• bone formation;
  • collagen;
  • osteoblast;
  • transgenic mice


  1. Top of page
  2. Abstract
  7. Acknowledgements

To characterize the function of type XIII collagen, a transmembrane protein occurring at cell adhesion sites, we generated transgenic mice overexpressing it. High transgene expression was detected in cartilage and bone. The overexpression mice developed an unexpected skeletal phenotype marked by a massive increase in bone mass caused by increased bone formation rather than impaired resorption.

Introduction: Type XIII collagen is a type II transmembrane protein that is expressed in many tissues throughout development and adult life. It is located in focal adhesions of cultured fibroblasts and other cells and in the adhesive structures of tissues. To further characterize the function of this protein, we generated transgenic mice overexpressing it. High transgene expression was detected in cartilage and bone in locations also containing the endogenous protein.

Materials and Methods:Col13a1 5′-flanking sequences were tested for their efficiencies to drive gene expression. Skeletal tissues of transgenic mice and wildtype littermates were compared using histological, immunohistochemical, and bone histomorphometrical analyses. Bone formation rate was measured by tetracycline double-labeling. Osteoclast number and resorption activity were determined using standard methods. RNA samples from transgenic and wildtype femurs were analyzed by Northern blotting and quantitative RT-PCR.

Results: There was no defect in early skeletal development, but the high bone mass phenotype became apparent in heterozygous mice at the age of 3-4 weeks. The changes were most noticeable in proximal long bones but were also detectable in calvarial bones. The cortical bone cross-sectional area and the volumetric BMD were highly increased, but the bone marrow was well formed. Histological and histomorphometric analysis showed that trabecular bone volume was not significantly altered. Because of the normal epiphyseal growth plates, the longitudinal growth was not affected. Bone formation rate was several times higher in the overexpression mice than in their normal littermates, whereas the osteoclast number and resorption activity were normal. RNA analysis revealed increased expression in the transcription factor Runx2 and IGF-II, both known to be involved in bone biology.

Conclusion: Overexpression of type XIII collagen in skeletal tissues leads postnatally to an abnormally high bone mass caused by increased bone formation rather than impaired resorption. The findings suggest that type XIII collagen has an important role in bone modeling, and in particular, it may have a function in coupling the regulation of bone mass to mechanical use.


  1. Top of page
  2. Abstract
  7. Acknowledgements

TYPES XIII, XXIII, AND XXV form the newest subgroup of structurally related molecules among the large family of collagens.(1–3) The most extensively characterized molecule within this subgroup is type XIII collagen, a type II transmembrane protein that is expressed in many tissues such as bone, cartilage, heart, muscle, and skin throughout development and adult life.(4, 5) Type XIII collagen is located in focal adhesions of cultured fibroblasts and other cells and in the adhesive structures of tissues, such as the myotendinous junctions in muscle, intercalated discs in the heart, and the cell-basement membrane interphases.(5–7) The type XIII collagen ectodomain can bind to fibronectin, heparin, the basement membrane components nidogen-2 and perlecan, and the α1 subunit of integrin.(8–10) Because of its location in tissues and cultured cells and its binding properties, type XIII collagen is thought to be involved in cellular adhesion and migration.

Type XIII collagen α chains form disulfide-bonded homotrimers.(11) The primary structure is composed of three collagenous domains (COL1-COL3), which are flanked and interrupted by noncollagenous domains (NC1-NC4). The short cytosolic domain and the transmembrane domain form part of the NC1 domain, whereas the other domains form the ectodomain, which is a rod of length about 150 nm with two flexible hinges coinciding with the NC2 and NC3 domains.(10) It has been shown that the sequences needed for the association of the three type XIII collagen α-chains reside in their N-terminal portions and that triple helix formation seems to proceed from the N terminus to the C terminus.(9) Because of complex alternative splicing, the primary structures of COL1, NC2, COL3, and NC4 can vary.(12–15)

We have previously used homologous gene targeting to generate a mouse line, Col13a1N/N, expressing modified type XIII collagen that lacks the extreme 96 N-terminal residues, including the cytosolic, transmembrane, and association domains.(16) These N-terminally truncated type XIII collagen molecules are transported to the roughly correct location despite their lack of a transmembrane domain, but they are functionally impaired, leading to a mild muscular phenotype in the gene-targeted mice, including abnormalities in the sarcolemma-basement membrane interphase.(16) In contrast, homozygous transgenic mice with a 90 amino acid deletion in the COL2 domain die before birth, at E10.5, because of a lack of chorion-allantois fusion and a functioning placenta, or at E12.5, because of cardiac defects.(17)

To further characterize the roles of type XIII collagen, we generated transgenic mouse lines overexpressing type XIII collagen α chains. Use of a 1-kb fragment derived from the Col13a1 promoter linked with type XIII collagen cDNA sequences led to strong expression of the transgene in skeletal tissues. This overexpression resulted in massive increase in bone mass, which suggests that type XIII collagen has a role in bone development and growth.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Constructs for promotor analysis

Constructs consisting of different lengths of 5′-flanking sequences of the mouse type XIII collagen gene were made and subcloned into the pGL2-basic vector (Promega) upstream from the luciferase gene. Plasmids containing Col13a1 sequences (M Sund, unpublished results, 1997) were first cleaved by BglII (Luc 1), ScaI (Luc 1.5), or NsiI (Luc 5) restriction enzymes. A linker primer containing restriction sites NotI, SpeI, SalI, and BglII was attached to the 5′ end of Luc 1, and a linker with NotI, SpeI, SalI, and NsiI restriction sites was attached to the 5′ end of Luc 5. A BstB1 restriction site was used as a common 3′ end for all three constructs. Finally, the ends of the promoter fragments were made blunt, and a SmaI site in the pGL2-basic vector was used for subcloning.

Cell culture and transfection assays

NIH/3T3 cells were cultured in DMEM supplemented with 10% calf serum, penicillin, and streptomycin at 37°C in a 5% CO2 atmosphere. NIH/3T3 cells were transfected with FuGENE 6 Transfection Reagent (Roche) according to manufacturer's instructions. Twenty-four hours after transfection, cell extracts were collected, and luciferase activity was measured using the Luciferase assay system (Promega). All transfections were done together with the pCMV-β-galactosidase plasmid (Clontech) to normalize transfection efficiencies. The β-galactosidase activity was measured using the β-galactosidase enzyme assay system (Promega). The pGL2-Basic vector was used as negative control and the pGL2-Control vector as positive control (Promega).

Transgene constructs

The mouse P40 cDNA clone(15) containing exons 2–41 was used to generate transgenic mice. A genomic DNA fragment containing the 159-bp extreme 3′ sequences of the first intron and the first 91 bp of the second exon was generated by PCR using the plasmid 3HA(17) and ligated to the 5′ end of the cDNA clones. To drive expression of the transgene, a 2.5-kb BglII-digested genomic DNA fragment containing 1 kb of promoter and 5′ flanking sequences, the complete first exon, and 0.8 kb of the first intron was further cloned at the 5′ end of the cDNA. A hemagglutinin (HA) tag, sequences for transcription termination, and the SV40 Poly-A DNA fragment were cloned at the 3′ end of the cDNA. In a second construct, a point mutation changing a glycine in exon 32 of the COL3 domain of type XIII collagen to a tryptophan was generated in the cDNA clone P40(15) using the Pharmacia U.S.E. kit system (Pharmacia Biotech). The oligonucleotides used were Hindmut 2 5′-AGAAAGGAGAAGCTTGGGAGAAAGGCGA-3′ and Hindmutrev 2 5′-TCGCCTTTCTCCCAAGCTTCTCCTTTCT-3′ (a new HindIII restriction site underlined, with the altered base marked with an asterisk). The mutated cDNA clone was used for subsequent cloning of the transgene construct as described above.

Generation and identification of transgenic mice

The 4.7-kb inserts were released from the pSP72 vector by SpeI digestion. The transgene construct was microinjected into one-cell stage embryos obtained from B6D2F1 (hybrid of C57BL/6J × DBA/2J) mice and implanted into NMRI pseudopregnant foster mothers. The transgene-positive mice were identified by PCR and Southern blot analysis. The Col13a1oe mice were genotyped using the following primers: MutScreen2 5′-GGTTTACCGGGGCCTCCTGGACCAAAGGG-3′ and MutScreen2rev 5′-GGCCTGCTTGTCCTGTCTCCCCTTTCTCC-3′. The tail genomic DNA of the Col13a1oew mice was amplified using the following primers: Mut 2 forward 5′-GCCAGGGACGCCAGGAACCAAGGG-3′ and Mut 2 reverse 5′-CCAGGCAATCCCAGAGGCCCCCGG-3′. After amplification, the 740-bp fragment was digested with the HindIII restriction enzyme to detect the additional cleavage site caused by the mutation. The genotypes of the founder mice were also verified by Southern blot analysis using the KpnI and XbaI restriction sites, as described earlier.(17)

Analysis of transgene expression by Western blotting and immunofluorescence staining

Tissues were rapidly frozen in liquid nitrogen and stored at −70°C until use for immunohistochemical analysis or protein extraction.

Skeletal tissues were ground in a mortar in liquid nitrogen followed by homogenization in sample buffer (0.06 M Tris, pH 6.8, 0.5 M urea, 10% glycerol) and addition of SDS (2%) and bromphenol blue. β-Mercaptoethanol (5%) was added to some of the samples. The samples were incubated for 5 minutes at +95–100°C and centrifuged for 30 minutes. The supernatants were analyzed by SDS-PAGE followed by Western blotting with antibodies specific to type XIII collagen(18) and the HA-tag (Santa Cruz Biotechnology).

Tissues for immunofluorescence staining were embedded in Tissue Tek O.C.T compound (Sakura Finetek) and rapidly frozen in liquid nitrogen. Samples were sectioned at 5 μm, fixed in methanol, and stained with the type XIII collagen and HA-tag antibodies.

Analysis of skeletal tissues by Northern blotting and quantitative RT-PCR

Total RNA was extracted from mouse tissues as previously described.(19) Skeletal tissues were ground in a mortar in liquid nitrogen before RNA extraction. For Northern blot analysis, 20-μg aliquots from the total RNA samples were electrophoresed in 1% agarose-5.5% formaldehyde gels and transferred to Hybond-N membranes (Amersham), and the filters were hybridized with32P-labeled cDNA probes.

The RT reaction was performed using the SuperScript First-Strand Synthesis system for RT-PCR kit (Invitrogen) according to the manufacturer's instructions. The samples were measured with the ABI 7700 Sequence Detection System using TaqMan chemistry. The forward and reverse primers and fluorogenic probes used were (1) type XIII collagen, GGGAAGCCCCGAAGATGT, TCTTCCAGTGGGACCAGGAG, and 5′-Fam-TCCAGGATGTAACTGCCCACCAGGA-Tamra-3′, (2) Runx2, ACTGGCGGTGCAACAAGAC, CTGGTACCTCTCCGAGGGC, and 5′-Fam-CTGCCCGTGGCCTTCAAGGTTGT-Tamra-3′, (3) IGF-II, GAGCTTGTTGACACGCTTCAGT, CGGCTTGAAGGCCTGCT, and 5′-Fam-TGTCTGTTCGGACCGCGGCTTCTA-Tamra-3′. The results were normalized to 18S RNA quantified from the same samples using the forward and reverse primers TGGTTGCAAAGCTGAAACTTAAAG and AGTCAAATTAAGCCGCAGGC, respectively. The probe for the 18S amplicon was 5′-Vic-CCTGGTGGTGCCCTTCCGTCA-Tamra-3′.

X-ray analysis and staining of skeletons

X-ray examinations were performed on anesthetized 6-month-old wildtype and transgenic mice. Skeletons of E16.5, newborn, and 1- to 6-month-old mice were dissected, eviscerated, and stained with Alizarin red and Alcian blue as described earlier.(20, 21)

Histological and immunohistochemical analysis

For histology, tissues were dissected and fixed in 10% phosphate-buffered formalin, pH 7.0. Skeletal tissues were either decalcified in 0.5 M EDTA, pH 7.4, or embedded in methyl-methacrylate and processed for hard tissue sections. Decalcified bones were embedded in paraffin and sectioned at 5 μm. The paraffin sections were stained with hematoxylin and eosin and analyzed under light microscopy for basic histology. Antibodies against type XIII collagen(18) and the HA-tag (Santa Cruz Biotechnology) were used for immunohistochemical analysis with Histomouse SP bulk kit (Zymed Laboratories) according to the manufacturer's instructions. Hard tissue samples were sectioned at 10 μm and analyzed unstained or with Masson Goldner trichrome staining.

pQCT measurements

Tibias and femurs were dissected out from 3-month-old mice and analyzed by pQCT (XCT 960A, Stratek) as shown previously.(22) The bones were inserted into glass tubes and scanned at the mid-diaphysis. BMC, BMD, cortical area (CSA), and cross-sectional moment of inertia (CSMI) were analyzed. The cortical attenuation threshold was 0.700 cm−1. The voxel size for wildtype bones was 0.092 × 0.092 × 1.25 (mm3) and that for transgenic bones was 0.148 × 0.148 × 1.25 (mm3).

Histomorphometry of trabecular bone

Trabecular bone volume was analyzed from paraffin sections of the distal femoral metaphysis of 1-month-old mice. An area located 0.5–1 mm from the epiphyseal cartilage and extending across the marrow cavity was measured. A Nikon Optiphot II microscope was used, and the images were digitized with a Sony DXC-930P 3CCD camera and analyzed with MCID-M4 image analysis software (Imaging Research).

Measurement of bone formation rate

Double labeling was performed on 3-month-old mice, which were injected intraperitoneally with 20 mg/kg of tetracycline (Terramycin 200 mg/ml; Pfizer Animal Health) 10 days and again 3 days before death. Their femurs and tibias were dissected out, fixed in 70% ethanol, and embedded in methyl methacrylate. The samples were sectioned and analyzed by fluorescence microscopy.

Analysis of osteoclast number and resorption activity

Paraffin sections from wildtype and transgenic femurs were stained for TRACP using a leukocyte acid phosphatase kit (Sigma Diagnostics, St Louis, MO, USA). The numbers of TRACP-stained osteoclasts were calculated for the distal metaphyseal area using a light microscope. The osteoclast resorption activity was analyzed by culturing bone marrow cells from 6-day-old wildtype and transgenic mice on bovine bone slices. Multinuclear TRACP+ cells were counted as osteoclasts, and resorption pits were visualized by WGA-Lectin and active osteoclasts by phalloidin staining as previously described.(23, 24) The osteoclast resorption activity was expressed as resorbed bone area/osteoclast.

Statistical analysis

Data are shown as means ± SD. Statistical significance was determined by Student's t-test. Values of p < 0.05 were considered statistically significant (p < 0.05,p < 0.01,p < 0.001).


  1. Top of page
  2. Abstract
  7. Acknowledgements

Generation of transgenic mice overexpressing type XIII collagen

Our aim was to test the effect of overexpression of type XIII collagen in mice. In case mere overexpression would not cause phenotypic changes, we also introduced the substitution of a glycine in the collagenous sequence to a bulky amino acid residue likely to impair collagen triple helix formation known in the case of fibrillar collagens.(25, 26) Thus, transgene constructs for misexpression of normal (Col13a1oe) and mutant (Col13a1oew) type XIII collagen were produced. The promoter of the mouse type XIII collagen gene has been predicted to locate between nucleotides −834 and −584 with respect to the initiation ATG.(27) We tested the efficiency of nucleotides −984 to −231, −1754 to −231, and −4878 to −231 of the Col13a1 5′-flanking sequences for stimulating the expression of a luciferase reporter gene in transiently transfected NIH/3T3 cells. There were no significant differences in luciferase activity between the different constructs (Fig. 1A). Thus, the shortest 5′ fragment was found to be sufficient to drive gene expression, and it was further used in transgene constructs.

thumbnail image

Figure FIG. 1.. Analysis of the mouse Col13a1 promoter and 5′-flanking sequences and generation of transgenic mice overexpressing normal and mutant type XIII collagen. (A) The effect of the Col13a1 promoter and different lengths of 5′-flanking sequences on luciferase gene expression in NIH/3T3 cells. The promoter constructs with their 3′ end at nucleotide-231 are shown on the left. The luciferase activities were normalized to β-galactosidase activities of cotransfected cells. The values represent the mean ± SD of five independent experiments, each run in duplicate. No significant differences were detected in the relative luciferase activities between the different constructs. (B) Schematic presentation of the transgene constructs. Col13a1oe, overexpression of normal type XIII collagen chains, and Col13a1oew, overexpression of type XIII collagen chains with a Gly to Trp point mutation. The mouse p40 cDNA clone(15) was used for the cloning of both constructs. A point mutation changing glycine to tryptophan and causing an additional Hind III restriction site in the COL3 domain was introduced by PCR mutagenesis. The restriction sites used for cloning the constructs are shown. Arrows (Col13a1oe) and arrowheads (Col13a1oew) indicate the locations of the PCR primers used for genotyping.

Download figure to PowerPoint

In addition to the promoter and 5′ flanking sequences, the transgene constructs contained the first exon and part of intron 1 cloned in front of the rest of the type XIII collagen cDNA (with and without a glycine to tryptophan point mutation; Fig. 1B). Six of the ensuing transgenic mouse lines (two for Col13a1oe and four for Col13a1oew) showed a skeletal phenotype and were characterized further. Identical phenotypes were observed in both types of mutant mouse line, suggesting that the overexpression rather than the point mutation in the COL3 domain was the cause of the phenotype. In the presentation of the consequences of the overexpression of normal or mutant type XIII collagen, we will show pictures mainly for Col13a1oe, but in some cases for Col13a1oew, the data being applicable to both types of mouse lines.

Expression of the transgene and endogenous type XIII collagen in mice

Western blot analysis was performed on several tissues from 2-month-old control and Col13a1oe mice with HA-tag and type XIII collagen antibodies to detect the transgene and endogenous type XIII collagen expression. Type XIII collagen expression was similar in the transgenic lines to that in the control samples, except in the skin, cartilage, and skeletal tissues, where markedly strong bands were found for type XIII collagen (Fig. 2A). In fact, in all of the six lines, the highest level of expression of the transgene was detected in skeletal tissues. Analysis of protein samples from calvariae and femurs using the HA-tag antibody under reduced and nonreduced conditions showed that the transgene-derived type XIII collagen α-chains formed disulfide-bonded trimers (Figs. 2B and 2C), as has previously been observed for type XIII collagen.(11)

thumbnail image

Figure FIG. 2.. Western blot analysis and immunofluorescence staining of wildtype and transgenic mouse tissues. (A) Increased type XIII collagen expression was found in adult transgenic skin, cartilage, and bone compared with wildtype littermates. The migration of a 123-kDa marker protein is indicated. (B and C) Western blot analysis of a 2-month-old femur (F) and calvaria (C). (B) Expression of type XIII collagen (arrow) was detected at higher levels in the transgene-positive skeletal tissues (TG) than in the corresponding wildtype tissues (WT). Western blotting with a HA-tag antibody (C) showed high expression of the transgene. The protein samples were analyzed under both reduced and nonreduced conditions. The transgene-derived type XIII collagen α1(XIII) chains formed trimers. The upper arrow points to trimeric α1(XIII) chains and the lower arrow to monomeric chains. (D-G) Immunofluorescence staining of developing bone at E17.5 with antibodies to type XIII collagen (D and E) and the HA-tag (F and G). High skeletal transgene expression was detected in the periosteum (arrows) and the proliferative/prehypertrophic zone of the growth plate (arrowheads). The expression level seems to be higher in the transgene tissues (TG). Bars: 20 μm.

Download figure to PowerPoint

Immunofluorescence stainings of adult and developing mouse tissues with HA-tag and type XIII collagen antibodies showed that the transgene expression follows closely that of endogenous type XIII collagen, with the same pattern as previously found for type XIII collagen expression during normal mouse development,(5) although at higher levels in the bone and cartilage of the transgenics (Figs. 2D-2G). Staining with the HA-tag antibody, which specifically detects the transgene construct, highlighted the overexpression of type XIII collagen in the proliferative zone of the growth plate and periosteum (Figs. 2E and 2G). In developing cartilage, the proliferating chondrocytes already showed strong transgene expression before mineralization started (data not shown).

Northern blotting and quantitative RT-PCR analysis of total RNA from skeletal tissues of heterozygous mice revealed strong type XIII collagen expression in the transgenic bone compared with wildtype littermates (Figs. 3A and 3B). In contrast, Northern blotting analysis for bone and cartilage collagen types I, II, and X and the noncollagenous osteoblast marker osteocalcin did not reveal any differences in the expression level of these genes between the control and transgenic mice (Fig. 3A). Interestingly, the osteoblast-specific transcription factor Runx2 and one of the most abundant growth factors in bone, IGF-II, were found to be upregulated (Fig. 3B).

thumbnail image

Figure FIG. 3.. Expression of selected genes in skeletal tissues (A) Northern blot analysis of 2-month-old wildtype (WT) and transgenic (TG) bones. The expression of collagen types II and X was analyzed in femurs, whereas other samples were from calvariae. Type XIII collagen expression is clearly increased in the mutant skeletal tissues. There were no significant differences in the expression of collagen types I, II, and X or of osteocalcin between the transgenic mice and their wildtype littermates. β-actin hybridization reveals equal loading of samples. (B) Quantitative RT-PCR analysis of transgenic and control femurs. RNA samples from 1-month-old wildtype (WT; n = 4) and transgenic (TG; n = 4) femurs were analyzed by quantitative real-time RT-PCR. Type XIII collagen, Runx2, and IGF-II expression was significantly increased in the transgenic femur.

Download figure to PowerPoint

Excessive bone formation in transgenic mice overexpressing type XIII collagen

Mice heterozygous for the transgene were mated, and transgene positive offspring were born in a normal ratio (data not shown). We could not detect by eye or microscopically any differences between the transgenic mice and their wildtype littermates at birth, but the transgenic mice began to walk abnormally at the age of 2 months, and their hind limbs were affected. X-ray analysis showed that the heterozygous overexpression mice had striking overgrowth of the femur and humerus (Fig. 4). This massive increase in bone mass can already be seen in heterozygous mice, and no further alterations were noted in homozygous mice. Closer examination also revealed abnormally high mass in other long bones, in the rib-vertebrae region, and in the calvaria (Figs. 4A-4D). The phenotype in the heterozygous mice could be visualized by Alizarin red/Alcian blue staining at the age of 3–4 weeks, the changes being most noticeable in the proximal long bones, and the skeletal abnormalities were very severe by the age of 2 months (Figs. 4E-4J).

thumbnail image

Figure FIG. 4.. X-ray examination and Alizarin red/Alcian blue staining of wildtype and transgenic mice. (A-D) X-ray examination of 6-month-old wildtype and transgenic mice. Increased bone mass is clearly visible in the long bones of the transgenic mouse (C and D), and the calvarial bones are also clearly thicker and highly mineralized (B) relative to those of its wildtype littermate (A). (E and F) Alizarin red/Alcian blue staining of newborn (Nb) mice. (G-J) Skeletons of older mice were stained only with Alizarin red. The skeletal phenotype became evident in 1-month-old transgenic mice (G and H) and progressed with age (I and J). Bars: 1.5 mm (G-J).

Download figure to PowerPoint

Despite the overgrowth of the bones, the mutant mice were fertile and had a normal life span. All six individual mouse lines (two Col13a1oe and four Col13a1oew lines) showed increased bone mass, confirming that the phenotype is caused by the misexpression of type XIII collagen. The severity of the phenotype correlated with the transgene expression level in bone (data not shown).

Increased mineral apposition and active osteoblast zones in transgenic mice

Analysis of the femurs and tibias of 3-month-old overexpression and control mice by pQCT showed that BMC and BMD were significantly elevated in the transgene cases. The moment of inertia at mid-diaphysis was 100 (tibia) to 1000 (femur) times higher in mutant mice (Table 1).

Table Table 1.. pQCT Analysis of Mutant and Control Bones
Thumbnail image of

Histological analysis of the tibias and femurs showed a massive increase in cortical bone in the overexpression mice (Figs. 5A-5D), and abnormally high bone mass was also detected in histological sections of calvariae (Figs. 5E and 5F), clavicles, and pelvic bones (data not shown). The bone marrow space was formed normally (Figs. 5A, 5B, and 6). Histomorphometric analysis on the distal femur of 1-month-old mice further showed that trabecular bone volume was not significantly changed in transgenic mice (Fig. 6). The proportional area of trabecular bone from the marrow space was 7.0 ± 1.3% (n = 3) in wildtype and 4.7 ± 1.1% (n = 3) in transgenic mice (p > 0.05). The growth plates were intact throughout development, and no obvious cartilage defects could be detected (Figs. 5G-5J). The basic histology of the other organs seemed normal, although the transgene and endogenous type XIII collagen molecules were widely expressed in tissues (Fig. 2).

thumbnail image

Figure FIG. 5.. Histological analysis of skeletal tissues. Hematoxylin/eosin stainings of decalcified bone sections. (A and B) Tibias of 1-month-old (A) wildtype and (B) transgenic mice. (B) The cortex of the bone is clearly thicker in the mutant tibia. The bone marrow (BM) is formed normally. (C and D) Femurs of 2-month-old (C) wildtype and (D) transgenic mice. Note the thick layer of periosteal osteoblasts (PO and arrow in the inset) in the mutant bone. EO, endosteal osteoblast layer. (E and F) Calvariae of 2-month-old (E) wildtype and (F) transgenic mice. Increased bone formation was also detected in the calvaria. (G-J) Epiphyseal growth plates of (G and H) 1- and (I and J) 2-month-old (G and I) wildtype and (H and J) transgenic mice. The growth plates seemed to be intact in the transgenic mice throughout their development. Bars: 25 (A, B, and E-J) and 50 μm (C and D).

Download figure to PowerPoint

thumbnail image

Figure FIG. 6.. Histological analysis of femurs from 1-month-old mice. The sections show a marked increase in cortical bone in the transgenic mouse (TG), while the bone marrow and the trabecular bone seem similar to the wildtype (WT) mouse. The volume of the trabecular bone was assessed by histomorphometric analysis of the distal femurs at regions indicated by the arrows. Bar: 1 mm.

Download figure to PowerPoint

Immunohistochemical analysis of the femurs of 2-month-old wild type mice revealed endogenous type XIII collagen expression in periosteal and endosteal osteoblasts (Fig. 7A). The long bones of age-matched transgenic mice exhibited thickening of the periosteum, with several layers of cells (Fig. 7B), whereas only one cell layer of bone-lining osteoblasts was detected in the controls (Fig. 7A). The periosteal osteoblasts were shown by electron microscopy to be larger, with an enlarged endoplasmic reticulum, suggesting increased activity, and they also had more protrusions toward the bone matrix (data not shown). The endosteal cell layer appeared similar both in transgenic (Fig. 7C) and wildtype bone (Fig. 7A). Both the periosteal and endosteal cells expressed the transgene (Figs. 7B and 7C). Moreover, hard tissue sections revealed a large amount of osteoid, indicating increased bone formation combined with an increased mineral apposition rate (data not shown).

thumbnail image

Figure FIG. 7.. High transgene expression in periosteal osteoblasts. (A-C) Immunohistochemical analysis of (A) wildtype and (B and C) transgenic femurs with type XIII collagen and HA-tag antibodies. (A) Type XIII collagen is endogenously expressed both in periosteal (arrows) and endosteal (arrowheads) osteoblasts. (B) The periosteal layer of osteoblasts is markedly thicker in the mutant femur than in that of a wildtype littermate and shows strong transgene expression. High transgene expression was also detected in endosteal osteoblasts of mutant femur (arrowheads in C). Bars: 100 (A and C) and 25 μm (B).

Download figure to PowerPoint

High bone mass is caused by increased bone formation rather than decreased bone resorption

Because histological analysis suggested increased bone deposition, we measured the bone formation rate using tetracycline double-labeling. The 3-month-old overexpression and control mice received injections of tetracycline twice, 10 days and again 3 days before death, so that the distance between the two labels could be taken to represent the extent of bone formation in 7 days (Fig. 8). This showed bone formation and the mineralization rate to be several times higher in the overexpression mice than in their wildtype littermates.

thumbnail image

Figure FIG. 8.. Osteoblast activity and increased bone formation rate in transgenic mice. (A-D) Tetracycline double-labeling of 3-month-old (A) wildtype (n = 5) and (B) transgenic (n = 6) mice. The distance between the two labels (bar) represents the rate of bone formation in 7 days. Both (C) endosteal and (D) periosteal mineral deposition were calculated. Bone formation is clearly increased in the transgenic mouse (TG) relative to its wildtype (WT) littermate.

Download figure to PowerPoint

To find out whether impaired bone resorption contributed to the accumulation of bone in the overexpression mice, we performed TRACP stainings to detect osteoclasts in the distal metaphyseal areas of the mutant and control femurs. The overexpression mice had 7.0 ± 0.7 osteoclasts/mm of trabecular bone surface and the controls had 7.7 ± 1.6 osteoclasts/mm (p > 0.05). Thus no significant differences in the numbers of osteoclasts could be detected between the overexpression mice and their wildtype littermates at the virtually normal metaphyses. However, high numbers of osteoclasts were also seen locally in the diaphyseal area of the transgenic bone compared with the bones of their wildtype littermates (data not shown), indicating increased resorption activity and modeling in the cortical areas where the phenotype with high bone formation was most evident. The resorption activity of osteoclasts was analyzed by culturing isolated osteoclasts on bovine bone slices. Osteoclasts and the areas of resorption pits were calculated, and the resorption activity was expressed as the resorbed bone area/osteoclast (μm2), being 498.9 ± 266.4 (SD) for wildtype and 493.6 ± 381.9 for transgenic (p > 0.05). The number of osteoclasts, their activity shown by formation of actin rings, and resorption activity in the culture model were not significantly changed between the overexpression mice and their wildtype littermates (Table 2).

Table Table 2.. Osteoclast Resorption Activity in Transgenic Mice Overexpressing Type XIII Collagen
Thumbnail image of

Some parameters of bone metabolism were also analyzed in serum samples from 6-week-old overexpression and control mice. Serum alkaline phosphatase (S-ALP) activity was significantly higher in the overexpression mice, indicating increased osteoblast activity and bone formation, but serum calcium (S-Ca) and phosphate (S-Pi) levels did not differ. Furthermore, PTH, which is secreted by the parathyroid glands and is known to increase serum calcium levels by accelerating bone resorption, did not differ between overexpression and normal littermates (data not shown).


  1. Top of page
  2. Abstract
  7. Acknowledgements

To explore the function of type XIII collagen, we generated transgenic mice overexpressing wildtype and missense mutant α1(XIII) chains. High transgene expression was detected in bone, cartilage, and skin, all tissues known to synthesize this collagen endogenously.(4, 5) Surprisingly, the overexpression of α1(XIII) chains led to an enormous increase in bone mass in both types of mouse line. The abnormally high bone mass was most conspicuous in long bones, but it was also found in calvarial bones. Interestingly, the mice were indistinguishable from their littermates at birth, although high transgene expression was already detectable in the developing bone and cartilage. Thus, the overexpression of type XIII collagen allows normal skeletal development, but the changes in long bones, most readily noticeable in the femur, begin to appear in a rapidly progressive manner at the age of 3–4 weeks. The changes in the calvarial bones develop more slowly. The symmetrical increase in cortical bone throughout the skeleton, together with the histological analysis, suggests benign bone overgrowth rather than malignant transformation of the bone cells. Thus, the data suggest that type XIII collagen affects bone modeling processes.

Analysis of bone tissue indicated that the transgene-derived α1(XIII) chains formed disulfide-bonded trimers in the manner previously reported for endogenously produced and recombinant type XIII collagen.(7, 11) This is a prerequisite for triple helix formation, which is consequently likely to occur. We also compared the tissue expression pattern of the transgenic type XIII collagen with that of endogenous molecules using antibodies against type XIII collagen and against the HA-tag located at the extreme C-terminal end of the transgene product. The transgene products followed the staining pattern of endogenous type XIII collagen closely in the developing cartilage and bone. In adult bone, type XIII collagen is found in periosteum in mice and humans.(4) Moreover, recent analysis of a type XIII collagen knock-in mouse line where the lacZ gene was placed under control of the type XIII collagen promoter has indicated that type XIII collagen is strongly expressed in developing and adult bone, more specifically in periosteal osteoblasts (A Latvanlehto, A Koski, R Sormunen, M Ilves, J Tuukkanen, A-P Kvist, and T Pihlajaniemi, unpublished results, 2004). Thus the 1-kb promoter and 5′ flanking sequences derived from the mouse Col13a1 gene and used here to drive transgene expression are capable of directing synthesis correctly in bone and cartilage. In summary, Western blotting and immunofluorescence staining of tissues suggested that the transgene type XIII collagen is properly folded and located in a number of tissues. Thus, the observed phenotype reflects overexpression of type XIII collagen in the skeletal tissues, where this collagen also occurs normally.

Abnormalities in bone modeling may lead to increased bone mass, as described for osteopetrosis, which is caused by a failure of osteoclast formation or function resulting in increased cancellous bone formation in the medullary cavity, and osteosclerosis, which is caused by increased osteoblast activity.(28) Furthermore, defects in chondrocyte formation and differentiation can alter the growth plate, leading to chondrodysplasias.(29, 30) The present overexpression mice showed a striking increase in the amount of cortical bone, but the bone marrow was normally formed (Fig. 6). The bone formation rate, as detected by tetracycline double-labeling, was clearly increased. Osteoclasts could be detected in the trabecular bone regions of the mutants at amounts comparable with wildtype mice, but their amount was increased in the cortical bone areas (data not shown), suggesting intracortical bone modeling. We also analyzed the resorption activity of cultured osteoclasts from transgenic and control mice, but found no difference in the osteoclast resorption activity (Table 2), so that impaired osteoclast function and an osteopetrotic phenotype could be excluded. Osteoblasts have an important role in osteoclastogenesis. RANKL expressed on the surfaces of osteoblasts interacts with its receptor RANK on osteoclasts and their precursors, resulting in fusion of pre-osteoclasts into multinucleated bone-resorbing cells.(31) Microarray analysis from femurs of 1-month-old wildtype and transgenic mice did not reveal any differences in the expression level of RANKL or RANK between these two (R Ylönen, unpublished data, 2005), suggesting that osteoblasts overexpressing type XIII collagen are capable of supporting normal osteoclastogenesis. Furthermore, the transgenic mice did not have clear cartilage defects, and the growth plates were essentially intact throughout development. Thus, the phenotype observed in the type XIII collagen overexpression mice resembles osteosclerosis.

Osteosclerotic phenotypes have been described in mice overexpressing the fos proteins Fra-1(32) and ΔFosB,(33) and lack of the osteoblast-specific marker osteocalcin also results in increased cortical bone thickness.(34) Other mouse models with high bone mass include leptin or leptin receptor-deficient mice.(35) It should be noted that the increase in bone mass in the type XIII collagen overexpression mice was much more pronounced than in the other osteosclerotic models. Furthermore, whereas there was a marked increase in cortical bone, trabecular bone was not affected, which differs from the situation in the known osteosclerotic phenotypes. The long bones increase in width because of the deposition of new matrix by periosteal osteoblasts.(36) The normal numbers of osteoclasts and the increased bone formation detected by tetracycline double-labeling suggest that the type XIII collagen overexpression phenotype is caused by altered osteoblast function or proliferation. Both endosteal and periosteal osteoblasts showed strong transgene expression, but only the periosteal zones were clearly thicker in the transgenic bone (Fig. 7). The strong effect on cortical bone may be explained by higher sensitivity of periosteal cells to type XIII collagen, leading to enhanced function of those cells.

It has been shown that mechanical loading increases bone formation and that immobilization can cause osteoporosis because of decreased bone formation.(37–39) The factors mediating these processes are still unknown. Recent data suggest that osteocytes may have a role as primary targets in this process.(40) Osteocytes form a network in the bone matrix, communicating with each other and with the bone surface osteoblasts by means of gap junctions. Mechanical stimulation of osteocytes causes them to secrete signals such as prostaglandins and NO.(41) The appearance of increased bone formation in the long bones of the transgenic mice overexpressing type XIII collagen at the age of 3–4 weeks coincides with the age when the mice start to move about more actively and their long bones are exposed to increasing mechanical stress. It is clear that the observed 100- to 1000-fold increase in bone inertia moment, indicating resistance to mechanical loading, is not under control. The results are suggestive of a role for type XIII collagen as a mediator of mechanical stress, so that its overexpression may affect its interaction with other matrix components or interfere with certain critical signaling events. We plan to study this further in a cell culture system where primary osteoblasts are exposed to mechanical strain.

A central regulator of osteoblast differentiation and function is the osteoblast-specific transcription factor Runx2.(28) Lack of this in mice leads to an absence of osteoblasts and bone.(42, 43) It has been shown that, in addition to its role in osteoblast differentiation and skeletal development, Runx2 is also expressed in osteoblasts postnatally(44) and that it regulates bone matrix deposition by differentiated osteoblasts after birth.(45) The present quantitative RT-PCR showed that Runx2 expression was increased in the femurs of the mice overexpressing type XIII collagen. Cbfa-1 sites are known to occur in genes expressed in osteoblasts,(44, 46, 47) and interestingly, computer analysis revealed that the type XIII collagen promoter has five possible binding sites for this transcription factor, one of which is included in the promoter sequence of the transgene constructs. We suggest that type XIII collagen may have a role in bone formation by differentiated osteoblasts after birth, through interaction with other osteoblast-specific bone markers such as Runx2.

A number of cytokines, growth factors, hormones, and transcription factors are known to regulate bone formation and modeling, the two most abundant growth factors found in human bone being TGF and IGF-II.(48) Interestingly, we detected upregulation of IGF-II RNA in transgenic femurs compared with controls. Also, antibody staining of tissues revealed stronger IGF-II signals in the mutant osteoblasts than in the wildtype ones (data not shown). In view of the markedly increased number of osteoblasts in the overexpression mice, it is possible that the increased IGF-II expression contributes to osteoblasts being released from their normal growth control functions, leading to increased activity of those in a periosteal position.

We were surprised to find that strong bone formation in the mice overexpressing type XIII collagen was not associated with increases in mRNA encoding type I collagen, the major organic constituent of bone, or the cartilage collagen types II, X, and XI. We thus consider it likely that the high bone mass develops because there are more osteoblasts and osteocytes, as witnessed by the expanded osteoblast zones and the thick cortical bone, and thus more cells capable of producing matrix components in the mutant mice. Moreover, IGF-II has been reported to induce bone formation in rats in vivo,(49) so that its upregulation may enhance matrix deposition and thereby lead to the high bone mass phenotype.

All in all, type XIII collagen is an authentic component of developing and adult bone, and the studies presented here suggest that changes in its expression can greatly affect bone biology. The skeletal phenotype of the mice overexpressing type XIII collagen is distinct from any other phenotype previously described and is caused by increased bone formation rather than decreased bone resorption. RNA analysis of transgenic mice suggested that the expression of this collagen type may be linked to IGF-II and Runx2, the possible effects of which on type XIII collagen function will be further studied in the future. Osteoporosis and bone fracturing are worldwide health problems and objects of ongoing research and development work aimed at providing better care. Because overexpression of type XIII collagen leads to increased bone mass with high BMD, it could be considered a new target molecule with therapeutic potential.


  1. Top of page
  2. Abstract
  7. Acknowledgements

The authors thank Maija Seppänen, Jaana Väisänen, Sirpa Kellokumpu, and Minna Vanhala for expert technical assistance, Docent Raija Sormunen for expertise in electron microscopy, Professor Eero Vuorio for histological analysis of cartilage, and Professor Irma Thesleff for critical reading of the manuscript and valuable comments. This work was supported by grants from the Academy of Finland (Centre of Excellence programme 44843, 202873), EU programme QLK3-CT-2000–00084, and the Sigrid Jusélius Foundation.


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • 1
    Pihlajaniemi T, Rehn M 1995 Two new collagen subgroups: Membrane associated collagens and types XV and XVIII. Prog Nucleic Acid Res Mol Biol 50: 225262.
  • 2
    Hashimoto T, Wakabayashi T, Watanabe A, Kowa H, Hosoda R, Nakamura A, Kanazawa I, Arai T, Takio K, Mann DM, Iwatsubo T 2002 CLAC: A novel Alzheimer amyloid plaque component derived from a transmembrane precursor, CLAC-P/collagen type XXV. EMBO J 21: 15241534.
  • 3
    Banyard J, Bao L, Zetter BR 2003 Type XXIII collagen, a new transmembrane collagen identified in metastatic tumor cells. J Biol Chem 278: 2098920994.
  • 4
    Sandberg M, Tamminen M, Hirvonen H, Vuorio E, Pihlajaniemi T 1989 Expression of mRNAs coding for the α1 chain of type XIII collagen in human fetal tissues: comparison with expression of mRNAs for collagen types I, II, and III. J Cell Biol 109: 13711379.
  • 5
    Sund M, Väisänen T, Kaukinen S, Ilves M, Tu H, Autio-Harmainen H, Rauvala H, Pihlajaniemi T 2001 Distinct expression of type XIII collagen in neuronal structures and other tissues during mouse development. Matrix Biol 20: 215231.
  • 6
    Peltonen S, Hentula M, Hägg P, Ylä-Outinen H, Tuukkanen J, Lakkakorpi J, Rehn M, Pihlajaniemi T, Peltonen J 1999 A novel component of epidermal cell-matrix and cell-cell contacts: transmembrane protein type XIII collagen. J Invest Dermatol 113: 635642.
  • 7
    Hägg P, Väisänen T, Tuomisto A, Rehn M, Tu H, Huhtala P, Eskelinen S, Pihlajaniemi T 2001 Type XIII collagen: a novel cell adhesion component present in a range of cell-matrix adhesions and in the intercalated discs between cardiac muscle cells. Matrix Biol 19: 727742.
  • 8
    Nykvist P, Tu H, Ivaska J, Käpylä J, Pihlajaniemi T, Heino J 2000 Distinct recognition of collagen subtypes by alpha(1)beta(1) and alpha(2)beta(1) integrins. Alpha(1)beta(1) mediates cell adhesion to type XIII collagen. J Biol Chem 275: 82558261.
  • 9
    Snellman A, Tu H, Väisänen T, Kvist AP, Huhtala P, Pihlajaniemi T 2000 A short sequence in the N-terminal region is required for the trimerization of type XIII collagen and is conserved in other collagenous transmembrane proteins. EMBO J 19: 50515059.
  • 10
    Tu H, Sasaki T, Snellman A, Gohring W, Pirilä P, Timpl R, Pihlajaniemi T 2002 The type XIII collagen ectodomain is a 150-nm rod and capable of binding to fibronectin, nidogen-2, perlecan, and heparin. J Biol Chem 277: 2309223099.
  • 11
    Snellman A, Keränen MR, Hägg PO, Lamberg A, Hiltunen JK, Kivirikko KI, Pihlajaniemi T 2000 Type XIII collagen forms homotrimers with three triple helical collagenous domains and its association into disulfide-bonded trimers is enhanced by prolyl 4-hydroxylase. J Biol Chem 275: 89368944.
  • 12
    Juvonen M, Pihlajaniemi T 1992 Characterization of the spectrum of alternative splicing of α1(XIII) collagen transcripts in HT-1080 cells and calvarial tissue resulted in identification of two previously unidentified alternatively spliced sequences, one previously unidentified exon, and nine new mRNA variants. J Biol Chem 267: 2469324699.
  • 13
    Juvonen M, Sandberg M, Pihlajaniemi T 1992 Patterns of expression of the six alternatively spliced exons affecting the structures of the COL1 and NC2 domains of the α1(XIII) collagen chain in human tissues and cell lines. J Biol Chem 267: 2470024707.
  • 14
    Juvonen M, Pihlajaniemi T, Autio-Harmainen H 1993 Location and alternative splicing of type XIII collagen RNA in the early human placenta. Lab Invest 69: 541551.
  • 15
    Peltonen S, Rehn M, Pihlajaniemi T 1997 Alternative splicing of mouse α1(XIII) collagen RNAs results in at least 17 different transcripts, predicting α1(XIII) collagen chains with length varying between 651 and 710 amino acid residues. DNA Cell Biol 16: 227234.
  • 16
    Kvist AP, Latvanlehto A, Sund M, Eklund L, Väisänen T, Hägg P, Sormunen R, Komulainen J, Fässler R, Pihlajaniemi T 2001 Lack of cytosolic and transmembrane domains of type XIII collagen results in progressive myopathy. Am J Pathol 159: 15811592.
  • 17
    Sund M, Ylönen R, Tuomisto A, Sormunen R, Tahkola J, Kvist AP, Kontusaari S, Autio-Harmainen H, Pihlajaniemi T 2001 Abnormal adherence junctions in the heart and reduced angiogenesis in transgenic mice overexpressing mutant type XIII collagen. EMBO J 20: 51535164.
  • 18
    Hägg P, Rehn M, Huhtala P, Väisänen T, Tamminen M, Pihlajaniemi T 1998 Type XIII collagen is identified as a plasma membrane protein. J Biol Chem 273: 1559015597.
  • 19
    Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate- phenol-chloroform extraction. Anal Biochem 162: 156159.
  • 20
    McLeod MJ 1980 Differential staining of cartilage and bone in whole mouse fetuses by alcian blue and alizarin red S. Teratology 22: 299301.
  • 21
    Selby PB 1987 A rapid method for preparing high quality alizarin stained skeletons of adult mice. Stain Technol 62: 143146.
  • 22
    Jämsä T, Jalovaara P, Peng Z, Väänänen HK, Tuukkanen J 1998 Comparison of three-point bending test and peripheral quantitative computed tomography analysis in the evaluation of the strength of mouse femur and tibia. Bone 23: 155161.
  • 23
    Lakkakorpi P, Tuukkanen J, Hentunen T, Järvelin K, Väänänen K 1989 Organization of osteoclasts microfilaments during the attachment to bone surface in vitro. J Bone Miner Res 4: 817825.
  • 24
    Selander K, Lehenkari P, Väänänen HK 1994 The effects of biphosphonates on the resorption cycle of isolated osteoclasts. Calcif Tissue Int 37: 646650.
  • 25
    Myllyharju J, Kivirikko KI 2001 Collagens and collagen-related diseases. Ann Med 33: 721.
  • 26
    Prockop DJ, Kivirikko KI 1995 Collagens: Molecular biology, diseases, and potentials for therapy. Annu Rev Biochem 64: 403434.
  • 27
    Kvist AP, Latvanlehto A, Sund M, Horelli-Kuitunen N, Rehn M, Palotie A, Beier D, Pihlajaniemi T 1999 Complete exon-intron organization and chromosomal location of the gene for mouse type XIII collagen (col13a1) and comparison with its human homologue. Matrix Biol 18: 261274.
  • 28
    Karsenty G 1999 The genetic transformation of bone biology. Genes Dev 13: 30373051.
  • 29
    Mundlos S, Olsen BR 1997 Heritable diseases of the skeleton. Part I: Molecular insights into skeletal development-transcription factors and signaling pathways. FASEB J 11: 125132.
  • 30
    Mundlos S, Olsen BR 1997 Heritable diseases of the skeleton. Part II: Molecular insights into skeletal development-matrix components and their homeostasis. FASEB J 11: 227233.
  • 31
    Chambers TJ 2000 Regulation of the differentation and function of osteoclasts. J Pathol 192: 413.
  • 32
    Jochum W, David JP, Elliott C, Wutz A, Plenk H Jr, Matsuo K, Wagner EF 2000 Increased bone formation and osteosclerosis in mice overexpressing the transcription factor Fra-1. Nat Med 6: 980984.
  • 33
    Sabatakos G, Sims NA, Chen J, Aoki K, Kelz MB, Amling M, Bouali Y, Mukhopadhyay K, Ford K, Nestler EJ, Baron R 2000 Overexpression of DeltaFosB transcription factor(s) increases bone formation and inhibits adipogenesis. Nat Med 6: 985990.
  • 34
    Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, Smith E, Bonadio J, Goldstein S, Gundberg C, Bradley A, Karsenty G 1996 Increased bone formation in osteocalcin-deficient mice. Nature 382: 448452.
  • 35
    Ducy P, Amling M, Takeda S, Priemel M, Schilling AF, Beil FT, Shen J, Vinson C, Rueger JM, Karsenty G 2000 Leptin inhibits bone formation through a hypothalamic relay: A central control of bone mass. Cell 100: 197207.
  • 36
    Erlebacher A, Filvaroff EH, Gitelman SE, Derynck R 1995 Toward a molecular understanding of skeletal development. Cell 80: 371378.
  • 37
    Weinreb M, Rodan GA, Thompson DD 1989 Osteopenia in the immobilized rat hind limb is associated with increased bone resorption and decreased bone formation. Bone 10: 187194.
  • 38
    Hillam RA, Skerry TM 1995 Inhibition of bone resorption and stimulation of formation by mechanical loading of the modeling rat ulna in vivo. J Bone Miner Res 10: 683689.
  • 39
    Jämsä T, Tuukkanen J, Jalovaara P 1998 Femoral neck strength of mouse in two loading configurations: Method evaluation and fracture characteristics. J Biomech 31: 723729.
  • 40
    Mullender MG, Huiskes R 1997 Osteocytes and bone lining cells: Which are the best candidates for mechano-sensors in cancellous bone? Bone 20: 527532.
  • 41
    Chow JW, Fox SW, Lean JM, Chambers TJ 1998 Role of nitric oxide and prostaglandins in mechanically induced bone formation. J Bone Miner Res 13: 10391044.
  • 42
    Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K, Shimizu Y, Bronson RT, Gao YH, Inada M, Sato M, Okamoto R, Kitamura Y, Yoshiki S, Kishimoto T 1997 Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell 89: 755764.
  • 43
    Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell IR, Stamp GW, Beddington RS, Mundlos S, Olsen BR, Selby PB, Owen MJ 1997 Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell 89: 765771.
  • 44
    Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G 1997 Osf2/Cbfa1: A transcriptional activator of osteoblast differentiation. Cell 89: 747754.
  • 45
    Ducy P, Starbuck M, Priemel M, Shen J, Pinero G, Geoffroy V, Amling M, Karsenty G 1999 A Cbfa1-dependent genetic pathway controls bone formation beyond embryonic development. Genes Dev 13: 10251036.
  • 46
    Harada H, Tagashira S, Fujiwara M, Ogawa S, Katsumata T, Yamaguchi A, Komori T, Nagatsuka M 1999 Cbfa1 isoforms exert functional differences in osteoblast differentiation. J Biol Chem 274: 69726978.
  • 47
    Jimenez MJ, Balbin M, Lopez JM, Alvarez J, Komori T, Lopez-Otin C 1999 Collagenase 3 is a target of Cbfa1, a transcription factor of the runt gene family involved in bone formation. Mol Cell Biol 19: 44314442.
  • 48
    Strong DD, Beachler AL, Wergedal JE, Linkhart TA 1991 Insulinlike growth factor II and transforming growth factor beta regulate collagen expression in human osteoblast-like cells in vitro. J Bone Miner Res 6: 1523.
  • 49
    Ishibe M, Ishibashi T, Kaneda K, Koda T, Rosier RN, Puzas JE 1998 Stimulation of bone formation in vivo by insulin-like growth factor-II in rats. Calcif Tissue Int 63: 3638.