Abstract
Context:
Type VIII osteogenesis imperfecta (OI; OMIM 601915) is a recessive form of lethal or severe OI caused by null mutations in P3H1, which encodes prolyl 3-hydroxylase 1.
Objectives:
Clinical and bone material description of non-lethal type VIII OI.
Design:
Natural history study of type VIII OI.
Setting:
Pediatric academic research centers.
Patients:
Five patients with non-lethal type VIII OI, and one patient with lethal type VIII OI.
Interventions:
None.
Main Outcome Measures:
Clinical examinations included bone mineral density, radiographs, and serum and urinary metabolites. Bone biopsy samples were analyzed for histomorphometry and bone mineral density distribution by quantitative backscattered electron imaging microscopy. Collagen biochemistry was examined by mass spectrometry, and collagen fibrils were examined by transmission electron microscopy.
Results:
Type VIII OI patients have extreme growth deficiency, an L1–L4 areal bone mineral density Z-score of −5 to −6, and normal bone formation markers. Collagen from bone and skin tissue and cultured osteoblasts and fibroblasts have nearly absent 3-hydroxylation (1–4%). Collagen fibrils showed abnormal diameters and irregular borders. Bone histomorphometry revealed decreased cortical width and very thin trabeculae with patches of increased osteoid, although the overall osteoid surface was normal. Quantitative backscattered electron imaging showed increased matrix mineralization of cortical and trabecular bone, typical of other OI types. However, the proportion of bone with low mineralization was increased in type VIII OI bone, compared to type VII OI.
Conclusions:
P3H1 is the unique enzyme responsible for collagen 3-hydroxylation in skin and bone. Bone from non-lethal type VIII OI children is similar to type VII, especially bone matrix hypermineralization, but it has distinctive features including extremely thin trabeculae, focal osteoid accumulation, and an increased proportion of low mineralized bone.
Type VIII OI bone has hypermineralization like type VII, but distinct histology, with thin, sparse trabeculae and focal osteoid. P3H1 is the unique enzyme for α1(I)P986 hydroxylation in skin and bone.
Osteogenesis imperfecta (OI) is a heritable bone dysplasia with hallmark features of bone fragility and deformity, as well as growth deficiency. OI is a collagen-related disorder, with most cases (≈85%) caused by defects in type I collagen itself, whereas the rare forms of OI are caused by defects in genes whose protein products interact with type I collagen (1). Each OI type identified a gene whose role in bone formation was not previously appreciated (1). The first group of genes identified to cause recessive OI encoded the three components of the collagen prolyl 3-hydroxylation complex: CRTAP (cartilage associated protein), the helper protein of the complex; P3H1 (prolyl 3-hydroxylase 1); and CyPB (cyclophilin B), encoded by PPIB (2–5). These proteins form a 1:1:1 complex in the endoplasmic reticulum (ER) that modifies discrete proline residues of type I collagen post-translationally (6–8). Subsequent demonstration that CRTAP and P3H1 were mutually stabilizing in the ER underlined the high similarity of type VIII OI, caused by defects in P3H1, with type VII OI, caused by defects in CRTAP (9).
Mutations in P3H1 (NM_022356) were first reported as the cause of recessive type VIII OI in 2007 (3). There are now 48 distinct mutant P3H1 alleles reported in patients, from West Africa, North America, Europe, India, China, Egypt, and the Middle East, most of which are null mutations. The overall incidence of type VIII OI in North America can be estimated at one in 130 000 births (10). Clinically, type VIII OI is a severe to lethal skeletal dysplasia. Many affected individuals die in the perinatal period from respiratory causes, including essentially all individuals homozygous for the West African founder mutation (c.1080+1G>T). The oldest reported living type VIII OI patients are now in their mid-20s (3, 11–13). However, the range of residual type I collagen 3-hydroxylation overlaps broadly in lethal and non-lethal type VIII OI, suggesting that residual enzyme activity is not the basis for non-lethal outcomes.
Although bone studies have recently been reported from a P3h1-null mouse (14, 15), minimal bone histology has been described from three type VIII OI patients (16–18). Both irregular and sharp demarcations between bone and cartilage at the costochondral junction were reported. All patients have thin hypercellular trabeculae lined by osteoblasts and a fibrotic marrow. Histomorphometry measurements have not been reported, nor is it known whether there is redundancy for P3H function in bone tissue.
We provide here the first insight into the effect of null mutations in P3H1 on patient bone tissue. We report bone histology and histomorphometry, bone mineralization density distribution (BMDD), and measurements of procollagen 3-hydroxylation in bone and skin tissue of non-lethal type VIII OI patients.
Patients and Methods
Patients
Probands 1 and 2 were seen at the National Institutes of Health (NIH) Clinical Center under an Institutional Review Board (IRB)-approved protocol. The P3H1 mutation for proband 1 was previously reported as case 5 by Cabral et al (3) and as proband 10 by Baldridge et al (13). The mutation of proband 2 was reported as case 5 by Chang et al (9), but radiographs and clinical features were not presented. Clinical tests on serum and urine were performed by the NIH Clinical Center laboratory; bone-specific alkaline phosphatase (BSAP) was assayed by Esoterix, Inc. Patient 1 received oral alendronate for 3 years starting at age 9; he received 5 mg daily for 2 years, then 35 mg once weekly for an additional year. His elective iliac crest biopsy was obtained at Shriners Hospital for Children in Chicago at age 9 years, before starting treatment, but blood and urine were obtained at age 17 years, 5 years after cessation of bisphosphonate treatment. Patient 2 received 1 year of pamidronate iv, starting at age 10 years (3 mg/kg/d × 3 days per cycle, every 3 months) at a community hospital in Florida. His elective iliac crest biopsy was obtained before treatment (age 10 years), whereas blood and urine samples were obtained at age 13, 2 years after treatment cessation.
Probands 3 and 4 were assessed at Shriners Hospital for Children in Montreal, under a study approved by the McGill University IRB. Proband 3 has a homozygous deletion of P3H1 exon 9, which has been reported in another case (13). He began pamidronate treatment at 5 weeks of age, at 9 mg/kg/y administered every 3 months. An iliac crest bone sample was obtained at age 4 years. Proband 4 is homozygous for the P3H1 West African founder mutation (3, 10). Pamidronate treatment was begun at 16 months of age, at 9 mg/kg/y administered every 3 months. An iliac bone sample was obtained at age 5 years.
Proband 5 was assessed at the Hospital for Special Surgery (HSS) in New York City; inclusion of her case in this series was approved by the HSS IRB. She is also homozygous for the West African founder mutation. She began bisphosphonate treatment at age 2 weeks, receiving iv pamidronate every 2 months (9 mg/kg/y) for 1 year, then every 3 months (12 mg/kg/y). Her bone sample was obtained as surgical discard during a distal femoral osteotomy at age 6 years. After this, she resumed iv pamidronate at a lower dose that continues currently (3 mg/kg/y, divided every 3 months). Clinical data were extracted by retrospective chart review. Proband 6 was a newborn male with lethal type VIII OI who was delivered at a community hospital in Northern Virginia. His parents elected comfort care only, with the approval of the Hospital Ethics Board, and he died at about 1 month of age. This lethal case is briefly presented for comparison with probands 4 and 5, who have the same mutation.
Cell culture
Human dermal fibroblasts were cultured in DMEM (Life Technologies) supplemented with 10% fetal bovine serum, 2 mm L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37°C and 5% CO2. Conditioned medium was collected as described (3).
Mass spectrometry
Pro986 3-hydroxylation in α1(I) chains in conditioned media and tissue was determined using mass spectrometry, as described previously (3). Electrospray mass spectrometry was performed on collagen tryptic peptides using an LCQ Deca XP iontrap mass spectrometer equipped with inline liquid chromatography (ThermoFinnigan), using a C8 capillary column (300 μm × 150 mm; Grace Vydac 208MS5.315; W. R. Grace & Co.) eluted at 4.5 mL/min.
Transmission electron microscopy of proband fibrils
Dermal punch biopsies were obtained from probands 1 and 2 and age-matched unaffected controls. The samples were fixed in 2.5% glutaraldehyde and then processed as previously described (19). Then, 600- to 800-Å sections were obtained with an AO Reichert Ultracut ultramicrotome mounted on copper grids, stained with lead citrate, and examined in a Zeiss EM10CA transmission electron microscope (JFE Enterprises). The diameters of 200 fibrils from each proband were measured and plotted with Microsoft Excel.
Bone samples
Iliac bone samples were obtained from probands 1–4. Tetracycline double labeling was performed prior to biopsy on proband 2 and single labeling on proband 1. From proband 5, a residual bone sample was obtained from a distal femoral osteotomy during rodding. Femoral tissue was obtained from proband 6, with lethal type VIII OI, at autopsy. All samples were fixed in 70% ethanol and embedded in polymethylmethacrylate.
Bone histomorphometry
Histomorphometry measurements were performed on transiliac bone biopsies from probands 1 and 2, but not on children who were receiving bisphosphonate at the time of iliac bone biopsy because histomorphometric results in children with OI are strongly influenced by bisphosphonates (20) and the aim of the present study was to elucidate disease-related rather than treatment-related skeletal characteristics. For bone histomorphometry analyses, 3 μm-thick sections were cut from the tissue block with a hard microtome (Leica SM2500; Nussloch). Bone histomorphometric analysis was performed according to Dempster et al (21) on the whole area of the bone sections. Digital images of the sections were analyzed using ImageJ software version 1.46r (Wayne Rasband, NIH). Nomenclature follows the recommendation of the American Society for Bone and Mineral Research. Results are compared to previously published reference data from healthy children and children with OI type I, as well as with individuals having OI type VII (22–24).
Quantitative backscattered electron imaging analyses
We used quantitative backscattered electron imaging (qBEI), as described elsewhere, to assess the BMDD in all bone samples (25). In contrast to bone histomorphometry, BMDDs in children with OI are not detectably altered by bisphosphonate treatment (26). Consequently, we present qBEI results from all patients for cancellous and cortical bone. The derived BMDD parameters were the mean calcium concentration (CaMean; weighted mean), the most frequently occurring calcium concentration (CaPeak; the peak position of the BMDD) in the sample, the width of the BMDD distribution (CaWidth; full width at half maximum) reflecting the heterogeneity in matrix mineralization, the fraction of low mineralized bone (CaLow; the percentage of the area below 17.68 wt% Ca, corresponding to the fifth percentile of adult reference BMDD and reflecting the portion of the bone area undergoing primary mineralization), and the fraction of highly mineralized bone matrix (CaHigh; the percentage of bone area that is mineralized above 25.30 wt% Ca, corresponding to the 95th percentile of the adult reference range) (27).
Results obtained in the present study are compared with an established BMDD pediatric reference database (28), as well as to the BMDD from children with OI with classical collagen gene mutations causing OI types I (29) and VII (30). We also present unpublished cortical bone BMDD from the cohort of children with OI type I on whom trabecular BMDD was previously published (29). Due to growth spurt and pubertal changes, the interindividual variability of Young Reference BMDD parameters is higher than in adults (28, 31). Because average bone matrix mineralization is determined by the rate of bone turnover (27), matrix mineralization is lower in growing children who have increased bone turnover than in adults who have completed growth.
Statistics
Statistical analysis was performed using GraphPad Prism 6.0f (GraphPad Software, Inc). Normality of the data was tested by D'Agostino-Pearson omnibus test. Normally distributed data are given as mean (SD); non-normally distributed data are given by median [25th percentile; 75th percentile]. Differences in trabecular and cortical BMDD in OI type VIII children (from whom a transiliac bone biopsy was available, n = 4) vs the Young Reference population (28), children with OI types I and VII (29, 30), were tested using Kruskal-Wallis test followed by Dunn's multiple comparison test. Additionally, we performed a Mann-Whitney test to directly compare CaLow between cases with OI type VIII and OI type VII.
Results
P3H1 mutations and clinical features of type VIII OI
We studied five individuals with type VIII OI and severe non-lethal outcome. Proband 1 is homozygous for a premature stop codon in P3H1 and has attained the third decade of life. Probands 2 and 3 have frameshift mutations in P3H1, causing functionally null alleles. Probands 4 and 5 are homozygous for the West African P3H1 allele, which is generally lethal in infancy. However, they were treated with bisphosphonate since infancy and have attained the second decade of life with a very severe osteochondrodysplasia.
All patients have white sclerae, normal dentition, relative macrocephaly, and triangular facies. Skeletal findings include rhizomelia, radiographic “popcorn” calcifications at epiphyses, and extremely low bone density (dual-energy x-ray absorptiometry Z-score = −5 to −6 before treatment). Bone deformities and long bone fractures show continued deterioration of the skeleton with age. Patients have severe and progressive scoliosis despite bisphosphonate treatment, similar to patients with COL1A1/COL1A2 mutations (32). All patients have extreme growth deficiency; at ages 22, 18, 15, 15, and 12 years, they have the heights of average 3, 1.2, 8, 5, and 1.2-year-old children of the same gender, respectively. Abnormal pulmonary function tests and cardiac valvular regurgitation were detected in some patients. Clinical features are compared in Table 1. For detailed case reports and radiographs, see Supplemental Figures 1–4.
Table 1.
Clinical Features of Type VIII OI
| Proband 1 | Proband 2 | Proband 3 | Proband 4 | Proband 5 | Proband 6 | |
|---|---|---|---|---|---|---|
| P3H1 mutations | c.1656C>A/c.1656C>A | c.1383_1389dupGAACTCC/c.1924_1934delCAGCCTAGTG | c.1346–340_1473+36del/c.1346–340_1473+36del | c.1080+1G>T/c.1080+1G>T | c.1080+1G>T/c.1080+1G>T | c.1080+1G>T/c.1080+1G>T |
| Growth data | ||||||
| Birth | Term, 2.73 kg | Term, 3.01 kg | Term, 2.975 kg | Term, 2.98 kg | 34 wk, 2.38 kg | 34 wk, 1.893 kg |
| Current age | 22 y | 18 y | 15 y | 15 y | 12 y | Died, 3 wk |
| Length (50%)a | 3 y | 14 mo | 8 y | 5 y | 14 mo | |
| Weight (50%)a | 6.5 y | 11 mo | 11 y | 7 y | 6 y | 33 wk |
| Head circumference (50%)a | 9 y | 22 mo | Appropriate for age | Appropriate for age | 3 y | 34 wk |
| Facial | ||||||
| Facial shape | Triangular | Triangular | Triangular | Triangular | Triangular | |
| Scleral hue | White | White | White | White | White | White |
| Dentinogenesis imperfecta | No | No | No | No | No | |
| Skeletal | ||||||
| DXA (L1–L4) Z-score | −6.1 | −5.8 | −0.8 (pBP) | −2.9 (pBP) | −9.7 (preBP)/−5.3 (pBP) | |
| Scoliosis | Yes | Yes | No | Yes | Yes | |
| Vertebrae | Multiple T-L compression | Multiple T-L compression | Multiple T-L compression | Multiple T-L compression | Multiple compression fractures | Intact |
| Rhizomelia | Yes | Yes | Yes | Yes | Yes | Yes |
| Popcorn LE+UE | Diffuse | Diffuse | None | Diffuse | None | |
| Pulmonary (PFTs) | Normal | Mild restrictive | Not done | Not done | Moderate obstructive FVC 139% FEV1 73% |
|
| FVC 110% | Moderate obstructive | |||||
| FEV1 91% | FVC 76% | |||||
| FEV1 60% | ||||||
| Cardiac | Trace mitral and tricuspid regurgitation | Pulmonary pressure, 32 mm Hg | Not done | Not done | ECG/EKG normal | |
| Mild tricuspid regurgitation | ||||||
| Neurology | Mild BI | BI | MRI not done | MRI not done | MRI not done | |
| Prominent ventricles + sulci | Prominent ventricles + sulci | |||||
| Audiology | Normal | Slight hearing loss | Not done | Not done | Normal | |
| Requiring preferential seating | ||||||
| Intellect | Above average student | Average student | Normal | Normal | Delayed, in special education classes |
Abbreviations: BI, basilar invagination; DXA, dual-energy x-ray absorptiometry; ECG/EKG, electrocardiogram; FEV1, Forced expiratory volume in 1 second; FVC, Forced vital capacity; LE+UE, Lower extremities + Upper extremities; MRI, magnetic resonance imaging; preBP, before bisphosphonate therapy; pBP, post bisphosphonate therapy; PFTs, Pulmonary function tests; T-L, thoracic-lumbar.
50% indicates the age at which proband measurements would be in the 50th percentile for a child of the same sex.
Proband bone metabolites
Serum markers of bone metabolism were obtained from probands 1 and 2 (Table 2). Markers of bone formation were within age and gender limits for both probands (33). Serum tartrate-resistant acid phosphatase (TRACP-5b), a measure of osteoclast activity, was elevated in both probands (2 SD above matched controls) (33). However, osteoclast number is not elevated in either proband on histomorphometry (below). Urine metabolites, including NTX crosslinks, were largely within normal limits. Their kidneys were normal in echogenicity; renal size was less than the fifth percentile vs age norms, but at the 75th percentile (proband 1) or 50th percentile (proband 2) vs height norms (34).
Table 2.
Bone Markers in Serum and Urine in Type VIII OI
| Analyte | Normal Range | Proband 1 | Proband 2 | Proband 5 |
|---|---|---|---|---|
| Age, y | 17 | 13 | 11 | |
| Serum chemistries | ||||
| Alkaline phosphatase, U/L | 98–618 | 287 | 277 | 231 |
| BSAP, ng/mL | <20.2 | 51 | 91 | 85.4 |
| Osteocalcin, ng/mL | 9–40 (ref 33) | 8.3 | 18.2 | 75 |
| Acid phosphatase, U/L | 3.1–7.0 | 9.3 | 11.2 | |
| 25-OH vitamin D, ng/mL | 20–100 | 7 | 19 | |
| 1,25(OH)2 vitamin D3, pg/mL | 27–51 | 53 | 47 | |
| Intact PTH, pg/mL | 54 (9–69) | 11.2 (16–87) | ||
| Calcium, mg/dL | 9.0–10.7 | 9.4 | 9.1 | 9.7 |
| Phosphate, mg/dL | 3.1–5.1 | 3.6 | 4.8 | 5.5 |
| Urine chemistries | ||||
| Urine pH | 4.5–8.0 | 6.0 | 7.0 | |
| Calcium, mmol/d | 1.25–7.5 | 1.95 | 2.68 | |
| Phosphate, g/d | 0.40–1.3 | 0.36 | 0.21 | |
| Anion gap (Na + KCl) | + | + | ||
| GFR (Schwartz formula) | 102 | 102 | ||
| Protein, mg/m2/L | <4 | 5.2 | 2 | |
| NTX telopeptide/Cr,a nmol/mmol | 307–1367 | 376 | 625 | 98 |
| Electrophoresis | WNL | WNL | ||
| Immunofixation | No anomalous Ig | No anomalous Ig | ||
| Renal ultrasound | ||||
| Sizeb | 13 y, 8.5–11; 17 y, 9.5–10.5 | R, 7.9 × 4.5; L, 9.7 × 3.5 | R, 5.9 × 2.8; L, 6.0 × 3.4 | |
| Echogenicity | WNL | WNL |
Abbreviations: GFR, glomerular filtration rate; NTX, N-telopeptide of type I collagen; WNL, Within normal limits; R, right; L, left.
Urinary NTX in multiple studies (44) were within the normal range in OI children despite their smaller size.
Normal range for age, but proband renal size is 50th percentile (proband 2) or 75th percentile (proband 1) vs norms for height (34).
Proband serum and urine electrolytes were normal.
Comparison of α1(I)Pro986 3-hydroxylation in proband 1 tissues and cells
We compared collagen from cultured fibroblasts and osteoblasts with collagen present in tissues of proband 1. Minimal α1(I) Pro986 3-hydroxylation was detected in collagen secreted from cultured cells [3–4% of total α1(I) chains] and extracted from skin and bone tissues [1–3% of total α1(I) chains], whereas normal controls have almost complete hydroxylation (95–98%) at this site. These data demonstrate that P3H1 is the responsible isoform in both tissues.
Proband dermal collagen fibrils have abnormal morphology
Proband dermal fibrils differ from age-matched controls in size and morphology. Both probands had more large fibrils than controls, likely representing fused fibrils. Also, proband 2 had more small fibrils, with significantly increased size variability (P < .001). Fibril borders of both probands had little irregularities, rather than a normal sharp and round appearance (Figure 1, left panel). Each proband had occasional fragmented or bizarre fibrils (Figure 1, arrows).
Figure 1.
Dermal fibril diameters of probands 1 and 2 and age-matched normal controls. Left and middle panels, Transmission electron microscopy images of collagen fibrils of probands 1 and 2 and their age-matched controls; right panel, distribution plots of diameters of 200 fibrils measured per sample. Some irregular fibrils exist in both probands (white arrows). Both probands have an increase in the number of large fibrils; proband 2 also has an increase in small fibrils, compared to the respective control. Scale bars, 500 nm.
Bone histology and histomorphometry in type VIII OI
Bone histomorphometry was performed on biopsy samples from probands 1 and 2, obtained before bisphosphonate treatment (Table 3). In both specimens, trabeculae were remarkably thin and isolated (Figure 2, A–D). Trabecular thickness was half that of healthy controls and was markedly smaller than in OI types I and VII (Figure 2, A and D). In proband 2, trabecular number could not be assessed (because one cortex was not available), but the calculated increase in trabecular surface per bone volume indicates decreased trabecular thickness, rather than low trabecular number. Also, the cortices appeared rather thin, trabecularized, and porous (Figure 2F). Average osteoid thickness was somewhat lower in probands vs controls, but osteoid surface and volume were increased, reflecting scattered focal osteoid accumulation (Figure 2, B and E). Bone formation indices were within the normal range (proband 1) or increased (proband 2). The number of osteoclasts on the bone surface was markedly decreased in both patients, but the extent of eroded surface generated by resorbing osteoclasts was within the normal range (proband 1) or increased (proband 2).
Table 3.
Results of Bone Histomorphometry
| Controls | Proband 1 | Proband 2 | OI Type Ia | OI Type VIIb | |
|---|---|---|---|---|---|
| n | 10 | 32 | 4 | ||
| Age at biopsy, y | 7.0–10.9 | 9 | 10 | 7.9 ± 2.9 | 3.6 ± 0.7 |
| Structural parameters | |||||
| Cortical width, mm | 0.97 ± 0.37 | 0.25 | 0.42c | 0.52 ± 0.20 | 0.38 ± 0.13 |
| Bone volume per tissue volume, % | 22.4 ± 4.2 | 8.14 | NA | 11.0 ± 5.2 | 12.3 ± 1.5 |
| Trabecular thickness, μm | 129 ± 17 | 58.2 | 65.08 | 105 ± 25 | 101 ± 15 |
| Trabecular number, 1/mm | 1.73 ± 0.17 | 1.39 | NA | 1.03 ± 0.39 | 1.24 ± 0.17 |
| Bone surface/bone volume, mm2/mm3 | 16.8 ± 2.5 | 34.36 | 30.73 | 19.8 ± 4.7 | 20.2 ± 2.7 |
| Bone formation parameters | |||||
| Osteoid thickness, μm | 5.9 ± 1 | 5.3 | 4.6 | 5.5 ± 1.7 | 5.3 ± 0.8 |
| Osteoid volume per bone volume, % | 2.64 ± 1.04 | 7.7 | 6.3 | 5.2 ± 2.6 | 5.5 ± 1.0 |
| Osteoid surface per bone surface, % | 29.1 ± 12.9 | 54.0 | 51.9 | 48 ± 14 | 53 ± 4 |
| Osteoblast surface per bone surface, % | 8.2 ± 4.4 | 8.9 | 22.8 | 19.4 ± 9.5 | 24.1 ± 2.3 |
| Osteoblast surface per osteoid surface, % | 28.8 ± 15.1 | 11.7 | 44.1 | 39 ± 14 | 45 ± 3 |
| Mineral apposition rate, μm/d | 0.95 ± 0.07 | 0.7 | 0.73 ± 0.18 | 0.71 ± 0.12 | |
| Mineralizing surface per bone surface, % | 14.9 ± 4.5 | 28.8 | 23.1 ± 9.7 | 26 ± 5 | |
| Bone formation rate per bone surface, mm3/mm2/y | 51.8 ± 6.1 | 77.3 | 77 ± 34 | 135 ± 45 | |
| Adjusted apposition rate, μm/d | 0.47 ± 0.18 | 0.4 | 0.35 ± 0.14 | 0.37 ± 0.12 | |
| Mineralization lag time, d | 14.1 ± 4.3 | 11.2 | 16.5 [12.5–19.8] | 15 ± 4 | |
| Bone resorption parameters | |||||
| No. of osteoclasts per bone surface/mm | 0.36 ± 0.16 | 0.1 | 0.1 | 0.47 ± 0.29 | Not defined |
| Osteoclast surface per bone surface, % | 1.29 ± 0.62 | 0.4 | 0.4 | 1.37 [1.05–1.70] | 2.3 ± 0.6 |
| Eroded surface per bone surface, % | 17. 0 ± 6.0 | 12.9 | 21.9 | 15.6 [13.7–21.8] | 22.6 ± 6 |
Abbreviation: NA, not available. Control data are expressed as mean ± SD for healthy children from Glorieux et al (22).
Reference data from 19 children with OI type I (29) selected from a larger cohort published by Rauch et al (23).
OI type VII data from Ward et al (24).
Cortical width is calculated as mean width from two cortices. In the bone biopsy sample from proband 2, only one cortex was available.
Figure 2.
Light microscopic analysis of transiliac bone sections and BMDD measurements. Goldner's trichrome staining represents mineralized bone matrix in green and osteoid in purple. A–C, Proband 1; D–F, proband 2. In both samples, thin isolated trabecular features are seen (overviews, A and D). B and E, At higher magnification, both probands have generally thin osteoid, covering the mineralized (green) trabecular surface. C, Proband 1 also shows focal areas with thicker osteoid formation. F, The cortex from proband 2 appears trabecularized and consequently very porous. E, Note the high number of osteoblasts covering the trabecular surface in the bone sample from proband 2, in line with the extended area of tetracycline double-labeling seen the fluorescence image (inset in panel D). G and H, Backscattered electron images from the entire transiliac bone biopsy sample from proband 1 (G, no bisphosphonate treatment) and proband 3 (H, bisphosphonate treated). Note the increase in trabecular bone and the much thicker cortex in proband 3 in comparison to proband 1. I, The BMDD curves of cancellous bone from both probands are shifted toward higher mineral content of the bone matrix compared to the Young Reference cohort (28), irrespective of bisphosphonate treatment.
BMDD shows high matrix mineralization of type VIII OI bone
The qBEI results from all bone samples are given in Table 4. In both trabecular and cortical bone, CaMean and CaPeak were elevated in each proband when compared to normal controls but were similar to OI types I and VII. Bone samples from probands 5 and 6 are femoral rather than transiliac bone. The proband 5 sample was from a femoral osteotomy. Proband 6 bone was obtained at autopsy and is instructive for comparison to non-lethal samples. As expected for infant bone, it is largely primary woven bone, known to be more highly mineralized than remodeled or lamellar bone tissue (35). Thus, in comparison with our non-lethal probands, this sample showed the highest values for CaPeak and CaHigh.
Table 4.
qBEI Results
| BMDD | Controls* | Proband 1 | Proband 2 | Proband 3 | Proband 4 | Proband 5 | Proband 6 | OI Type I** | OI Type VII*** |
|---|---|---|---|---|---|---|---|---|---|
| Cancellous bone | n = 54 | n = 19 | n = 4 | ||||||
| CaMean [wt% Ca] | 20.95 (0.57) | 21.79 | 21.02 | 23.02 | 21.72 | 22.70 | 22.43 | 22.43 (0.63) | 21.65 [21.48; 21.75] |
| CaPeak [wt% Ca] | 21.66 (0.52) | 23.74 | 23.40 | 24.09 | 23.05 | 24.24 | 25.13 | 23.39 (0.57) | 22.43 [22.30; 22.49] |
| CaWidth [Δwt% Ca] | 3.47 [3.12; 3.64] | 2.95 | 4.33 | 2.78 | 3.64 | 3.15 | 4.33 | 3.08 (0.28) | 3.21 [3.04; 3.29] |
| CaLow [% bone area] | 6.14 [4.90; 7.99] | 11.89 | 15.09 | 8.16 | 9.66 | 10.19 | 12.79 | 5.94 (2.05) | 6.04 [5.72; 6.46] |
| CaHigh [% bone area] | 0.89 [0.43; 1.47] | 10.45 | 12.42 | 21.33 | 9.21 | 20.04 | 33.95 | 7.54 [5.00; 11.82] | 2.52 [1.96; 3.20] |
| Cortical bone | n = 53 | n = 19 | n = 4 | ||||||
| CaMean [wt% Ca] | 20.45 [19.68; 21.04] | 21.90 | 22.34 | 22.32 | 22.46 | NA | NA | 22.51 (0.46) | 21.72 [21.71; 21.87] |
| CaPeak [wt% Ca] | 21.14 [20.62; 21.75] | 23.57 | 23.74 | 23.22 | 23.57 | NA | NA | 23.29 (0.48) | 22.75 [22.40; 23.11] |
| CaWidth [Δwt% Ca] | 3.81 [3.38; 4.38] | 3.29 | 3.47 | 3.29 | 3.30 | NA | NA | 3.28 (0.25) | 4.10 [3.64; 4.57] |
| CaLow [% bone area] | 9.06 [6.22; 15.00] | 9.95 | 8.20 | 6.72 | 6.78 | NA | NA | 4.60 (0.80) | 6.01 [5.41; 7.87] |
| CaHigh [% bone area] | 0.46 [0.28; 1.22] | 10.84 | 13.62 | 10.32 | 12.13 | NA | NA | 8.60 (4.00) | 8.11 [4.83; 11.14] |
Figure 2, G—I, shows the backscattered images from the iliac bone samples of proband 1 (untreated, Figure 2G) and proband 3 (pamidronate treated, Figure 2H). Proband 3 has much thicker trabeculae and cortices than proband 1, consistent with histology of treated OI children (20, 36). Nevertheless, the degree of mineralization of the bone matrix was similar in both cases (Figure 2I) (26). We therefore grouped the four transiliac bone samples from probands 1–4 and compared their BMDD outcomes with reference values, as well as with OI types I and VII.
In cancellous bone, CaPeak and CaHigh were significantly increased in OI type VIII cases (+8.8% and more than 12-fold higher, respectively; both P < .001) compared to healthy controls, but not significantly different from OI types I and VII. Interestingly, CaLow was significantly increased in OI type VIII vs both normal and OI type I (+75.6%, P < .05; +90.1%, P < .01, respectively). Direct comparison of CaLow in type VIII and type VII OI (Mann-Whitney test) also showed increased CaLow in type VIII OI (+78%; P = .03). The increase in CaLow may reflect the foci of increased osteoid formation observed in type VIII bone.
Analysis of cortical bone revealed significantly increased CaMean [+9.2%; P < .01], CaPeak (+11.5%; P < .001), and CaHigh (25-fold increase; P < .001) in OI type VIII vs healthy controls. Additionally, CaLow from all four type VIII OI samples was increased vs OI type I (+63.3%; P < .05) but not vs controls or OI type VII. However, it is noteworthy that CaLow is increased by about one-third in cortical and cancellous bone of probands 1 and 2, who were untreated with bisphosphonate, vs treated probands 3 and 4.
Taken together, the qBEI results indicate that the bone matrix from our type VIII OI probands has significantly higher mineralization than healthy controls, but not higher than children with OI types I and VII. In addition, the CaLow of untreated type VIII OI bone is increased in both cortical and cancellous compartments, in comparison to types VII and I OI.
Discussion
This report focuses on the clinical features and bone material properties of five patients with non-lethal and one with lethal type VIII OI, caused by null mutations in the P3H1 gene. Absence of P3H1 causes a lethal or severe osteochondrodystrophy, first described in 2007 (3). Probands 1–3 have null P3H1 alleles and currently range in age from 15–22 years. Probands 4 and 5 are homozygous for the West African P3H1 allele; these patients generally die in the first several months of life from respiratory failure (3, 13). It is not clear to what extent the survival of probands 4 and 5 to their second decade, albeit with extremely severe bone dysplasia with respiratory and neurological deficiency, can be attributed to early pamidronate treatment, which may have impacted thorax stiffness sufficiently to enable critical respiratory function. Although treatment of proband 5 was initiated in the newborn nursery, proband 4 was first referred to specialty care and treatment initiation at age 16 months. Probands 1, 2, 4, and 5 show progressive skeletal deterioration with age, despite bisphosphonate treatment (Supplemental Figures 1–4). No renal or urinary abnormalities were detected, despite the occurrence of P3H1 in renal matrix as the proteoglycan leprecan (37). Serum bone formation markers, including osteocalcin, alkaline phosphatase, and BSAP, were also not remarkable in probands 1 and 2.
We investigated the redundancy of P3H1 function by examining skin and bone tissues and cultured fibroblasts and osteoblasts from proband 1. Type I collagen prolyl 3-hydroxylation was essentially absent [1–4%, vs normal 96–98% α1(I) 3-hydroxylation] in skin and bone tissues, fibroblasts, and osteoblasts, leading to collagen fibrils with irregular cross-sectional borders and to a subpopulation of fused or fragmented fibrils. These data establish the unique role of P3H1 for collagen hydroxylation in bone, with no redundancy evident from other prolyl hydroxylases. Total absence of collagen 3-hydroxylation was previously reported in tail tendon, skin, and bone of P3h1 knockout (KO) mice and in bone, skin, kidney, and cartilage of Crtap KO mice (3, 5, 15, 38). The 55% 3-hydroxylation of α1(I) Pro986 reported from osteoblasts with the West African P3H1 allele (13) presumably represents residual P3H1 activity in bone because this allele generates multiple alternatively spliced transcripts (3), rather than compensation by P3H2 or P3H3.
Because P3H1 and CRTAP form a mutually stabilizing complex in the ER, it would be expected a priori that the histology of type VII and VIII OI patients and the KO murine models for P3H1 and CRTAP would be similar to each other but might differ from OI with a collagen structural abnormality. However, CRTAP is in part a secreted molecule, whereas P3H1 is retained in the ER by a KDEL sequence, so these proteins do not have identical biochemistry (6, 9). Histomorphometry of untreated children with type VIII OI revealed that reduced bone volume was predominantly due to very thin trabeculae (about half the thickness of type I and VII OI), as well as to the typical OI reduction of trabecular number. Trabecular thickness is reduced in femora of P3h1 KO mice (14) but not as severely as in patients, consistent with a less severe phenotype in P3h1-null rodents than children. Type VIII OI bone also showed scattered focal osteoid accumulation, despite a normal average osteoid thickness. This has not been reported in type VII OI, although only histology from a hypomorphic mutation is available (24). The Crtap (5) and P3h1 (14) KO mice have reduced osteoid thickness, whereas children with type I, VII, and VIII OI have normal values. Focal osteoid accumulation is apparently a primary finding of type VIII OI, rather than osteomalacia caused by the patients' vitamin D deficiency because mineral lag time was not increased. Next, kinetic parameters of bone formation in type VIII OI were similar to the elevated values found in both type I and VII OI, again delineating these findings as typical for OI. The porous and trabecularized cortex of proband 2 might also suggest active cortical bone (re)modeling, supported by elevated serum TRACP-5b and increased eroded surface, although osteoclast number and surface were not elevated, and urinary crosslinks were normal at the time of biopsy. The elevation of TRACP-5b in type VIII serum may reflect the activity of large multinucleated osteoclasts (39) and/or dysfunctional osteoclasts at the chondro-osseous junction formed during bisphosphonate treatment, as previously reported (40, 41). Interestingly, neither the Crtap (5) nor P3h1 (14) KO mice have high bone turnover. Instead, they have reduced bone formation and normal to low osteoblast and osteoclast indices, suggestive of an osteoblast defect.
Bone matrix mineralization is more consistent among OI types and murine models. BMDD in type I, III, IV, and VII OI has revealed high matrix mineralization (29, 30). Both cortical and cancellous bone from type VIII OI are hypermineralized as well, making hypermineralization an essentially defining feature of OI. In cancellous and cortical bone, CaPeak and CaHigh were significantly higher than normal and were similar in value to type I and VII OI. CaLow was increased with respect to OI type I, reflecting the foci of unmineralized osteoid and the higher osteoid surface in the untreated type VIII OI probands. In contrast, bone from bisphosphonate-treated patients had reduced CaLow, mirroring the decrease in bone formation (27). Interestingly, CaLow was not significantly increased in either P3h1 (14) or Crtap KO mice (30), consistent with their low osteoid indices. Besides these variations in CaLow, the qBEI analyses revealed that BMDD in OI type VIII was not altered by bisphosphonate treatment, as previously noted in classical OI (26) and the oim mouse (42). This might be due to the fact that in OI bone, the altered mineralization kinetics result in more dense packing of mineral particles and an inherent saturation of the matrix (43). Hence, bisphosphonate treatment does not further increase mineral content of the matrix but instead leads to an increase in bone volume (20, 26).
Taken together, bone from patients with non-lethal type VIII OI has properties generally similar to type VII OI bone. Distinctive features of type VIII OI bone histology include extremely thin and sparse trabeculae and focal accumulation of unmineralized osteoid. Type VIII OI bone shares the hypermineralization of classical and recessive OI, although CaLow, reflecting bone undergoing primary mineralization, is increased in type VIII OI bone, in comparison to type VII and I OI.
Acknowledgments
The authors thank Daniela Gabriel, Petra Keplinger, Sonja Lueger, and Phaedra Messmer for careful sample preparations and qBEI measurements.
This study was supported by the Allgemeine Unfallversicherungsanstalt (research funds of the Austrian workers compensation board) and the Wiener Gebietskrankenkasse (Viennese sickness insurance funds) (to N.F.-Z., P.R., K.K.), National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR037318 (to D.R.E.), the Shriners of North America (to P.A.S., F.R., and F.H.G.), and National Institute of Child Health and Human Development intramural funds (to J.C.M.). F.R. received support from the Chercheur-Boursier Clinicien program of the Fonds de Recherche du Québec-Santé.
Authors' roles included: data collection—N.F.-Z., A.M.B., M.W., E.C., T.E.H., G.P., W.C., P.A.S., F.R., and J.C.M.; data analysis—N.F.-Z., A.M.B., M.W., W.C., P.R., K.K., D.R.E., F.R., J.C.M.; data interpretation—N.F.-Z., A.M.B., M.W., E.C., T.E.H., G.P., W.C., P.R., K.K., D.R.E., C.R., F.R., J.C.M.; drafting the manuscript—J.C.M. and N.F.-Z.; and revising the manuscript content—N.F.-Z., A.M.B., M.W., E.C., D.R.E., C.R., F.R., and J.C.M. All authors approved the final version of the manuscript. J.C.M. and N.F.-Z. take responsibility for the integrity of the data analysis.
Current address for W.C.: OpGen, Inc., Gaithersburg, Maryland 20878.
Disclosure Summary: The authors have nothing to disclose.
Footnotes
- BMDD
- bone mineralization density distribution
- BSAP
- bone-specific alkaline phosphatase
- CRTAP
- cartilage associated protein
- ER
- endoplasmic reticulum
- KO
- knockout
- OI
- osteogenesis imperfecta
- P3H1
- prolyl 3-hydroxylase 1
- qBEI
- quantitative backscattered electron imaging
- TRACP-5b
- serum tartrate-resistant acid phosphatase.
References
- 1. Forlino A, Marini JC. Osteogenesis imperfecta. Lancet. 2016;387(10028):1657–1671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Barnes AM, Chang W, Morello R, et al. Deficiency of cartilage-associated protein in recessive lethal osteogenesis imperfecta. N Engl J Med. 2006;355:2757–2764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Cabral WA, Chang W, Barnes AM, et al. Prolyl 3-hydroxylase 1 deficiency causes a recessive metabolic bone disorder resembling lethal/severe osteogenesis imperfecta. Nat Genet. 2007;39:359–365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Barnes AM, Carter EM, Cabral WA, et al. Lack of cyclophilin B in osteogenesis imperfecta with normal collagen folding. N Engl J Med. 2010;362:521–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Morello R, Bertin TK, Chen Y, et al. CRTAP is required for prolyl 3- hydroxylation and mutations cause recessive osteogenesis imperfecta. Cell. 2006;127:291–304. [DOI] [PubMed] [Google Scholar]
- 6. Vranka JA, Sakai LY, Bächinger HP. Prolyl 3-hydroxylase 1, enzyme characterization and identification of a novel family of enzymes. J Biol Chem. 2004;279:23615–23621. [DOI] [PubMed] [Google Scholar]
- 7. Marini JC, Cabral WA, Barnes AM, Chang W. Components of the collagen prolyl 3-hydroxylation complex are crucial for normal bone development. Cell Cycle. 2007;6:1675–1681. [DOI] [PubMed] [Google Scholar]
- 8. Hudson DM, Eyre DR. Collagen prolyl 3-hydroxylation: a major role for a minor post-translational modification? Connect Tissue Res. 2013;54:245–251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Chang W, Barnes AM, Cabral WA, Bodurtha JN, Marini JC. Prolyl 3-hydroxylase 1 and CRTAP are mutually stabilizing in the endoplasmic reticulum collagen prolyl 3-hydroxylation complex. Hum Mol Genet. 2010;19:223–234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Cabral WA, Barnes AM, Adeyemo A, et al. A founder mutation in LEPRE1 carried by 1.5% of West Africans and 0.4% of African Americans causes lethal recessive osteogenesis imperfecta. Genet Med. 2012;14:543–551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Willaert A, Malfait F, Symoens S, et al. Recessive osteogenesis imperfecta caused by LEPRE1 mutations: clinical documentation and identification of the splice form responsible for prolyl 3-hydroxylation. J Med Genet. 2009;46:233–241. [DOI] [PubMed] [Google Scholar]
- 12. Pepin MG, Schwarze U, Singh V, Romana M, Jones-Lecointe A, Byers PH. Allelic background of LEPRE1 mutations that cause recessive forms of osteogenesis imperfecta in different populations. Mol Genet Genomic Med. 2013;1:194–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Baldridge D, Schwarze U, Morello R, et al. CRTAP and LEPRE1 mutations in recessive osteogenesis imperfecta. Hum Mutat. 2008;29:1435–1442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Fratzl-Zelman N, Bächinger HP, Vranka JA, Roschger P, Klaushofer K, Rauch F. Bone matrix hypermineralization in prolyl-3 hydroxylase 1 deficient mice. Bone. 2016;85:15–22. [DOI] [PubMed] [Google Scholar]
- 15. Vranka JA, Pokidysheva E, Hayashi L, et al. Prolyl 3-hydroxylase 1 null mice display abnormalities in fibrillar collagen-rich tissues such as tendons, skin, and bones. J Biol Chem. 2010;285:17253–17262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Moul A, Alladin A, Navarrete C, Abdenour G, Rodriguez MM. Osteogenesis imperfecta due to compound heterozygosity for the LEPRE1 gene. Fetal Pediatr Pathol. 2013;32:319–325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Takagi M, Ishii T, Barnes AM, et al. A novel mutation in LEPRE1 that eliminates only the KDEL ER-retrieval sequence causes non-lethal osteogenesis imperfecta. PLoS One. 2012;7:e36809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. van Dijk FS, Nikkels PG, den Hollander NS, et al. Lethal/severe osteogenesis imperfecta in a large family: a novel homozygous LEPRE1 mutation and bone histological findings. Pediatr Dev Pathol. 2011;14:228–234. [DOI] [PubMed] [Google Scholar]
- 19. Cabral WA, Makareeva E, Colige A, et al. Mutations near amino end of α1(I) collagen cause combined osteogenesis imperfecta/Ehlers-Danlos syndrome by interference with N-propeptide processing. J Biol Chem. 2005;280:19259–19269. [DOI] [PubMed] [Google Scholar]
- 20. Rauch F, Travers R, Plotkin H, Glorieux FH. The effects of intravenous pamidronate on the bone tissue of children and adolescents with osteogenesis imperfecta. J Clin Invest. 2002;110:1293–1299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Dempster DW, Compston JE, Drezner MK, et al. Standardized nomenclature, symbols, and units for bone histomorphometry: a 2012 update of the report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res. 2013;28:2–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Glorieux FH, Travers R, Taylor A, et al. Normative data for iliac bone histomorphometry in growing children. Bone. 2000;26:103–109. [DOI] [PubMed] [Google Scholar]
- 23. Rauch F, Travers R, Parfitt AM, Glorieux FH. Static and dynamic bone histomorphometry in children with osteogenesis imperfecta. Bone. 2000;26:581–589. [DOI] [PubMed] [Google Scholar]
- 24. Ward LM, Rauch F, Travers R, et al. Osteogenesis imperfecta type VII: an autosomal recessive form of brittle bone disease. Bone. 2002;31:12–18. [DOI] [PubMed] [Google Scholar]
- 25. Roschger P, Fratzl P, Eschberger J, Klaushofer K. Validation of quantitative backscattered electron imaging for the measurement of mineral density distribution in human bone biopsies. Bone. 1998;23:319–326. [DOI] [PubMed] [Google Scholar]
- 26. Weber M, Roschger P, Fratzl-Zelman N, et al. Pamidronate does not adversely affect bone intrinsic material properties in children with osteogenesis imperfecta. Bone. 2006;39:616–622. [DOI] [PubMed] [Google Scholar]
- 27. Roschger P, Paschalis EP, Fratzl P, Klaushofer K. Bone mineralization density distribution in health and disease. Bone. 2008;42:456–466. [DOI] [PubMed] [Google Scholar]
- 28. Fratzl-Zelman N, Roschger P, Misof BM, et al. Normative data on mineralization density distribution in iliac bone biopsies of children, adolescents and young adults. Bone. 2009;44:1043–1048. [DOI] [PubMed] [Google Scholar]
- 29. Roschger P, Fratzl-Zelman N, Misof BM, Glorieux FH, Klaushofer K, Rauch F. Evidence that abnormal high bone mineralization in growing children with osteogenesis imperfecta is not associated with specific collagen mutations. Calcif Tissue Int. 2008;82:263–270. [DOI] [PubMed] [Google Scholar]
- 30. Fratzl-Zelman N, Morello R, Lee B, et al. CRTAP deficiency leads to abnormally high bone matrix mineralization in a murine model and in children with osteogenesis imperfecta type VII. Bone. 2010;46:820–826. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Parfitt AM, Travers R, Rauch F, Glorieux FH. Structural and cellular changes during bone growth in healthy children. Bone. 2000;27:487–494. [DOI] [PubMed] [Google Scholar]
- 32. Sato A, Ouellet J, Muneta T, Glorieux FH, Rauch F. Scoliosis in osteogenesis imperfecta caused by COL1A1/COL1A2 mutations - genotype-phenotype correlations and effect of bisphosphonate treatment. Bone. 2016;86:53–57. [DOI] [PubMed] [Google Scholar]
- 33. Rauchenzauner M, Schmid A, Heinz-Erian P, et al. Sex- and age-specific reference curves for serum markers of bone turnover in healthy children from 2 months to 18 years. J Clin Endocrinol Metab. 2007;92:443–449. [DOI] [PubMed] [Google Scholar]
- 34. Han BK, Babcock DS. Sonographic measurements and appearance of normal kidneys in children. AJR Am J Roentgenol. 1985;145:611–616. [DOI] [PubMed] [Google Scholar]
- 35. Jones SJ, Glorieux FH, Travers R, Boyde A. The microscopic structure of bone in normal children and patients with osteogenesis imperfecta: a survey using backscattered electron imaging. Calcif Tissue Int. 1999;64:8–17. [DOI] [PubMed] [Google Scholar]
- 36. Munns CF, Rauch F, Travers R, Glorieux FH. Effects of intravenous pamidronate treatment in infants with osteogenesis imperfecta: clinical and histomorphometric outcome. J Bone Miner Res. 2005;20:1235–1243. [DOI] [PubMed] [Google Scholar]
- 37. Lauer M, Scruggs B, Chen S, Wassenhove-McCarthy D, McCarthy KJ. Leprecan distribution in the developing and adult kidney. Kidney Int. 2007;72:82–91. [DOI] [PubMed] [Google Scholar]
- 38. Baldridge D, Lennington J, Weis M, et al. Generalized connective tissue disease in Crtap−/− mouse. PLoS One. 2010;5:e10560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Cheung MS, Glorieux FH, Rauch F. Large osteoclasts in pediatric osteogenesis imperfecta patients receiving intravenous pamidronate. J Bone Miner Res. 2009;24:669–674. [DOI] [PubMed] [Google Scholar]
- 40. Brizola E, Shapiro JR. Bisphosphonate treatment of children and adults with osteogenesis imperfecta: unanswered questions. Calcif Tissue Int. 2015;97:101–103. [DOI] [PubMed] [Google Scholar]
- 41. Halasy-Nagy JM, Rodan GA, Reszka AA. Inhibition of bone resorption by alendronate and risedronate does not require osteoclast apoptosis. Bone. 2001;29:553–559. [DOI] [PubMed] [Google Scholar]
- 42. Misof BM, Roschger P, Baldini T, et al. Differential effects of alendronate treatment on bone from growing osteogenesis imperfecta and wild-type mouse. Bone. 2005;36:150–158. [DOI] [PubMed] [Google Scholar]
- 43. Fratzl-Zelman N, Schmidt I, Roschger P, et al. Mineral particle size in children with osteogenesis imperfecta type I is not increased independently of specific collagen mutations. Bone. 2014;60:122–128. [DOI] [PubMed] [Google Scholar]
- 44. Dwan K, Phillipi CA, Steiner RD, Basel D. Bisphosphonate therapy for osteogenesis imperfecta. Cochrane Database Syst Rev. 2014;7:CD005088. [DOI] [PubMed] [Google Scholar]