Elsevier

Bone

Volume 36, Issue 5, May 2005, Pages 902-908
Bone

In vitro functional assay of alleles and haplotypes of two COL1A1-promoter SNPs

https://doi.org/10.1016/j.bone.2004.12.012Get rights and content

Abstract

Osteoporosis is a common disease with a strong genetic component. We previously described two polymorphic sites in the COL1A1 gene promoter, −1997 G/T and −1663indelT, which have been associated with bone mineral density (BMD), a surrogate trait for osteoporosis. Here, we explore the molecular mechanisms underlying this association by performing transient transfections in MG-63 cells of constructs bearing different COL1A1 promoter regions, containing different alleles or haplotypes of the polymorphic sites. These promoter regions drove the transcription of a luciferase reporter gene. The main differences in transcriptional activity relied on an inhibitory region localized to the −1284 to −254 interval. Regarding the polymorphisms, reproducible differences were observed between the alleles of each of them: the G allele at −1997 showed a higher transcriptional activity than the T allele, as did the 7T allele of −1663 as compared with 8T. Accordingly, the T-8T haplotype was the weakest transcriber. A functional interaction was found between the −1997 and −1663 polymorphisms, in that the difference in transcriptional activity between the 7T and 8T alleles was dependent on the allele at −1997. This different transcriptional activity of the two −1663indelT alleles correlated with different binding capacities of the corresponding oligonucleotides to osteoblast nuclear proteins. Supershift assays allowed us to identify one of these proteins as the architectural transcription factor Nmp4/CIZ, a protein known to be an inhibitor of BMP/Smad signalling.

Introduction

Type I collagen is the most abundant component of the extracellular matrix of connective tissues and, in particular, of bone. Its transcriptional regulation is highly complex because of the influence of developmental, environmental and hormonal factors, and overall it plays an important role in many physiological as well as pathological events [13], [28]. The protein is encoded by two genes, COL1A1 and COL1A2, which, under physiological conditions, are expressed and regulated in a coordinated fashion to produce the α1 and α2 polypeptide chains, respectively. Two α1 chains interact with one α2 to form a characteristic triple-helix.

The control of collagen gene transcription is believed to involve complex interactions among DNA-binding proteins recognizing separate cis-acting elements in the 5′ regulatory and proximal promoter elements. Several studies of the promoter function of the mouse, rat, or human COL1A1 genes have been performed, both in vitro and using transgenic mice [2], [3], [14], [16], [21], [22]. A minimal promoter within the 220-bp region upstream of the transcriptional start site of COL1A1 contains sequences sufficient for its basal and tissue-specific transcription [7], [26]. In addition, Alvarez et al. [1] have characterized an architectural transcription factor that binds to regulatory regions located more than 1.5 kb upstream of the transcriptional start site.

The two type I collagen genes constitute good candidates to partly explain the genetic variability underlying osteoporosis or bone mineral density (BMD, a surrogate marker of the disease). A G to T polymorphism within an Sp1 binding element in intron 1 (position +1245) of the COL1A1 gene has been described as being associated with osteoporotic fracture [10].

In a previous study, we identified two single-nucleotide polymorphisms (SNPs) in the promoter of the COL1A1 gene: −1663indelT and −1997G/T [8]. The G to T transversion at −1997 was significantly associated with lumbar spine BMD in Spanish postmenopausal women, as was the interaction between the two polymorphisms. The association between the −1997G/T polymorphism and BMD was later replicated in a cohort of British women [19]. Interestingly, the −1663indelT polymorphism was in strong linkage disequilibrium with the +1245G/T (Sp1) polymorphism of intron 1. Gel retardation assays to test binding of osteoblast nuclear proteins to the PCOL1 and PCOL2 sites, which contained the two polymorphisms (−1663indelT and −1997G/T, respectively), revealed specific binding to single-stranded oligonucleotides harboring either of the sites [8]. In the case of the −1997G/T polymorphism, allelic differences in binding were observed, the G allele exhibiting stronger binding than the T.

To assess the involvement of these two COL1A1-promoter SNPs in transcriptional regulation, we have fused a 2.5-kb DNA fragment comprising bases −2483 to +40 of the human COL1A1 promoter to a luciferase reporter gene and assayed the four haplotypes in transiently transfected MG-63 osteosarcoma cells. We have also studied the influence of several hormones, cytokines, and vitamins on COL1A1 promoter function. Finally, we have identified one of the proteins binding to the PCOL1site.

Section snippets

Cell culture and nuclear protein extracts

Human bone cells were obtained from surgical specimens of healthy subjects undergoing surgery for acute traumatic conditions and without any other bone disease. The protocol used was based on a method described by Marie et al. [18] with some modifications [9], [20]. A primary osteoblast culture was established by pooling cells from four donors. Cells were grown in Dulbecco's modified Eagle medium (DMEM, Gibco-BRL, Paisley, Scotland, UK) supplemented with 20% fetal calf serum (FCS, Biological

Characterization of osteosarcoma proteins that bind to the PCOL1 or PCOL2 probes

Osteosarcoma cell lines such as MG-63 have been widely used for transfection assays, while human primary osteoblasts (HPO) are poorly transfected. To validate the use of the former in place of the latter, we have compared their protein profiles by gel shift assays with the PCOL1 and PCOL2 probes. A similar set of complex bands was observed when using nuclear extracts from either cell type (not shown). We have further compared the HPO and MG-63 cell lines by performing Southwestern analyses. The

Discussion

Association studies have been widely used to identify genes responsible for complex diseases (reviewed in Refs. [11], [12]). Following a positive result, the question remains as to whether the studied polymorphism is the functional cause of the associated phenotype or, instead, it is in linkage disequilibrium with another site, which is the functional one. In a previous study [8], we described two new polymorphisms in a COL1A1 promoter region known to be involved in gene expression in

Acknowledgments

The authors are grateful to Dr. M. Alvarez and Dr. J.P. Bidwell for kindly supplying us with Nmp4/CIZ antibodies and expression constructs and to Dr. S.H. Ralston for critical review of the manuscript. The authors also thank the Serveis Científico-Tècnics, Universitat de Barcelona, for automated sequencing resources, and R. Rycroft for revising the English. N. Garcia-Giralt and M. Bustamante were recipients of fellowships from the CIRIT (Generalitat de Catalunya) and the Spanish Ministerio de

References (28)

  • Z. Bogdanovic et al.

    Parathyroid hormone inhibits collagen synthesis and the activity of rat col1a1 transgenes mainly by a cAMP-mediated pathway in mouse calvariae

    J. Cell. Biochem.

    (2000)
  • D.A. Brenner et al.

    Analysis of the collagen alpha 1(I) promoter

    Nucleic Acids Res.

    (1989)
  • C. Büttner et al.

    Transcriptional activation of the type I collagen genes COL1A1 and COL1A2 in fibroblasts by interleukin-4: analysis of the functional collagen promoter sequences

    J. Cell. Physiol.

    (2004)
  • N. García-Giralt et al.

    Two new single-nucleotide polymorphisms in the COL1A1 upstream regulatory region and their relationship to bone mineral density

    J. Bone Miner. Res.

    (2002)
  • Cited by (41)

    • Association of the Sp1 binding site and -1997 promoter variations in COL1A1 with osteoporosis risk: The application of meta-analysis and bioinformatics approaches offers a new perspective for future research

      2020, Mutation Research - Reviews in Mutation Research
      Citation Excerpt :

      In addition to the SNVs mentioned above, another variation in the COL1A1 promoter (-1663indelT) also plays a considerable role in osteoporosis risk [40,75]. Previous studies have shown that Nmp4 and Sp1 TFs bind to the sites harboring -1663indelT and +1245 G > T, respectively [75,120]. However, the TF that interacts with the -1997 G > T site remains unknown [75].

    • Genetics of musculoskeletal soft tissue injuries: Current status, challenges, and future directions

      2019, Sports, Exercise, and Nutritional Genomics: Current Status and Future Directions
    • Gene variants within the COL1A1 gene are associated with reduced anterior cruciate ligament injury in professional soccer players

      2013, Journal of Science and Medicine in Sport
      Citation Excerpt :

      This COL1A1 −1997 G/T polymorphism has been associated with bone mineral density (BMD), and has also been reported to interact with the COL1A1 Sp1 +1245G/T polymorphism to regulate BMD.17–19 Furthermore, a functional significance of the COL1A1 −1997 G/T polymorphism was demonstrated in the in vitro regulation of the COL1A1 gene in osteoblasts; the G allele showed a higher transcriptional activity than the T allele.20 Despite its functionality, thus far, the potential influence of the COL1A1 −1997 G/T polymorphism on the incidence of ACL rupture injury has never been studied.

    • Genetic Determinants of Bone Mass and Osteoporotic Fracture

      2008, Principles of Bone Biology: Volume 1-2, Third Edition
    • Type I Collagen Structure, Synthesis, and Regulation

      2008, Principles of Bone Biology: Volume 1-2, Third Edition
    View all citing articles on Scopus
    1

    Present address: URFOA-IMIM, Hospital del Mar, Universitat Autònoma de Barcelona, Barcelona, Spain.

    View full text