As originally proposed by Albright,1 defective target organ response to PTH should lead to the same consequences, hypocalcemia and hyperphosphatemia, as deficiency of the hormone (Fig. 164-1). In theory, resistance could involve either of the target organs of the hormone (kidney and bone), or both. Renal resistance to PTH would cause loss of the phosphaturic response and lead to hyperphosphatemia. The latter and/or primary resistance to PTH would lead to diminished formation of 1,25-dihydroxyvitamin D. 1,25-dihydroxyvitamin deficiency, by reducing gastrointestinal absorption of calcium, also contributes to hypocalcemia. Because the skeletal calcium-mobilizing effect of PTH is dependent on 1,25-dihydroxyvitamin D,16 a blunted calcemic response to PTH could be caused by renal resistance to PTH with resultant reduction in 1,25-dihydroxyvitamin D formation, without invoking primary skeletal resistance to PTH. Primary skeletal resistance to PTH should cause loss of both the calcemic and bone-remodeling responses to the hormone.
Resistance to PTH action could, in theory, be due to defects at any one of the multiple steps in the pathway of hormone action. If PTH actions are mediated through second messengers other than, or in addition to, stimulation of cAMP formation (Fig. 164-2), selective lesions causing incomplete forms of PTH resistance are possible. Heterogeneity of PTH receptors, either in a given target organ or between kidney and bone, could provide another theoretical basis for selectivity in hormone resistance.
Within the cAMP pathway (Fig. 164-2), certain lesions (eg, specific PTH receptor antagonists, abnormal receptors, abnormal substrate(s) for cAMP-dependent protein kinase, or tissue-specific Gs defects) would be expected to cause isolated resistance to PTH. In contrast, defects in Gs, adenylyl cyclase, cAMP phosphodiesterase, or cAMP-dependent protein kinase could cause generalized resistance to agonists acting through the cAMP pathway.25,47 Even for these "general" components of the cAMP transduction pathway, cDNA cloning has revealed substantial heterogeneity at the gene and/or mRNA level. Differences in range of expression and function for subtypes of each of these components have not yet been fully defined, so that it is possible that defects in one or more of these components could cause more selective forms of hormone resistance.
Genetic defects in several components of the cAMP transduction pathway have been identified in eukaryotic cells. Discrete mutations in the α subunit of Gs in mouse S49 lymphoma cells can lead to partial or total deficiency of Gs activity, to receptor uncoupling, or to failure to undergo activation by GTP.48 Mutations in cAMP phosphodiesterase causing excessive activity in S49 cells have also been described.49 Mutations causing deficient cAMP phosphodiesterase activity50 and calmodulin-sensitive adenylyl cyclase activity51 have been identified in Drosophila , and interestingly, both are associated with learning defects. A naturally occurring mouse mutant, barrelless, disrupts patterning of the somatosensory cortex, and was identified as an adenylyl cyclase type I mutation.52 Mutants deficient in adenylyl cyclase activity or showing constitutive activation of cAMP-dependent protein kinase due to a regulatory subunit defect have been described in yeast,53 whereas a catalytic subunit mutation abolishes kinase activity in S49 cells.48 Each of these mutations causes generalized abnormalities in cellular response to extracellular signals.
Available evidence suggests that defects at different sites in the PTH response pathway are responsible for distinct forms of PHP. The majority of patients with PHP tested show a markedly lower increase in urinary cAMP excretion in response to PTH infusion than do normal controls or patients with PTH-deficient forms of hypoparathyroidism (Fig. 164-4). These data indicate that a defect (or defects) in the pathway that is proximal to cAMP generation is responsible for the majority of cases of PHP. Rarely, patients with PHP6 show a normal rise in urinary cAMP excretion in response to PTH. Their defect is presumptively located distal to cAMP generation.
Deficient urinary cAMP excretion in response to PTH infusion in PHP. Two hundred USP units of purified bovine PTH were infused intravenously between 9:00 and 9:15 AM. Urine was collected hourly (one-half-hourly immediately after infusion) before and after infusion for measurement of cAMP. The individual responses in a group of 13 subjects with PHP and a group of normal controls are shown. (From Chase et al.5 Used by permission.)
In the sections that follow, PHP is classified according to the presumed locus of the defect that causes PTH resistance. PHP is divided into type I (defect proximal to cAMP production) and type II (defect distal to cAMP generation). PHP type I may be further subdivided into Ia (generalized hormone resistance due to a defect in a common component of the receptor-cyclase complex, such as Gs) and Ib (resistance limited to PTH). Pseudopseudohypoparathyroidism (PPHP) is the term applied to individuals who are first-degree relatives of subjects with PHP Ia and who show Gs deficiency and features of Albright osteodystrophy, but who are not overtly resistant to hormones.
Pseudohypoparathyroidism Type IA
This is a familial form of PHP characterized by generalized resistance to agents that stimulate adenylyl cyclase and by the phenotypic features of Albright osteodystrophy. Patients with PHP Ia show a blunted rise in urinary cAMP excretion in response to PTH.5,54 This implicates a defect in the receptor-G protein-adenylyl cyclase complex. Clinical observations compatible with resistance to hormones other than PTH suggested the possibility of a defect in a general (e.g., Gs) rather than a specific (e.g., PTH receptor) component of the complex.25 Thus, primary hypothyroidism, particularly a subtle form detectable by thyrotropin-releasing factor stimulation of thyrotropin, is extremely common in this form of PHP.55 Primary hypothyroidism is not associated with thyroid enlargement in these patients, nor are thyroid antibodies detectable.55 Primary thyroid resistance to thyrotropin could explain these results. Similarly, a patient with PHP and phenotypic features of Albright osteodystrophy was reported to show resistance to gonadotropins and glucagon.56
Renal tissue obtained at autopsy from a woman with a familial form of PHP associated with Albright osteodystrophy showed qualitatively normal stimulation of adenylyl cyclase activity by PTH.57 Studies of renal biopsy material from a similar patient confirmed that the PTH receptor and adenylyl cyclase were qualitatively normal, but revealed a subtle abnormality compatible with a defect in Gs.58 Because Gs is ubiquitously distributed, the hypothesis that Gs deficiency causes generalized hormone resistance in certain patients with PHP could be tested by assays of accessible tissues rather than of relatively inaccessible PTH target organs, bone and kidney.
Two groups,7,8 using a functional assay based on stimulation of adenylyl cyclase activity by Gs in membrane extracts and an assay based on cholera toxin-catalyzed incorporation of radioactive ADP-ribose into Gs-α, measured Gs in membranes from patients with PHP and normal controls. An approximately 50 percent reduction in Gs activity was found in red cell membranes from almost all patients with PHP and Albright osteodystrophy (Fig. 164-5), and normal activity was seen in membranes from patients with PHP lacking the physical features of Albright osteodystrophy.7,8 Platelet,59 fibroblast,60,61 and lymphoblast62 membranes from subjects with PHP and Albright osteodystrophy are also deficient in Gs activity. Renal membranes from the patient with PHP that showed a subtle defect in adenylyl cyclase stimulation58 were also found to contain about half the Gs activity of renal membranes taken from three normal controls.63
Gs activity in erythrocyte membranes of control subjects and subjects with PHP with (+AHO) or without (−AHO) the phenotypic features of Albright hereditary osteodystrophy. Membrane extracts to be assayed for Gs activity were added to a preparation of avian erythrocyte membranes containing adenylyl cyclase. Resultant activity (referred to as “adenylate cyclase” in the figure) is a function of Gs activity in the added extract of human erythrocyte membranes. Activity is expressed as percent of the activity (defined as 100 percent) of a pooled membrane standard from normal subjects. (From Levine et al.90 Used by permission.)
Reduced Gs activity in membranes from patients with PHP, in particular those showing features of Albright osteodystrophy, has been confirmed in studies of patients in Japan64 and Europe.65 Gi, the G protein associated with inhibition of adenylyl cyclase (Fig. 164-2), is a gene product distinct from Gs24 and a substrate for ADP-ribosylation by pertussis toxin. Two studies, using cholera and pertussis toxin-catalyzed ADP-ribosylation, indicate that Gs is deficient in red cell membranes from patients with PHP and Albright osteodystrophy, but that Gi concentration is normal.66,67 Deficient Gs activity is also reflected in reduced formation of the high affinity ternary complex between agonist, receptor, and Gs in membranes from patients with PHP Ia.68
Deficient Gs activity in membranes from patients with PHP Ia could be due to reduced synthesis of Gs protein or to synthesis of a defective form of one of the subunits of Gs. Immunoblot analysis of cell membranes from patients with PHP Ia shows reduced Gs-α protein in most patients.38 In most cases, reduction in amount of Gs-α protein appears to be due to reduction at the mRNA level. Blot hybridization analysis of RNA from cultured skin fibroblasts shows a significant reduction (to less than 50 percent of normal) in steady state content of Gs-α mRNA in most patients with PHP Ia.69,70 No abnormality in mRNA size was observed; also, S-1 nuclease analysis showed that all forms of Gs-α mRNA (see above) are proportionately reduced in patients with PHP Ia.69 Genomic blots show no gross deletions or rearrangements of the Gs-α gene in most subjects studied.69,70 The data are compatible with subtle lesions in the gene that either reduce transcription or prevent formation of stable mRNA.
The genetic defect in patients with PHP Ia (and PPHP) has been confirmed by the identification of multiple heterozygous loss-of-function mutations within the GNAS1 gene (see Table 164-1). Mutations are spread throughout the Gs-α coding exons with the exception of exon 3, which can be spliced out and still produce a functional Gs-α. In many cases, the mutations result in abnormal RNA processing and lack of expression of the mutant allele (i.e., base substitution at a splice junction site; coding frameshift mutations, premature stop codons, and large deletions; for examples, see references10,71–82 Table 164-1 and Fig. 164-6). One specific 4 bp deletion within exon 7 of GNAS1 was identified in affected members from multiple unrelated PHP Ia kindreds.71,77–80 (Aldred & Trembath, 2000; Walden et al, 1999; Lim et al, 2002; Ahrens et al, 2001; Linglart et al, 2002; Mantovani et al, 2000; de Sanctis et al, 2000; Yeh et al, 2000; Shore et al, 2002) This deletion hotspot coincides with a consensus sequence for arrest of DNA polymerase α, and likely results from arrest of polymerization and slipped-strand mispairing during DNA replication.78 Another deletion hotspot is located in exon 5 (74) (Aldred & Trembath, 2000; Lim et al, 2002; Mantovani et al, 2000; Shore et al, 2002). Several other mutations also occur in at least two kindreds (Table 164-1).
Upper panel: Pedigree of kindred with PHP Ia with Alu I restriction analysis of genomic DNA. The genomic fragment including exons 10 and 11 of Gs-α was amplified using the polymerase chain reaction and digested with Alu I restriction endonuclease. The digests were electrophoresed in a nondenaturing 5% acrylamide gel. Symbols of pedigree members with PHP are blackened and that for the member with PPHP is stippled. The asterisk denotes an unaffected member with four unaffected sons who were not analyzed. One subject who died in infancy and whose disease status is unknown is listed with a question mark. The size of the restriction fragments is depicted on the left with the 500-base pair undigested fragment shown in lane U. Unaffected members show 403- and 97-base pair fragments, whereas affected subjects show 2 additional fragments of 319 and 84 base pairs, as predicted if these individuals are heterozygous for the mutation discussed below.
Lower panel: Genomic fragment including exons 10 and 11 and showing Alu I restriction sites. The 500-base pair genomic fragment contains an Alu I restriction site 97 base pairs from the 5′ end within the normal sequence of exon 10, predicting 97- and 403-base pair fragments in the normal allele as shown above the diagram. As shown below the diagram a G-to-C base substitution at the donor splice junction site of intron 10 creates a new Alu I restriction site, which leads to 319-, 97-, and 84-base pair fragments. (From Weinstein et al.10 Used by permission.)
Table 164-1: GNAS1 Mutations in PHP IA and PPHP |Favorite Table|Download (.pdf) Table 164-1: GNAS1 Mutations in PHP IA and PPHP
| ||Systematic name* ||Trivial Name† ||mRNA ||Protein ||Comments ||Reference |
|Exon 1 ||c.1A > G ||M1V ||NI‡ ||Abnormal ||Cyc-assay↓¶ ||9 |
| || || || ||70K form; || || |
| || || || ||Nl form↓ || || |
|Exon 1 ||c.92C > T ||N31X ||ND‡ ||↓ || ||82 |
|Exon 1 ||c.111C > G ||Y37X ||ND ||↓ || ||82 |
|Exon 1/Intron 1 ||g.119-156del ||Deletion ||ND ||ND ||Cyc-assay ↓ ||81 |
|Intron 3 ||IVS3-2A > G ||Acceptor splice junction ||ND ||ND || ||73 |
|Exon 4/Intron 4 ||c.278del43nt (including some of IVS4) ||Deletion ||ND ||ND || ||75 |
|Exon 4 ||c.292-294del ||N98del ||ND ||ND || ||Unpublished (LSW) |
|Exon 4 ||c.296T > C ||L99P ||ND ||↓ || ||72 |
|Exon 4 ||c.301-302del ||Frameshift ||↓ ||ND || ||191 |
|Exon 5 ||c.343C > T ||P115S ||ND ||ND || ||77 |
|Exon 5 ||c.348delC ||Frameshift ||ND ||ND ||Cyc- assay ↓ ||74 |
|Intron 5 ||IVS5+1G > A ||Donor splice junction ||ND ||ND || ||76 |
|Exon 6 ||c.493C> T ||R165C ||NI ||↓ || ||72 |
|Exon 7 ||c.565-568del ||Frameshift ||↓ ||ND ||Present in several kindreds; ||71,77–80 |
| || || || || ||Cyc-assay ↓ || |
|Exon 7 ||c.568T > G ||Y190D ||ND ||ND || ||73 |
|Exon 8 ||c.617-618del ||Frameshift ||ND ||ND || ||191 |
|Exon 8 ||c.640-643del ||Frameshift ||NI ||↓ || ||72 |
|Exon 9 ||c.692G > A ||R231H ||ND ||ND ||Defective receptordependent signaling ||86 |
|Exon 10 ||c.750C > G ||S250R ||NI ||↓ ||Cyc- assay ↓ ||83 |
| || || || || ||↑ GDP release || |
| || || || || ||↑ GTP hydrolysis || |
|Exon 10 ||c.772C > T ||R258W ||NI ||↓ ||Cyc-assay ↓ ||88,192 |
|Exon 10 ||c.776A > G ||E259V ||ND ||ND ||Decreased stability and signaling ||77,193 |
|Exon 10 ||c.798-799insC ||Frameshift ||ND ||ND ||Cyc- assay ↓ ||74 |
|Exon 10 ||c.814delC ||Frameshift ||ND ||ND || ||10 |
|Intron 10 ||IVS10+1G > C ||Donor splice junction ||↓ ||ND || ||10 |
|Exon 13 ||c.1096G > C ||A366S ||ND ||ND ||PHP la and testotoxicosis; Cyc-assay ↓ at 37°C; rapid GDP release and activation at low temp ||85 |
|Exon 13 ||c.1147-1149del ||1383del ||ND ||ND || ||73 |
|Exon 13 ||c.1154G > A ||R385H ||ND ||Moderate ↓ ||Cyc-assay ↓; uncoupled from receptor ||84 |
Missense mutations in GNAS1 have also been identified. In some cases, the encoded amino acid substitution (L99P and R165C72; S250R83 appears to globally alter tertiary structure, or intracellular trafficking of the Gs-α protein, because in each case, the level of Gs-α mRNA is normal, but the level of membrane Gs-α protein is decreased. In one kindred, a mutation at the translational start site codon results in the expression of an abnormally large form of Gs-α protein which is presumed to be nonfunctional.9 For several missense mutations the biochemical defect remains to be determined.
Some GNAS1 missense mutations associated with PHP Ia or PPHP produce specific biochemical defects in the Gs-α protein. Several missense mutations identified in the carboxy-terminal region [R385H84, Y391X (Linglart et al., 2002), I382del (Wu et al., 2001)] result in global or specific defects in receptor coupling.84 Another missense mutation (A366S) in a region encoding a portion of the highly conserved guanine nucleotide binding site was identified in two unrelated males who presented with Albright osteodystrophy and PTH resistance in association with gonadotropin-independent precocious puberty.85 This mutation was shown to decrease the protein's affinity for GDP. At internal body temperature (37°C), Gs-α unbound to guanine nucleotide is unstable, leading to decreased expression of Gs-α protein and the clinical expression of AHO and hormone resistance. At slightly lower temperatures (the ambient temperature of the testes), this mutation leads to Gs activation, increased intracellular cAMP, and gonadotropin-independent precocious puberty, since GDP release is normally the rate-limiting step in G protein activation. A mutation (R231H) within a region of Gs-α that undergoes a major conformational shift upon receptor-activation (switch 2 region) has been shown to have a specific defect in activation by receptor or the transition state analog aluminum fluoride.86,87 Mutation of the E259 residue in the switch 3 region produces a similar biochemical defect.193 A mutation (R258W) within the switch 3 region, has been demonstrated to have an increased rate of GDP release88 and an increased rate of GTP hydrolysis.192
Deficient Gs activity might be expected to reduce adenylyl cyclase responsiveness to stimulation by a wide variety of agonists. Indeed, the S49 mouse lymphoma mutant CYC-, which is totally deficient in Gs activity48 and in Gs-α mRNA,39 fails to respond to prostaglandins and β-adrenergic agonists, potent stimulators of adenylyl cyclase activity in wild-type cells. An S49 cell mutant with partial deficiency of Gs shows reduced adenylyl cyclase stimulation by agonists.48 Fibroblasts and platelets from patients with PHP Ia have shown reduction in agonist-stimulated cAMP formation in some,60,89 but not all, studies.59,61
Impaired adenylyl cyclase stimulation due to Gs deficiency should lead to generalized resistance to agents acting via cAMP. Clinical studies of patients with PHP Ia provide evidence for resistance to hormones other than PTH. Primary thyroid resistance to thyrotropin appears to be extremely common, if not universal, in patients with PHP Ia (Fig. 164-7).55,90 Defective adenylyl cyclase stimulation by GTP and thyrotropin in thyroid membranes obtained from a patient with PHP Ia, provides direct evidence for thyroid resistance to thyrotropin, presumably secondary to Gs deficiency, in this disease.91 Gonadotropin,56 90 92–94 glucagon,56 90,95) calcitonin, (Vlaeminck-Guillem et al., 2001) and isoproterenol96–97 resistance have also been documented in patients with PHP Ia, particularly when proximal responses such as plasma cAMP concentration are measured. Recent data suggest that the tissue-specific nature of hormone resistance in PHP Ia might be caused by tissue-specific imprinting of Gs-α (see section on Genetics105) Distal responses to hormones, for example, glycemic response to glucagon,90,95 free fatty acid response to isoproterenol,98 and antidiuretic response to arginine vasopressin,99 however, are often normal in patients with PHP Ia.
Basal and peak serum thyrotropin (TSH) concentrations following intravenous administration of 500 μg thyrotropin-releasing factor (TRF) to subjects with PHP. The normal range for basal and peak TSH is shown in the hatched boxes. Most subjects with deficient Gs activity (left) show hyperresponsiveness to TRF, indicative of primary hypothyroidism. With one exception, subjects with PHP and normal Gs activity (right) show normal responses. (From Levine et al.90 Used by permission.)
Olfactory dysfunction was documented by objective testing in patients with PHP Ia, but not in patients with PPHP or PHP Ib (normal Gs).100,101 This was attributed to a possible role for Gs in transduction of odorant stimuli, but it is now likely that a unique G protein (termed Golf, as it is expressed primarily in olfactory neuroepithelium) is responsible for primary odorant signal transduction. Apparent olfactory dysfunction, as well as sensorineural hearing dysfunction reported in PHP Ia,102 may actually reflect effects of neuronal Gs deficiency on higher-order signal processing rather than primary sensory impairment. The mild mental retardation often seen in patients with PHP Ia103 could also complicate interpretation of sensory testing.
While resistance to PTH, TSH, and gonadotropins is common in PHP Ia, resistance to other hormones, such as vasopressin or adrenocorticotropin, which also stimulate Gs-coupled pathways, is almost never observed in this disorder.90,99,104 The basis for variability in clinical expression of hormone resistance in PHP Ia is unclear. Differences in the absolute concentrations of cAMP needed to achieve physiologic responses to agonists could influence the impact of Gs deficiency. One may also speculate that the degree of clinically evident hormone resistance is a function of tissue- and hormone-specific variations in the components of the cAMP signal transduction pathway. Recent data suggests that the tissue-specific nature of hormone resistance in PHP Ia might be caused by tissue-specific imprinting of Gs-α (see section on Genetics and reference 105).
The phenotypic features of Albright osteodystrophy (Fig. 164-8) are an important component of PHP Ia, but their underlying basis is not understood. One possibility is that Gs deficiency, through limitation in cAMP formation or perhaps through some other mechanism, is the proximate cause of some or all of the features of Albright osteodystrophy. This is a plausible, but unproven, mechanism for the obesity and mental retardation often seen in patients with PHP Ia. Lipolytic factors act by stimulating cAMP formation97 (Carel et al, 1999) and Gs-α deficiency may promote adipocyte differentiation (Wang et al, 1992). Abnormalities in cAMP metabolism are associated with learning defects in invertebrate50,61 and vertebrate52,106 model systems. Short stature and metacarpal and metatarsal shortening are typical features of Albright osteodystrophy and may be due to premature skeletal maturation and closure of epiphyses.107 Defects in the PTH/PTHrP receptor are also associated with premature skeletal maturation.108 Growth plate chondrocytes with a Gs-α null allele differentiate more rapidly in vivo (Chung et al, 2000). Data on growth hormone secretion in PHP Ia are limited to anecdotal reports; some suggesting normal growth hormone secretion,104,109 others consistent with growth hormone deficiency.93,110 (Germaine-Lee et al, 2001) The relationship of either Gs or growth hormone deficiency to the skeletal abnormalities of Albright osteodystrophy, including the soft tissue ossification and calcification (Fig. 164-9), remains unclear. The presence of activating Gs-α mutations in fibrous dysplasia of bone (36) and Gs-α null mutations in AHO and progressive osseous heteroplasia (see below) suggests that Gs-α deficiency promotes osteoblast differentiation. Prolactin deficiency has also been reported in PHP Ia,111–113 but the cause of this abnormality has not been elucidated.
Mother (left) and daughter with PHP Ia. Several features of Albright osteodystrophy are evident, including obesity, short stature, round face, and short neck. Short fourth and fifth fingers (particularly on the right hand of both subjects) and short fourth toes on the left feet of both subjects are due to short metacarpal and metatarsal bones, respectively.
Features of Albright osteodystrophy (clockwise from upper left) include: (a) short stubby fingers, particularly the fourth and fifth and the distal phalanx of the thum; (b) shortening of the fourth metacarpal; (c) subcutaneous calcification, shown here in the abdominal wall (lower-right quadrant); and (d) shortening of the left fourth metatarsal on the right.
Since its initial description by Albright,2 the entity PPHP has generated much confusion. In part, this stems from the relative lack of objective specificity of certain features of the Albright phenotype, including short stature, obesity, “round face,” and even shortening of fourth metacarpals. Diagnosis of normocalcemic individuals with some or all of these features as having PPHP is almost certain to lead to inaccuracy; for example, patients with Turner syndrome may be included. Genetic linkage between PHP and PPHP was recognized early,3 and reemphasized more recently.114 The term PPHP is best restricted to relatives of patients with PHP Ia who show phenotypic features of Albright osteodystrophy and a normal rise in urinary cAMP excretion in response to PTH infusion.5,115
Diagnosis of PHP or PPHP based on serum calcium concentration (Albright's original criterion) may be misleading. Patients with clear-cut resistance to PTH (elevated serum PTH, blunted urinary cAMP response to PTH) may be normocalcemic.116 Hypocalcemia is not present from birth in patients with PHP Ia; instead, it may develop during the first decade and be preceded by other signs of PTH resistance, including hyperphosphatemia and elevated serum PTH.54,117 (Yu et al, 1999). Serum calcium, moreover, may fluctuate between low and normal concentrations in patients with PHP.118 Because Gs deficiency is likely present from birth, and unlikely to fluctuate during life, variations in serum calcium must reflect other factors such as 1,25-dihydroxyvitamin D synthesis. Estrogens and placental synthesis of 1,25-dihydroxyvitamin D in pregnancy are among the factors that can alter serum calcium in PHP.119
Within a family with PHP Ia, multiple members may show features of Albright osteodystrophy and yet show wide variation in degree of hormone resistance. Unaffected family members (no hormone resistance, no features of Albright osteodystrophy) show normal Gs activity,115,120 and lack mutations in the Gs-α gene (Fig. 164-6).10 Subjects with Albright osteodystrophy show similar reductions in Gs activity and identical mutations in the Gs-α gene whether (with PHP Ia) or not (with PPHP) they are overtly resistant to hormones.9,10,63,115,121 Recent clinical genetic studies,76,80,122 (Hayward et al, 2001) as well as studies in mice with a targeted mutation of the Gs-α gene,105 suggest that imprinting of GNAS1 is the likely basis for the variable presentation of patients with Albright osteodystrophy (PHP Ia vs. PPHP) (see “Genetics” below).
Pseudohypoparathyroidism with Albright Osteodystrophy, Generalized Hormone Resistance, and Normal Guanine Nucleotide-Binding Protein Activity
This form of PHP could be termed “Ia or Ib” depending on whether the Albright osteodystrophy phenotype or Gs activity is used for classification; PHP type “Ic” is an alternative term that has been proposed.12 Although most patients with generalized hormone resistance and Albright osteodystrophy are deficient in Gs, several groups have reported normal Gs activity in a small number of patients with this phenotype.65,90,120 It is possible that subtle defects in Gs in such individuals are not detected by available assays, but in at least two families, such a Gs defect (receptor uncoupling) was specifically excluded.120 Elevation in Gi, another theoretical explanation for this phenomenon, has not been observed using pertussis toxin-catalyzed ADP-ribosylation.66 A single patient was reported with this phenotype and biochemical data compatible with an adenylyl cyclase defect.124 With the cloning of cDNAs encoding adenylyl cyclase,42 molecular genetic studies of such patients should be feasible. Defects in other general components of the cAMP pathway, for example, cAMP phosphodiesterase, also should be sought, as these could lead to a similar phenotype. The occurrence of Albright osteodystrophy in this form of PHP suggests the possibility that different defects in the cAMP transduction pathway could cause this phenotype, with deranged synthesis, degradation, or action of cAMP as the underlying mechanism.
Pseudohypoparathyroidism Type IB
In this form of PHP, physical appearance is normal, resistance is generally limited to PTH, and Gs activity in peripheral (e.g., red blood cell) cell membranes is normal.90,125 The presumed site of the defect in the pathway is proximal to cAMP generation, as in PHP Ia, because urinary cAMP excretion in response to PTH infusion is abnormally low.90 This form of PHP may itself be heterogenous, with reports of both familial111,126 and apparently sporadic127 cases. In one study,127 cultured skin fibroblasts from 7 of 10 patients with PHP Ib showed selective resistance to PTH in terms of cAMP formation. The persistence of abnormal PTH response in cells cultured in vitro is consistent with an intrinsic abnormality in the PTH receptor, a candidate gene for PHP Ib. In a subsequent study of fibroblasts from subjects with PHP Ib, however, the reduction in cAMP response to PTH could be corrected by treatment of cells with dexamethasone and correlated with an increase in receptor mRNA after dexamethasone treatment.128 Subsequent molecular genetic studies failed to identify mutations in the PTH receptor gene in subjects with PHP Ib.129,130 Given the lethal phenotype of homozygous disruption of the PTH receptor gene in mice and in the human disease Blomstrand chondrodysplasia,108 it appears unlikely that PHP Ib could be caused by homozygous loss of function mutation of the PTH receptor. However, subtler abnormalities have yet to be completely excluded.129 An analysis of the promoter region of the PTH receptor gene revealed a general as well as a renal-specific promoter, but no specific defect in the promoter regions could be identified in seven PHP Ib subjects.131
In one study examining four PHP Ib kindreds, the disease was mapped to 20q13, in the vicinity of GNAS1 (Jüppner et al, 1998). Moreover, in these families maternal inheritance of the trait produced offspring with PTH resistance while paternal transmission produced silent carriers, similar to the inheritance pattern of PTH resistance in patients with AHO. However, the fact that Gs is unaffected in erythrocyte membranes from these patients makes it unlikely that PHP Ib is due to simple Gs-α null mutations. In virtually all cases PHP Ib is associated with a GNAS1 imprinting defect, in which the exon 1A promoter region has a paternal-specific imprinting pattern (unmethylated, transcriptionally active) on both alleles (Liu et al, 2000a; Bastepe et al, 2001b; Jan de Beur et al, 2001; J.Liu, L.S.W., unpublished data). Similar imprinting defects involving the upstream NESP55 and XLαs upstream regions are only present in a small subset of patients, making it unlikely that these regions or their respective gene products are important in the pathogenesis of PHP Ib. Paternal uniparental disomy (UPD) of chromosome 20, which leads to a similar GNAS1 imprinting abnormality, also resulted in PTH and TSH resistance in one patient (Bastepe et al, 2001a). Although the underlying genetic mutations that lead to the GNAS1 imprinting defect (at least in familial PHP Ib) are likely to be within the GNAS1 locus, there is some evidence to suggest that they may be at a distance from exon 1A or perhaps in some cases on other unlinked loci (Bastepe et al, 2001b).
The common exon 1A methylation defect in PHP Ib provides for a straightforward diagnostic assay and strongly implicates this defect in the pathogenesis of PHP Ib. The presence of a paternal-specific imprinting pattern on both GNAS1 alleles presumably leads to loss of Gs-α expression and PTH resistance in renal proximal tubules due to lack of an 'active maternal' allele. The imprinting defect should have no effect on Gs-α expression in the vast majority of other tissues where Gs-α is normally biallelically expressed. This would explain why PHP Ib patients have normal Gs-α expression in erythrocytes and lack the AHO phenotype. The mechanisms by which differential methylation of exon 1A might lead to tissue-specific imprinting of Gs-α are unknown, but several potential models have been proposed (Weinstein et al, 2001).
Three siblings with a PHP Ib phenotype were shown to have a Gs-α missense mutation that deletes residue I382 within the carboxyl-terminus, which results in selective uncoupling of Gs-α from the PTH receptor (Wu et al, 2001). As with other GNAS1 mutations, only maternal transmission of the I382del mutation results in PTH resistance.
Although in general PHP Ib patients do not show clinical evidence for TSH resistance, a few patients present with very mildly elevated TSH values (L.S.W., unpublished data) and TSH resistance was present in the patient with paternal UPD of chromosome 20 (Bastepe et al, 2001a). Hypothyroidism secondary to autoimmune thyroiditis was reported in one PHP Ib patient (111). In common with PHP type Ia, prolactin deficiency has been reported in PHP type Ib.111–113 Defective prolactin response to various stimuli occurs in some, but not all, patients with both PHP Ia and Ib.90,112
Pseudohypoparathyroidism Type II
This form of PHP involves a defect distal to PTH-stimulated cAMP production. It was originally described in a child of normal appearance who was hypocalcemic and hyperphosphatemic, who showed elevated serum PTH concentration, and who had normal renal function, all consistent with PTH-resistant hypoparathyroidism.6 Urinary cAMP response to PTH infusion, however, was completely normal, and the phosphaturic response was equivocal. Such a phenotype could be caused by a defect in a PTH and cAMP-sensitive renal phosphate transport mechanism, but this hypothesis has not been directly tested. Because the phosphaturic response to PTH is often equivocal even in normal subjects,6 the diagnosis rests on finding clear elevations in serum PTH in association with hypocalcemia, normal renal function, and a normal urinary cAMP response to PTH infusion.
PHP type II may be an acquired disease;132 only a single instance of familial PHP II (in two brothers) has been reported.132 Calcium infusion may normalize the phosphaturic response to PTH in patients with PHP type II.133 This effect is not due to suppression of endogenous PTH secretion, and may reflect calcium dependence of PTH-stimulated phosphaturia.133 Findings compatible with PHP type II also occur in some patients with vitamin D deficiency.132,134 Urinary cAMP excretion in response to PTH is normal in such patients, but the phosphaturic response is defective. Treatment with vitamin D restores a normal phosphaturic response.132,134 The mechanism of vitamin D action involves more than restoration of normocalcemia, because hypocalcemic patients with PTH-deficient hypoparathyroidism show a normal phosphaturic response to PTH.
PHP type II in association with Sjögren syndrome was reported in a patient with antirenal tubular plasma membrane autoantibodies.135 Whether such antibodies were relevant to the pathogenesis of the phosphate transport defect is not clear, but this case emphasizes that PHP type II may be an acquired defect caused by diverse mechanisms.
Progressive Osseous Heteroplasia
Some patients with heterozygous Gs-α null mutations develop progressive osseous heteroplasia (POH), a severe disorder of ectopic bone formation characterized by large plaque-like dermal and subcutaneous ossifications which invade the deep connective tissues and skeletal muscle (Kaplan & Shore, 2000). Similar to the more restricted ossifications typically observed in AHO, the ectopic ossifications in POH form primarily by intramembranous, rather than endochondral, ossification. In POH the ossifications often restrict joint mobility and constrain the growth of long bones in the extremities. POH affects both males and females and can be sporadic or familial. Although in most cases there are no other associated features of AHO, it may occasionally be associated with other features of AHO or PHP Ia (Eddy et al, 2000).
Recent studies have identified heterozygous germline Gs-α null mutations in POH patients, both those that present with POH alone and those that present with POH and other features of AHO or PHP Ia (Yeh et al, 2000; Eddy et al, 2000; Shore et al, 2002). The mutations in POH include typical frameshift, nonsense, or splice junction mutations that fully disrupt Gs-α mRNA and/or protein expression. Two mutations (see Table 164-1) have been identified in both AHO and POH patients (Yeh et al, 2000; Shore et al, 2002). While it has been claimed that POH results from paternal Gs-α mutations (Shore et al, 2002), maternal transmission of POH has been reported in a patient with reduced Gs activity and other features of PHP Ia (Eddy et al, 2000). In one family, paternal transmission of a Gs-α mutation resulted in 5 cases of POH, while maternal transmission of the identical mutation resulted in AHO in the next generation (Shore et al, 2002; hormonal studies were not performed in the AHO patients).
It is unclear why similar Gs-α null mutations result in either POH or AHO. It is possible that other 'modifying' genes involved in the regulation of Gs-α, another component of the cAMP pathway, or other pathways involved in bone formation, may determine the severity of ectopic ossification in any individual patient. Lower Gs-α expression and intracellular cAMP levels may promote osteoblast differentiation in ectopic locations by inducing the mRNA for the osteoblast-specific isoform of the osteogenic transcription factor Cbfa1/RUNX2 (Yeh et al, 2000) and/or inhibiting the degradation of Cbfa1/RUNX2 protein (Tintut et al, 1999).
Parathyroid Hormone Inhibitors as a Cause of Pseudohypoparathyroidism
Potent antagonists to PTH action on its target organs could, in theory, cause PHP. Inhibitors could be abnormal forms of PTH itself or unrelated molecules, such as antibodies to PTH receptors. An abnormal form of PTH, lacking biologic activity, could not by itself account for the resistance to exogenous PTH characteristic ofPHP. Thus, in cases with secretion of a putatively ineffective form of PTH, renal responsiveness to exogenous PTH was normal.136–138 To date, there is no evidence that an altered form of PTH could function as a potent inhibitor of normal PTH action in vivo.34 Molecular genetic analysis of the PTH gene, moreover, failed to reveal abnormalities in most families with hypoparathyroidism that were studied.139 In one family with isolated hypoparathyroidism, a mutation in the PTH gene was found, but this involved the signal peptide-encoding region and presumptively impaired processing of preproPTH to proPTH.140
Using an ultrasensitive renal cytochemical assay for PTH, one group reported a dissociation between PTH biologic activity and immunoreactivity in patients with PHP type I.141 Whereas immunoreactivity was supranormal, biologic activity was in the normal range. These results were interpreted to indicate the existence of an inhibitor of PTH action in the serum of patients with PHP type I. Another group, using the same assay, reported similar findings in vitamin D-treated patients with PHP, but observed that patients treated for nutritional vitamin D deficiency showed a similar discrepancy between immunoreactive PTH and biologic activity.142 Reports of improved PTH responsiveness after correction of hypocalcemia with vitamin D treatment143 or after parathyroidectomy (in one of Albright's original patients144) have prompted speculation that secretion of an abnormal form of PTH could lead to PTH resistance in PHP. Given the difficulty in synthesizing PTH analogues that function as potent antagonists invivo,34 it seems unlikely that secretion of an altered form of PTH could explain resistance to pharmacologic doses of exogenous PTH in patients with PHP. There is no evidence, moreover, that vitamin D treatment restores a normal urinary cAMP response to PTH in PHP. Also, the parathyroidectomized patient with PHP showed a deficient urinary cAMP response to PTH infusion even after parathyroidectomy.5 The hypothesis remains unproven thatan inhibitor of PTH action, related to PTH itself or not, could explain some cases of PHP. In any event, PTH inhibitors cannot explain the resistance to other hormones commonly seen in PHP Ia.
Skeletal Resistance to Parathyroid Hormone in Pseudohypoparathyroidism
Studies of vitamin D metabolites in patients with PHP suggest that the deficient calcemic response to PTH reflects 1,25-dihydroxyvitamin D deficiency rather than primary skeletal resistance to PTH. Bone cells isolated from two PHP patients demonstrated normal responsiveness to PTH in vitro.145,146 Serum 1,25-dihydroxyvitamin D is often low in untreated patients with PHP type I, and treatment with vitamin D restores calcemic responsiveness to PTH without changing the deficient urinary cAMP response.16,147–149 Selective deficiency of 1,25-dihydroxyvitamin D was invoked as the cause of isolated skeletal resistance to PTH in an unusual variant of PHP.150 Deficient 1,25-dihydroxyvitamin D formation in response to PTH was directly demonstrated in patients with PHP,151,152 and reflects renal resistance to PTH. Dibutyryl cAMP, a metabolically stable analogue, can bypass the site of resistance in PHP type I, increase 1,25-dihydroxyvitamin D formation, and correct other abnormalities secondary to renal resistance to PTH.153–155
The bone-remodeling response to PTH is relatively intact in many patients with PHP. Decreased bone density, increased urinary hydroxyproline excretion, and increased resorption onbone biopsy have been reported.16,156 These findings reflect the difference in requirement for normal 1,25-dihydroxyvitamin D concentration of the remodeling, as opposed to the calcemic response to PTH. Patients with PHP may actually show rickets157,158 or osteomalacia148 that improves on treatment with vitamin D.
Overt osteitis fibrosa cystica is rare in PHP, but some patients with hypocalcemia, hyperphosphatemia, normal renal function, renal resistance to PTH, and skeletal manifestations of excessive PTH secretion (so-called pseudohypohyperparathyroidism) have been reported.159,160 These have virtually all been patients with PHP type Ib. Although subtle findings compatible with bony demineralization can be seen in patients with PHP Ia,156 radiographically visible parathyroid bone disease is rare.160 One group has postulated that there is a spectrum of skeletal responsiveness to PTH in PHP.160 Potential differences in the transduction mechanisms of the skeletal calcemic and remodeling responses to PTH and in the PTH receptors of kidney and bone provide a theoretical basis for lesions that spare the bone-remodeling response to PTH. A more likely explanation for why PTH resistance is more severe in the kidney than in the bone is the tissue-specific difference in Gs-α imprinting, which would lead to more significant loss of Gs-α expression in the renal proximal tubules than in bone. PTH hypersecretion secondary to renal resistance could then lead to the characteristic findings of osteitis fibrosa cystica. This is consistent with the normal responsiveness to PTH of bone cells isolated from PHP patients (145,146). The finding of the same GNAS1 imprinting defect in both PHP Ib patients with and without overt bone disease suggest that these two presentations are due to the same underlying defect rather than being separate genetic entities (Liu et al, 2000a).