SEARCH

SEARCH BY CITATION

Keywords:

  • ossification of the posterior longitudinal ligament;
  • COL11A2;
  • polymorphism;
  • RNA splicing;
  • ectopic ossification

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Ossification of the posterior longitudinal ligament (OPLL) of the spine is the leading cause of myelopathy in Japan. In earlier studies, we provided genetic linkage and allelic association evidence of distinct differences in the human collagen α2(XI) gene (COL11A2) that might constitute inherited predisposition to OPLL.(1) In the present study, a strong allelic association with non-OPLL (p = 0.0003) was observed with an intron 6 polymorphism [intron 6 (−4A)], in which the intron 6 (−4A) allele is more frequently observed in non-OPLL subjects than in OPLL patients. In addition, a newly identified polymorphism in exon 6 [exon 6 (+28A)] was in linkage disequilibrium with the intron 6 (−4A). The functional impact of the polymorphisms was analyzed by comparing the differences in messenger RNA (mRNA) splicing by reverse-transcription polymerase chain reaction (RT-PCR) analysis in cultured cells from the interspinous ligament and an in vitro exon trapping study. The intron 6 (−4A) allele resulted in skipping exon 6 and retaining exon 7, while the exon 6 (+28A) allele was not associated with alteration in mRNA splicing. Similar mRNA species were observed in undifferentiated osteoblast (Ob) cells and in cells from posterior longitudinal ligament of non-OPLL subjects. The region containing exons 6-8 is an acidic subdomain presumably exposed to the surface that could interact with molecules of the extracellular matrix. Accordingly, retaining exon 7 together with removal of exon 6 observed in intron 6 (−4A) could play a protective role in the ectopic ossification process because the same pattern was observed in undifferentiated Ob cells and nonossified posterior longitudinal ligament cells.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

RECENTLY, THERE has been much interest in the molecular pathogenesis of the common metabolic disorders such as essential hypertension, diabetes mellitus, coronary heart disease, and obesity.(2–4) The etiological heterogeneity and multifactorial nature of these common diseases challenge our understanding at all levels. Even if the genetic susceptibilities are determined by current molecular genetics, the daunting task of examining these lifelong genetically determined disturbances by practical experimental procedures remains. Ossification of the posterior longitudinal ligament (OPLL) of the spine is a common disorder among Japanese and other Asian populations, with a reported prevalence of 1.9-4.3% of the general population over the age of 30 years.(5) Heterotopic ossification of the spinal ligament is the specific feature of OPLL, which compresses the spinal cord and leads to various degrees of myelopathy. In spite of the late-onset nature of the disease, it is well known that genetic determinants play an important role in OPLL. The possibility that OPLL has strong genetic determinants is supported by several lines of evidence, including the high rate of concordance among monozygotic/dizygotic (MZ/DZ) twins and the increased recurrence risk in siblings, 10 times the risk in the general population.(6, 7) OPLL is considered to be a state of hyperostosis. Another ossification disorder, diffuse idiopathic skeletal hyperostosis (DISH) appears to be related to OPLL.(8) Interestingly, OPLL is relatively uncommon among white populations, in whom DISH has a high frequency, 25% in males and 15% in females over the age of 50 years.(9)

In a previous report,(1) we provided genetic linkage evidence in Japanese affected sib-pairs showing that the genetic locus for OPLL is within or near the human leukocyte antigen (HLA) region of chromosome 6. Two genes in the region, the collagen α2(XI) (COL11A2) and retinoic X receptor β (RXRβ) genes are considered candidate genes for OPLL, and the patients have been screened extensively for molecular variants. Strong evidence of an allelic association to OPLL was observed with a T to A nucleotide substitution at position −4 from the acceptor site in intron 6 [denoting intron 6 (−4A)], suggesting a certain functional role of this polymorphism in the pathogenesis of OPLL.(1) In the RXRβ gene, three polymorphisms were detected, two of them showing positive associations with OPLL.(10) However, these polymorphisms were in linkage disequilibrium with the intron 6 (−4A) of the COL11A2 and showed less significant allelic association to OPLL than did the intron 6 (−4A). Together, it is strongly suggested that the intron 6 (−4A) of the COL11A2 constitutes a genetic determinant of OPLL. Because the intron 6 (−4A) was found to be more common in non-OPLL subjects than in OPLL patients, it is possible that the intron 6 (−4A) allele could be protective in the ossification process. Despite these genetic findings, determining the involvement of the COL11A2 in the etiology of OPLL is hampered by the lack of evidence regarding the functional significance of the COL11A2 polymorphisms. In this study, we identified a polymorphism in exon 6, which is in linkage disequilibrium with the intron 6 (−4A). We examined COL11A2 to find if the polymorphism in either intron 6 or exon 6 has potential impact on alternative splicing of the COL11A2 transcript that may be part of the molecular etiology of OPLL. Here, we report functional impact of the intron 6 polymorphism of the COL11A2 that results in altered splicing in the region containing exons 6-8. These exons encode an acidic subdomain, also called a variable region because of the complicated alternative splicing that depends on cell type or developmental stage.(11–14) The consequence of the altered splicing caused by intron 6 (−4A), especially the existence of exon 7, may be part of the biological mechanism by which individual differences in the COL11A2 are protective in the development of ectopic ossification in the cervical ligament.

It is well known that endochondral ossification plays a key role in the ectopic ossification process of OPLL, based on histopathological findings of chondrocyte proliferation around the ossified lesions.(15) There appear to be many cartilaginous cells with thickened matrices and nutritional vessels in the area of the active ossification front. However, a mixed mechanism involving both membranous and endochondral ossification processes has been proposed(16) and the mixed pattern is shown in Figs. 1C-1H. Currently, the model system for endochondral ossification in human tissues or cells has not been established. Because α2(XI)-collagen was immunodetected in the membranous ossification legion (Fig. 1H) as well as in cultured ligament cells, we used human normal osteoblast (Ob) cells and ligament cells to investigate the role of intron 6 (−4A) polymorphism of the COL11A2 in the ossification process.

thumbnail image

Figure FIG. 1.. (A) Tomogram and (B) MRI of the cervical spine from a typical mixed type of OPLL patient and pathohistological findings of ossified legion. (A) A huge heterotopic ossification of the posterior longitudinal ligament was observed between C2 and C6. (B) The spinal cord was severely compressed by the ossified lesion. (C) Hematoxylin and eosin staining of a sagittal section of posterior ligament of OPLL. The proliferation of cartilagenous (Ca) cells and osteoblastic (Ob) cells underneath the LBF were observed. (D and G) Higher magnifications of the boxed areas in panel C, Ca cells, cartilagenous extracellular matrix (asterisk), and Ob cells were shown. (E and F) The tissue in panel D was further subjected to immunohistochemistry with antibodies specific to α2(XI)-collagen and α1(II)-collagen. These antibodies positively stained Ca cells and cartilagenous extracellular matrix (asterisk). (H) The tissue in panel G was analyzed by immunohistochemistry with α2(XI)-collagen antibody. Ob cells were positively stained with α2(XI)-collagen antibody. (I) Specificity of the antibody used was confirmed by Western blot using protein extract from ligament cells. An antibody against exon 5-specific peptide of α2(XI)-collagen revealed a single band whereas α1(II)-collagen antibody (monoclonal antibody of commercial origin) showed two bands. These signals were eliminated after collagenase treatment (c-ase). (C) Scale bar = 500 μm (×10); (D, E, F, G, and H) scale bar = 100 μm (×50).

Download figure to PowerPoint

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

OPLL criteria and selection of subjects

We studied 195 OPLL patients and 187 non-OPLL subjects in Kagoshima, Japan, with written informed consent, which was approved by the Ethical Committee of Kagoshima University. OPLL was diagnosed by ectopic ossification in the posterior longitudinal ligament found by radiographic examination by expert orthopedic surgeons. All the non-OPLL subjects with no signs of spinal ossification were over 65 years old, thereby excluding most unmanifested disease. Genomic DNA was obtained from peripheral blood by a standard protocol. White samples were healthy controls from the Chicago area (a gift from G.I. Bell, University of Chicago).

Genotyping molecular polymorphisms of the human COL11A2 by mutagenically separated-polymerase chain reaction

Polymorphism at intron 6 (−4) or exon 6 (+28) in the COL11A2 was determined by mutagenically separated-polymerase chain reaction (MS-PCR) as described.(17) Three primers were included in one reaction. For the detection of the exon 6 (+28) allele, two allele-specific primers E6G(5′-ACCCCCCAATCCCACGCCACTGAGTCTCTCTACTATGACAACG-3′) and E6A (5′-GCCCACTGAGTCTCTCTACTATGACTAGA-3′) and the common complementary strand primer G70 (5′-GGGGTGCTGGGAGAAGGGAAG-3′) were used. For detection of the intron 6 (−4) allele, I6T (5′-TCTCCTTCCTTCCCACCCCGACCTACTCTTCTCCTTCTCTCCTTGCGGT-3′) and I6A (5′-TGCTACTCTTCTCCTTCTCTCCTTGGGCA-3′), together with the common complementary strand primer G72 (5′-GAGTGAGTCACAGGTGCCCACT-3′) were used. The PCR conditions were 40× (95°C for 30 s, 48°C for 30 s, and 72°C for 30 s) for exon 6 (+28) and 37× (95°C for 30 s, 50°C for 30 s, and 72°C for 30 s) for intron 6 (−4). The PCR products were analyzed by electrophoresis on a 4% NuSieve (3:1) agarose gel (FMC Corp., Rockland, ME, USA). In each PCR reaction, control DNAs of three known distinct genotypes and a negative control (water) were included.

Statistical analysis

Comparisons of the genotypic frequencies of polymorphisms between cases and controls were performed by use of contingency χ2 test. For multiple site comparison, haplotype frequencies were estimated for phase-unknown samples based on a maximum likelihood method, by use of a computer program GENEF (J.-M. Lalovel, University of Utah, Salt Lake City, UT, USA) as described in Jeunemaitre et al.(18) Relative risk of haplotype between cases and controls was statistically evaluated by χ2 test, in which the test of homogeneity was made with 1 degree of freedom (df) rather than a global test with multiple degrees of freedom. SPSS package version 8.0 (SPSS, Inc. Chicago, IL, USA) was used for statistical analysis.

Isolation and culture of spinal ligament cells

Interspinous ligament tissues at the cervical spine were obtained aseptically by surgical operation using expansive laminaplasty technique by posterior approach from 10 OPLL patients and 10 non-OPLL (cervical spondylotic myelopathy) patients as described previously.(10) Posterior longitudinal ligament tissues, the lesion directly involved in OPLL, were obtained aseptically using anterior interbody fusion technique by anterior approach from 1 OPLL patient and from 1 non-OPLL (cervical disc herniation) patient. Written informed consent was obtained from all the subjects, and the Ethical Committee of Kagoshima University approved the protocol. The ligament tissues were washed with phosphate-buffered saline several times, minced into about 1-mm2 pieces, and the explant cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) (Gibco BRL, Rockville, MD, USA) supplemented with 10% fetal bovine serum (Gibco BRL) and 0.2 mM L-ascorbic acid-2-phosphate.

Reverse-transcription-PCR and Southern blot analysis

Total RNA was extracted from cultured cells with TRIZOL Reagent based on the manufacturer's protocol (Gibco BRL). One microgram of total RNA was reverse-transcribed (RT) with an oligo (dT)12-18 primer using Moloney murine leukemia virus (M-MLV) RT (Gibco BRL) at 37°C for 1 h. After 20 minutes incubation at 37°C with RNAse H (Toyobo, Osaka, Japan) to remove hybrid messenger RNA (mRNA), the complementary DNA (cDNA) was amplified by the PCR using a 1/20 aliquot of cDNA as a template and AmpliTaq Gold DNA polymerase (PE Applied Biosystems, Urayasu, Japan). The following two specific oligonucleotide primers were used for the reactions: primer 5F, the forward primer on exon 5 (5′-CATGTGAACAGAAGGAGCTGGA-3′) and primer 9R, the reverse primer on exon 9 (5′-CTTTCTCTCCCTTCAGCCCTCGG-3′; Fig. 2A). Human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA also was amplified as a positive control (Stratagene, La Jolla, CA, USA). The PCR products were analyzed by electrophoresis on a 2% agarose gel, blotted onto nylon filters, and hybridized with exon-specific oligonucleotide probes. Oligonucleotide probes specific for exon 6 (probe Ex6, 5′-CCCACTGAGTCTCTCTACTATG-3′), exon 7 (probe Ex7, 5′-AGAAATCCTGGAGTCGAGCCTCT-3′), and exon 8 (probe Ex8, 5′-CAGCCGACAGGTTCCAGGCAGA-3′) were used to detect alternatively spliced COL11A2 mRNAs. An oligonucleotide probe (probe Ex5/9, 5′-CAGAGCCAGGCTGCCCAT-3′) corresponding to nine nucleotides at the 3′ end of exon 5 plus nine nucleotides at the 5′ end of exon 9 was used to detect mRNA species with deletion of exons 6-8. Primer 5F was applied to detect exon 5. The oligonucleotides were end-labeled with [γ-32P]-adenosine triphosphate (ATP; Amersham Pharmacia Biotech, Tokyo, Japan) by a kinase reaction.

thumbnail image

Figure FIG. 2.. Alternative splicing of the region containing exons 6-8 of COL11A2 mRNA in cultured spinal ligament cells. (A) PCR primers 5F and 9R and the oligonucleotide probe for Southern blot were shown. (B) Amplified products were separated on a 2% agarose gel. The genotypes at intron 6 (−4) of each sample were shown at top. As a positive control, PCR products of GAPDH were shown at bottom. Lanes 1-3 represented non-OPLL subjects and lanes 4-6 showed OPLL patients. (C) Southern blot analysis using exon-specific probes was performed to clarify the complicated exon combinations of the RT-PCR products.

Download figure to PowerPoint

In vitro splicing analysis

A genomic region of 2387 base pairs (bp) containing exons 5-7 was amplified from two distinct P1 clones containing the entire human COL11A2(10) and subcloned into the pSPL3 expression vector (Gibco BRL). The genotypes at exon 6 (+28) and intron 6 (−4) of one P1 clone were guanine and thymine, respectively, and the genotypes of the other two clones were adenine and adenine. The following two specific primers were applied to amplify the fragments: primer E567F (5′-GGGAGGCAAAGAATTCTGGAA-3′), locating on intron 4 with an additional EcoRI site, and primer E567R (5′-AAACTCGAGAATCATGGAAGGAGGCCTGG-3′), locating on intron 7 with an additional XhoI site. The products were directionally subcloned into the pSPL3 vector using the EcoRI and XhoI sites. All inserts and junctions were verified by direct sequencing. To generate four possible haplotypes, the intrinsic BglII sites on exon 5 and intron 6 of both clones were digested and the BglII fragments were exchanged and re-subcloned. Exon trapping to analyze exons 5-7 mRNA splicing in vitro was performed according to the manufacturer's protocol (Gibco BRL). The four distinct pSPL3 expression vectors were transfected into COS-7 cells with FuGENE 6 transfection reagent (Boehringer Mannheim, Indianapolis, IN, USA). Twenty-four hours after transfection, total RNA was isolated and subjected to RT-PCR analysis.

Normal human osteoblast cell culture and differentiation

Normal human osteoblast (NHOst) cells were obtained commercially (Clonetics Corp., Walkersville, MD, USA). The NHOst cells were cultured in Osteoblast Growth Medium (Clonetics Corp.) supplemented with SingleQuots (fetal bovine serum, ascorbic acid, and gentamicin/amphotericin-B; Clonetics Corp.) under 5% CO2 at 37°C. Differentiation (bone mineralization) was induced by Differentiation SingleQuots (hydrocortisone 21 hemisuccinate, β-glycerophosphate; Clonetics Corp.) as described in the manufacturer's protocol. The bone mineralization was confirmed directly by von Kossa staining,(19) or indirectly by measuring alkaline phosphatase (AP) activity.(20) The von Kossa staining was performed to detect mineralization using 2.5% silver nitrate. To determine AP activity, cells were lysed with 200 mM sodium carbonate/100 mM glycine/1 mM MgCl2/0.1% Triton X-100. The crude lysate was sonicated and cleared by centrifugation at 9000g for 30 minutes. The AP activity of the supernatants was analyzed by Fast p-Nitrophenyl Phosphate Tablet Sets (pH 9.9; Sigma, St. Louis, MO, USA). The protein concentration of each sample was determined by Bradford method.(21) To monitor changes in splicing before and after differentiation, RT-PCR also was performed using NHOst cell mRNA in the same conditions as described for ligament cells. NHOst cells were heterozygous at intron 6 (−4) position.

Antibodies production, affinity purification, and Western blot

Rabbit polyclonal antibody was raised against a synthetic peptide corresponding to amino acid sequences at exon 5 (peptide: RERPQNQQPHRAQRSPQQQP-C) of α2(XI)-collagen.(22) Additional cysteine residue at the carboxyl terminus was applied for coupling to the keyhole limpet of hemocyanin using a commercial kit (Pierce, Rockford, IL, USA). The antibody was affinity-purified with the antigenic peptide coupled to activated Thiol Sepharose 4B (Amersham Pharmacia Biotech). There is no homology between the α2(XI)-exon 5 peptide and any sequence in α1(XI)-collagen.

Cultured interspinous ligament cells were harvested at confluence of 80% with rubber policemen. The cells were treated with 0.1% Triton X-100/Tris-buffered saline, incubated at 4°C for 30 minutes, and centrifuged at 12,000g for 30 minutes. The cell pellets were subjected to immunoblot analysis. For collagenase digestion, the cell pellets were dialyzed against 50 mM Tris-HCl, pH 7.6, containing 0.2 M NaCl, 5 mM CaCl2, and 10 mM N-ethylmaleimide and digested with 34 U/ml collagenase S-1 (Nitta Gelatin Co., Osaka, Japan) for 5 h at 37°C, with occasional shaking. Digestion was stopped by addition of 5 volumes of cold acetone followed by precipitation for 1h at −20°C. Samples were separated on a 7% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis(23) under reducing condition and electrotransferred onto Immobilon-P membrane (Millipore, Bedford, MA, USA). The blots were blocked with 3% bovine serum albumin (BSA) and subsequently reacted with the polyclonal antibody specific for exon 5 of the α2(XI)-chain or a monoclonal antibody against the α1(II)-chain (Oncogene, Cambridge, MA, USA). Bound antibodies were reacted with horseradish peroxidase-labeled antibody and visualized by enhanced chemiluminescence (ECL) Western blotting detection reagent (Amersham Pharmacia Biotech.).

OPLL tissue preparation for immunohistochemistry

Specimens of ossified ligament tissue and nonossified ligament tissue were obtained en bloc from a typical cervical OPLL subject during surgery by anterior approach. The specimens were fixed by immersion in 10% neutrally buffered formalin and decalcified with 0.36 M EDTA (pH 7.0-7.2) before embedding in paraffin.(24) Paraffin sections of 3- to 5-μm thickness were prepared and subjected to hematoxylin and eosin staining and to immunohistochemistry. Immunohistochemistry was performed by avidin-biotin complex method using an ABC elite kit (Vector Laboratory, Burlingame, CA, USA) as previously reported.(24) The specimens were pretreated with 0.25% trypsin (Gibco BRL), 1.45 U/ml hyaluronidase (Sigma), and 0.25 U/ml Chondroitinase ABC (Sigma) after deparaffinization.(25) The 1.5 μg/ml of anti-α2(XI)-collagen (exon 5) polyclonal antibody or 33 μg/ml of α1(II)-collagen monoclonal antibody was used as primary antibody. Detection was performed using 3,3-diaminobenzidine tetrachloride (Dojindo, Kumamoto, Japan).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Clinical features of OPLL and mixed ossification process

Figure 1 shows a representative case of OPLL with a huge ossification of the ligament compressing the spinal canal between C2 and C6, shown in a tomogram and magnetic resonance imaging (MRI) (Figs. 1A and 1B). Hematoxylin and eosin staining of a sagittal section of OPLL shows proliferation of osteoblastic (Ob) cells and cartilaginous cells (Ca) underneath the lamellar bone formation (LBF) (Fig. 1C magnified in Figs. 1D and 1G). Immunohistochemistry ofα2(XI)-collagen using exon 5-specific antibody revealed that Ca cells (arrowhead in Fig. 1E), cartilagenous extracellular matrix (asterisk in Fig. 1E), and Ob cells (arrow in Fig. 1H) were positively stained. The α1(II)-collagen also was detected in Ca cells (arrowhead in Fig. 1F) and the cartilagenous extracellular matrix (asterisk in Fig. 1F). Specificity of the α2(XI)-collagen antibody was confirmed by Western blot analysis from protein extract of ligament cells as shown in Fig. 1I. The α2(XI)-collagen antibody detected a specific band above the 97-kDa protein marker that corresponds to a monomer form of the α2(XI)-chain. As a positive control, the monoclonal antibody for α1(II)-collagen was analyzed, and two bands around 90 kDa were detected. All the signals were sensitive to collagenase treatment (c-ase in Fig. 1I).

Association study and linkage disequilibrium between the intron 6 and the exon 6 polymorphisms

We previously documented multiple polymorphisms in the COL11A2, in which the statistically strongest association to OPLL was observed with the intron 6 (−4A) polymorphism.(1) Further screening of the gene by direct sequencing has identified a nucleotide substitution in exon 6. A guanine to adenine substitution at position 28 (nucleotide numbering is from the start of exon 6), which replaces glutamine at codon 272 by lysine, was in linkage disequilibrium with the intron 6 polymorphism. Increasing the sample size, we reexamined the allelic association of the polymorphism at intron 6 (−4A) with OPLL. The intron 6 (−4A) allele was more common among non-OPLL subjects than OPLL patients (22.5% vs. 12.0%; χ2=12.88 [df = 1]; p = 0.0003; Table 1). A similar but less significant association was observed in the exon 6 (+28A) polymorphism (24.0% vs. 14.7%; χ2=10.46 [df = 1]; p = 0.0012; Table 1), reflecting a linkage disequilibrium between the two polymorphisms. Pairwise linkage disequilibrium between two polymorphisms was estimated according to Thompson's formula (D/Dmax = 0.89; p < 0.0001 for non-OPLL controls).(26) Furthermore, the two polymorphisms were examined jointly by haplotype analysis based on the maximum likelihood method, and the estimated haplotype frequency was compared between OPLL patients and controls (Table 2). A haplotype carrying the more common alleles at both sites (H1) was overrepresented in OPLL patients (χ2=16.49 [df = 1]; p = 0.00005), while the other haplotype (H4), consisting of the rarer allele counterparts, was less common in OPLL patients (χ2 =6.56 [df = 1]; p = 0.010). These results strongly support the possibility that the COL11A2 constitutes a genetic susceptibility for OPLL and that the polymorphisms at intron 6 and exon 6 could independently or cooperatively have a physiological impact in the etiology of the disease. Because of the higher frequency of the intron 6 (−4A) and exon 6 (+28A) in non-OPLL subjects than in OPLL patients, the polymorphisms could be protective of the ectopic ossification.

Table Table 1.. Allelic Association Between Variants in COL11A2 and OPLL
Thumbnail image of
Table Table 2.. Estimated Haplotype Frequency in Japanese (Case and Control) and White (Control) Populations
Thumbnail image of

The incidence of OPLL in white populations is reported to be very low, with a prevalence of 0.01-1.7%.(27) We expected predominance of the protective allele of COL11A2 might partly explain the low incidence of OPLL in white populations, and the allelic frequencies of the intron 6 and exon 6 polymorphisms in white controls were determined. The allelic frequencies of intron 6 (−4A) and exon 6 (+28A) were significantly higher in white populations than in Japanese controls (32.4% vs. 22.5%, χ2=8.60 [df = 1], and p = 0.0034, for intron 6; 31.3% vs. 24.0%, χ2=4.50 [df = 1], and p = 0.034, for exon 6 [+28A]). The haplotype frequency was estimated using two polymorphisms, as shown in Table 2. The haplotype (H4) of both the rarer alleles was significantly higher in white populations than in Japanese controls (χ2=19.03 [df = 1]; p = 0.00001). Because of the very low incidence of OPLL in white populations, we could not obtain a large enough number of OPLL patients to perform a case-control study in this population.

Altered splicing of COL11A2 mRNA caused by the intron 6 (−4A) in cultured ligament cells

Because the region containing exons 6-8 of the COL11A2 transcript exhibits a complicated splicing pattern previously described in humans and mice,(11, 13, 28) the intron 6 (−4A) or exon 6 (+28A) might affect splicing of COL11A2 mRNA. We obtained tissues from the cervical interspinous ligament in 10 OPLL patients and 10 controls, which were then subjected to primary culture. Total RNA, prepared from cultured ligament cells of OPLL patients or controls, was subjected to RT-PCR analysis to determine if the splicing alterations of COL11A2 mRNA are a result of the genetic variation. PCR primers and oligonucleotide probes were designed to cover exons 6-8, as shown in Fig. 2A. Southern blot of RT-PCR products hybridized with exon-specific oligonucleotide probes to identify combinations of alternatively spliced exons was used (Fig. 2C). The identity of the PCR products was obtained by direct sequencing of the products (data not shown). All the samples were analyzed and representative data are shown in Fig. 2B. We detected no distinct differences in splicing pattern between the cells from OPLL patients and non-OPLL subjects; however, differences caused by the genetic variation at intron 6 were shown. Only 1 of the 20 samples obtained was homozygous for intron 6 (−4A), which showed mainly three bands (Fig. 2B, lane 3). The intron 6 (−4A) allele is associated with an increased amount of E5789 (lacking exon 6) and E579 (lacking exon 6 and 8). E59 (lacking exon 6, 7, and 8), observed predominantly in cartilage tissue,(11, 12) was almost absent in the cells of the intron 6 (−4A) homozygote. The intron 6 (−4T) homozygote contains mainly three bands: E56789 (all of the exon is present), E5679 (lacking exon 8), and E59 (Fig. 2B, lanes 1 and 6). As expected, mixed splicing patterns were observed in the heterozygous samples (lanes 2, 4, and 5, Fig. 2B) because of the participation of both alleles.

In vitro splicing analysis

Although RT-PCR analysis of the interspinous ligament cells showed splicing variation that varied according to the intron 6 (−4) genotype, the limitations of analysis using human tissue samples are considerable. First, only one intron 6 (−4A) homozygous sample was available in the present study, and intron 6 (−4T/A) heterozygote samples showed a quantitatively inconsistent pattern to some extent. Second, we were not able to show that the intron 6 genotype could be attributed only to the difference in splicing. Because the two polymorphisms at intron 6 and at exon 6 are in tight linkage disequilibrium, all the ligament samples heterozygous for intron 6 (−4) also were heterozygous for exon 6 (+28). To address these limitations, in vitro splicing analysis was performed. Expression vectors containing exons 5-7 with four different haplotypes (combination of two sites, Fig. 3A) were generated in the pSPL3 vector and transfected into COS-7 cells, followed by RT-PCR analysis (Figs. 3B and 3C). As shown in Fig. 3C, the genotype of exon 6, either +28G or +28A, failed to affect splicing. In contrast, we observed differences in splicing pattern caused by intron 6 (−4T or −4A). E57 (lacking exon 6) was observed in both genotypes, whereas E5 (lacking exon 6 and 7) was observed only in the intron (−4T) genotype. We could not detect an mRNA species containing all the exons probably because of the truncated constructs used in the in vitro experiment.

thumbnail image

Figure FIG. 3.. In vitro exon trapping analysis to determine if the intron 6 (−4) polymorphism or the exon 6 (+28) polymorphism is responsible for the alternative splicing. (A) pSPL3 vectors containing exons 5-7 with four different haplotypes (combination of two polymorphisms at exon 6 and intron 6) were generated. (B) The four vectors were transfected into COS-7 cells followed by RT-PCR using the vector-specific primers SD6 and SA2. (C) The amplified products were separated on a 2% agarose gel followed by Southern blot analysis to identify exon combinations of the RT-PCR products using oligonucleotide probes. V/V denotes vector/vector splicing.

Download figure to PowerPoint

Cultured osteoblast after differentiation

To investigate the splicing alterations of COL11A2 mRNA during the ossification process in more detail, primary NHOst cells before and after differentiation were studied. NHOst cells were subjected to differentiation by adding hydrocortisone and β-glycerophosphate, and bone mineralization was monitored by von Kossa staining or measuring AP activity. After 3 weeks of stimulation, an accumulation of von Kossa-positive particles became evident (Fig. 4A) and a 2.5-fold increase of cell-associated AP activity was observed (data not shown).

thumbnail image

Figure FIG. 4.. COL11A2 mRNA splicing pattern depends on the ossification status was examined in primary human osteoblast cells. Differentiation of osteoblast cells was induced by adding 40 nM hydrocortisone 21 hemisuccinate and 2 mM β-glycerophosphate (shown as +) into the culture medium for 3 weeks. (A) The differentiation was monitored by von Kossa staining. Without adding these reagents (shown as −), no von Kossa positive cells were observed. Hematoxylin and eosin stained cells were shown in the small box. (B) RT-PCR of exons 5-9 using primers 5F and 9R (Fig. 2A) with (+) or without (−) differentiation was performed. Control PCR products of GAPDH are shown below. Scale bar = 100 μm (×50).

Download figure to PowerPoint

Total RNA was prepared from both undifferentiated and differentiated NHOst cells, and subjected to RT-PCR analysis (Fig. 4B). In undifferentiated NHOst cells, the predominant E59 species and the minor E579 species were present. After differentiation, the E579 species disappeared and the E5679 species appeared. E5679 is a specific species associated with the intron 6 (−4T) allele in ligament cells.

Posterior longitudinal ligament cell culture

In OPLL, the ossification lesion is in the posterior longitudinal ligaments. Accordingly, the direct impact of the intron 6 polymorphism on the ossification can be shown only using the posterior ligament cells. However, patients requiring an anterior interbody fusion technique to remove posterior longitudinal ligament or intervertebral disc are very few, and we were able to obtain only two posterior longitudinal ligament samples: one from an OPLL patient and one from a non-OPLL (cervical disc herniation) patient undergoing surgery. The tissues were subjected to cell culture. Both subjects were heterozygous for intron 6 (−4). RT-PCR analysis was performed as described previously, and a comparison of cells from the OPLL patient and the non-OPLL patient was made (Fig. 5). E5789, which was associated with the intron 6 (−4A) allele (a protective allele) in interspinous ligament cells, was observed in nonossified ligament cells. Furthermore E59, which was associated with the intron 6 (−4T) allele in interspinous ligament cells, was detected in cells from the OPLL patient.

thumbnail image

Figure FIG. 5.. Alternative splicing of COL11A2 in the posterior longitudinal ligament cells of the cervical spine from an OPLL patient and a cervical disc herniation (CDH) subject as a control were analyzed. RT-PCR of COL11A2 exons 5-9 using primers 5F and 9R (Fig. 2A) was performed. Control PCR products of GAPDH also were shown.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The pathophysiological mechanism of OPLL remains a matter of speculation despite numerous investigations. Complex networks of genes together with environmental interactions all need to be considered. Our previous and ongoing genetic linkage and allelic association studies strongly support the possibility that molecular variation in the COL11A2 constitutes an inherited predisposition to OPLL. The finding that the intron 6 (−4A) allele is observed more frequently in non-OPLL subjects than in OPLL patients (χ2 = 12.88; p = 0.0003; Table 1) suggests that the intron 6 (−4A) might function as a protective allele for OPLL. This hypothesis is supported somewhat by the finding that a high frequency of the intron 6 (−4A) allele is present in the white populations, in whom a low frequency of OPLL has been reported. Similar results were obtained in the exon 6 (+28A) polymorphism because of linkage disequilibrium with the intron 6 (−4A). Because these data are statistical, the physiological significance of the COL11A2 variation on the ectopic ossification process must be established by experimental procedures. Because the polymorphism at position −4 of intron 6 is genetically associated with OPLL (Table 1), it seems likely that the best candidate functional alteration due to the polymorphism is the splicing in the exons 6-8 region. Whether or not the intron 6 or exon 6 polymorphism has a direct physiological effect or acts as a marker for another molecular variant yet to be identified is not known.

In the present study, we have provided experimental evidence that the intron 6 (−4A) polymorphism contains a qualitative alteration in the splicing of the COL11A2 transcript, whereas the exon 6 (+28A) polymorphism shows no impact in cells from interspinous ligament or in vitro splicing analysis (Figs. 2 and 3). Because very complicated splicing patterns occur in the region containing exons 6-8 of the COL11A2, as shown in Fig. 2, involvement of the intron 6 polymorphism in the splicing alteration must be interpreted carefully. Our observations can be characterized as follows: the presence of exon 7 together with the absence of exon 6 was shown to be associated with the intron 6 (−4A) polymorphism in interspinous ligament cells as well as in the exon trapping experiment (Figs. 2 and 3). The specific splicing pattern in ligament cells carrying the intron 6 (−4A) also was observed in cells in undifferentiated Ob or nonossified posterior longitudinal ligament. The occurrence of alternative splicing is known to depend on the availability of an acceptor site.(29, 30) A nucleotide substitution at intron 6 (−4), T to A, may provide a stronger acceptor site in exon 7 than that in exon 6, resulting in the skipping of the upstream exon. This splicing species, observed in the intron 6 (−4A) allele in interspinous ligament cells and nonossified osteoblast cells and posterior ligament cells, could be related closely to the protective role in ectopic ossification.

Type XI collagen, a fibril-forming minor collagen, is composed of three distinct molecules: α1(XI)-, α2(XI)-, and α3(XI)-chains, and is cross-linked by the lysyl oxidase-mediated mechanism. The α3(XI)-chain was revealed to be an overglycosylated form of α1(II)-collagen lacking an N-terminal cysteine-rich domain.(31) Type XI collagen coassembles with type II collagen in the integral component of the fibril network and may regulate the diameter of the collagen fibrils. This hypothesis is supported by the recent finding of an α1(XI)-null mutation in cho/cho (chondrodysplasia) mice, in which the cartilage has abnormally thick collagen fibrils.(32) The highly acidic subdomain encoded by exons 6-8 at the aminoterminal of the α2(XI)-chain provides potential sites for interaction of type XI collagen with other molecules and may prevent further deposition of collagen fibrils. These structure differences due to alternative splicing involving the acidic subdomain may be implicated in susceptibility to the ectopic ossification process through endochondral or membranous ossification. Interestingly, exon 7 contains eight acidic residues of the 21 amino acids and no basic residue indicating that the presence of exon 7 might increase cross-linking to each other or with other molecules. It is tempting to speculate that the presence of the exon 7-containing transcript, associating with the intron 6 (−4A) allele, could function protectively against the ectopic ossification process by interacting with various components of the extracellular matrix.

Taken together, these findings suggest potential involvement of the COL11A2 in the pathogenesis of the ectopic ossification seen in OPLL. Because subtle lifelong effects of genetic variation lead to heterotopic ossification in OPLL, ascertaining the physiological impact of genetic variation within a reasonable experimental time is difficult. The current data are a step in understanding molecular etiology of OPLL. How the difference in splicing of the COL11A2 leads directly to the heterotopic ossification observed in OPLL in the long term will be further investigated by in vivo experiments using a genetically manipulated mouse model.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We gratefully acknowledge the advice kindly provided by K. Hoshijima (University of Utah) and S. Matsufuji (Jikei Medical School) concerning this work. This work is supported by the Ministry of Public Health and Welfare Research Grant for Specific Diseases, Japan (I.I.) and the Japanese Ministry of Science, Education, Sports and Culture (J.T. and I.I.).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Koga H, Sakou T, Taketomi E, Hayashi K, Numasawa T, Harata S, Yone K, Matsunaga S, Otterud B, Inoue I, Leppert M 1998 Genetic mapping of ossification of the posterior longitudinal ligament of the spine. Am J Hum Genet 62:14601467.
  • 2
    Schork NJ, Cardon LR, Xu X 1998 The future of genetic epidemiology. Trends Genet 14:266272.
  • 3
    Lander ES, Schork NJ 1994 Genetic dissection of complex traits. Science 265:20372048.
  • 4
    Inoue I, Nakajima T, Williams CS, Quackenbush J, Puryear R, Powers M, Cheng T, Ludwig EH, Sharma AM, Hata A, Jeunemaitre X, Lalouel J-M 1997 A nucleotide substitution in the promoter of human angiotensinogen is associated with essential hypertension and affects basal transcription in vitro. J Clin Invest 99:17861797.
  • 5
    Matsunaga S, Sakou T 1997 Epidemiology of ossification of the posterior longitudinal ligament. In: YonenobuK, SakouT, OnoK (eds.) Ossification of the Posterior Longitudinal Ligament. Springer-Verlag, Tokyo, Japan, pp. 1117.
  • 6
    Terayama K 1989 Genetic studies on ossification of the posterior longitudinal ligament of the spine. Spine 14:11841191.
  • 7
    Sakou T, Taketomi E, Matsunaga S, Yamaguchi M, Sonoda S, Yashiki S 1991 Genetic study of ossification of the posterior longitudinal ligament in the cervical spine with human leukocyte antigen haplotype. Spine 16:12491252.
  • 8
    Trojan DA, Pouchot J, Pokrupa R, Ford RM, Adamsbaum C, Hill RO, Esdaile JM 1992 Diagnosis and treatment of ossification of the posterior longitudinal ligament of the spine: Report of eight cases and literature review. Am J Med 92:296306.
  • 9
    Weinfeld RM, Olson PN, Maki DD, Griffiths HJ 1997 The prevalence of diffuse idiopathic skeletal hyperostosis (DISH) in two large American Midwest metropolitan hospital populations. Skeletal Radiol 26:222225.
  • 10
    Numasawa T, Koga H, Ueyama K, Maeda S, Sakou T, Harata S, Leppert M, Inoue I 1999 Human retinoic X receptor β: Complete genomic sequence and mutation search for ossification of posterior longitudinal ligament of the spine. J Bone Miner Res 14:500508.
  • 11
    Tsumaki N, Kimura T 1995 Differential expression of an acidic domain in the amino-terminal propeptide of mouse pro-α2(XI) collagen by complex alternative splicing. J Biol Chem 270:23722378.
  • 12
    Matsui Y, Kimura T, Tsumaki N, Nakata K, Yasui N, Araki N, Hashimoto N, Uchida A, Ochi T 1998 Splicing patterns of type XI collagen transcripts act as molecular markers for osteochondrogenic tumors. Cancer Lett 124:143148.
  • 13
    Lui VCH, Ng LJ, Sat EWY, Nicholls J, Cheah KSE 1996 Extensive alternative splicing within the amino-propeptide coding domain of α2(XI) procollagen mRNA. J Biol Chem 271:1694516951.
  • 14
    Sugimoto M, Kimura T, Tsumaki N, Matsui Y, Nakata K, Kawahata H, Yasui N, Kitamura Y, Nomura S, Ochi T 1998 Differential in situ expression of α2(XI) collagen mRNA isoforms in the developing mouse. Cell Tissue Res 292:325332.
  • 15
    Ono K, Ota H, Tada K, Hamada H, Takaoka K 1977 Ossified posterior longitudinal ligament. A clinicopathologic study. Spine 2:126138.
  • 16
    Ono K, Yonenobu K, Miyamoto S, Okada K 1999 Pathology of ossification of the posterior longitudinal ligament and ligamentum flavum. Clin Orthop 359:1826.
  • 17
    Rust S, Funke H, Assmann G 1993 Mutagenically separated PCR (MS-PCR): A highly specific one step procedure for easy mutation detection. Nucleic Acids Res 21:36233629.
  • 18
    Jeunemaitre X, Inoue I, Williams C, Charru A, Tichet J, Powers M, Sharma AM, Gimenez-Roqueplo AP, Hata A, Corvol P, Lalouel J-M 1997 Haplotypes of angiotensinogen in essential hypertension. Am J Hum Genet 60:14481460.
  • 19
    Rungby J, Kassem M, Eriksen EF, Danscher G 1993 The von Kossa reaction for calcium deposits: Silver lactate staining increases sensitivity and reduces background. Histochem J 25:446451.
  • 20
    Terakado A, Tagawa M, Goto S, Yamazaki M, Moriya H, Fujimura S 1995 Elevation of alkaline phosphatase activity induced by parathyroid hormone in osteoblast-like cells from the spinal hyperostotic mouse TWY (twy/twy). Calcif Tissue Int 56:135139.
  • 21
    Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248254.
  • 22
    Vuristo MM, Pihlajamaa T, Vandenberg P, Prockop DJ, Ala-Kokko L 1995 The human COL11A2 gene structure indicates that the gene has not evolved with the genes for the major fibrillar collagens. J Biol Chem 270:2287322881.
  • 23
    Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680685.
  • 24
    Yonemori K, Imamura T, Ishidou Y, Okano T, Matsunaga S, Yoshida H, Kato M, Sampath TK, Miyazono K, ten Dijke P, Sakou T 1997 Bone morphogenetic protein receptors and activin receptors are highly expressed in ossified ligament tissues of patients with ossification of the posterior longitudinal ligament. Am J Pathol 150:13351347.
  • 25
    Kumagai J, Sarkar K, Uhthoff HK, Okawara Y, Ooshima A 1994 Immunohistochemical distribution of type I, II, and III collagens in the rabbit supraspinatus tendon insertion. J Anat 185:279284.
  • 26
    Thompson EA, Deeb S, Walker D, Motulsky AG 1988 The detection of linkage disequilibrium between closely linked markers: RFLPs at the AI-CIII apolipoprotein genes. Am J Hum Genet 42:113124.
  • 27
    Firooznia H, Benjamin VM, Pinto RS, Golimbu C, Rafii M, Leitman BS, McCauley DI 1982 Calcification and ossification of posterior longitudinal ligament of spine. Its role in secondary narrowing of spinal canal and cord compression. NY State J Med 82:11931198.
  • 28
    Lui VCH, Kong RYC, Nicholls J, Cheung ANY, Cheah KSE 1995 The mRNA for the three chains of human collagen type XI are widely distributed but not necessarily co-expressed: Implications for homotrimeric, heterotrimeric and heterotypic collagen molecules. Biochem J 311:511516.
  • 29
    Brunak S, Engelbrecht J, Knudsen S 1991 Prediction of human mRNA donor and acceptor sites from the DNA sequence. J Mol Biol 220:4965.
  • 30
    Oshima Y, Gotoh Y 1987 Signals for selection of a splice site in pre-mRNA. J Mol Biol 195:247259.
  • 31
    Morris NP, Bachinger HP 1987 Type XI collagen is a heterotrimer with the composition (1α, 2α, 3α) retaining non-triple-helical domains. J Biol Chem 262:1134511350.
  • 32
    Li Y, Lacerda DA, Warman ML, Beier DR, Yoshioka H, Ninomiya Y, Oxford JT, Morris NP, Andrikopoulos K, Ramirez F, Wardell BB, Lifferth GD, Teuscher C, Woodward R, Taylor BA, Seegmiller RE, Olsen BR 1995 A fibrillar collagen gene, Col11a1, is essential for skeletal morphogenesis. Cell 80:423430.