Conditional Inactivation of Noggin in the Postnatal Skeleton Causes Osteopenia

Abstract

Noggin is an antagonist of bone morphogenetic proteins (BMP), and its overexpression causes suppressed osteoblastogenesis and osteopenia. Global inactivation of Noggin results in severe developmental defects and prenatal lethality, but the consequences of the conditional inactivation of Noggin on the postnatal skeleton are not known. To study the function of noggin in osteoblasts, we generated tissue-specific null Noggin mice by mating Noggin conditional mice, where the Noggin allele is flanked by loxP sequences, with mice expressing the Cre recombinase under the control of the osteocalcin promoter (Oc-Cre). Noggin conditional null mice exhibited decreased weight, shortened femoral length, and generalized osteopenia. Bone histomorphometric and microarchitectural analyses of distal femurs revealed decreased bone volume due to a reduced number of trabeculae in 1- and 3-month-old Noggin conditional null mice. Vertebral microarchitecture confirmed the osteopenia observed in Noggin conditional null mice. Osteoclast number was increased in 1-month-old male Noggin conditional null mice, and bone formation was increased in 3-month-old mice, but female mice did not exhibit increased bone remodeling. In conclusion, Noggin inactivation causes osteopenia, suggesting that BMP in excess have a detrimental effect on bone or that noggin has a BMP-independent role in skeletal homeostasis.

Bone morphogenetic proteins (BMP) play a role in the regulation of endochondral bone formation, and when implanted sc, they induce ectopic bone formation (1, 2). BMP can induce the differentiation of mesenchymal cells toward the osteoblastic lineage, but the role of BMP in osteoblastogenesis has been challenged after the demonstration that BMP increase the expression of the Wnt antagonists sclerostin and dickkopf1 (DKK1), and as a consequence, they inhibit Wnt signaling (3–5). BMP initiate a signal transduction cascade activating the mothers against decapentaplegic (Smad) or the MAPK pathways (6, 7). In osteoblastic cells, BMP use primarily the mothers against decapentaplegic (Smad) signaling pathway (8, 9). The activity of BMP is controlled by a large group of secreted polypeptides that prevent BMP signaling by binding BMP and by precluding ligand-receptor interactions (1, 2, 10).

Noggin, a member of the Spemann organizer, is one of the BMP antagonists that has been studied in greater detail, because its function is restricted to the specific binding of BMP and as a consequence to decrease BMP activity. Noggin is a secreted glycoprotein and as a dimer has a molecular mass of 64 kDa. Noggin binds with various degrees of affinity BMP-2, -4, -5, -6 and -7, and growth and differentiation factors 5 and 6, but not other members of the TGFβ family of peptides (11–14). Consequently, noggin is considered a specific BMP antagonist. Noggin transcripts are prominently expressed in the central nervous system and are also expressed in skeletal muscle, lung, skin, cartilage, and bone (13). Osteoblasts express noggin, which decreases the effects of BMP in cells of the osteoblastic lineage (15, 16). BMP induce noggin in osteoblasts, and this effect appears to be a local feedback mechanism to temper BMP actions (16).

Transgenic overexpression of noggin in the bone environment causes osteopenia secondary to impaired bone formation (17). Noggin overexpression in cells of the osteoblastic lineage suppresses their differentiation, whereas noggin down-regulation in these cells enhances the expression of osteogenic differentiation markers (18, 19). In vivo, noggin down-regulation enhances the regeneration of bone defects (19). Homozygous null mutations of Noggin in mice result in serious developmental skeletal abnormalities and embryonic lethality not allowing the study of the adult Noggin null skeletal phenotype (20, 21). The dual conditional inactivation of Noggin and Gremlin1 during somite patterning has confirmed a role of these two BMP antagonists in axial skeletal development (22). However, the consequences of Noggin inactivation in the adult skeleton have not been reported. The intent of the present study was to define the function of noggin in the postnatal skeleton in vivo. We determined the general characteristics, the body composition, and the histomorphometric and skeletal microarchitectural properties of conditional Noggin mutant mice from 1 month to 6 months of age. Noggin was inactivated in osteoblasts after the recombination of loxP sequences flanking the Noggin allele. For this purpose, mice where the Noggin allele was flanked by loxP sequences were crossed with transgenics expressing the Cre recombinase under the control of the osteocalcin promoter.

Materials and Methods

Noggin conditional null mice

Mice where the Noggin allele was flanked by loxP sequences were provided by R. Harland (Berkeley, CA) and studied in a 129Sv/C57BL/6 genetic background after the deletion of the neomycin selection cassette (22). The loxP sites were located 252 bp upstream of ATG and 1959 bp downstream of ATG, so that Cre recombination would result in the deletion of the entire Noggin coding sequence (Fig. 1A) (22). To study the inactivation of Noggin in mature osteoblasts, transgenic mice expressing the Cre recombinase under the control of a 3.9-kb fragment of the human osteocalcin promoter (Oc-Cre), created in a Friend leukemia virus strain B genetic background, were obtained from T. Clemens (Baltimore, MD) (23). Transgenics expressing Cre were bred to Noggin^+/− heterozygous null mice in a C57BL/6 background, provided by R. Harland, and intermated to create homozygous Oc-Cre mice in a heterozygous Noggin^+/− background (20). These were mated with homozygous Noggin^loxP/loxP mice, generating an experimental cohort, where Cre deletes the loxP flanked sequences from the Noggin^loxP allele and where a Noggin null allele is retained (Noggin^Δ/−), and a control littermate cohort carrying a Cre-deleted Noggin^loxP allele and a wild-type allele (Noggin^Δ/+). To ensure that the latter were appropriate controls, nonconditional Noggin heterozygous null mice were compared with wild-type littermate C57BL/6 mice for their skeletal phenotype. Male and female experimental and control littermate mice were compared at 1, 3 and 6 months of age. Genotyping of Oc-Cre, Noggin^loxP and Noggin^+/− mice was carried out in tail DNA by PCR using specific primers (Table 1). Deletion of loxP flanked sequences by the Cre recombinase was documented by PCR in DNA extracted from either calvariae or femurs (because of limited sample availability) of 1- to 6-month-old mice using specific primers (Table 1). Noggin deletion was confirmed by the determination of Noggin mRNA levels by real-time RT-PCR in calvarial and femoral extracts from 1-month-old mice and in calvarial extracts from 3 and 6-month-old mice. Animal experiments were approved by the Animal Care and Use Committee of Saint Francis Hospital and Medical Center.

Fig. 1.

Targeting construct, validation of loxP recombination, skeletal x-rays, weight, and femoral length of male and female Oc-Cre;Noggin^Δ/− conditional null mice (filled circles and black bars) and Noggin^Δ/+ littermate controls (open circles and white bars). A, The targeting construct used for the conditional inactivation of Noggin. B, DNA extracted from either calvariae or femurs from Noggin conditional null and control mice before and after loxP recombination by Cre expressed under the control of the osteocalcin promoter. A 400-bp band is detected for the Noggin^Δ allele. C, Noggin mRNA levels in total calvarial extracts expressed as Noggin copy number corrected for Rpl38 and control values normalized to 1.0. D, A representative skeletal x-ray of 1-month-old Oc-Cre;Noggin^Δ/− conditional null mouse and Noggin^Δ/+ littermate control. Arrows in the inset point to articular deformities in femoral-tibial joint. E and F, Weight in grams (E) and femoral length in millimeters (F). Values in C, E, and F are means ± sem; n = 4. *, Significantly different from control mice, P < 0.05.

Table 1.

Primers used for genotyping, loxP recombination, and real time RT-PCR

Allele or gene	Strand	Primer sequence (5′–3′)	Size (bp)
Genotyping
Oc-Cre transgene	Forward	CAAAGAGCCCTGGCAGAT	300
	Reverse	TGATACAAGGGACATCTTCC
NogginloxP allele	Forward	GGGTGTCAGAGCGCGCCCAC	202 wild type
	Reverse	TCCCTAGTCAGTTGTGGGTC	300 NogginloxP
Noggin− allele	Forward	GCATGGAGCGCTGCCCCAGC	160
	Reverse	GAGCAGCGAGCGCAGCAGCG
LoxP recombination
Noggin	Forward	AGGCTCCGCACAGAGAAACAAG	400
	Reverse	GAGTCCATGTCCAGGGAAGGGG
Real time RT-PCR
Noggin	Forward	GTCACCCTGTACGCCCTGGT
	Reverse	CGGCTGTGTAGATAGTGCTGGC[FAM]G
Dkk1	Forward	CCCTCCCTTGCGCTGAAGATGAGGAGT
	Reverse	CGCTTTCGGCAAGCCAGAC
Sost	Forward	AGGAATGATGCCACAGAGGTC
	Reverse	CTGGTTGTTCTCAGGAGGAGGCTC
Rpl38	Forward	CGAACCGGATAATGTGAAGTTCAAGGTT[FAM]G
	Reverse	CTGCTTCAGCTTCTCTGCCTTT

X-ray analysis, bone mineral density (BMD), and femoral length

X-rays were performed on eviscerated mice at an intensity of 30 kV for 20 sec on a Faxitron x-ray system (model MX 20; Faxitron X-Ray Corp., Wheeling, IL). Total, femoral, and vertebral areal BMD (grams per square centimeter) were measured on anesthetized mice using the PIXImus small-animal DEXA system (GE Medical System/LUNAR, Madison, WI) (24). Calibrations were performed with a phantom of defined value, and quality assurance measurements were performed before each use. The coefficient of variation for total BMD was less than 1% (n = 9). Femoral images acquired with microcomputed tomography (μCT) were used to determine femoral length.

Microcomputed tomography

Bone microarchitecture of vertebrae and femurs from Noggin-null and control mice was determined using a μCT instrument (μCT 40; Scanco Medical AG, Bassersdorf, Switzerland) calibrated weekly using a phantom provided by the manufacturer (25). The vertebral body of L3 and femurs were scanned in 70% ethanol at high resolution, energy level of 55 kVp, intensity of 145 μA, and integration time of 200 msec. Trabecular bone volume fraction and microarchitecture were evaluated starting approximately 1.0 mm from the cranial side of the vertebral body or 1.0 mm proximal from the femoral condyles. For L3 and femurs, a total of 100 and 160 consecutive slices, respectively, were acquired at an isotropic voxel size of 6 μm³ and a slice thickness of 6 μm and chosen for analysis. Contours were manually drawn every 10 slices a few voxels away from the endocortical boundary to define the region of interest for analysis. The remaining slice contours were iterated automatically. Trabecular regions were assessed for total volume, bone volume, bone volume fraction (bone volume/total volume), trabecular thickness, trabecular number, trabecular separation, connectivity density, structure model index (SMI) and material density (expressed in milligrams hydroxyapatite per cubic centimeter) using a Gaussian filter (σ = 0.8; support = 1) and user-defined threshold. SMI calculates the amount of plates (ideal value closer to 0) or rods (ideal value closer to 3) by means of three-dimensional image analysis based on a differential analysis of the triangulated bone surface. For analysis of femoral cortical bone, contours were iterated across 100 slices along the cortical shell, excluding the marrow cavity at the midshaft. Cortical bone was analyzed at the femoral midshaft for bone volume fraction, cortical thickness, cortical porosity, periosteal and endosteal perimeters, and material density using a Gaussian filter (σ = 0.8; support = 1) and user-defined threshold.

Bone histomorphometric analysis

Static and dynamic histomorphometry were carried out on Noggin-null and control mice after they were injected with calcein, 20 mg/kg, and demeclocycline, 50 mg/kg, at an interval of 2 d for 1-month-old animals and 7 d for 3- and 6-month-old animals. Mice were killed by CO₂ inhalation 2 d after the demeclocycline injection. Femurs were dissected and fixed in 70% ethanol, dehydrated, and embedded undecalcified in methyl methacrylate. Longitudinal femoral sections, 5 μm thick, were cut on a microtome (Microm; Richards-Allan Scientific, Kalamazoo, MI) and stained with 0.1% toluidine blue or von Kossa. Static parameters of bone formation and resorption were measured in a defined area between 360 and 2160 μm from the growth plate, encompassing 2.59 mm² of area, using an OsteoMeasure morphometry system (Osteometrics, Atlanta, GA). For dynamic histomorphometry, mineralizing surface per bone surface and mineral apposition rate were measured on unstained sections under UV light, using a diamidino-2-phenylindole fluorescein isothiocyanate Texas red filter, and bone formation rate was calculated. The terminology and units used are those recommended by the Histomorphometry Nomenclature Committee of the American Society for Bone and Mineral Research (26).

Osteoblast cell cultures

Primary osteoblasts were isolated from parietal bones of 3- to 5-d-old Friend leukemia virus strain B wild-type mice by sequential collagenase digestion, as described (27). Cells were cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with nonessential amino acids, 20 mm HEPES, 100 μg/ml ascorbic acid, and 10% fetal bovine serum (Atlanta Biologicals, Norcross, GA) at 37 C in a humidified 5% CO₂ incubator. Cells were cultured to confluence and exposed to DMEM or BMP-2 (Wyeth Research, Collegeville, PA) in the absence of serum for 24 h. To inactivate Noggin in vitro, osteoblasts from Noggin^loxP/loxP mice were transferred to medium containing 2% fetal bovine serum for 1 h and exposed overnight to 100 multiplicity of infection of replication-defective recombinant adenoviruses. An adenoviral vector expressing Cre recombinase under the control of the cytomegalovirus (CMV) promoter (Ad-CMV-Cre; Vector Biolabs, Philadelphia, PA) was delivered to induce recombination of the loxP sequences and inactivate Noggin. An adenoviral vector expressing green fluorescent protein (GFP) under the control of the CMV promoter (Ad-CMV-GFP; Vector Biolabs) was used as control.

Real-time RT-PCR

RNA was extracted from calvariae, femurs, or osteoblasts and mRNA levels determined by real-time RT-PCR (28, 29). RNA was reverse-transcribed using SuperScript III Platinum Two-Step qRT-PCR kit (Invitrogen) or iScript kit (Bio-Rad, Hercules, CA) according to the manufacturer's instructions and amplified in the presence of specific primers for Noggin, Dkk1, Sost (sclerostin), and ribosomal protein L38 (RpL38), and Platinum Quantitative PCR SuperMix-UDG (Invitrogen) or iQ-SYBR Green Supermix (Bio-Rad) at 60 C for 45 cycles (Table 1). Transcript copy number was estimated by comparison with a standard curve constructed using Noggin (Regeneron Pharmaceuticals, Tarrytown, NY), Dkk1 (Christof Niehrs, Heidelberg, Germany), Sost (Thermo Fischer Scientific, Huntsville, AL), or Rpl38 (American Type Culture Collection, Manassas, VA) cDNA (13, 30). Reactions were conducted in a CFX96 RT-PCR detection system (Bio-Rad), and fluorescence was monitored during every PCR cycle at the annealing step. Data are expressed as copy number corrected for Rpl38.

Statistical analysis

Data are expressed as means ± sem. Statistical differences were determined by unpaired Student's t test or ANOVA.

Results

General characteristics of Noggin conditional null mice

To study Noggin conditional null mice, homozygous Oc-Cre transgenics in a heterozygous null Noggin background (Oc-Cre/Oc-Cre;Noggin^+/−) were crossed with homozygous Noggin^loxP/loxP mice to create Oc-Cre;Noggin^Δ/− experimental and Oc-Cre;Noggin^Δ/+ controls. To ensure the validity of controls, Noggin heterozygous mice and Noggin^loxP/loxP were compared with wild-type control mice. Noggin^+/− and Noggin^loxP/loxP mice appeared normal and healthy up to 6 months of age. Noggin^loxP/loxP male and female mice at 1, 3, and 6 month of age, and Noggin^+/− male and female mice at 1 and 6 month of age (3-month-old mice were not available for μCT) did not exhibit changes in trabecular bone μCT parameters when compared with wild-type controls. Modest and inconsistent changes in cortical thickness were noted; a 12% decrease was observed in 1-month-old female Noggin^+/− mice and a 10% increase in 6-month-old male Noggin^loxP/loxP mice (not shown). Consequently, Noggin^Δ/+ heterozygous littermate mice were considered appropriate controls for Noggin^Δ/− conditional null mice.

Because the experimental mating scheme generated Oc-Cre;Noggin^Δ/− and Oc-Cre;Noggin^Δ/+ controls, deletion of loxP-flanked sequences was detected in either femoral or calvarial extracts from the experimental cohort, Noggin^Δ/−, and in extracts from littermate controls, Noggin^Δ/+ (Fig. 1B). Noggin mRNA levels measured in calvarial extracts from 1- and 3-month-old Noggin^Δ/− conditional null mice were suppressed in relationship to those measured in littermate controls Noggin^Δ/+ (Fig. 1C). A smaller decrease in Noggin mRNA levels was observed in calvariae from 6-month-old mice, possibly due to a decline in the activity of the osteocalcin promoter (17, 31). The down-regulation of noggin transcripts was verified in femurs from 1-month-old mice. Noggin mRNA levels were decreased from (values are means ± sem; n = 4) 0.24 ± 0.02 in femurs from Noggin^Δ/+ control male mice to 0.10 ± 0.02 Noggin/Rpl38 copy number in Noggin^Δ/− conditional null male mice (P < 0.01) and from 0.15 ± 0.01 in femurs from control Noggin^Δ/+ female mice to 0.08 ± 0.01 Noggin/Rpl38 copy number in Noggin^Δ/− conditional null female mice (P < 0.01). The appearance of Noggin conditional null mice was not different from controls except for modest articular deformities in the femoral-tibial joint, which were confirmed by contact radiography (Fig. 1D). The weight and femoral length were not decreased substantially in Noggin^Δ/− mice when compared with littermate controls, although at 1 month of age and in males at 6 months of age, the differences in both parameters were statistically significant (Fig. 1, E and F). Areal BMD was decreased in 1-month-old Noggin conditional null mice of both sexes, and the osteopenia was generalized because total BMD as well as vertebral and femoral BMD were decreased (Table 2).

Table 2.

Areal BMD of 1-month-old Noggin conditional null mice and littermate controls

	Total		Femoral		Vertebral
	Control	NogginΔ/−	Control	NogginΔ/−	Control	NogginΔ/−
Males	342 ± 3	295 ± 3a	420 ± 12	338 ± 23a	381 ± 14	313 ± 12a
Females	339 ± 4	305 ± 7a	419 ± 4	349 ± 11a	361 ± 3	308 ± 13a

BMD in grams per square centimeter × 10⁴ was obtained from 1-month-old male and female Oc-Cre;Noggin^Δ/− conditional null mice and Noggin^Δ/+ littermate controls. Values are means ± sem; n = 5–8.

^aSignificantly different from controls, P < 0.05 by unpaired t test.

Skeletal microarchitecture of Noggin conditional null mice

μCT of the distal femur revealed that male and female Noggin conditional null mice had decreased trabecular bone volume at 1 and 3 months of age due to a decreased number of trabeculae, although this decrease did not reach statistical significance in 3-month-old male mice (Table 3 and Fig. 2). The osteopenia was associated with decreased connectivity, and SMI revealed a tendency toward rod-like trabeculae. The trabecular microarchitectural changes were similar in male and female Noggin conditional null mice. Six-month-old male Noggin conditional null mice had a nonsignificant decrease in trabecular bone volume when compared with control littermate mice. Cortical bone volume and thickness, determined at the femoral midshaft, were decreased and cortical porosity was increased in 1- and 3-month-old Noggin-null mice, although in 3-month-old male mice, only the decrease in cortical thickness was statistically significant when compared with controls. Analysis of L3 was performed to assess whether the osteopenic phenotype observed was generalized. Confirming the femoral μCT analysis, both male and female Noggin conditional null mice exhibited vertebral osteopenia at 1 month of age (Table 4).

Table 3.

Femoral microarchitecture assessed by μCT of 1-, 3-, and 6-month-old Noggin conditional null mice and littermate controls

	1 Month		3 Months		6 Months
	Control	NogginΔ/−	Control	NogginΔ/−	Control	NogginΔ/−
Males
Trabecular bone
Bone volume/tissue volume (%)	5.6 ± 0.3	2.9 ± 0.6a	4.4 ± 0.3	2.8 ± 0.5a	2.7 ± 0.4	1.7 ± 0.2
Trabecular separation (μm)	205 ± 7	289 ± 26a	248 ± 11	287 ± 17	349 ± 20	379 ± 22
Trabecular number (1/mm)	4.9 ± 0.2	3.6 ± 0.3a	4.1 ± 0.2	3.5 ± 0.2	2.9 ± 0.1	2.7 ± 0.2
Trabecular thickness (μm)	25.2 ± 0.3	24.0 ± 0.7	32.1 ± 0.9	29.9 ± 1.2	37.3 ± 0.1	38.6 ± 1.9
Connectivity density (1/mm3)	222 ± 20	65 ± 21a	93 ± 15	56 ± 21	32 ± 6	16 ± 5
SMI	2.7 ± 0.03	3.1 ± 0.09a	3.1 ± 0.08	3.4 ± 0.17	3.0 ± 0.11	3.6 ± 0.07a
Density of material (mg HA/cm3)	950 ± 5	901 ± 14a	1068 ± 5	1045 ± 11	1093 ± 8	1107 ± 8
Cortical bone
Bone volume/tissue volume (%)	93.5 ± 0.3	89.6 ± 0.2a	94.3 ± 0.3	93.8 ± 0.3	94.5 ± 0.1	94.9 ± 0.1
Porosity (%)	6.5 ± 0.3	10.4 ± 0.2a	5.7 ± 0.3	6.2 ± 0.3	5.5 ± 0.1	5.1 ± 0.1a
Cortical thickness (μm)	114 ± 2	82 ± 3a	189 ± 3	165 ± 6a	184 ± 4	196 ± 3a
Periosteal perimeter (mm)	4.5 ± 0.03	4.8 ± 0.15	4.9 ± 0.04	5.5 ± 0.07a	5.3 ± 0.07	6.1 ± 0.06a
Endosteal perimeter (mm)	3.8 ± 0.02	4.3 ± 0.13a	3.6 ± 0.05	4.4 ± 0.06a	4.1 ± 0.08	4.7 ± 0.07a
Density of material (mg HA/cm3)	1114 ± 4	1040 ± 7	1352 ± 13	1328 ± 10	1365 ± 4	1384 ± 8a
Females
Trabecular bone
Bone volume/tissue volume (%)	4.3 ± 0.2	2.5 ± 0.3a	3.2 ± 0.6	1.8 ± 0.2a	1.2 ± 0.1	0.9 ± 0.2
Trabecular separation (μm)	252 ± 11	314 ± 9a	312 ± 14	355 ± 8a	484 ± 39	476 ± 28
Trabecular number (1/mm)	4.0 ± 0.2	3.2 ± 0.1a	3.3 ± 0.2	2.8 ± 0.1a	2.1 ± 0.1	2.2 ± 0.1
Trabecular thickness (μm)	26.1 ± 0.4	25.3 ± 0.3	35.4 ± 1.5	36.0 ± 2.2	37.0 ± 1.5	35.1 ± 3.3
Connectivity density (1/mm3)	127 ± 12	50 ± 6a	49 ± 15	12 ± 2a	13 ± 2	10 ± 6
SMI	2.9 ± 0.04	3.1 ± 0.05a	3.3 ± 0.15	3.9 ± 0.09a	3.3 ± 0.22	3.8 ± 0.07
Density of material (mg HA/cm3)	974 ± 6	934 ± 5a	1094 ± 10	1064 ± 10	1122 ± 8	1073 ± 23
Cortical bone
Bone volume/tissue volume (%)	93.4 ± 0.4	89.4 ± 0.5a	94.5 ± 0.1	93.4 ± 0.2a	94.8 ± 0.1	94.6 ± 0.1
Porosity (%)	6.6 ± 0.4	10.6 ± 0.5a	5.5 ± 0.1	6.6 ± 0.2a	5.2 ± 0.1	5.4 ± 0.2
Cortical thickness (μm)	112 ± 1	82 ± 3a	180 ± 1	153 ± 4a	193 ± 3	185 ± 5
Periosteal perimeter (mm)	4.4 ± 0.05	4.7 ± 0.07a	4.7 ± 0.06	5.2 ± 0.15a	4.9 ± 0.07	56 ± 0.10
Endosteal perimeter (mm)	3.7 ± 0.04	4.1 ± 0.06a	3.5 ± 0.06	4.2 ± 0.1a	3.7 ± 0.06	4.4 ± 0.08
Density of material (mg HA/cm3)	1120 ± 5	1064 ± 6a	1320 ± 8	1298 ± 15	1394 ± 4	1393 ± 10

Bone μCT was performed on femurs from 1-, 3-, and 6-month-old male and female Oc-Cre;Noggin^Δ/− conditional null mice and Noggin^Δ/+ littermate controls. Voxel size used for analysis was 6 μm³. Trabecular and cortical bone was analyzed in the distal femur and midshaft, respectively. Values are means ± sem; n = 5–7. HA, Hydroxyapatite.

^aSignificantly different from controls, P < 0.05 by unpaired t test.

Fig. 2.

Representative femoral microarchitecture assessed by μCT scanning of 1-, 3-, and 6-month-old male and female Oc-Cre;Noggin^Δ/− conditional null mice and Noggin^Δ/+ littermate controls. Trabecular bone was analyzed in the distal femur and cortical bone in the femoral midshaft.

Table 4.

Vertebral microarchitecture assessed by μCT of 1-month-old Noggin conditional null mice and littermate controls

	Control	NogginΔ/−
Males
Bone volume/tissue volume (%)	7.1 ± 0.6	5.1 ± 0.2a
Trabecular separation (μm)	201 ± 8	221 ± 8
Trabecular number (1/mm)	5.1 ± 0.2	4.6 ± 0.2
Trabecular thickness (μm)	24.2 ± 0.6	22.1 ± 0.5a
Connectivity density (1/mm3)	379 ± 31	281 ± 12a
SMI	2.4 ± 0.1	2.5 ± 0.1
Density (mg/cm3)	952 ± 12	912 ± 10a
Females
Bone volume/tissue volume (%)	6.5 ± 0.5	4.7 ± 0.5a
Trabecular separation (μm)	214 ± 7	231 ± 5
Trabecular number (1/mm)	4.8 ± 0.1	4.3 ± 0.1a
Trabecular thickness (μm)	24.8 ± 0.7	22.2 ± 0.8a
Connectivity density (1/mm3)	298 ± 27	225 ± 35
SMI	2.5 ± 0.1	2.6 ± 0.1
Density (mg/cm3)	958 ± 14	909 ± 11a

Bone μCT was performed on vertebrae from 1-month-old male and female Noggin^Δ/− conditional null and Noggin^Δ/+ littermate controls. Voxel size used for analysis was 6 μm³. Values are means ± sem; n = 5–6.

^aSignificantly different from controls, P < 0.05 by unpaired t test.

Femoral histomorphometry of Noggin conditional null mice

Bone histomorphometric analysis of femurs from Noggin conditional null mice at 1 month of age confirmed the microarchitectural findings (Table 5 and Fig. 3). At 1 month of age, male and female Noggin conditional null mice had decreased bone volume/tissue volume secondary to a decrease in trabecular number and thickness. There was no change in osteoblast number/perimeter or parameters of bone formation. An approximately 40% significant increase in osteoclast number and an approximately 30% statistically nonsignificant increase in eroded surface were noted in male but not in female mice. The osteopenia was less pronounced in 3- and 6-month-old Noggin conditional null mice and was observed in male but not in female mice. Bone formation rate was increased in 3-month-old male Noggin-null mice as a result of increased mineralizing surface but was not increased in Noggin conditional null female mice. This was possibly related to a decrease in the number of osteoblasts in female mice, which also displayed decreased osteoid surface.

Table 5.

Femoral histomorphometry of 1-, 3-, and 6-month-old Noggin conditional null mice and littermate controls

	1 Month		3 Months		6 Months
	Control	NogginΔ/−	Control	NogginΔ/−	Control	NogginΔ/−
Males
Bone volume/tissue volume (%)	12.1 ± 1.1	5.9 ± 0.9a	8.7 ± 0.4	6.2 ± 0.5a	7.3 ± 0.8	4.4 ± 0.9a
Trabecular separation (μm)	213 ± 17	416 ± 65a	254 ± 15	334 ± 20a	445 ± 44	742 ± 127
Trabecular number (1/mm)	4.3 ± 0.3	2.4 ± 0.4a	3.6 ± 0.2	2.9 ± 0.2a	2.2 ± 0.2	1.5 ± 0.2a
Trabecular thickness (μm)	28.3 ± 0.8	24.0 ± 0.5a	24.0 ± 0.9	21.4 ± 0.8a	33.4 ± 1.5	28.3 ± 2.8
Osteoblast surface/bone surface (%)	31.8 ± 1.4	37.5 ± 4.5	23.7 ± 1.9	21.1 ± 1.2	22.1 ± 3.8	19.4 ± 1.1
Number of osteoblasts/bone perimeter (1/mm)	34.2 ± 2.0	40.4 ± 6.3	19.8 ± 1.7	16.7 ± 1.1	19.4 ± 3.5	18.1 ± 1.1
Osteoid surface/bone surface (%)	4.2 ± 0.7	4.3 ± 0.9	3.3 ± 0.5	2.2 ± 0.4	3.7 ± 0.9	2.3 ± 0.2
Osteoclast surface/bone surface (%)	11.6 ± 0.7	16.4 ± 1.4a	8.7 ± 0.8	7.3 ± 0.8	4.9 ± 0.4	6.0 ± 0.5
Number of osteoclasts/bone perimeter (1/mm)	5.9 ± 0.4	8.4 ± 0.7a	3.7 ± 0.3	3.2 ± 0.3	2.5 ± 0.2	3.1 ± 0.2
Eroded surface/bone surface (%)	24.5 ± 1.4	31.1 ± 3.9	17.2 ± 1.3	14.3 ± 1.4	10.0 ± 0.7	11.3 ± 0.7
Mineral apposition rate (μm/d)	3.54 ± 0.19	3.49 ± 0.20	0.90 ± 0.03	0.77 ± 0.11	0.73 ± 0.05	0.74 ± 0.06
Mineralizing surface/bone surface (%)	3.2 ± 0.3	2.8 ± 0.4	2.4 ± 0.4	6.7 ± 1.2a	6.1 ± 0.7	4.9 ± 0.3
Bone formation rate (μm3/μm2 · d)	0.114 ± 0.011	0.096 ± 0.013	0.021 ± 0.003	0.058 ± 0.010a	0.046 ± 0.008	0.036 ± 0.004
Females
Bone volume/tissue volume (%)	8.4 ± 0.7	6.0 ± 0.7a	4.9 ± 0.5	3.9 ± 0.2	2.1 ± 0.3	2.2 ± 0.5
Trabecular separation (μm)	323 ± 34	374 ± 44	435 ± 37	470 ± 24	1408 ± 410	2385 ± 1431
Trabecular number (1/mm)	3.0 ± 0.3	2.7 ± 0.2	2.3 ± 0.2	2.1 ± 0.1	0.8 ± 0.2	0.9 ± 0.2
Trabecular thickness (μm)	28.5 ± 0.6	22.4 ± 1.0a	20.7 ± 0.8	18.5 ± 0.5a	26.0 ± 1.3	27.8 ± 3.6
Osteoblast surface/bone surface (%)	32.0 ± 0.9	34.8 ± 2.7	36.1 ± 1.8	28.7 ± 2.4a	29.1 ± 7.6	34.5 ± 3.0
Number of osteoblasts/bone perimeter (1/mm)	33.2 ± 1.3	37.0 ± 3.9	29.9 ± 1.8	23.0 ± 2.1a	27.2 ± 7.7	31.6 ± 3.5
Osteoid surface/bone surface (%)	3.7 ± 0.5	4.3 ± 0.6	5.9 ± 0.5	2.8 ± 0.7a	2.4 ± 0.9	4.5 ± 1.1
Osteoclast surface/bone surface (%)	14.9 ± 0.7	14.1 ± 0.9	8.7 ± 0.7	9.1 ± 0.4	7.8 ± 1.1	8.4 ± 1.6
Number of osteoclasts/bone perimeter (1/mm)	7.5 ± 0.3	7.2 ± 0.5	3.9 ± 0.3	4.0 ± 0.2	4.0 ± 0.5	4.4 ± 0.9
Eroded surface/bone surface (%)	29.8 ± 1.2	27.9 ± 1.7	16.8 ± 1.2	17.0 ± 1.1	15.7 ± 1.8	14.5 ± 2.8
Mineral apposition rate (μm/d)	3.52 ± 0.15	3.40 ± 0.18	1.32 ± 0.05	1.40 ± 0.04	1.20 ± 0.09	1.24 ± 0.07
Mineralizing surface/bone surface (%)	3.7 ± 0.6	2.7 ± 0.5	5.0 ± 0.6	5.7 ± 1.0	15.8 ± 2.0	13.2 ± 3.3
Bone formation rate (μm3/μm2 · d)	0.130 ± 0.023	0.091 ± 0.019	0.066 ± 0.009	0.080 ± 0.013	0.184 ± 0.011	0.164 ± 0.044

Bone histomorphometry was performed on distal femurs from 1-, 3-, and 6-month-old male and female Oc-Cre;Noggin^Δ/− conditional null mice and Noggin^Δ/+ littermate controls. Values are means ± sem; n = 4–10.

^aSignificantly different from controls, P < 0.05 by unpaired t test.

Fig. 3.

Representative histological sections and calcein/demeclocycline labeling of femoral sections from 1-, 3-, and 6-month-old male and female Oc-Cre;Noggin^Δ/− conditional null mice and Noggin^Δ/+ littermate controls. Sections from the distal femur were stained with von Kossa without counterstain (final magnification, ×40) at all ages or unstained and examined under fluorescence microscopy at 1, 3, and 6 months of age (final magnification, ×100).

Osteoblast cultures

To explore possible mechanisms responsible for the phenotype observed, calvarial osteoblasts were treated with BMP-2 and examined for the expression of the Wnt antagonists DKK1 and sclerostin. BMP-2 at 100 ng/ml for 24 h induced DKK1 mRNA levels from (means ± sem; n = 6) 0.81 ± 0.08 in control to 1.30 ± 0.10 Dkk1/Rpl38 copy number (P < 0.01) in BMP-2-treated cultures. Sclerostin transcripts were not detectable either in the absence or presence of BMP-2. To determine whether Noggin down-regulation sensitizes osteoblasts to the effect of BMP-2, osteoblasts from Noggin^loxP/loxP mice were transduced with the adenoviral vector Ad-CMV-Cre to deliver Cre recombinase to delete Noggin or with Ad-CMV-GFP as a control vector. Noggin transcripts were down-regulated in Ad-CMV-Cre transduced cultures by 85–95% (n = 4; P < 0.05) when compared with Ad-CMV-GFP-transduced cultures, both under basal and BMP-2-treated conditions. BMP-2 at 100 ng/ml for 24 h induced Dkk1 mRNA levels from (means ± sem; n = 4) 0.58 ± 0.05 to 3.83 ± 0.36 Dkk1/Rpl38 copy number in control cultures (P < 0.01) and from 1.00 ± 0.05 to 6.03 ± 0.40 Dkk1/Rpl38 copy number (P < 0.01) in Noggin down-regulated cultures.

Discussion

Our findings indicate that Noggin inactivation in the postnatal skeleton causes osteopenia. Noggin conditional null male and female mice exhibited generalized osteopenia at 1 month of age, and this persisted for a 3 month period and was less evident in 6-month-old mice. This is possibly because the activity of the osteocalcin promoter declines with age, and in accordance, calvarial Noggin mRNA levels were 30% lower in conditional Noggin-null 6-month-old mice but not significantly different when compared with heterozygous controls (17, 31). The decreased bone volume was secondary to a decrease in trabecular number and thickness. Although both male and female Noggin conditional null mice exhibited osteopenia, the cause seemed to be different. Male mice had an increase in the number of osteoclasts, a nonsignificant increase in eroded surface, and a subsequent increase in bone formation, suggesting that bone remodeling was enhanced. The phenotype observed in conditional Noggin null male mice indicates an increase in bone resorption by BMP consistent with what has been reported in transgenics overexpressing BMP-4 under the control of a 2.3-kb fragment of the type 1 collagen promoter (32). In contrast, female Noggin conditional null mice exhibited decreased osteoblast number and osteoid surface but no changes in osteoclast number. It is possible that a resorptive phenotype was not evident in female mice because of the inhibitory effect of estrogens on bone resorption (33–35). It is noteworthy that the cellular changes were observed in 3-month-old female mice, when the osteopenia was detectable by μCT and not by histomorphometry. This may reflect differences in the sensitivity of the two methodologies. The early osteopenic phenotype observed at 1 month of age may also be secondary to a late developmental defect because the osteocalcin promoter is active at 18.5 d of embryonic life, and Noggin conditional null mice displayed minor joint abnormalities resembling those reported after the global inactivation of Noggin (20). In this study, we confirm earlier observations demonstrating a more pronounced decline in trabecular bone volume in C57BL/6 female than in male mice with age, possibly explaining some of the differences in the male and female phenotypes (36).

Noggin conditional null male mice exhibited an increase in bone remodeling, but the increased bone formation was not sufficient to normalize trabecular bone volume. The increased bone formation was secondary to an increase in mineralizing surface and not in mineral apposition rate. This indicates an increase in areas actively forming bone and not in the function of individual osteoblasts. Changes in mineralizing surface responsible for changes in bone formation are not unusual and were found in Cebp homologous protein, Gremlin, and Sost-null mice (37–39). They were also reported in Cynomolgus monkeys treated with sclerostin antibodies, which exhibited a pronounced enhancement in mineralizing surface and a modest change in mineral apposition rate (40). The transient increase in bone formation observed in Noggin conditional null mice confirms results observed after the conditional deletion of the BMP antagonist Gremlin1 in osteoblasts and is in agreement with the suppression of bone formation observed in transgenic mice overexpressing noggin in osteoblasts (17, 41). However, the decreased number of osteoblasts observed in female Noggin conditional null mice is not in accordance with studies demonstrating that down-regulation of Noggin enhances osteoblastogenesis (19).

The osteopenic phenotype of the Noggin conditional inactivation resembles that observed after the inactivation of connective tissue growth factor (Ctgf), a member of the Cyr61, CTGF, Nov (CCN) family of proteins known to bind and antagonize BMP action, and that of the global inactivation of Gremlin1 (37, 42–44). It is of interest that, like in the case of noggin, transgenic overexpression of CTGF or Gremlin causes osteopenia, whereas their inactivation reveals that they are necessary for skeletal homeostasis or that BMP in excess is detrimental to bone (37, 44–46). These observations suggest that when in excess BMP antagonists inhibit BMP actions and bone formation, but basal levels of BMP antagonists are required to maintain bone homeostasis, possibly by tempering selected BMP actions.

The in vivo phenotype described indicates that noggin is required not only for the previously reported role in skeletal development but also for normal postnatal bone remodeling because the inactivation of Noggin leads to osteopenia in adult male and female mice (20, 21). A plausible explanation for the osteopenia observed in male Noggin-null mice is a sensitization of BMP signaling and an increase in bone resorption caused by BMP-2 and -4 (32, 47, 48). An explanation for the reduced osteoblast number observed in female Noggin-null mice is enhanced BMP activity leading to an increased expression of the Wnt antagonists DKK1 and possibly sclerostin and suppressed Wnt signaling and osteoblastogenesis (3, 4). In the present studies, we have confirmed that BMP-2 induces DKK1 expression in osteoblasts. The levels of Dkk1 mRNA were modestly increased in osteoblasts from Noggin-inactivated osteoblasts under basal and BMP-2-treated conditions. The higher basal levels may be a reflection of an effect of endogenous BMP (49). We were unable to detect sclerostin transcripts in either control or BMP-2-exposed osteoblasts. This is possibly due to the fact that sclerostin is mostly expressed in osteocytes, and the cell isolation and culture conditions used favor the enrichment of osteoblasts but not of osteocytes (50). It is also possible that in the absence of noggin, other BMP antagonists, such as gremlin, acquire an overcompensatory function, because gremlin mRNA expression is induced by BMP in osteoblasts (51).

Although the phenotype observed could imply direct functions of noggin in the skeleton, the main activity of noggin is the binding of BMP indicating that the osteopenic phenotype observed is most probably due to excessive BMP activity. However, a BMP-independent activity of noggin in the skeleton has not been ruled out and could be responsible for the phenotype observed. The results observed in Noggin conditional null mice differ from those reported after the global inactivation of genes encoding for the Wnt antagonists sclerostin and DKK1 (38, 52). The deletion of Sost, encoding for sclerostin, or haploinsufficiency of Dkk1 causes increased bone formation and bone mass, indicating that these antagonists could be targeted to enhance Wnt activity as possible therapeutic anabolic approaches in the management of osteoporosis (53, 54). The results we describe in Noggin conditional null mice suggest that neutralization of noggin to enhance BMP activity would cause osteopenia and would not yield a sustained increase in bone formation and bone mass. Data derived from these mouse models suggest that the suppression of Wnt antagonists may result in a skeletal anabolic effect, whereas the suppression of noggin may not have this desirable outcome.

In conclusion, our studies reveal that Noggin inactivation in the postnatal skeleton causes osteopenia, indicating that excessive exposure to BMP may have a detrimental effect in bone.