Progression of PTH Resistance in Autosomal Dominant Pseudohypoparathyroidism Type Ib Due to Maternal STX16 Deletions

Abstract

Context

Maternally inherited STX16 deletions that cause loss of methylation at GNAS exon A/B and thereby reduce Gsα expression are the most frequent cause of autosomal dominant pseudohypoparathyroidism type Ib (AD-PHP1B). Early identification of these disease-causing variants in the children of affected and unaffected female carriers would prompt treatment with calcium and calcitriol once parathyroid hormone (PTH) levels increase, thereby preventing hypocalcemia and associated complications.

Objective

This study aimed to determine when PTH and calcium abnormalities develop after birth if a STX16 deletion is inherited maternally.

Methods

Forty-four children of affected (n = 7) or unaffected (n = 7) females with a STX16 deletion were investigated for the presence of these variants. If a deletion was identified, measurement of PTH, calcium, phosphate, and thyrotropin (TSH) was advised.

Results

The STX16 deletion that causes AD-PHP1B was identified in 25 children. Pretreatment laboratory results were available for 19 of those cases. Elevated PTH levels were detected by 2 years of age, and these were progressively higher if laboratory testing was first performed after establishing the genetic defect later in life. Total serum calcium levels remained within normal limits until about 5 years of age. TSH levels showed no consistent rise over time.

Conclusion

Establishing whether a STX16 deletion is inherited from a female carrier of a disease-causing variant rapidly establishes the diagnosis of AD-PHP1B. Several years before overt hypocalcemia developed, PTH levels increased, thereby establishing the onset of PTH resistance. Our findings provide diagnostic guidance and when treatment with calcium and calcitriol should be considered in order to prevent hypocalcemia and associated sequelae.

Pseudohypoparathyroidism type Ib (PHP1B) is characterized by resistance to parathyroid hormone (PTH), leading to hypocalcemia and hyperphosphatemia, and in some cases resistance toward additional hormones, including thyrotropin (thyroid-stimulating hormone; TSH) and calcitonin (1-3). The autosomal dominant form of PHP1B (AD-PHP1B) is most frequently caused by maternal deletions in STX16, the gene encoding syntaxin 16, that lead in cis to loss of methylation at the differentially methylated GNAS exon A/B (4-8) (Fig. 1). This epigenetic change reduces, through as-yet unknown mechanisms, expression of the stimulatory G protein α-subunit (Gsα) from the maternal allele in several different tissues, including proximal renal tubules and thyroid.

Figure 1.

Schematic presentation of the STX16-GNAS locus. The GNAS complex gives rise to several sense transcripts and 1 antisense transcript. Gsα, which is encoded by exons 1-13, is biallelically expressed in most tissues, except for few tissues that include the proximal renal tubules, where paternal Gsα expression is partially or completely silenced through yet-undefined mechanisms later in life. The promoters for the antisense transcript (AS), for the XL transcript encoding an extra-large form of Gsα and for the A/B transcript are methylated on the maternal allele and are thus transcribed exclusively from the paternal allele. The NESP promoter is methylated on the paternal allele and transcripts are derived only from the maternal allele. NESP, XL, and A/B transcripts have unique first exons that splice onto GNAS exon 2-13. Boxes represent exons and splicing patterns are represented by solid or stippled angled lines. Arrows show the direction of transcription and * indicate sites that undergo methylation on either the maternal (XL, A/B, and AS) or the paternal (NESP) allele. STX16, the gene encoding syntaxin 16, is depicted by a single box although it comprises several exons. Deletions in this gene that lead to loss of methylation at the maternal exon A/B alone (red X) and thus cause AD-PHP1B are represented by the red bracket.

Pseudohypoparathyroidism type Ia (PHP1A) is caused by variants involving the maternal GNAS exons 1-13 that encode Gsα. PHP1A patients often have elevated TSH levels during neonatal screening and later in life (1-3), while PHP1B patients show less frequently evidence for resistance toward this hormone. PTH resistance in the proximal renal tubules does not become apparent for both PHP variants until Gsα expression declines during childhood, but the progression of hormonal resistance has only been reported anecdotally. For example, a patient with an inactivating Gsα mutation on the maternal GNAS allele had a normal increase in urinary cyclic adenosine monophosphate (cAMP) excretion after PTH infusion at the age of 7 months, yet a completely blunted response by the age of 3.9 years (9). These findings are consistent with those in mice lacking Gnas exon 1 on the maternal allele that show normal PTH levels for the first weeks of life, after which a gradual increase becomes apparent (10).

Even in the absence of symptomatic hypocalcemia, PHP1A is typically diagnosed during childhood because of Albright hereditary osteodystrophy (AHO) features, while PHP1B is frequently not diagnosed until the second decade of life or later. In fact, in families in which maternal STX16 deletions cause PHP1B, the age at which elevated PTH levels are first noted seem to be highly variable, numerous affected members are not diagnosed until adulthood, and some individuals have no evidence for hypocalcemia or secondary hyperparathyroidism, when evaluated later in life (11, 12). Thus, despite the biallelic lack of methylation at GNAS exon A/B, Gsα appears to be expressed in PHP1B patients at sufficient levels to prevent hormonal resistance for extended periods of time. However, it is unknown whether paternal and maternal Gsα transcription declines simultaneously in different tissues, whether this reduction occurs at the same age, and whether it involves similar as-yet unknown mechanisms for both alleles.

After the discovery of the 3-kb STX16 deletion as the most frequent cause of AD-PHP1B (4), a familial maternal 4.4-kb deletion in that gene was identified, which also leads to isolated loss of methylation at GNAS exon A/B (5). Discovery of the latter disease-causing genetic variant made it possible to prospectively follow 1 AD-PHP1B patient since birth because her diagnosis had been established through GNAS methylation studies of cord blood DNA. Repeated testing revealed that her PTH was significantly elevated by the age of 21 months, despite normal calcium and phosphate levels. These laboratory findings indicated that renal PTH resistance was not obvious until the second year of life. However, once Gsα protein levels had declined in the proximal renal tubules sufficiently to impair PTH-dependent 1,25(OH)₂ vitamin D generation, PTH levels increased, thereby enhancing calcium reabsorption in the distal renal tubules. In this portion of the kidney and in bone, PTH resistance does not develop at all or not to the same extent, thus allowing PTH-dependent calcium reabsorption. Furthermore, calcium is mobilized from bone, which also helped maintain her blood calcium levels within normal limits.

If an inherited genetic defect in STX16 (or GNAS) is known, the diagnosis of AD-PHP1B can be readily established or excluded by a simple polymerase chain reaction test using genomic DNA extracted from blood leukocytes or buccal scrapings (13). Thus, it is feasible to noninvasively determine whether females with a STX16 deletion had passed the disease-causing variant to their children and would therefore be at risk of developing symptomatic hypocalcemia and/or persistently elevated PTH levels. This makes it possible to rapidly establish whether an offspring is affected by PHP1B without the need for laboratory testing, which can provide false reassurance that an individual is unaffected if only calcium and/or phosphate are determined (5). Furthermore, unless PTH is elevated, repeated measurements are required to ascertain whether an individual is unaffected, since hormonal resistance typically does not develop until later in life (5, 11, 12). Noninvasive genetic testing can thus help avoid the need for recurrent laboratory studies of affected and unaffected individuals. Children who test positive for a STX16 (or GNAS) mutation can be monitored in regular intervals until PTH levels are elevated, at which point treatment with calcium and activated vitamin D analogs should be considered, thereby preventing clinically significant sequelae from hypocalcemia and hyperphosphatemia, and possibly other manifestations of abnormal mineral ion homeostasis (2).

We now describe our experience with offering genetic testing for the offspring of females who are known carriers of STX16 deletions, that is, females who are affected by PHP1B (maternal STX16 deletions) or are unaffected carriers of such a mutation (paternal STX16 deletions). If a mutation was identified, measurement of PTH, calcium, and phosphate, and TSH was encouraged to determine whether evidence for PTH- and/or TSH resistance and mineral ion abnormalities were apparent. The ages at which PTH levels were elevated for individuals with maternal STX16 deletions provided insights into the progression of PTH resistance, thus helping to guide the initiation of treatment with 1,25(OH)₂ vitamin D and calcium.

Methods

Laboratory Data and Physical Findings of the Investigated PHP1B Patients

We investigated offspring of 14 females from 3 unrelated AD-PHP1B kindreds, who are either affected by PHP1B (maternal mutation, n = 7) or unaffected carriers (paternal mutation, n = 7) of the 3-kb (4, 13, 14) or the 4.4-kb STX16 deletion (5) (Table 1). The STX16 deletion that causes AD-PHP1B was identified in 25 of their 44 children (57%), which prompted measurement of calcium, phosphate, and PTH levels, as well as TSH and free thyroxine (fT4) for 20 of these children, including the previously reported case W-12 (5). No laboratory measurements have yet been obtained for 5 of previously not reported cases, who are carriers of a maternal STX16 deletion.

Table 1.

Investigated children of affected or unaffected mothers with STX16 deletions

Information for the mothers			Girls		Boys
Code	STX16 deletion	Affected	Affected	Unaffected	Affected	Unaffected	References
F25	3-kb	yes		1	1		4, 13
F35	3-kb	no	1	1		1	4, 13
F42	3-kb	yes	1			1	4, 13
F45	3-kb	yes	3		2		4, 13
F47	3-kb	yes	1			1	4, 13
36/II-1	3-kb	no	1	1	3	2	13, 14
36/II-2	3-kb	no	2	4		3	13, 14
36/III-3	3-kb	yes		1	2		13, 14
36/III-4	3-kb	yes	1		1		13, 14
W-II/5	4.4-kb	no			1	1	5, 13
W-II/6	4.4-kb	no	2				5, 13
W-II/8	4.4-kb	yes			1		5, 13
W-II/9	4.4-kb	no	1			1	5, 13
W-II/10	4.4-kb	no		1	1		5, 13
		Number of children	13	9	12	10

Codes of the mothers are provided, as well as the size of their STX16 deletion and whether they are affected or unaffected. For each mother, the number of affected and unaffected boys and girls is indicated.

Genetic Analysis

Genomic DNA, extracted from either blood cells or buccal scraps, was used for multiplex polymerase chain reaction amplification employing previously reported primers and conditions (4, 5).

Results

Genetic testing was performed for 44 offspring of affected or unaffected females with maternal or paternal STX16 deletions, respectively. These studies revealed for 25 children a maternally inherited STX16 deletion, thus establishing the diagnosis of AD-PHP1B. Based on these findings, measurement of PTH, calcium, and phosphate, as well as TSH and fT4, was encouraged. Results were obtained for 19 previously not reported individuals, who were diagnosed with AD-PHP1B before treatment with oral calcium and 1,25(OH)₂ vitamin D was initiated.

The 3-kb STX16 deletion, initially reported for 2 females in kindred 36, 36/III-3 and 36/III-4 (14), is the most frequent genetic defect leading to AD-PHP1B if inherited maternally (1-3). During childhood, both patients had suffered from frequent paroxysmal attacks during which their legs would stiffen, causing them to fall forward. They furthermore experienced facial spasms, tremors, and choreoathetotic movements of hands and feet, which worsened during pubertal development; the older girl furthermore revealed radiographic evidence for brown tumors. After establishing the diagnosis of AD-PHP1B and initiating treatment with calcitriol and calcium symptoms resolved in both cases.

Because of the severe, long-standing abnormalities encountered in 36/III-3 and 36/III-4 that worsened in part because of the considerable delay in establishing the diagnosis of AD-PHP1B, the children born to 36/III-3 underwent genetic testing on day 2 of life and at 4.5 months of age, respectively. Furthermore, the 2 boys for whom the mutation was identified (36/IV-1 and 36/IV-2) underwent frequent laboratory evaluations in an effort to avoid hypocalcemia and secondary hyperparathyroidism that had presumably started early in the life of their mother.

An increased PTH level was first detected for 36/IV-1 at the age of 2 years (Fig. 2A). At that time, he was receiving 400 IU vitamin D daily, which was increased by the age of 3 years and 4 months to 900 IU vitamin D daily. Initially, he had 3 to 5 servings of dairy, later 2 servings of dairy on most days; because of the elevated PTH, the dietary calcium intake of patient 36/IV-1 was furthermore supplemented with elemental calcium (250 mg/day). However, despite increasing the doses of calcium carbonate and vitamin D, his total calcium levels continued to decline but remained within normal limits over the first years of life (Fig. 2B). His serum phosphate levels also remained within normal limits (Fig. 2C). PTH levels, which had remained elevated above 200 pg/mL after the age of 4 years, were strongly correlated with calcium (R² = 0.71), but less with phosphate (R² = 0.48) (Fig. 2E and 2F). Therapy with levothyroxine (L-T4) (25 mcg) was initiated at the age of 13 months, when the TSH level was 16.3 mIU/L (Fig. 2D) and treatment with 1,25(OH)₂ vitamin D (0.25 mcg/day) was started at the age of 5 years and 6 months. By the age of 5 years and 11 months, when his elemental calcium intake was 500 mg twice daily, calcitriol was escalated to 0.5 mcg/day, which gradually reduced PTH levels to the upper end of the normal range (data obtained after initiation of treatment with 1,25(OH)₂ vitamin D are not shown in Fig. 2).

Figure 2.

Laboratory findings before 6 years of age in 4 children with AD-PHP1B due to maternally inherited 3-kb STX16 deletions. Laboratory measurements for PTH (panel A), calcium (panel B), phosphate (panel C), and TSH (panel D) were obtained before treatment with 1,25(OH)₂ vitamin D was started. Relationship between PTH and calcium (panel E), and between PTH and phosphate (panel F). The gray areas represent the normal ranges for each analyte; individual patients are indicated by different symbols and colors.

The STX16 deletion was identified for patient 36/IV-2 at the age of 4.5 months and he therefore underwent, as his older brother, frequent laboratory testing. The PTH level was first elevated at 25 months of age (see Fig. 2A), but his calcium and phosphate levels remained within normal limits. Because of the prior experience with his older brother, treatment with 1,25(OH)₂ vitamin D was initiated at the age of 3 years and 11 months. Patient F-58 had normal PTH, calcium, and phosphate levels at the ages of 14 and 18 months, respectively, but after being lost to follow-up his PTH levels had increased to 449 pg/mL by the age of 58 months (see Fig. 2A), when treatment with calcium and calcitriol was started. Patient F-63 had slightly elevated PTH levels at 27 and 29 months of age, with normal calcium and phosphate.

Because of the delayed onset of PTH resistance in 36/IV-1 and 36/IV-2, which is similar to our previously reported case (5), we evaluated numerous additional AD-PHP1B patients after a STX16 deletion was identified. As anticipated for a marker that increases as PTH resistance develops postnatally, PTH levels were elevated in most cases for whom a STX16 deletion had been identified by the age of 2 years and thus had prompted laboratory evaluations (Fig. 3A). After repeat laboratory testing to confirm elevated PTH levels, therapy was initiated. Individuals who underwent laboratory testing for a STX16 deletion after the age of 5 years showed, when first measured, more profoundly increased PTH levels.

Figure 3.

First laboratory measurements after identification of a STX16 deletion in untreated AD-PHP1B patients. Measurements of parathyroid hormone (PTH, panel A), calcium (Ca, panel B), phosphate (P, panel C), and thyroid-stimulating hormone (TSH, panel D) for 19 cases with AD-PHP1B due to STX16 deletions. All laboratory measurements were obtained before treatment with calcium and 1,25(OH)₂ vitamin D was initiated. Relationship between PTH and calcium (panel E), and between PTH and phosphate (panel F). The gray areas represent the normal ranges for each analyte; for each patient only the last laboratory measurement before initiation of treatment is shown; individual patients are indicated by different symbols and colors.

Calcium levels were all within the normal range for children younger than 5 years (Fig. 3B), even though PTH levels were already elevated. This suggested that normocalcemia had been maintained because of enhanced PTH-dependent bone resorption combined with a reduction in urinary calcium excretion, since increased PTH-dependent calcium reabsorption in the distal tubules. Much lower calcium levels were measured for those AD-PHP1B cases, who had been diagnosed between the ages of 5 and 10 years; 1 patient had a calcium level of only 5.5 mg/dL. Two of the patients older than 10 years had calcium levels within the normal range, while the 4 other older cases had levels that were moderately to severely reduced. Calcium and PTH levels were correlated (R² = 0.46) (Fig. 3E).

Phosphate levels were at the upper end of the normal range or slightly elevated in children younger than 5 years, but they were considerably elevated in several patients between the ages of 5 and 10 years, as well as during adolescence, while adult patients revealed levels within the normal range (Fig. 3C). Phosphate and PTH levels in patients younger than 10 years were less well correlated (R² = 0.24; Fig. 3F). TSH levels were elevated only in 3 cases (Fig. 3D).

The levels for PTH, TSH, calcium, and phosphate were not statistically different for males and females with a maternally inherited STX16 deletion, and there were no laboratory differences for the offspring of affected and unaffected mothers.

Discussion

Maternal STX16 deletions are the most frequent cause of AD-PHP1B (1-8). In the current studies, we investigated numerous children born to women who had previously been shown to be either affected or unaffected carriers of a 3.0-kb or a 4.4-kb STX16 deletion (4, 5, 13, 14). Because their genetic deletions were known, several of these women approached us to help determine whether their children had inherited the genetic defect. Alternatively, we alerted known female carriers of a STX16 deletion to the possibility of genetically testing their children in order to prevent sequelae from hypocalcemia and prolonged PTH elevations in case they had inherited the mutant allele. If feasible, genomic DNA was extracted from buccal scrapings, which offered a noninvasive method for obtaining genetic material in sufficient amounts to allow testing for the previously established disease-causing familial mutation; this was particularly helpful for testing infants or young children (13). If a family-specific mutation was identified, it was suggested to pursue laboratory testing.

Offspring of an equal number of females from 3 unrelated kindreds were investigated, who were carriers of a maternally or paternally inherited STX16 deletion. Laboratory abnormalities were indistinguishable for male and female offspring and whether their mothers were affected or unaffected. None of the children who had inherited the mutant allele had been overtly symptomatic, although some of the affected children had shown profound hypocalcemia when laboratory testing was performed after establishing that the disease-causing deletion had been inherited.

Total blood calcium levels were within the normal range for all children who had been tested before the age of 5 years. However, their PTH levels were considerably elevated, which is similar to our previously reported findings in a single AD-PHP1B patient (5). These findings suggest that PTH-dependent bone resorption, possibly in combination with maximized calcium reabsorption in the distal renal tubules, can maintain normocalcemia in PHP1B patients for prolonged periods of time. As predicted, total calcium and PTH levels were strongly correlated for the entire cohort. Phosphate levels were at the upper end of normal or frankly elevated in some patients. However, only longitudinal phosphate measurements for 2 young AD-PHP1B patients correlated well with PTH levels.

Six cases were diagnosed with AD-PHP1B after the first decade of life, with 1 patient showing a total calcium level below 6 mg/mL in association with much elevated PTH levels. Importantly, several infants had normal total calcium and PTH levels when evaluated during the first 2 years of life, indicating that sufficient amounts of Gsα protein can be maintained in the proximal renal tubules for some time after birth despite the STX16 deletion and the associated loss of methylation at GNAS exon A/B. Subsequently, paternal Gsα expression appears to decline in those tissues or cells that rely, later in life, predominantly on maternally derived Gsα. It remains unknown, however, whether the reduction of Gsα expression follows different temporal, possibly tissue- or cell-specific patterns for the parental alleles. Given that STX16 deletions lead to an indistinguishable lack of methylation at the exon A/B differentially methylated region (DMR), it appears unlikely that the progressive reduction of Gsα expression involves distinct, parent-specific silencing mechanisms. These underlying incompletely defined mechanisms may involve biallelic expression of A/B transcripts from the nonmethylated exon A/B promoters, since the mRNA derived from exon 1A, that is, the murine equivalent of A/B, was shown to reduce Gsα expression in mice (15-17).

Prolonged maintenance of normocalcemia and normal PTH levels during infancy indicates that laboratory testing alone can provide falsely reassuring results. Our findings emphasize the importance of genetic testing to establish the diagnosis of AD-PHP1B and the need for repeat measurements of PTH, calcium, and phosphate to determine when treatment with activated vitamin D analogs should be initiated. Furthermore, so shown for infants 36/IV-1 and 36/IV-2, treatment with calcium carbonate and ergocalciferol alone failed to increase blood calcium sufficiently to prevent the continuous rise of their PTH levels. In fact, blood calcium increased and PTH normalized only after treatment with 1,25(OH)₂ vitamin D had been started, thus further highlighting the inability of the proximal tubules to generate sufficient amounts of this biologically active vitamin D analog despite elevated PTH levels. TSH levels were largely within normal limits, indicating that Gsα expression in the thyroid is typically not as profoundly impaired as in the proximal renal tubules. However, as observed for patient 36/IV-1, TSH levels can fluctuate significantly (3.6 to 10.6 mIU/mL) with low-normal fT4, suggesting that treatment with L-T4 may need to be started only if TSH levels are persistently or progressively elevated.

In summary, children of affected or unaffected female carriers of known STX16 mutations should be evaluated genetically to determine whether they are affected by AD-PHP1B. If the disease-causing variant is identified, an elevated PTH level helps determine when treatment with calcium and calcitriol should be initiated, even if calcium levels are still within normal limits.

Glossary

Abbreviations

AD-PHP1B

autosomal dominant pseudohypoparathyroidism type Ib

fT4

free thyroxine

Gsα

G protein α-subunit

L-T4

levothyroxine

PHP1A

pseudohypoparathyroidism type Ia

PHP1B

pseudohypoparathyroidism type Ib

PTH

parathyroid hormone

TSH

thyrotropin (thyroid-stimulating hormone)

Additional Information

Disclosures: The authors have no conflict of interest.

Data Availability: Some or all data generated or analyzed during this study are included in this published article or in the data repositories listed in References.