residue 13 in the ligand and Arg186 at the boundary of the N-domain and TM1.⁴³⁸ These and other data combined suggest that the ligand’s N-terminal domain binds as an α-helix to lie in a groove formed at the extracellular surface of the receptor’s juxtamembrane region and in the same plane as the outer layer of the cell membrane.^441,451 Such a binding mode could allow Val2, as well as other key ligand residues that define the agonist pharmacophore, such as Ile5 and Met8,^50,452 to make intimate contacts with the membrane embedded core region of the receptor and thereby induce the conformational changes involved in receptor activation.^59,453

The cognate interaction pocket in the receptor’s juxtamembrane region that accommodates and responds to the key pharmacophoric elements in the ligand have only been approximately defined. The development of highly potent N-terminal PTH fragment analogs, based on the weakly active native PTH(1-14) peptide scaffold, provides a class of minimized ligand structures that can used to specifically probe this component of the interaction process more effectively than is possible with PTH(1-34) (see Fig. 1-4). These minimized ligands, referred to generally as the M class of PTH ligands, maintain affinity and efficacy on the PTHR, as well as on a PTHR construct that lacks the N-terminal domain.^{50–53,57,429} Furthermore, the M-PTH analogs have highly stabilized N-terminal helical backbone structures, in part due to the incorporation of conformationally constrained amino acid analogs, such as α-amino-isobutyric (Aib), at positions 1 and 3.^51,54,55,454 A recent use of such modified M-PTH(1-14) analogs for compound screening resulted in the identification of a new class of PTH nonpeptide mimetic ligands that bind specifically to the PTH1R J domain.^455,456 Although these compounds bind with micromolar affinities and function as competitive antagonists, they represent potential leads towards new mimetic compounds that function as PTH1R agonists and are orally active.

Ligand-Induced Conformational Changes

Given the absence of a high-resolution, three-dimensional molecular structure for the intact PTH/PTHR bimolecular complex, it is difficult to describe the conformational changes that occur in the complex upon ligand binding, receptor activation, and G-protein coupling. Problems that need to be solved include determining how the N and J domains of the receptor are oriented relative to each other, whether PTH(1-34) adopts a linear or U-shaped structure as it is bound to the receptor,62–65 and the molecular movements that occur in the complex during the ligand binding, activation, and G-protein coupling cycle. Several new approaches, however, are beginning to help illuminate these aspects of PTH receptor function.

Kinetic binding assays performed in membranes with various PTH and PTHrP radioligand analogs has revealed that the PTH1R can adopt different conformations that display differential affinities for structurally diverse ligands and are differentially coupled to heterotrimeric G proteins.^439,457 In particular, the studies show that PTH(1-34) can form a complex with the receptor that remains stable upon the addition of GTPγS, a reagent that at the high concentrations used is expected to fully activate the G protein and hence dissociate the receptor–G protein heterotrimer complex and thus shift the receptor into a low-affinity state.⁴⁵⁷ Moreover, the ligand can bind efficiently to the receptor in membranes prepared from cells lacking G_αs. In contrast, PTH(1-14) forms a complex that rapidly dissociates upon GTPγS addition and binds only poorly in the absence of G_αs.⁴⁵⁷

The data thus suggested that certain ligands (e.g., PTH[1-34]) preferentially bind to or induce a novel high-affinity PTH1R conformation, called R⁰,⁴⁵⁸ that can stably hold the ligand in place even upon G-protein activation/dissociation, whereas other ligands (e.g., M-PTH[1-14]) preferentially bind to or induce a different high-affinity conformation, called RG, that is dependent on G-protein coupling and thus reverts to a low-affinity form upon G-protein activation/dissociation. These findings predicted that ligands that bind to the R⁰ conformation would produce protracted signaling responses, owing to multiple or prolonged rounds of G-protein coupling. This prediction was indeed supported by data showing that PTH(1-34) produced extended cAMP responses after ligand washout, as compared to M-PTH(1-14).⁴⁵⁷ It was also found that PTHrP(1-36), in contrast to PTH(1-34), binds with relative selectivity to the RG versus R⁰ conformation.⁴⁵⁹ These observations suggested the hypothesis that altered modes of receptor conformational selectivity at the PTH1R, as exhibited by different ligands, may translate into altered biological actions for those ligands in vivo and may even explain the relatively deficient capacity of PTHrP(1-36), versus PTH(1-34), to induce vitamin D synthesis and increase blood calcium levels when injected into human subjects.⁴⁶⁰ This general hypothesis was tested further using a new analog, M-PTH(1-34), that has even stronger selectivity for the R⁰ conformation than does PTH(1-34). The analog indeed produced markedly prolonged cAMP responses in cells and markedly prolonged, even as compared to PTH(1-34), hypercalcemic (Fig. 1-18) and hypophosphatemic responses when injected into mice.⁴⁶¹

That the PTH1R can adopt multiple conformational states and thereby display altered modes of biological action is consistent with the finding of certain PTH ligand analogs that have capacities to selectively activate one signal pathway over another58–60,247 or, when altered by certain point mutations, to mediate ligand-independent signaling activity.⁴⁶² The notion of multiple PTH1R conformations and altered signaling directionality is also in line with the generally accepted views of how most if not all G protein–coupled receptors operate in their respective biological milieus.⁴⁶³ The recent use of fluorescent (Forster) resonance energy transfer (FRET) methodologies, utilizing PTH receptors tagged with spectral variants of green fluorescent protein together with fluorescently labeled PTH ligands, has opened new routes for defining with high temporal resolution and precision the kinetics of receptor conformational change and ligand-receptor interaction processes. Such studies show that indeed the structurally different ligands, such as PTH(1-34), PTHrP(1-36), and M-PTH(1-14), form kinetically distinguishable complexes with the PTH1R and confirm the differences in dissociation rates seen in the radioligand-based membrane binding studies (Fig. 1-19A).^418,453,459

Mechanisms of Prolonged Signaling at The PTH1R

The subcellular mechanisms and pathways that underlie the differences in the duration of the signaling responses seen for the different analogs remains to be established, but it has been found, again using the FRET approach as well as fluorescence confocal microscopy methods, that these long-acting ligands induce long-lived complexes of ligand, receptor, and G protein (see Fig. 1-19B) that remain intact and associated with adenylyl cyclase, even as the proteins move to the intracellular endosomal domain.⁴⁶⁴ This raises the intriguing possibility that in addition to, or as an extension of, the receptor-phosphorylation and arrestin-dependent internalization, desensitization and recycling processes that are known to regulate PTH1R actions,^59,465–468 there may be a component of the mechanism by which the ligand•receptor•G-protein complex formed by PTH but not PTHrP remains assembled within the endosomes and catalytically active for longer time intervals than previously appreciated.⁴⁶⁴ Further studies are needed to test this hypothesis. The results could provide additional clues for designing new PTH analogs that have time-limited or prolonged actions on the PTH1R; such ligands could have improved efficacy, compared to PTH(1-34), as therapies for the diseases of osteoporosis^277,469 and hypoparathyroidism,⁴⁷⁰ respectively.

Other Receptors for C-Terminal PTH or PTHrP

There is evidence for the presence of additional receptors and/or binding proteins that interact with portions of PTH or PTHrP C-terminal of the PTH(1-34), but thus far no cDNA encoding such a receptor has been identified. Considerable information has accumulated that indicates that regions of PTH(1-84) other than the N-terminal residues may be responsible for novel biological actions (see Table 1-2), although many of these effects are still largely limited to in vitro studies. A series of synthetic peptides forming the central region of PTH were reported to delineate a core domain, PTH(30–34), that stimulates the proliferation of chondrocytes but does not appear to involve the cAMP/PKA pathway.^{244,471–474} Early competition binding assays with renal plasma membranes and rat osteosarcoma cells demonstrated binding sites for C-terminal PTH(53-84) that are discrete from those that bind PTH(1-34).^475–478 The observation that N-terminal and C-terminal fragments of PTH elicited contrasting effects on alkaline phosphatase activity^479–481 supports the existence of a receptor with specificity for the C-terminal portion of PTH. Considerable evidence suggests that the C-terminal portion of PTH(1-84) has biological activity. For example, C-terminal PTH fragments such as PTH(39-84) and PTH(53-84) stimulate the formation of bone-resorbing osteoclasts from precursor cells,⁴⁸² and the same PTH fragments were shown to enhance the influx of ⁴⁵Ca into SaOS-2 cells.⁴⁸³ Several different clonal cell lines, including bone- and cartilage-derived cells in which the common PTH/PTHrP receptor had been deleted through homologous recombination, showed increased proliferative activity in response to intact PTH and C-terminal fragments.^484,485 Furthermore, PTH(1-84) was more potent than PTH(1-34) in increasing the concentration of fibronectin in vivo, ⁴⁸⁶ and C-terminal PTH fragments increased type I procollagen expression.⁴⁸⁷

In direct support of a biological role for C-terminal PTH fragments, ROS 17/2.8 cells, rat PT-r3 cells, osteolytic cell lines lacking the PTH1R were shown to specifically interact with the C-terminal portion of PTH.^488,489 With the use of radiolabeled [Tyr34]hPTH(19-84), which does not bind to cells that express high concentrations of the PTH/PTHrP receptor, these cells were shown to bind PTH(1-84) and PTH(19-84) with equivalent high affinity, whereas fragments of PTH(19-84) that were truncated at the N-terminal end showed a progressive loss of binding affinity.⁴⁸⁸ However, even PTH(53-84) showed some displacement of the radiolabeled [Tyr34]hPTH(19-84), whereas the midregional fragment PTH(44-68) failed to do so. Other studies using ROS 17/2.8 cells, but with [³⁵S-Met]hPTH(1-84) instead of ¹²⁵I-labeled [Tyr34]hPTH(19-84), showed similar binding characteristics of C-terminal PTH fragments, and these studies indicated that residue 84 contributes to the overall ligand-binding affinity.⁴⁸⁹ Photoaffinity cross-linking experiments led to the identification of approximately 80-kD and 30-kD proteins that show high-affinity interaction with the C-terminal portion of PTH but not with the N-terminal PTH(1-34).⁴⁸⁸ These findings suggest that the C-terminal portion of PTH interacts specifically with a novel receptor/binding protein that is distinct from the PTH1R. Complementary DNA clones encoding this putative receptor/binding protein have not yet been isolated, and little or nothing is known about the signal transduction systems that might mediate the actions of these midregional/C-terminal PTH fragments.

The development of antibodies that can distinguish truly intact PTH(1-84) from variants that are N-terminally truncated or otherwise posttranslationally modified has led to the realization that such PTH variants can be present in the circulation, and, in cases of primary or secondary hyperparathyroidism, potentially at levels exceeding those of PTH(1-84).172,191,192,255 How such PTH variants might affect calcium homeostasis and other physiologic processes remains to be determined, but in a rat model study, N-terminally truncated PTH was shown to effectively inhibit the capacity of PTH(1-34) to stimulate synthesis of 1,25(OH)₂-vitamin-D₃.^173,490

There is also evidence to suggest that non-PTH1R receptors can mediate actions induced by the midregion of PTHrP and may thus play an important role in placental calcium transport^309,325 (see earlier discussion), as well as modulate biological functions in skin.³²⁴ Likewise, actions of the C-terminal PTHrP portion on osteoclasts and osteoblasts^{317–319,321} and on the central nervous system⁴⁹¹ appear to be mediated through a specific receptor, and other receptors have been characterized that interact equivalently with the N-terminal portions of PTH and PTHrP but signal only through changes in intracellular free calcium.^492–494