| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |



From the Section of Orthodontics,*
UCLA School of
Dentistry, Los Angeles, California; the Departments of Restorative
Dentistry
and
Orthodontics,
Harvard School of Dental
Medicine, Boston, Massachusetts; the Department of Cell
Biology,§
Harvard Medical School, Boston,
Massachusetts; and the Jane and Jerry Weintraub Center for
Reconstructive Biotechnology,¶
Division of
Advanced Prosthodontics, Biomaterials and Hospital Dentistry, UCLA
School of Dentistry, Los Angeles, California
| Abstract |
|---|
|
|
|---|
1(IX) collagen
(Col9a1) in alveolar bone wound healing, while the long isoform
of Col9a1 transcript was absent. Type IX collagen was immunolocalized
in the preliminary matrix organized in granulation tissue before
trabecular bone formation in tooth extraction socket. In Col9a1-null
mutant mice, there were considerable variations in alveolar
bone wound healing with the absence of or abnormally organized
trabecular bone. Occasionally, unusual apposition of
cortical-bone-like layers in bone marrow space was observed. The
Col9a1-null mice indicated no growth retardation, and their
facial and long bones maintained the normal size and shape.
However, the primary spongiosa region of adult Col9a1 mutant
mice showed an abnormal trabecular bone structure associated with
abnormal immunostaining with the hypertrophic cartilage specific type X
collagen antibody. These data suggest that type IX collagen short
transcriptional variant is involved in the restoration and remodeling
processes of trabecular bone.
| Introduction |
|---|
|
|
|---|
Wounds created in the bone marrow space by bone ablation in the long bone3 and by tooth extraction in the alveolar bone4,5 undergo a sequential healing process that restores trabecular bone structure without cartilage precursor tissue formation. Histological and ultrastructural studies report the formation of a preliminary matrix composed of thin collagen fibers that are synthesized during the initial stages of wound healing before trabecular bone restoration.6,7 The preliminary matrix is formed within the granulation tissue and connects the neighboring bone and soft tissues. Disturbance of the preliminary matrix results in the lack of trabecular bone restoration. However, the composition and functional role of this initial preliminary matrix have not yet been fully elucidated.
Studies of the rat tooth extraction model has suggested the temporal
expression of cartilage collagen types II and IX at the steady-state
transcriptional level during the early stages of wound
healing.4,5
The
1(IX) collagen (Col9a1) gene has two
independent promoter/transcriptional start sites that are used to
synthesize long and short isoforms8
in a tissue-specific
fashion, including osteoprogenitor cells.9
In this study,
we focused on type IX collagen as a candidate component of the
preliminary matrix appearing before trabecular bone restoration. We
demonstrated that the short type IX collagen isoform was likely to be
involved in the preliminary matrix during trabecular bone restoration,
and that the null mutation of the Col9a1 alleles in mice significantly
disturbed the restoration and remodeling of trabecular bone with
minimal effects on the cortical bone.
| Materials and Methods |
|---|
|
|
|---|
A polymerase chain reaction (PCR) product spanning the entire intron sequence between exons 6 and 7 of Col9a1 was amplified from rat genomic DNA using two flanking primers derived from the coding sequences of exon 6 (5'-CUA CUA CUA CUA CGG AGA GAG ACG TGC-3') and exon 7 (5'-CAU CAU CAU CAU CTC ATC GGT GGT CTG-3'). A PCR product was cloned and sequenced.
The downstream transcriptional start site was characterized using a combination of the RNase protection assay and the primer extension assay. The isolated PCR fragment of genomic DNA (HR1) containing tentative exon 1* was used to generate anti-sense and sense riboprobes. An extension primer was designed to encode the reverse and complementary sequence of the 3' end of the alternative exon 1* (5'-TTG AGC AGC ACA TAA CG-3'). RNase protection and primer extension experiments were performed using poly(A)+ RNA isolated from rat calvarial osteoprogenitor cells, which contains only short isoform of Col9a1.4,9 The protected riboprobe and the primer extended products were analyzed on a 5% sequence gel. The downstream transcriptional start site in the HR1 sequence was identified as simultaneously indicated by RNase protection and primer extension assays.
Expression of Col9a1 Transcripts during Alveolar Bone Wound Healing
Forty-day-old Sprague-Dawley rats were anesthetized by an intramuscular injection of ketamine (1.5 mg/100 g body weight) and rompun (0.824 mg/100 g body weight). Maxillary molars were extracted with a dental explorer under a dissection microscope.10 The alveolar bones containing the healing extraction socket were harvested before tooth extraction and at the 4-day and 7-day postextraction periods, and total RNA for each sample was prepared separately. A total RNA sample was also isolated from adult rat epiphyseal tissues containing hyaline cartilage and subarticular bone.
Long and short transcripts of the Col9a1 gene were identified by reverse transcriptase PCR using the combinations of a common 3' primer (5'-CCG GAA CTC CAG GAG GC-3') and one of the following specific 5' primers: 5'-GTA GAC TTC AGG ATT CCA-3' for the long Col9a1 transcript or 5'-TTG CAA CCA CTA CCC TG-3' for the short Col9a1 transcript (based on the rat alternative exon 1* sequence determined in the present study). The PCR products representing the long (583 bp) and the short (261 bp) Col9a1 transcripts were identified by Southern blotting using a radiolabeled oligonucleotide probe 5'-AGA GGT CCT CCG GGT GAG CAG GGG-3'. For a housekeeping control gene, GAPDH primers were used in a PCR experiment.
In Situ Hybridization
The healing maxillary alveolar bone tissues were harvested at day 4 postextraction and immediately fixed in 4% paraformaldehyde in phosphate buffered saline (PBS), pH 7.4, for 2 days at 4°C. The specimens were treated for conventional paraffin embedding after decalcification. Some histological sections stained with hematoxylin and eosin (H&E) were examined for wound healing and bone formation. For in situ hybridization, the optimal method for the rat alveolar bone tissue has been previously established.4 An 800-bp cDNA encoding rat Col9a1, EKT101, was radiolabeled with 35S dATP by nick translation and served as probe. After counterstaining with Harris alum-hematoxylin, in situ hybridization sections were examined under a light microscope.
Immunohistology
Eight-micron sagittal sections of the 4-day postextraction specimen were cut and mounted on silane-prep slides. Deparaffinized sections were treated with 10 mg/ml hyaluronidase in 0.1 mol/L sodium acetate buffer at 37°C for 1 hour. This was followed by an incubation in 1% hydrogen peroxide-methanol at room temperature for 30 minutes to block endogenous peroxide. Sections were treated with Tris-buffered saline (TBS) and then blocked with 10% normal goat serum in TBS at room temperature for 30 minutes in a humid chamber. Monoclonal antibodies (mAbs) B31 and D19 were applied and incubated at 4°C for 2 hours in a humid chamber. mAbs B31 and D19 were generated against the bovine low molecular weight pepsin-digested triple helical fragment containing Col9a1, Col9a2, and Col9a3 (kindly provided by Dr. Ye, University of Tennessee at Memphis and VA Medical Center, Memphis, TN). B31 and D19 cross-react with type IX collagen of different species including chicken, bovine, rat, and human.11 The slides were rinsed with TBS, and secondary antibody was applied. A standard avidin-biotin complex/immunoperoxidase protocol was used to visualize the immunolocalization of rat type IX collagen.
Genotyping of Mice with Inactivated
1(IX) Alleles
Type IX collagen knockout mice have been generated and described.12 The neo gene was inserted into exon 8 of the Col9a1 gene by homologous recombination, resulting in inactivation of the Col9a1 alleles. Homozygous knockout mice have lost the ability to synthesize both the short and long forms of the Col9a1 chain. The skeletal development and gross abnormality of mice were evaluated by lateral radiographs.
Genotype was established using 1 cm of tail tendon harvested from each offspring (at about 3 weeks of age) of heterozygous mutant parents. The tail tendon was transferred into a 700-µl lysis mix containing 10 mg/ml proteinase K in 100 mmol/L Tris-HCl, pH 8.5, 5 mmol/L EDTA, 0.2% sodium dodecyl sulfate, and 200 mmol/L NaCl at 55°C for 6 hours. Supernatant was collected after 30 minutes centrifugation and transferred to 700 µl of ice-cold isopropanol. Precipitated DNA was removed and suspended in 10 mmol/L Tris, pH 7.5, 0.1 mmol/L EDTA. Thirty-microliter samples were digested with EcoRI overnight. Genomic fragments were separated by electrophoresis run at 30 V for 6 hours. Southern blot analysis was performed with a radiolabeled DNA probe located in Col9a1 between the HindIII and AbaI sites.
Histological and Immunohistological Examinations of Trabecular Bone Restoration and Remodeling
Littermates of 4- to 5-week-old wild-type mice (n = 10) and homozygous Col9a1-null mutant mice (n = 10) were used. Animals received an intramuscular injection of ketamine (1.5 mg/100 g body weight) and rompun (0.824 mg/100 g body weight) for anesthesia, and unilateral maxillary molars were extracted with a dental explorer. Animals were sacrificed at 7 and 14 days after extraction. Maxillae were harvested and prepared for histological evaluation. Coronal and sagittal sections were cut. In coronal sections, maxillary tissues were oriented to provide standardized sections perpendicular to the midpalatal suture mesial to the remaining second molar. Sagittal sections were made along the midpalatal suture. Sections were stained with H&E and examined under a light microscope.
Mandibular condyles of 8-week-old homozygous Col9a1-null mutants and normal mice were selected for examination of trabecular bone remodeling. The histological sections were prepared as above, stained with Alcian blue, and counterstained with H&E. The mandibular condyles were further examined immunohistochemically with anti-type X collagen antibody.13 Harvested condyles were embedded in OCT media and frozen at -80°C. Ten-micron sections were prepared in a cryostat and subjected to an immunohistological protocol as described above.
| Results |
|---|
|
|
|---|
|
|
|
Immunohistochemistry experiments with anti-type IX collagen antibody
indicated that type IX collagen was a component of the preliminary
matrix (Figure 3)
. Within this amorphous region of the preliminary
matrix, type IX collagen immunostaining resembled the trabecular bone
structure and the immunostaining-negative areas were associated with
small blood vessels. The Type IX collagen immunostaining was also found
over the new trabecular bone and on the surface of newly formed
trabecular bone matrix. However, in old bone and cortical bone, type IX
collagen immunostaining was not observed (data not shown).
Col9a1-null mice have been generated by homologous recombination of a
neo gene inserted into exon 8 (Figure 4A)
, resulting in the lack of expression
of both the long and short Col9a1 mRNAs.12
The
Col9a1-null mice exhibit minimal developmental and pathological
phenotypes, and hyaline cartilage was developed and maintained
normally. Wild-type (+/+) and homozygous Col9a1-null (-/-) mice, as
determined by genomic Southern blot analyses (Figure 4B)
, indicated
similar skeletal features on radiographs (Figure 4C)
. The size and
shape of bones including the mandible did not show significant
difference between the mutants and controls.
|
|
| Discussion |
|---|
|
|
|---|
Examination of the rat tooth extraction model undergoing bone wound healing has suggested expression of type II collagen and short type IX collagen transcripts without formation of cartilage tissue. The lack of the NC4 domain in the short type IX collagen molecule has been postulated to prevent, in part, the formation of cartilage tissue in the extraction socket.4 However, because type IX collagen is not a matrix component of mature bone of either endochondral or intramembranous origin,22-25 the role of short type IX collagen in bone wound healing has not been established. This study further characterized the expression pattern of the short Col9a1 collagen transcript during the early bone remodeling stages. A combination of in situ hybridization and immunohistological studies indicated the strong association between type IX collagen and the relatively unorganized amorphous fibrous matrix in the tooth extraction socket. The histological characteristics of type IX collagen-containing matrix were unlike either bone or cartilage; rather, they resembled the description of preliminary matrix that has been reported during bone marrow ablation wound healing as a prerequisite for intramembranous trabecular bone restoration.7 The nature of the preliminary matrix has not been well understood. This study demonstrated, for the first time, that type IX collagen was a protein component of the preliminary matrix. The monoclonal antibodies used in this study recognize the cognitive epitopes of the type IX collagen heterotrimer structure containing Col9a1, Col9a2, and Col9a3. Because type IX collagen is found covalently associated with type II collagen fibril in hyaline cartilage,26,27 type IX collagen in the preliminary matrix may also form a composite collagen fibril with type II collagen. The missing amino-terminal globular domain from the short type IX collagen molecule due to Col9a1 alternative transcription may determine the unique surface topography and chemistry of the preliminary matrix. It is tempting to speculate that the preliminary matrix may contribute to the osteoblast homing through the type II/IX collagen composite fibril.
Mice carrying abnormal Col9a1 chains exhibit early onset of osteoarthritic cartilage degradation.28,29 Immortalized chondrocytes lacking the Col9a1 chain synthesize thicker, or fused, type II collagen fibrils in vitro.30 However, investigation of Col9a1-null mutant mice has shown only minimal pathological phenotypes in cartilage tissues.12 The young adult Col9a1-null mutant mice (48 weeks old) did not show observable growth retardation or abnormalities in skeletal size and shape, including craniofacial bones. In the present study, Col9a1-null mutant mice were combined with the tooth extraction model to address the role of type IX collagen in bone wound healing. Unlike the consistent trabecular bone formation in the wild-type mice, the mutant tooth extraction socket showed various healing patterns. Trabecular bone restoration was often disturbed, and, in some cases, unusual new cortical-bone-like formation took place in the tooth extraction socket. Due to alternative mRNA processing, the primary structure of the short Col9a1 chain resembles that of the Col9a2 and Col9a3 chains.8,31 It is conceivable that the missing short Col9a1 chain can be effectively substituted with the Col9a2 or Col9a3 chain. Although this potential compensation phenomenon must be addressed, our data support that the Col9a1-null mutation primarily affects the sound restoration of trabecular bone.
Mandibular condyles of 8-week-old homozygous Col9a1-null mutants and normal mice were selected for examination of trabecular bone remodeling. Mandibular condyle is initially organized as a mass of secondary cartilage, which lacks an active proliferation. Chondrocyte hypertrophy occurs at around the birth in rats and mice, and type X collagen is synthesized similarly to primary cartilage.32 The formation of trabecular bone rapidly replaces hypertrophic cartilage, which becomes only several cell layers by postnatal day 14. Type X collagen is an exclusive product of hypertrophic chondrocytes.33 Therefore, the extended immunostaining of type X collagen found in both the hypertrophic cartilage zone and the primary spongiosa of Col9a1-null mice is a unique observation, and has not been reported previously.
Type X collagen has been reported to form hexagonal matrix aggregate in vitro.34 Type X collagen shares the structural homology with type VIII collagen,35 which forms the hexagonal matrix of cornea endothelial Descemets membrane.36 The postulated functions of type X collagen during endochondral ossification span from a structural component rigidly supporting the hypertrophic cartilage matrix column32,37 to an angiogenic factor.38 Although the exact role of type X collagen is not yet fully elaborated, it is well established that type X collagen is localized predominantly at the matrix/cell interface within the hypertrophic cartilage zone.39 Removal of hypertrophic chondrocytes exposes the type X collagen membrane to the invading osteogenic cells, and the immediate disappearance of type X collagen during formation of the primary spongiosa bone may therefore be explained by an active matrix remodeling.
Our tooth extraction study suggests the involvement of Col9a1 in the preliminary matrix, which may provide a temporary scaffold and homing site for osteoblasts responsible for trabecular bone formation. The Col9a1 mutation may alter this homing mechanism of the osteoblasts. The interference of osteoblast homing may then limit the recruitment of osteoclasts, resulting in the abnormal retention of hypertrophic cartilage matrix containing type X collagen. Further studies are needed to address these hypotheses.
Differentiated osteoblasts contribute differently to adult cortical and trabecular bones, which vary in matrix components40-42 and metabolic reactions such as postmenopausal osteoporosis and diabetes.43,44 To date, little is known about how adult trabecular and cortical bones undergo distinct remodeling processes. Our data suggest that there may be heterogeneous bone remodeling processes and that the transient expression of short type IX collagen by its osteoblasts is required for trabecular bone remodeling.
| Footnotes |
|---|
Supported in part by National Institutes of Health grants EY08219, DE10870, and AR36820.
Accepted for publication August 24, 1999.
| References |
|---|
|
|
|---|
1(IX) collagen develop noninflammatory degenerative joint disease. Proc Natl Acad Sci USA 1994, 91:50705074
1(IX) gene on the human chromosome 6. Eur J Biochem 1989, 179:71-78[Medline]
1 chain of type IX collagen (COL9A1). Genomics 1993, 17:694-698[Medline]
2(IX) collagen cDNA and assignment of the human COL9A2 gene to chromosome 1. FEBS Lett 1993, 319:177-180[Medline]
2(IX) collagen (COL9A2) map to human chromosome 1p32.3-p33 and mouse chromosome 4. Genomics 1994, 23:158-162[Medline]
3 chain of human type IX collagen: Linkage of the gene COL9A3 to chromosome 20q13.3. Genomics 1995, 30:329-336[Medline]
1(IX) collagen chain and its expression during limb development. Biochim Biophys Acta 1993, 1204:61-67
1(IX) collagen chain with a central deletion. Proc Natl Acad Sci USA 1993, 90:2870-2874
1(IX) collagen chains. Exp Cell Res 1995, 219:257-265[Medline]
3(IX) collagen chain completes the primary structure of type IX collagen. Eur J Biochem 1992, 205:443-449[Medline]
1(VIII) collagen cDNAs demonstrate that type VIII collagen is a short chain collagen and contains triple-helical and carboxyl-terminal non triple-helical domains similar to those of type X collagen. J Biol Chem 1989, 264:16022-16029This article has been cited by other articles:
![]() |
P. J. McClive and A. H. Sinclair Type II and Type IX Collagen Transcript Isoforms Are Expressed During Mouse Testis Development Biol Reprod, May 1, 2003; 68(5): 1742 - 1747. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |