Alice Arabian

Targeted Inactivation of the 25-Hydroxyvitamin D3-1α-Hydroxylase Gene (CYP27B1) Creates an Animal Model of Pseudovitamin D-Deficiency Rickets1

Pseudovitamin D-deficiency rickets is caused by mutations in the cytochrome P450 enzyme, 25-hydroxyvitamin D(3)-1alpha-hydroxylase (1alpha-OHase). Patients with the disease exhibit growth retardation, rickets, and osteomalacia. Serum biochemistry is characterized by hypocalcemia, secondary hyperparathyroidism, and undetectable levels of 1alpha,25-dihydroxyvitamin D(3). We have inactivated the 1alpha-OHase gene in mice after homologous recombination in embryonic stem cells. Serum analysis of homozygous mutant animals confirmed that they were hypocalcemic, hypophosphatemic, hyperparathyroidic, and that they had undetectable 1alpha,25-dihydroxyvitamin D(3). Histological analysis of the bones from 3-week-old mutant animals confirmed the evidence of rickets. At the age of 8 weeks, femurs from 1alpha-OHase-ablated mice present a severe disorganization in the architecture of the growth plate and marked osteomalacia. These results show that we have successfully inactivated the 1alpha-OHase gene in mice and established a valid animal model of pseudovitamin D-deficiency rickets.

Reduced chondrocyte proliferation and chondrodysplasia in mice lacking the integrin-linked kinase in chondrocytes

Chondrocyte proliferation and differentiation requires their attachment to the collagen type II–rich matrix of developing bone. This interaction is mediated by integrins and their cytoplasmic effectors, such as the integrin-linked kinase (ILK). To elucidate the molecular mechanisms whereby integrins control these processes, we have specifically inactivated the ILK gene in growth plate chondrocytes using the Cre-lox methodology. Mice carrying an ILK allele flanked by loxP sites (ILK-fl) were crossed to transgenic mice expressing the Cre recombinase under the control of the collagen type II promoter. Inactivation of both copies of the ILK-fl allele lead to a chondrodysplasia characterized by a disorganized growth plate and to dwarfism. Expression of chondrocyte differentiation markers such as collagen type II, collagen type X, Indian hedgehog and the PTH-PTHrP receptor was normal in ILK-deficient growth plates. In contrast, chondrocyte proliferation, assessed by BrdU or proliferating cell nuclear antigen labeling, was markedly reduced in the mutant growth plates. Cell-based assays showed that integrin-mediated adhesion of primary cultures of chondrocytes from mutant animals to collagen type II was impaired. ILK inactivation in chondrocytes resulted in reduced cyclin D1 expression, and this most likely explains the defect in chondrocyte proliferation observed when ILK is inactivated in growth plate cells.

Chondrocyte-Specific Modulation of Cyp27b1 Expression Supports a Role for Local Synthesis of 1,25-Dihydroxyvitamin D3 in Growth Plate Development

The Cyp27b1 enzyme (25-hydroxyvitamin D-1alpha-hydroxylase) that converts 25-hydroxyvitamin D into the active metabolite, 1,25-dihydroxyvitamin D(3) [1,25(OH)(2)D(3)], is expressed in kidney but also in other cell types such as chondrocytes. This suggests that local production of 1,25(OH)(2)D(3) could play an important role in the differentiation of these cells. To test this hypothesis, we engineered mutant mice that do not express the Cyp27b1 gene in chondrocytes. Inactivation of both alleles of the Cyp27b1 gene led to decreased RANKL expression and reduced osteoclastogenesis, increased width of the hypertrophic zone of the growth plate at embryonic d 15.5, increased bone volume in neonatal long bones, and increased expression of the chondrocytic differentiation markers Indian Hedgehog and PTH/PTHrP receptor. The expression of the angiogenic marker VEGF was decreased, accompanied by decreased platelet/endothelial cell adhesion molecule-1 staining in the neonatal growth plate, suggesting a delay in vascularization. In parallel, we engineered strains of mice overexpressing a Cyp27b1 transgene in chondrocytes by coupling the Cyp27b1 cDNA to the collagen alpha(1)(II) promoter. The transgenic mice showed a mirror image phenotype when compared with the tissue-specific inactivation, i.e. a reduction in the width of the hypertrophic zone of the embryonic growth plate, decreased bone volume in neonatal long bones, and inverse expression patterns of chondrocytic differentiation markers. These results support an intracrine role of 1,25(OH)(2)D(3) in endochondral ossification and chondrocyte development in vivo.

Insights into Vitamin D metabolism using cyp24 over-expression and knockout systems in conjunction with liquid chromatography/mass spectrometry (LC/MS).

The development of novel gene expression systems for cytochrome P450s (CYPs) together with a revolution in analytical mass spectrometry with the emergence of liquid chromatography/mass spectrometry (LC/MS) has opened the door to answering some long-standing questions in Vitamin D metabolism. Our studies focused on: (1) elucidating the role of CYP24 in 25-OH-D3 and 1alpha,25-(OH)2D3 metabolism; (2) exploring how DBP influences this process; (3) measuring 25-OH-D3 metabolism in CYP24-knockout (CYP24-XO) cells and; (4) comparing 1alpha-OH-D2 metabolism in the CYP24-XO mouse in vivo and in vitro. Methodology employed CYP24 over-expression and knockout systems in conjunction with state-of-the-art analytical LC/MS, diode array, and radioisotopic detection methods. We found that CYP24 metabolizes 25-OH-D3 and 1alpha,25-(OH)2D3 at similar rates in vitro, but that for 25-OH-D3 but not 1alpha,25-(OH)2D3, this rate is strongly influenced by the concentration of DBP. Unlike their wild type littermates, the administration of 25-OH-D3 to CYP24-XO mice results in no measurable 24,25-(OH)2D3 production. When neonatal murine keratinocytes are prepared from wild type and CYP24-XO mice there was no measurable production of 24,25-(OH)2D3 or 1alpha,24,25-(OH)2D3 in CYP24-XO mice. Similar experiments using the same wild type and CYP24-XO animals and cells and [3H] 1alpha-OH-D2 resulted in the apparent paradox that the Vitamin D prodrug was 25-hydroxylated in vivo but 24-hydroxylated in vitro.

Osteoclast-specific inactivation of the integrin-linked kinase (ILK) inhibits bone resorption

Bone resorption requires the adhesion of osteoclasts to extracellular matrix (ECM) components, a process mediated by the αvβ3 integrin. Following engagement with the ECM, integrin receptors signal via multiple downstream effectors, including the integrin‐linked kinase (ILK). In order to characterize the physiological role of ILK in bone resorption, we generated mice with an osteoclast‐specific Ilk gene ablation by mating mice with a floxed Ilk allele with TRAP‐Cre transgenic mice. The TRAP‐Cre mice specifically excised floxed alleles in osteoclasts, as revealed by crossing them with the ROSA26R reporter strain. Osteoclast‐specific Ilk mutant mice appeared phenotypically normal, but histomorphometric analysis of the proximal tibia revealed an increase in bone volume and trabecular thickness. Osteoclast‐specific Ilk ablation was associated with an increase in osteoclastogenesis both in vitro and in vivo. However, the mutant osteoclasts displayed a decrease in resorption activity as assessed by reduced pit formation on dentin slices in vitro and decreased serum concentrations of the C‐terminal telopeptide of collagen in vivo. Interestingly, compound heterozygous mice in which one allele of Ilk and one allele of the β3 integrin gene were inactivated (ILK+/−; β  3+/− ) also had increased trabecular thickness, confirming that β3 integrin and Ilk form part of the same genetic cascade. Our results show that ILK is important for the function, but not the differentiation, of osteoclasts. J. Cell. Biochem. 110: 960–967, 2010.

The PTH-Gαs-protein kinase A cascade controls αNAC localization to regulate bone mass.

ABSTRACT The binding of PTH to its receptor induces Gαs-dependent cyclic AMP (cAMP) accumulation to turn on effector kinases, including protein kinase A (PKA). The phenotype of mice with osteoblasts specifically deficient for Gαs is mimicked by a mutation leading to cytoplasmic retention of the transcriptional coregulator αNAC, suggesting that Gαs and αNAC form part of a common genetic pathway. We show that treatment of osteoblasts with PTH(1–34) or the PKA-selective activator N6-benzoyladenosine cAMP (6Bnz-cAMP) leads to translocation of αNAC to the nucleus. αNAC was phosphorylated by PKA at serine 99 in vitro. Phospho-S99-αNAC accumulated in osteoblasts exposed to PTH(1–34) or 6Bnz-cAMP but not in treated cells expressing dominant-negative PKA. Nuclear accumulation was abrogated by an S99A mutation but enhanced by a phosphomimetic residue (S99D). Chromatin immunoprecipitation (ChIP) analysis showed that PTH(1–34) or 6Bnz-cAMP treatment leads to accumulation of αNAC at the Osteocalcin (Ocn) promoter. Altered gene dosages for Gαs and αNAC in compound heterozygous mice result in reduced bone mass, increased numbers of osteocytes, and enhanced expression of Sost. Our results show that αNAC is a substrate of PKA following PTH signaling. This enhances αNAC translocation to the nucleus and leads to its accumulation at target promoters to regulate transcription and affect bone mass.

1α,24(S)(OH)2D2 normalizes bone morphology and serum parathyroid hormone without hypercalcemia in 25-hydroxyvitamin D-1-hydroxylase (CYP27B1)-deficient mice, an animal model of vitamin D deficiency with secondary hyperparathyroidism.

Background: Vitamin D compounds are effective in managing elevated PTH levels in secondary hyperparathyroidism (SHPT) of renal failure. However, undesired increases in serum calcium and phosphorus associated with compounds such as calcitriol [1,25(OH)2D3] has prompted a search for compounds with improved safety profiles. 1α,24(S)(OH)2D2 (1,24(OH)2D2) is a vitamin D2 metabolite with low calcium-mo bilizing activity in vivo. We studied the efficacy of 1,24(OH)2D2 in mice lacking the CYP27B1 enzyme [25-hydroxyvitamin D-1α-hydroxylase (1α-OHase)], a novel vitamin D deficiency model with SHPT. Materials and methods: 1α-OHase-deficient (−/−) mice and normal (+/−) heterozygous littermates received 1,24(OH)2D2 (100, 300, 1000, and 3000 pg/g/day) or 1,25(OH)2D3 (30, 300, and 500 pg/g/day) for 5 weeks via daily sc injection. Control groups received vehicle. Results: Vehicle-treated 1α-OHase-deficient mice were hypocalcemic and had greatly elevated serum PTH. 1,24(OH)2D2 at doses above 300 pg/g/day normalized serum calcium, serum PTH, bone growth plate morphology, and other bone parameters. No hypercalcemia was observed at any dose of 1,24(OH)2D2 in normal or 1α-OHase-deficient animals. In contrast, 1,25(OH)2D2 at only 30 pg/g/day normalized calcemia, serum PTH, and bone parameters, but at higher doses completely suppressed PTH and caused hypercalcemia in both 1α-OHase-deficient and normal mice. Treatment with 500 pg/g/day of 1,25(OH)2D3 also induced osteomalacia in normal animals. Conclusion: 1,25(OH)2D3 was maximally active at 10-fold lower doses than 1,24(OH)2D2, but induced hypercalcemia and osteomalacia at high doses. 1,24(OH)2D2 normalized serum calcium, serum PTH, and bone histomorphometry without hypercalcemia in 1α-OHase-deficient mice with SHPT.

Inactivation of the integrin-linked kinase (ILK) in osteoblasts increases mineralization.

In osteoblasts, Integrin-Linked Kinase (ILK)-dependent phosphorylation of the cJUN transcriptional coactivator, αNAC, induces the nuclear accumulation of the coactivator and potentiates cJUN-dependent transcription. Mutation of the ILK phosphoacceptor site within the αNAC protein leads to cytoplasmic retention of the coactivator and cell-autonomous increases in osteoblastic activity. In order to gain further insight into the ILK-αNAC signaling cascade, we inactivated ILK using RNA knockdown in osteoblastic cells and engineered mice with specific ablation of ILK in osteoblasts. ILK knockdown in MC3T3-E1 osteoblast-like cells reduced phosphorylation of its downstream target glycogen synthase kinase 3β (GSK3β), which led to cytoplasmic retention of αNAC and increased mineralization with augmented expression of the osteoblastic differentiation markers, pro-α1(I) collagen (col1A1), Bone Sialoprotein (Bsp) and Osteocalcin (Ocn). Cultured ILK-deficient primary osteoblasts also showed increased cytoplasmic αNAC levels, and augmented mineralization with higher Runx2, Col1a1 and Bsp expression. Histomorphometric analysis of bones from mutant mice with ILK-deficient osteoblasts (Col1-Cre;Ilk(-/fl)) revealed transient changes, with increased bone volume in newborn animals that was corrected by two weeks of age. Our data suggest that the ILK-αNAC cascade acts to reduce the pace of osteoblast maturation. We propose that in vivo, functional redundancy is able to compensate for the loss of ILK activity, leading to the absence of an obvious phenotype when osteoblast-specific Ilk-deficient mice reach puberty.

Differential Effects of Oral Doxercalciferol (Hectorol®) or Paricalcitol (Zemplar®) in the Cyp27b1-Null Mouse Model of Uremia

Background/Aims: Kidney disease patients experience declining calcitriol levels and develop secondary hyperparathyroidism (SHPT). Animal models of uremia based on 5/6 nephrectomy (NTX) do not consistently reproduce this calcitriol deficiency. We developed an animal model, the NTX Cyp27b1-null mouse, which completely lacks endogenous calcitriol, and examined the suitability of this model for evaluation of treatment with vitamin D analogs in uremia. Methods: NTX was performed at 2 months of age. One week post-NTX, animals were treated for 4 weeks with vehicle; doxercalciferol at 30, 100 or 300 pg/g body weight (b.w.); or paricalcitol at 100, 300 or 1,000 pg/g b.w. by gavage 3 times per week. Results: Serum blood urea nitrogen and creatinine were elevated. Vehicle-treated NTX null mice had hypocalcemia and SHPT. Doxercalciferol at 100 or 300 pg/g b.w. normalized serum calcium and parathyroid hormone (PTH) levels. Paricalcitol at 300 or 1,000 pg/g normalized serum calcium, but PTH levels remained elevated. Osteomalacia was corrected by 100 pg/g b.w. of doxercalciferol or 1,000 pg/g b.w. of paricalcitol. The highest dose of doxercalciferol, but not of paricalcitol, significantly reduced osteitis fibrosa. Conclusion: Our results reveal the differential efficacy of doxercalciferol and paricalcitol in this novel animal model incorporating both calcitriol deficiency and renal insufficiency.

Optimal bone fracture repair requires 24R,25-dihydroxyvitamin D3 and its effector molecule FAM57B2

The biological activity of 24R,25-dihydroxyvitamin D3 [24R,25(OH)2D3] remains controversial, but it has been suggested that it contributes to fracture healing. Cyp24a1–/– mice, synthesizing no 24R,25(OH)2D3, show suboptimal endochondral ossification during fracture repair, with smaller callus and reduced stiffness. These defects were corrected by 24R,25(OH)2D3 treatment, but not by 1,25-dihydroxyvitamin D3. Microarrays with Cyp24a1–/– callus mRNA identified FAM57B2 as a mediator of the 24R,25(OH)2D3 effect. FAM57B2 produced lactosylceramide (LacCer) upon specific binding of 24R,25(OH)2D3. Fam57b inactivation in chondrocytes (Col2-Cre Fam57bfl/fl) phenocopied the callus formation defect of Cyp24a1–/– mice. LacCer or 24R,25(OH)2D3 injections restored callus volume, stiffness, and mineralized cartilage area in Cyp24a1-null mice, but only LacCer rescued Col2-Cre Fam57bfl/fl mice. Gene expression in callus tissue suggested that the 24R,25(OH)2D3/FAM57B2 cascade affects cartilage maturation. We describe a previously unrecognized pathway influencing endochondral ossification during bone repair through LacCer production upon binding of 24R,25(OH)2D3 to FAM57B2. Our results identify potential new approaches to ameliorate fracture healing.