| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CLINICAL CASE SEMINAR |
Departments of Pediatrics (M.H.W., C.E.-T., A.L.) and Pathology (C.T.), University of Texas Southwestern Medical Center, Dallas, Texas 75390; Orthopedic Pathology (F.H.G.), Armed Forces Institute of Pathology, Washington, D.C. 20306-6000; Division of Bone and Mineral Diseases (X.Z., S.M., M.P.W.), Washington University School of Medicine, Barnes-Jewish Hospital, St. Louis, Missouri 63110; and Center for Metabolic Bone Disease and Molecular Research (S.M., M.P.W.), Shriners Hospitals for Children, St. Louis, Missouri 63131
Address all correspondence and requests for reprints to: Dr. Michael P. Whyte, Shriners Hospitals for Children, 2001 South Lindbergh Boulevard, St. Louis, Missouri 63131. E-mail: mwhyte{at}shrinenet.org.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Here, we detail the clinical, radiographic, biochemical, molecular, and autopsy histopathological findings of a neonate of normal body length manifesting a distinctive disarray of both skeletal patterning and ossification associated with fatal respiratory insufficiency. The cumulative observations, including low serum levels of alkaline phosphatase (ALP) and osteocalcin (OC), suggested a primary osteoblast defect, perhaps due to hypophosphatasia (HPP) [Mendelian inheritance in man (MIM) no. 241500] or cleidocranial dysplasia (CCD) (MIM no. 119600). However, these diagnoses were not supported by molecular studies. Investigation of the genes encoding OC and the osteoblast transcription factor osterix (OSX) was also unrevealing. Instead, the disorder seems to be a unique osteochondrodysplasia, perhaps associated with parathyroid insufficiency, that disrupts both intramedullary and endochondral bone formation.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This term, 2.9-kg boy was born to a black 14-yr-old primigravida after little prenatal care. She presented to labor and delivery with intact amniotic membranes and evidence of pregnancy-induced hypertension. Ultrasonography revealed severe oligohydramnios, but no fetal abnormalities were noted. Labor was induced, and the baby was delivered vaginally using low-outlet forceps. Apgar scores were depressed at 1 min (Apgar score, 1) but responded at 5 and 10 min of postnatal life (Apgar scores, 6 and 7, respectively) to tracheal intubation and assisted ventilation for poor respiratory effort.
Physical examination showed body size appropriate for gestational age (weight, 75th percentile; length, 90th percentile; and fronto-occipital circumference, 30th percentile), yet the ponderal index of 0.015 (<10th percentile) indicated intrauterine growth retardation. Gestational age was approximately 36 wk by Ballard assessment (7). The skull was extremely soft, pliable, and felt like an inflated rubber balloon. Bone was palpable only in the facial bones and base of the occiput. There was frontal bossing, midface hypoplasia, and hypertelorism with prominent eyes (Fig. 1
, A and B). Ears were posteriorly rotated. The broad chest had widely spaced nipples. Clavicles were not palpable. A dimple was noted over the right lateral tibia.
|
The baby remained ventilator dependent, seemingly due to chest wall instability associated with defective thoracic ossification seen on radiographs. No glucocorticoid therapy was administered. He died at 16 d of age after respiratory support was withdrawn.
Radiographic findings
Radiographic survey revealed a variety of skeletal abnormalities. Ossification of the head was limited to the facial and basal portions of the skull (Fig. 2, A and B). The calvarium appeared as though it was a fibrous membrane. The mandible had a midline vertical cleft. Clavicles were absent. Scapulae were incompletely ossified. Humeral and tibial lengths measured at just the 3rd percentile, but the limbs were not short on physical examination due to radii and ulnas that were longer than the 97th percentile. No specific measurements of the femurs were taken. No fractures were noted, and the epiphyses and metaphyses of long bones appeared normal (Fig. 2C). The right tibia and fibula were bowed anteriorly. The tali and calcanei were not ossified (Fig. 2C). Coronal clefts involved all vertebral bodies, and the interpediculate distances were wide (Fig. 2D). The pubis and ischium were not ossified. Chest x-ray showed normal lung fields and volumes with an unremarkable cardiothymic shadow (Fig. 2E). However, only 11 pairs of ribs were present.
|
Low serum ALP activity, ranging from 6–28 IU/liter (normal range, 80–270 IU/liter), prompted further biochemical studies using a separate but limited serum specimen shipped frozen to the Research Center, Shriners Hospitals for Children (St. Louis, MO).
Serum total ALP activity was again markedly low at 39 IU/liter [laboratory reference range for children, 133–347 IU/liter (± 2 SD of the mean)] (8), with subnormal activity of the bone isoform of ALP at 17 IU/liter (normal range, 47–181 IU/liter) (8). Serum OC (Diagnostic Products Corp., Los Angeles, CA) was also strikingly low at 2 ng/ml (normal range for age, 10–64 ng/ml), with the assay repeated and result verified (Table 1).
|
). Vitamin D studies could not be performed.
Because the inborn error of metabolism HPP is characterized by low serum ALP activity (hypophosphatasemia) and the perinatal form of HPP manifests lethal respiratory insufficiency due to skeletal undermineralization at birth (9), we assayed two substrates for ALP that accumulate endogenously in HPP, phosphoethanolamine (PEA) and pyridoxal 5'-phosphate (PLP) (9). Urinary PEA (Children’s Medical Center, Dallas, TX) seemed increased, but precise quantification was precluded by coelution of taurine. However, serum PLP concentration was unremarkable at 21 nM [pediatric reference range, 5–107 nM (± 2 SD of the mean)] (8) and not markedly elevated, as invariably occurs in severe HPP (Table 1) (9, 10). Additional patient serum or urine was not available for further study.
The mother’s biochemical findings, studied 6 months after parturition, showed no evidence that she was a carrier for the disorder. Her serum total ALP activity was unremarkable at 76 IU/liter (normal range, 30–114 IU/liter); bone ALP isoform activity was 24 IU/liter (normal range, 3–38 IU/liter); and PLP concentration was 7 nM (normal range, 20–125 nM). No additional serum was available for OC assay.
Histopathology
Postmortem gross examination showed apparent absence of both the bony calvarium and clavicles. The skull consisted of a fibrous membrane. Gliosis and cystic degeneration of the right occipital lobe of the brain suggested prior hypoxic-ischemic damage. Although serum PTH had been detected at a normal level (see Biochemical investigations), no parathyroid glands were observed, but this is not uncommon in neonatal autopsies. No other soft tissue defects were encountered.
Portions of a proximal and distal tibia were processed at the Armed Forces Institute of Pathology (Washington, D.C.) according to standard protocols using a method that does not require decalcification. Four micron-thick sections were cut and used for Masson trichrome, Goldner trichrome, and Villaneuva stains. All were examined with a light microscope.
Chondrocytes in the growth plates had unremarkable appearances in the resting and in the proliferative zones. However, these cells were mildly to moderately disorganized in the hypertrophic zone and in the zone of provisional calcification (Fig. 3). The hypertrophic zones undulated with the primary spongiosa. Chondrocytic lacunae occurred in nests. Aggregates of hypertrophic chondrocytes were bunched up cells lacking the normal stacked appearance. Zones of provisional calcification followed the same pattern. Nevertheless, growth plate and primary spongiosa alterations were not those classically observed in rickets.
|
Trabecular bone seemed to form and to resorb normally. There was no increased osteoid deposition to indicate rickets, either reflected by total bone surfaces covered by osteoid or by osteoid seam width. There was no evidence of excessive bone degradation.
The marrow was unremarkable and retained cellular trilineage maturation.
Genetic studies
Karyotype was 46,XY. After informed written consent approved by the Washington University School of Medicine (St. Louis, MO), mutation searches for HPP [tissue nonspecific ALP isoenzyme gene (TNSALP)] or CCD [core-binding factor A1 gene (CBFA1, also called RUNX2)] or for defects involving the genes encoding OC or the human ortholog of the mouse gene for OSX [MIM no. 606633] were performed using the patient’s genomic DNA obtained from peripheral blood leukocytes. OSX is an osteoblast transcription factor and osx knockout mice have profound defects in skeletal mineralization (see Discussion).
TNSALP. In HPP, global deficiency of activity of the TNSALP (liver/bone/kidney) isoenzyme of ALP reflects deactivating mutations in the TNSALP gene (9). TNSALP coding exons (exons 2–12) and the adjacent splice sites were screened using two methodologies, denaturing gradient gel electrophoresis (DGGE) and denaturing high-performance liquid chromatography (dHPLC) (see Ref.11 , and methods not shown).
CBFA1. All eight CBFA1 exons (exons 0–7) and adjacent splice sites were screened for alterations using PCR amplification and DNA sequencing with published primers (12, 13). In addition, a recently identified upstream exon, referred to as exon –1, which encodes the 5'flanking region, was PCR amplified and sequenced using the following primers: forward primer, 5'-CAGTTTATCAAAGAATCATAC C-3'; and reverse primer, 5'-CCTCTCCAGTAATAGTGCTTG C-3' (14). The CBFA1 promoter region was PCR amplified and sequenced using the following primers: forward primer, 5'-CTCTGTTGGTCTCGGTGGCTGGTA-3'; and reverse primer, 5'-AATTTCACACAGACTCTTGAGCC-3'. For all instances, both strands of the PCR products were sequenced (15).
OC. The gene structure for OC (bone gla protein) was determined by aligning the cDNA (accession no. NM_000711) with the genomic (X04143) DNA sequence. PCR/sequencing primers were designed to cover the entire coding sequence. Two primer sets were developed to amplify exons 1 and 2 as well as exons 3 and 4 with their intervening sequences. Primers for exons 1 and 2 were as follows: forward, 5'-ACAGTGCTGGAGGCTGGCGG-3'; and reverse, 5'-GAATGAGACTGAGGGACCAGGG-3'. The primers for exons 3 and 4 were as follows: forward, 5'-TGATCCTCCCAAACCCAGAGC-3'; and reverse, 5'-GTGCCTGGAGAGGAGCAGAAC-3'. Both sets required 10% dimethylsulfoxide in the PCR amplification that included a 94 C denaturation for 4 min, followed by 32 cycles of 94 C for 20 sec, 63 C for 30 sec, and 72 C for 45 sec.
OSX. To identify the human ortholog of the mouse osx gene, we performed a BLAST search of GenBank using the murine osx cDNA sequence (GenBank accession no. AF184902). This revealed a human cDNA (GenBank accession no. XM_062600) that, when theoretically translated, generates a protein sequence homologous to the product of the mouse osx gene. Subjected to BLAST search, this located the human genomic OSX sequence on chromosome 12 (accession no. NT_009563). From the alignment of the human cDNA and genomic sequences, we determined the OSX gene structure and designed PCR primers to amplify and sequence the coding regions and splice sites.
Resembling the mouse osx gene, two putative OSX exons were identified. PCR primers for exon 1 were as follows: forward, 5'-CCGCTGGGAAAGCTGTAATTAGA-3'; and reverse, 5'-GACAAGCACCACAGCTACTATC-3'. PCR amplification included a 94 C denaturation for 4 min, followed by 32 cycles of 94 C for 20 sec, 56 C for 30 sec, and 72 C for 30 sec. Due to the large size of exon 2, two overlapping primer sets were developed. Exon 2 (set 1) primers were as follows: forward, 5'-GTTGCCCTTCATTTTGTACC-3'; and reverse, 5'-GCCTTGGGTTTATAGACATC-3'. Exon 2 (set 2) primers were as follows: forward, 5'-CTCAGCCTCCACTGAACC-3'; and reverse, 5'-GTGATTGGCAAGCAGTGGTC-3'. PCR amplification included a 94 C denaturation for 4 min, followed by 32 cycles of 94 C for 20 sec, 62 C for 30 sec, and 72 C for 30 sec.
Serum ALP and OC data review
Serum ALP levels are measured routinely, using standard automated methods (Vitrus 250; Johnson & Johnson, Rochester, NY), for all patients at the Research Center, Shriners Hospitals for Children (8).
Serum OC levels, reflecting 31 of our patient contacts with relatively severe HPP (six infantile and 25 childhood cases) (9) and three patients with CCD, were quantitated in our laboratory from November 1994 to August 1998 using the kit marketed by Nichols Institute Diagnostics (San Clemente, CA) and after April 2001 using the kit from Diagnostic Products Corp.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Although all of our 31 HPP patients studied had characteristically subnormal serum ALP activity, none had a low serum OC concentration. The HPP OC values were in or above the reference range for age. Therefore, the patient’s serum OC value was distinctly low compared with HPP patients and healthy controls.
The three children with CCD, all documented with mutations in CBFA1, had serum ALP levels that were low or low to normal at 185, 150, and 90 IU/liter at ages 9 yr, 7 yr, and 9 months (uncorrected by age 3 yr), respectively (normal range, 133–347 IU/liter). Serum OC levels had been measured in two of these CCD patients and were unremarkable at 65 and 26 ng/ml (normal range, 10–64 ng/ml). Therefore, both the serum ALP and OC levels in our patient were distinctly lower than in these three CCD patients.
Molecular studies
We examined four candidate genes (TNSALP, CBFA1, OC, and OSX) for mutations that might cause our patient’s skeletal disorder (see Discussion). He manifested some skeletal features of HPP, caused by defects in TNSALP, and CCD, caused by mutations in CBFA1. Because our patient had low circulating levels of OC, which is an early expression marker for osteoblast differentiation, we also examined the OC gene. Finally, based on some similarities with the phenotype of the osx knockout mouse and the role of osx in osteoblast differentiation, we also examined the OSX gene.
No defect was detected in the coding exons or splice sites of the TNSALP, CBFA1, OC, or OSX genes in our patient. We did find in TNSALP exons 5 and 9, by sequencing, two silent polymorphisms that predict a nucleotide change in the mRNA [C330T (Ser93Ser) and A876C (Pro275Pro)], but no amino acid change in the TNSALP isoenzyme itself. The Ser93Ser polymorphism has a frequency of 5%, and the Pro275Pro polymorphism has a frequency of 47% in normal populations (16, 17). Notably, these two polymorphisms were heterozygous, indicating that the patient was not homozygous for a single TNSALP mutant allele [homozygous mutations, or a deletion encompassing an entire exon (or larger), can sometimes escape detection by DGGE or dHPLC] (11).
For the CBFA1 gene, in addition to the coding exons (exon –1 and exons 0–7), we sequenced the 5'flanking region and the promoter regions without finding any mutation (14, 15).
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
HPP
HPP, characterized biochemically by hypophosphatasemia, features impaired skeletal mineralization due to globally deficient activity of TNSALP, one of four ALP isoenzymes in humans (9, 18, 19). All cases of HPP studied molecularly have revealed one or two mutations of TNSALP (http://www.sesep.uvsq.fr/Database.html) (9, 11, 17, 20, 21). Currently, 138 distinctive TNSALP defects have been identified in patients worldwide (http://www.sesep.uvsq.fr/Database.html). Consequently, deficient TNSALP activity on cell surfaces diminishes hydrolysis of PEA, inorganic pyrophosphate, and PLP, which accumulate extracellularly (9, 18, 19). Rickets or osteomalacia in HPP seems to reflect inorganic pyrophosphate blockade of hydroxyapatite crystal formation and growth (18, 19, 22, 23). Nevertheless, HPP expressivity varies remarkably, ranging from intrauterine death with essentially an unmineralized skeleton to a remote history in seemingly healthy adults of premature loss of their primary dentition during early childhood (9). Generally, the clinical severity of HPP correlates inversely with patient age at diagnosis and directly with the magnitude of the biochemical disturbances (9).
For our patient, the most pernicious form of HPP seemed possible because perinatal HPP manifests with profoundly low serum ALP activity and generalized skeletal hypomineralization, typically leading to lethal respiratory compromise soon after birth (9, 23). The radiographic manifestations are well characterized (24). Affected fetuses and newborns have deformed, soft skulls with large "fontanelles," reflecting deficient ossification of the calvarium. Ribs are poorly mineralized. The spine can contain unossified vertebral bodies and may be sagittally or coronally clefted. Long bones are almost always shortened and frequently bowed. Midshaft spurs are mentioned in some reports (23, 25, 26, 27), and distal spurring is mentioned in others (27). Severe growth plate and metaphyseal defects appear like rickets (24), but they occur despite normal or high extracellular levels of calcium and inorganic phosphate (9).
Nevertheless, our patient did not have HPP. Polyhydramnios, not oligohydramnios, typifies perinatal HPP (9). His birth length was not compromised (>90th percentile), although his humeral and tibial lengths were less than 3rd percentile. He otherwise had relatively normal-appearing humeri and femora (24). Ossification of clavicles is characteristically preserved to some degree in HPP, and absence has not been reported (24). There was neither rachitic change on radiographs nor evidence of impaired mineralization of skeletal matrix on histopathological studies. Furthermore, his plasma PLP concentration (a highly specific marker for HPP) was normal despite hypophosphatasemia (10). Finally, homozygous or heterozygous combinations of two deactivating TNSALP mutations cause the severe clinical forms of HPP (9, 11, 20, 21). All but one of 19 severely affected HPP patients that we studied (screening with DGGE and dHPLC) (11) were homozygotes or compound heterozygotes for TNSALP defects at splice sites or within exons. Our only exception was a single TNSALP mutation detected in one such patient (11), perhaps reflecting a second defect located outside the coding region or splice sites but more likely reflecting additional medical problems (Mumm, S. and M. P. Whyte, unpublished observations). Hence, together with the cumulative findings, absence of any TNSALP mutation on DGGE or dHPLC in our patient effectively excludes HPP.
CCD
CCD was suggested by our patient’s poorly ossified calvarium and apparent absence of clavicles, which are hallmarks of CCD (2, 4, 5). Additional CCD features included frontal bossing, a depressed nasal bridge, hypertelorism, hypoplastic scapulae, ribs of variable number with short ossified portions, and hypoplasia and delayed ossification of pelvic structures (28). Patients with CCD can, like ours, have normal birth lengths but later develop short stature (29).
Notably, hypophosphatasemia can occur in CCD. We first described this in a preliminary communication in 1992 (30), which predated discovery by others in 1997 of CBFA1 deactivation in CCD (12, 28). Simultaneously with our abstract and also in a preliminary report, Bull et al. (31) followed a CCD patient with hypophosphatasemia throughout infancy who suffered significant respiratory failure requiring mechanical ventilation at 1 month of age. From this (and additional unpublished experience), we have cautioned that CCD resembles severe HPP in infancy because hypophosphatasemia occurs with osteopathy, including a hypomineralized skull (9, 18, 19). In fact, this clinical-radiographic-biochemical overlap between CCD and HPP was emphasized in 2002 in separate case reports concerning hypophosphatasemia in CCD patients with CBFA1 defects and intact TNSALP genes (32, 33). However, these publications did not mention serum OC levels (see Unique disorder). For completeness here, we also caution (9) that the most severe (type II) form of osteogenesis imperfecta (OI) can cause transient neonatal hypophosphatasemia (34). Several other severe diseases sometimes engender low serum ALP activity but do not feature skeletal disease and, therefore, were not suspected for our patient (9, 18, 19).
However, CCD in our patient seemed unlikely because of his lethal outcome and especially severe skeletal radiographic abnormalities (4, 5). Neonatal death can occur in CCD from central nervous system insult related to poor skull ossification (35). Our patient’s cerebral injury noted at autopsy would unlikely be fatal. Neonatal respiratory distress has not been described in CCD. Instead, his pulmonary insufficiency was attributable to an unstable chest wall, despite an overall rib cage size that was not reduced and a thorax that was not bell shaped. In addition to the patient of Bull et al. (31), Chitayat et al. (36) described a 5-month-old CCD infant with pulmonary distress attributed to a narrow thorax. However, there is no other mention of ventilator dependence in CCD. Hence, our patient’s disorder seems more severe than previously reported for CCD.
In fact, we tested our patient for CBFA1 mutation but found none. CCD, an autosomal dominant trait (4, 5), was mapped to chromosome 6p21 in 1995 (37, 38, 39). CCD gene defects (40, 41), first elucidated in 1997 (12), involve the transcription factor CBFA1, which governs osteoblast differentiation (12, 40, 41). CBFA1 mutations in CCD include small deletions and insertions that disrupt the reading frame as well as missense and nonsense changes (12, 13, 42). Reports of affected siblings, originally thought to support autosomal recessive inheritance, instead suggest gonadal mosaicism (43). Now, sporadic cases of CCD are considered spontaneous CBFA1 mutations (43).
Notably, four patients with CBFA1 mutations differ from the classic CCD phenotype. In one patient, expansion of a polyalanine stretch from 17 to 27 residues accompanied only minor craniofacial features of CCD and brachydactyly (40). Condensation of this polyalanine tract from 17 to 11 residues, however, represents a normal polymorphism in CBFA1 (40). The three other unusual CCD phenotypes involved hypomorphic mutations (R391X, T200A, and 90insC) (42) but also caused mild CCD and/or only a dental phenotype.
Negative mutational analysis of CBFA1 in our patient goes against a diagnosis of CCD (12, 13, 42). However, our molecular methodology does not exclude heterozygous deletion of CBFA1. In fact, other investigators, using the same DNA sequencing technique, identified CCD-causing CBFA1 mutations in only 18 (43%) of 42 of affected individuals (12). Possibly, our patient represents the extreme of CCD with remarkable, extensive involvement of the thorax. However, complete deletion of one CBFA1 allele has been reported in at least three cases of CCD (40), yet the phenotype was typical for CCD and similar to other loss of function mutations. These defects include many nonsense or insertion mutations leading to frame shifts or small deletions, all of which engender a partial loss of the CBFA1 protein. CCD is also commonly caused by missense mutations, primarily in the RUNT domain (a DNA binding motif), but other defects can also occur throughout the protein. Indeed, missense mutations also cause typical CCD via heterozygous loss of function of the protein without evidence of dominant negative effects (41). Notably, mice that are homozygous deficient for CBFA1 (cbfa1–/–) die at birth due to respiratory failure secondary to absence of a rib cage (44, 45). Furthermore, they have no osteoblasts or osteoclasts. Their skeleton is normally patterned and of usual size but entirely cartilaginous (44, 45). We have no DNA-based evidence for autosomal recessive CCD in our patient.
OC defect
OC is one of the most abundant noncollagenous proteins in bone matrix (46). Nevertheless, mice homozygous for oc gene knockout (–/–) have normal skeletal patterning and high bone mass (47). Despite our patient’s markedly low serum OC level, we found no defect in his OC gene.
OSX deficiency
Because our patient’s disorder seems to be one of osteoblast dysfunction and his phenotype had similarities to both cbfa1 (44, 45) and tnsalp (48) null mice, we investigated OSX, which is the gene that encodes OSX. OSX is an osteoblast transcription factor that regulates differentiation of mesenchymal cells to osteoblasts (49). Osx null (–/–) mice have normal skeletal patterning but profoundly defective mineralization of both membranous and endochondral bone (49). At embryonic d 15.5, there is absence of mineralization in all facial and skull bones. Death occurs within 15 min of birth with difficulty breathing and cyanosis. No bone is formed (49). Clavicles and other skeletal elements derived from endochondral ossification are hypoplastic. Biochemical markers of osteoblast differentiation are low or undetectable, including OC, osteonectin, osteopontin, and bone sialoprotein (49). However, CBFA1 expression is normal. Therefore, osteoblast differentiation is arrested at an early stage but downstream of CBFA1 function. In our patient, osteoblasts were present in slightly decreased numbers but with normal histomorphology. DNA sequencing of his OSX coding regions and splice sites revealed no defect. Although we cannot exclude heterozygous deletion of an OSX allele, heterozygous osx+/– mice are normal and fertile (49), indicating that a heterozygous OSX mutation does not explain our patient’s disease.
Unique disorder
Our patient seems to have had a unique disorder featuring defects in chondrocyte organization in growth plates and osteoblast function, including those attributable to CBFA1 (12, 13, 50). In CCD, bone patterning and regional ossification are altered without complete disruption of osseous tissue. In cbfa1–/– mice (see CCD), the missing bony elements are those formed by intramembranous ossification (calvaria and clavicles), yet very early markers of the osteoblast phenotype (e.g. type I collagen and ALP) are present (44, 45). In our patient, osteoblasts were found at autopsy, but disturbed skeletal formation was also accompanied by low serum levels of both bone ALP and OC. In 1981, three siblings were reported with a previously undescribed perinatal lethal disorder featuring a poorly mineralized calvarium, absence of clavicles and cervical vertebrae, and small scapulae; however, no biochemical studies were reported, and additional features differed from our patient (51).
Although TNSALP gene defects causing HPP can sometimes disrupt intracellular processing of TNSALP and probably additional proteins in osteoblasts (52), reports of two young adults (53) and one infant (54) with convincing diagnoses of HPP mention normal serum levels of OC. Low serum OC levels have been described in only one but less certain case of HPP (55). Our 31 severely affected patients effectively exclude low circulating levels of OC as a feature of HPP.
OC is regarded as the most osteoblast-specific gene product (46). Because CBFA1 regulates all major genes expressed by osteoblasts (50), it acts as a heterodimer with core-binding factor B (50). Nevertheless, no cases of CCD have involved mutations in core-binding factor B, and we did not study this gene (50). Furthermore, no defects have been reported within the promoters of TNSALP, CBFA1, OC, or OSX, and we examined the promoter of CBFA1 in our patient and found it to be unaltered.
Elevated plasma PLP level in patients not taking vitamin B6 is a sensitive and specific marker for HPP (9, 10). When CCD and OI patients manifest hypophosphatasemia, PLP does not accumulate endogenously (Whyte, M. P., unpublished observation). Similarly, our patient’s plasma PLP concentration was normal despite marked hypophosphatasemia. When the TNSALP gene is intact, TNSALP expression in the liver and elsewhere (rather than a global deficiency) may explain normal plasma PLP levels and, ultimately, the absence of rickets. Normal levels of the liver isoform of TNSALP seem able to control TNSALP substrate accumulation in CCD, OI, and other disorders manifesting low serum bone ALP activity. Hypophosphatasemia in CCD and OI could reflect the paucity of skeletal tissue, although impaired biosynthesis or intracellular sequestration of TNSALP are also possible (34, 52). Reports of PEA and PLP accumulation with low serum ALP activity in CCD in 2002 do not address vitamin supplementation (32), and perhaps the higher reference range for PEA in children was not recognized (33).
In summary, our patient’s unique disorder with features of osteochondrodysplasia and metabolic bone disease appears attributable to disruption of some important regulator of osteoblast function that conditions skeletal patterning and ossification.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online November 23, 2004
Abbreviations: ALP, Alkaline phosphatase; CBFA1, core-binding factor A1; CCD, cleidocranial dysplasia; DGGE, denaturing gradient gel electrophoresis; dHPLC, denaturing HPLC; HPP, hypophosphatasia; MIM, Mendelian inheritance in man; OC, osteocalcin; OI, osteogenesis imperfecta; OSX, osterix; PEA, phosphoethanolamine; PLP, pyridoxal 5'-phosphate; TNSALP, tissue nonspecific ALP isoenzyme gene.
Received February 11, 2004.
Accepted November 15, 2004.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A/CBFA1 mutations in Japanese cleidocranial dysplasia patients. Gene 244:21–28[CrossRef][Medline]
Thr mutation associated with lethal hypophosphatasia. J Biochem (Tokyo) 123:968–977
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
