HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 21, 19019-19026, May 24, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/21/19019    most recent M200125200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Terraz, C. Articles by Rossert, J. Search for Related Content PubMed PubMed Citation Articles by Terraz, C. Articles by Rossert, J. Social Bookmarking                      What's this?

A Combination of cis-Acting Elements Is Required to Activate the Pro-1(I) Collagen Promoter in Tendon Fibroblasts of Transgenic Mice^*

Catherine Terraz, Gaelle Brideau, Pierre Ronco, and Jérôme Rossert^§

From the INSERM U489 and Université Paris VI, 75020 Paris, France

Received for publication, January 5, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The genes encoding the two type I collagen chains are selectively activated in few cell types, including fibroblasts and osteoblasts. By generating transgenic mice, we have previously shown that the activity of the mouse pro-1(I) promoter was controlled by separate cell-specific cis-acting elements. In particular, a sequence located between 3.2 and 2.3 kb was needed to induce expression of the reporter gene at high levels in tendon fibroblasts. In the present work, by using the same transgenic approach, we have identified two short elements in this sequence, named tendon-specific element (TSE) 1 and TSE2, that were necessary to direct reporter gene expression selectively in tendon fibroblasts. Gel shift assays showed that TSE1 and TSE2 bound proteins specifically present in nuclear extracts from tendon fibroblasts and that the sequence of TSE2 binding a tendon-specific protein corresponded to an E-box. Analysis of transgenic mice further indicated that TSE1 and TSE2 needed to cooperate not only with each other but also with other cis-acting elements of the proximal promoter to activate reporter gene expression in tendon fibroblasts. Similarly, it pointed out that the so-called osteoblast-specific element had to interact with downstream sequences to drive reporter gene expression in osteoblasts of transgenic mice. Thus, expression of the mouse pro-1(I) collagen gene in tendon fibroblasts appears to be the result of a unique combination of different cis-acting elements, including TSE1 and TSE2.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Type I collagen is a fibrillar collagen composed of two 1 chains and one 2 chain coiled around each other in a triple helix. It is a major protein of mammalian bodies and an essential component of most extracellular matrices. In the extracellular space, type I collagen molecules self-assemble into highly organized fibrils and then into fibers, which largely contribute to the high tensile strength of the framework supporting body structures (reviewed in Ref. 1). Evidence for a role of type I collagen in providing mechanical strength to tissues comes from pathophysiological analyses of genetic diseases resulting from mutations in one of the genes encoding type I collagen, such as osteogenesis imperfecta and Ehlers-Danlos syndromes type VIIA and VIIB (2). The hallmark of osteogenesis imperfecta is brittle bones, but the disease can also involve other tissues rich in type I collagen, such as ligaments, tendons, fascia, sclerae, and teeth (3). Ehlers-Danlos syndromes are characterized by skin hyperextensibility, vascular fragility, and increased ligament elasticity responsible for joint hypermobility (4).

Type I collagen is synthesized by a discrete subset of cells of mesenchymal origin, which contrasts with its wide distribution throughout the body. These cells are mostly fibroblasts, osteoblasts, and odontoblasts. Studies using transgenic mice harboring various fragments of the mouse and rat pro-1(I) collagen promoters have shown that a modular arrangement of separate cell-specific cis-acting elements was responsible for the expression of the pro-1(I) collagen gene in different subsets of type I collagen-producing cells (reviewed in Ref. 5). Analyses of transgenic mice showed that the mouse pro-1(I) promoter contained at least three different cell-specific regulatory elements. An element located within 900 bp of the proximal promoter induced lacZ and luciferase reporter gene expression in some skin fibroblasts. A second element located between 2.3 and 0.9 kb conferred high-level expression of these two reporter genes in osteoblasts and odontoblasts. A third element located between 3.2 and 2.3 kb was responsible for high-level expression of the lacZ gene in tendon and fascia fibroblasts (6). Similarly, in the rat pro-1(I) promoter, a 13-bp sequence located between 1683 and 1670 bp was necessary to drive chloramphenicol acetyltransferase reporter gene expression in bones, whereas an upstream sequence, located between 1.7 and 3.5 kb, was required to obtain high-level expression of the reporter gene in tendon fibroblasts (7, 8). The human pro-1(I) collagen promoter probably displays such a modular organization because the first 2.3 kb of this promoter induced growth hormone reporter gene expression in bones, tendons, and fascia of transgenic mice, but not in other type I collagen-containing tissues, such as perichondrium and skeletal muscles (9, 10).

Analysis of the rat and mouse pro-1(I) promoters led to the precise identification of an osteoblast-specific element (7, 11). Introduction of short mutations or deletions in this cis-acting element abolished the expression of the reporter gene in osteoblastic cells of transgenic mice (7), whereas transgenic mice harboring the mouse osteoblast-specific element multimerized four times and cloned upstream of a minimal promoter and the lacZ gene displayed X-gal¹ staining selectively in osteoblasts (11). In contrast, no fibroblast-specific element has been precisely identified thus far in either the pro-1(I) or pro-2(I) collagen genes (reviewed in Ref. 5).

We report the identification of two short cis-acting sequences of the mouse pro-1(I) promoter that are necessary for reporter gene expression at high levels specifically in tendon fibroblasts of transgenic embryos and bind nuclear proteins selectively present in tendon fibroblasts. We also show that these two elements, as well as the osteoblast-specific element, need to cooperate with other elements of the proximal promoter to drive expression of the lacZ reporter gene in tendons and ossification centers, respectively. This suggests that type I collagen expression in tendon fibroblasts and osteoblasts is finely tuned by unique combinations of cis-acting elements and complex interactions between trans-acting factors.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmid Construction-- For plasmid construction, we used placH as an expression vector. It contains the lacZ reporter gene cloned upstream of a sequence of the mouse protamine gene that supplies an intron and a polyadenylation signal and downstream of a polycloning site (6). pC2310 was generated by cloning a HindIII/XbaI fragment of the mouse pro-1(I) collagen promoter extending from 2310 bp to +110 bp in pLacH. pC2741 was obtained by cloning a SacI/HindIII fragment extending from 2741 bp to 2310 bp into pC2310. PCR products corresponding to segments of the pro-1(I) collagen promoter extending from 3150 to 2520 bp, 3050 to 2679 bp, 2879 to 2310 bp, 2879 to 2679 bp, 2845 to 2310 bp, and 2845 to 2364 bp were generated using Pfu polymerase and primers containing a HindIII restriction site, digested with HindIII, and cloned into the HindIII site of pC2310, generating pC3150210, pC3050369, pC2879, pC2879369, pC2845, and pC284554, respectively. pC4M was obtained by subcloning four copies of the segment extending from 2879 to 2679 bp upstream of four copies of the 117-bp osteoblast-specific element and 220 bp of the pro-1(I) proximal promoter in pA-ClacZ (11). pC28791317 was generated by subcloning a PCR fragment extending from 2879 to 1537 bp into pJ22, which contains 220 bp of the pro-1(I) proximal promoter cloned upstream of the lacZ gene in placH. pC320TSE1 was generated by site-directed mutagenesis using the QuikChange Site-directed Mutagenesis Kit (Stratagene, La Jolla, CA) and primers containing an 18-bp deletion extending from 2762 to 2745 bp (5'-AGAGAGACAGAACTCAGA-3'), following the manufacturer's instructions. Deletions were confirmed by sequencing pC320TSE1 (Genome Express, Paris, France).

Generation and Analysis of Transgenic Mice-- Transgenic embryos were generated using standard procedures (12). -Galactosidase activity was assessed on embryonic day 15.5 (E15.5) embryos as described previously (6). To analyze transgene integration, genomic DNA was extracted from yolk sacs with the DNeasy Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer's instructions, and a sequence of the lacZ gene was amplified by PCR, as described previously (13). For histology, embryos were dehydrated in ethanol, fixed in xylene, and embedded in paraffin. Microtome sections (4-5 µm) were counterstained with solid red.

Cell Lines-- FV6M cells are primary cultures of tendon fibroblasts that were kindly provided by F. Ruggiero (Institut de Biologie et Chimie des Protéines, Lyon, France). They were obtained from dorsal tendons of calf fetuses and cultured in Dulbecco's modified Eagle's medium supplemented with 4.5 g/liter glucose and 10% fetal calf serum (Invitrogen).

ROS17/2.8 cells are rat osteoblastic cells, which produce type I collagen. They were cultured in 50% Dulbecco's modified Eagle's medium and 50% Ham's F-12 (Invitrogen) supplemented with 10% fetal calf serum.

NIH3T3 cells are mouse fibroblastic cells. They were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum.

The RC.SVtsA58 cell line is a rabbit collecting duct cell line that has been generated in our laboratory, and it does not produce type I collagen (14). These cells were cultured in 50% Dulbecco's modified Eagle's medium and 50% Ham's F-12 supplemented with 2 mM glutamine, 5 mg/liter insulin, 50 nM dexamethasone, 5 mM transferrin, 30 nM selenium, 20 mM HEPES, and 2% fetal calf serum.

Electrophoretic Mobility Shift Assay-- Nuclear extracts from ROS 17/2.8, RC.SVtsA58, FV6M, and NIH3T3 cells were prepared as described previously (15). The probes were end-labeled by filling-in with [-³²P]dCTP using the Klenow fragment of the Escherichia coli DNA polymerase I. Then, 0.3 ng of each probe was incubated with 10 µg of nuclear extracts at room temperature for 30 min in 20 µl of binding buffer containing 100 mM Tris-HCl, 100 mM NaCl, 20 µM ZnSO₄, 1 mM dithiothreitol, and 500 ng of poly(dI-dC)·poly(dI-dC). Competition experiments were performed using a 60-fold molar excess of unlabeled double-stranded oligonucleotides (Table I). The complexes were resolved by electrophoresis in 7% polyacrylamide gels containing 22 mM Tris-borate (pH 8.0) and 0.5 mM EDTA.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

A Tendon-specific Enhancer Is Located between 2844 and 2741 bp-- Previously published data suggested that a tendon-specific element was located between 3.2 and 2.3 kb, in the mouse pro-1(I) promoter (6). To more precisely identify this regulatory element, we generated transgenic mice harboring different subsegments of the fragment extending from 3.2 to 2.3 kb. Each of these subsegments was cloned upstream of 2.3 kb of the mouse pro-1(I) proximal promoter and lacZ. This gene was used as a reporter because X-gal staining allows easy detection of tissues and cells expressing the transgene during embryonic development. Foster mothers were sacrificed at 15.5 days of gestation because at this time the endogenous gene is expressed in most type I collagen-producing cells of the embryo, including tendon fibroblasts and osteoblasts (16).

We first generated transgenic mice harboring a construct containing 2310 bp of the pro-1(I) proximal promoter cloned upstream of lacZ (pC2310; Fig. 1A) to confirm that this segment of the promoter induced an expression of the reporter gene restricted to osteoblastic cells. As expected, only ossification centers were stained with X-gal, and histological sections confirmed that the staining was restricted to osteoblasts (data not shown). We then generated transgenic mice harboring a construct containing 2741 bp of the pro-1(I) proximal promoter cloned upstream of lacZ (pC2741; Fig. 1A). Analysis of whole-mount embryos showed that five embryos expressed the lacZ reporter gene, and in all five cases, X-gal staining was restricted to ossification centers (Fig. 1B), a pattern of expression exactly similar to the one observed in E15.5 embryos harboring pC2310. Because these results suggested that an element located upstream of 2741 bp was required for lacZ expression in tendon fibroblasts, we then generated transgenic mice using a construct that contained 2845 bp of the pro-1(I) proximal promoter (pC2845; Fig. 1A). All four whole-mount embryos that showed X-gal staining expressed the transgene in ossification centers and also in tendons (Fig. 1, C and D). Histological sections confirmed that the reporter gene was expressed selectively in osteoblastic cells (data not shown) and tendon fibroblasts (Fig. 1, E and F). Similarly, the six embryos harboring a construct containing 2879 bp of the pro-1(I) proximal promoter (pC2879; Fig. 1A) showed X-gal staining in ossification centers and tendons (Fig. 1, G and H). Thus, taken together, these results showed that an element located between 2845 and 2741 bp was required for high-level expression of the lacZ reporter gene in tendon fibroblasts. We named this element TSE1.

View larger version (52K): [in this window] [in a new window]   Fig. 1.   Analysis of the expression of the lacZ reporter gene in E15.5 transgenic embryos. A, schematic representation of the constructs used to generate transgenic embryos. The constructs are not drawn to scale. B, representative pattern of expression of pC2741 in a whole-mount transgenic embryo after X-gal staining. Only ossification centers were stained. C and D, representative pattern of expression of pC2845 in a whole-mount transgenic embryo after X-gal staining. The arrows indicate tendons stained by X-gal. E and F, histological sections of embryos showing X-gal staining of tendon fibroblasts (arrows). G and H, representative pattern of expression of pC2879 in a whole-mount transgenic embryo after X-gal staining. Tendons and ossification centers were stained. The arrows indicate tendons stained by X-gal.

TSE1 Can Bind a DNA-binding Protein Expressed Specifically in Tendon Fibroblasts-- Comparison of TSE1 with sequences of the rat, human, and bovine pro-1(I) promoters showed that TSE1 is highly conserved between mouse ant rat but that only 36 bp of TSE1 are highly similar to the corresponding sequences of the human and bovine promoters (Fig. 2). These sequences are located between 2857 and 2823 bp in the human promoter, 2750 and 2715 bp in the rat promoter, and 2782 and 2747 bp in the bovine promoter (Fig. 2).

View larger version (8K): [in this window] [in a new window]   Fig. 2.   Sequence analysis of TSE1. Comparison of the sequence of TSE1 (GenBankTM accession number U38307) with sequences of the rat, human, and bovine pro-1(I) collagen promoters (GenBankTM accession numbers JO4464, X_98705, and AJ307029, respectively) shows that the 38-bp segment located at its 3' end is highly similar to corresponding sequences of the human, rat, and bovine promoters. Dash, identical nucleotide; dot, missing nucleotide.

To study the nuclear proteins able to bind TSE1, we focused on the 36-bp sequence that is well conserved between species, and we performed electrophoretic mobility shift assays using nuclear extracts from RC.SVtsA58, ROS17/2.8, NIH3T3, and FV6M cells. One complex, named complex A, was observed with all four different nuclear extracts (Fig. 3A). A second complex, named complex B, was only observed with nuclear extracts obtained from type I collagen-producing cells (Fig. 3A). Interestingly, a third complex, named complex C, was only observed with nuclear extracts obtained from tendon fibroblasts (Fig. 3A). Competition experiments showed that formation of this complex was completely abolished by a 60-fold molar excess of unlabeled probe (Fig. 3A). To more precisely define the sequence of TSE1 involved in the formation of complex C, we performed competition experiments using double-stranded oligonucleotides harboring short deletions within TSE1 (Table I). All competitors inhibited the formation of complex C, but competitor F had a much weaker effect (Fig. 3B). Competitor F contains a deletion that destroys a GAACT motif (Table I). Computer analysis of TSE1 did not disclose any putative DNA-binding protein that could interact with the sequence encompassing the GAACT motif (17). Because of this lack of candidate DNA-binding protein, we decided to confirm the role of the sequence identified with electrophoretic mobility shift assays by generating transgenic mice harboring a deletion encompassing the GAACT motif. A 3.2-kb segment of the pro-1(I) proximal promoter containing an 18-bp deletion extending from 2762 to 2745 bp was cloned upstream of the lacZ reporter gene and used to generate transgenic mice. Four E15.5 transgenic embryos expressed the lacZ reporter gene at high levels in ossification centers but not in tendons (Fig. 3C), confirming that the sequence binding complex C is essential to induce high-level expression of the reporter gene in tendons.

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Analysis of the proteins binding to TSE1. A, gel shift experiments were performed using a double-stranded oligonucleotide corresponding to the 3' end of TSE1 (2779 to 2741 bp) and nuclear extracts from different cell lines. Complex C (arrow) was selectively present when nuclear extracts from FV6M cells were used (lane 4). It disappeared in competition experiments with a 60-fold molar excess of unlabeled probe (lane 5). B, competition experiments were performed using double-stranded oligonucleotides harboring short deletions within the 2779 to 2741 bp sequence (Table I). Formation of complex C was inhibited by all competitors, but competitor F had a much weaker effect. C, representative pattern of expression of the lacZ gene in an E15.5 transgenic embryo harboring a 3.2-kb segment of the pro-1(I) proximal promoter that contains an 18-bp deletion extending from 2762 to 2745 bp (pC320TSE1). The lacZ reporter gene was expressed in ossification centers but not in tendons. The construct is not drawn to scale.

TSE1 Cooperates with Another Tissue-specific Element to Induce High-level Expression of the lacZ Gene in Transgenic Embryos-- To test whether TSE1 was sufficient to drive high-level expression of the reporter gene in tendon fibroblasts, it was multimerized four times and cloned upstream of four copies of the osteoblast-specific element and 220 bp of the pro-1(I) proximal promoter (pC4M; Fig. 4A). We have previously shown that this 220-bp segment of the promoter can be considered a minimal promoter (6). All four embryos expressing pC4M showed X-gal staining of ossification centers, but, on the contrary, none of them expressed the reporter gene in tendons and tendon fibroblasts (Fig. 4B; data not shown), suggesting that TSE1 needs to cooperate with other element(s) of the pro-1(I) proximal promoter to induce high-level expression of the lacZ gene in tendon fibroblasts. To identify these elements, we generated transgenic mice harboring a segment of the pro-1(I) promoter extending from 2879 to 2679 bp cloned upstream of 2310 bp of the pro-1(I) proximal promoter and the lacZ gene (pC2879369; Fig. 4A). Analysis of the five embryos that expressed the reporter gene showed that X-gal staining of ossification centers was present in all cases but that there was no staining of tendons (Fig. 4C). Histological analysis of these transgenic embryos also did not disclose any staining of tendon fibroblasts or of other fibroblastic cells (data not shown). Similarly, the five embryos expressing pC284554 (Fig. 4, A and D), the six embryos expressing pC3050369 (Fig. 4, A and E), and the three embryos expressing pC3150210 (Fig. 4, A and F) showed X-gal staining in ossification centers but not in tendons, and histological sections confirmed the absence of staining in tendon fibroblasts (data not shown).

View larger version (38K): [in this window] [in a new window]   Fig. 4.   Analysis of the expression of the lacZ reporter gene in transgenic embryos. A, schematic representation of the constructs used to generate transgenic embryos. The constructs are not drawn to scale. Four lines indicate four copies of the DNA sequence. BF, representative pattern of expression of pC4M (B), pC2879369 (C), pC284554 (D), pC3050369 (E), and pC3150210 (F) in whole-mount transgenic embryos after X-gal staining. In all cases, the staining was restricted to ossification centers. Inlets are detailed views of the forelimbs.

Taken together, these data suggested that, besides TSE1, a sequence located between 2363 and 2316 bp was necessary to drive reporter gene expression in tendon fibroblasts. We named this second tendon-specific cis-acting element TSE2.

TSE2 Binds a Tendon-specific DNA-binding Protein-- Comparison of TSE2 with sequences of the rat, human, and bovine pro-1(I) promoters showed that it was highly similar to sequences located between 2447 and 2401 bp in the human promoter, 2345 and 2298 bp in the rat promoter, and 2350 and 2303 bp in the bovine promoter (Fig. 5).

View larger version (8K): [in this window] [in a new window]   Fig. 5.   Sequence analysis of TSE2. Comparison of the sequence of TSE2 (GenBankTM accession number U38307) with sequences of the rat, human, and bovine pro-1(I) collagen promoters (GenBankTM accession numbers JO4464, X_98705, and AJ307029, respectively) shows that TSE2 is highly similar to corresponding sequences of the rat, human, and bovine promoters. Dash, identical nucleotide; dot, missing nucleotide.

To study the nuclear proteins able to bind TSE2, we performed electrophoretic mobility shift assays using nuclear extracts from RC.SVtsA58, ROS17/2.8, NIH3T3, and FV6M cells and a probe corresponding to the 48-bp sequence of TSE2. Several retarded complexes were observed in nuclear extracts obtained from all type I collagen-producing cells (FV6M, ROS17/2.8, and NIH3T3 cells), but interestingly, one additional complex, named complex E, was observed selectively with nuclear extracts obtained from FV6M cells (Fig. 6A). Complex E was partially competed by competition with a 60-fold molar excess of an unlabeled probe (Fig. 6A). To more precisely define the sequence of TSE2 responsible for the formation of complex E, we performed competition experiments using double-stranded oligonucleotides harboring short deletions within TSE2 (Table I). Formation of complex E was inhibited by all competitors exept competitor P, which contains a deletion that destroys an E-box (Fig. 6B; Table I). Thus, it is likely that complex E corresponds to the binding of basic helix-loop-helix proteins to TSE2.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Electrophoretic mobility shift analysis of the proteins binding to TSE2. A, gel shift experiments were performed using a double-stranded oligonucleotide corresponding to TSE1 (2364 to 2310 bp) and nuclear extracts from different cell lines. Complex E (arrow) was selectively present when nuclear extracts from FV6M cells were used (lane 4). This complex, like the others, was competed by a 60-fold molar excess of an unlabeled probe corresponding to TSE2 (lane 2). B, competition experiments were performed using double-stranded oligonucleotides harboring short deletions within the TSE2 sequence (Table I). Formation of complex E was inhibited by all competitors, except competitor P.

TSE1 and TSE2 Need to Interact with Downstream Elements-- The 1.6-kb segment of the mouse pro-1(I) proximal promoter has been reported to be involved in expression of the lacZ reporter gene in tendon fibroblasts of transgenic mice (18). To test whether TSE1 and TSE2 need to cooperate with this segment to induce expression of the lacZ gene at high levels in tendon fibroblasts, we generated transgenic mice harboring a 2879-bp segment of the pro-1(I) proximal promoter with an internal deletion between 1537 and 220 bp (pC28791317). None of the 12 transgenic embryos expressed the reporter gene in tendons (data not shown), suggesting that TSE1 and TSE2 need to cooperate with a cis-acting element located between 1537 and 220 bp. Moreover, these transgenic embryos also did not express the reporter gene in ossification centers, even though the transgene contained the osteoblast-specific element (data not shown). Thus, the osteoblast-specific element also needs to cooperate with other cis-acting elements located between 1537 and 220 bp when it is not multimerized.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Molecular mechanisms governing type I collagen gene expression are still quite elusive; in particular, the cis-acting elements that induce expression of the pro-1(I) collagen gene in fibroblastic cells are unknown (reviewed in Ref. 5). Previous studies performed using transgenic mice harboring various segments of the mouse pro-1(I) promoter have shown that the 3.2 to 2.3 kb segment contains cis-acting elements that are necessary to induce high-level expression of the lacZ reporter gene in tendon fibroblasts (6). To identify these cis-acting elements, we have generated transgenic embryos harboring different subsegments of the 3.2 to 2.3 kb fragment cloned upstream of a 2.3-kb segment of the pro-1(I) proximal promoter and the lacZ reporter gene. The lacZ gene allows precise identification of the cell types in which the transgene is active. The use of the 2.3-kb segment of the pro-1(I) proximal promoter permits easy identification of embryos that express the lacZ gene because this sequence induces high-level expression of the reporter gene in ossification centers (6). Foster mothers were sacrificed at 15.5 days of gestation because at this time the expression of the lacZ gene can be easily detected in ossification centers and developing tendons (6). At least three embryos expressing the lacZ gene were obtained for each construct to eliminate a site of integration effect. Our deletion analysis demonstrated that two short cis-acting elements located between 3.2 and 2.3 kb were necessary to induce high-level expression of the reporter gene specifically in tendon fibroblasts of transgenic mice. These elements, which we have named TSE1 and TSE2, are located about 2.8 and 2.3 kb upstream of the transcription start site, respectively. Mouse embryos harboring a transgene containing both TSE1 and TSE2 cloned upstream of a 2.3-kb segment of the pro-1(I) proximal promoter consistently co-expressed the reporter gene in tendon fibroblasts and ossification centers. In contrast, embryos harboring a transgene containing either TSE1 alone or TSE2 alone, cloned upstream of 2.3 kb of the pro-1(I) proximal promoter, expressed the reporter gene in ossification centers at high levels, but no staining of tendons could be detected in whole-mount embryos or in histological sections. Comparison of the sequences of TSE1 and TSE2 and the sequences of the rat, bovine, and human pro-1(I) promoters showed that the most 3' part of TSE1 and the whole TSE2 have highly similar counterparts, as seen for the osteoblast-specific element (11). Gel shift experiments showed that both TSE1 and TSE2 can bind proteins that are present in nuclear extracts from tendon fibroblasts, but not proteins from other type I collagen-producing cells or from epithelial cells. Competition experiments performed using double-stranded oligonucleotides harboring 5-bp deletions within the TSE2 sequence showed that the tendon-specific nuclear proteins binding TSE2 actually bound a consensus E-box (CACGTG) located at 2325 bp. E-boxes are known to bind basic helix-loop-helix transcription factors, and it is very tempting to speculate that the E-box located within TSE2 binds a basic helix-loop-helix transcription factor named scleraxis. Schweitzer et al. (19) have recently shown that in chick embryos, the expression of scleraxis becomes rapidly specific to the developing tendons and ligaments and that in the developing mouse limb, scleraxis transcripts are selectively found in tendons and their progenitors. Furthermore, they have also shown that induction of excess of scleraxis-positive mesenchymal cells did not result in the production of extra tendons (19). This last finding is fully consistent with the fact that activation of the pro-1(I) proximal promoter in tendon fibroblasts required the presence of TSE2 and also of TSE1. Analysis of TSE1 using electrophoretic mobility shift assays in the presence of competitors also led to the identification of a short sequence containing a GAACT motif that bound a tendon-specific nuclear protein. Computer analysis of this sequence did not enable us to identify a potential DNA-binding protein that could interact with it. Nevertheless, the role of this sequence has been confirmed by generating transgenic harboring an 18-bp deletion that encompasses the GAACT motif within TSE1. Whereas E15.5 transgenic embryos expressed the lacZ reporter gene at high levels in ossification centers, no X-gal staining could be detected in tendons.

Interestingly, when TSE1 was multimerized four times and cloned upstream of four copies of the osteoblast-specific element and 220 pb of the pro-1(I) proximal promoter, the lacZ reporter gene was expressed in ossification centers but not in tendons, and no staining of tendon fibroblasts could be detected by histological analysis. This shows that unlike the osteoblast-specific element, multimerization of TSE1 cannot overcome the absence of other cis-acting elements (11). Moreover, generation of transgenic mice harboring a deletion between 1537 and 220 bp (i.e. between the osteoblast-specific element and the minimal promoter) showed that TSE1 and TSE2 need to cooperate with elements located within this sequence to induce expression of the lacZ gene in tendon fibroblasts of transgenic mice. Similarly, the absence of staining of ossification centers in embryos harboring this deleted construct showed that the osteoblast-specific element also needs to cooperate with downstream elements. The need for such cooperativity had previously been masked by the systematic usage of four copies of the osteoblast-specific element (11).

The existence of cis-acting elements located between 1537 and 220 bp that are able to induce expression of reporter genes in tendons and ossification centers is consistent with data obtained from other groups with the human, rat, and mouse pro-1(I) promoters (9, 10, 18, 20, 21). Slack et al. (10) and Liska et al. (9) have shown that a transgene containing 2.3 kb of the human pro-1(I) collagen promoter cloned upstream of the growth hormone gene drove expression of this reporter gene in ossification centers and also in tendons. Bedalov et al. (20) have shown that a sequence of the rat promoter located between 1670 and 944 is responsible for reporter gene expression in tendons at low levels. Similarly, Houglum et al. (18) have shown that transgenic mice harboring a construct containing 1626 bp of the mouse pro-1(I) collagen promoter cloned upstream of lacZ expressed the reporter gene in tendons at low levels. More recently, Kern et al. (21) have shown the existence of a second osteoblast-specific element in the pro-1(I) promoter. They showed that four copies of a sequence of the mouse pro-1(I) collagen promoter extending from 1347 to 1338 bp cloned upstream of 220 bp of the pro-1(I) proximal promoter drove expression of the luciferase reporter gene in developing bones (21). Thus, expression of the lacZ gene in tendon fibroblasts and also in ossification centers appears to be the result of a unique combination of different cis-acting elements. Whereas some of these elements bind tissue-specific transcription factors, others may bind factors that are ubiquitously expressed or expressed in both osteoblasts and tendon fibroblasts.

Tissue-specific expression induced by a unique combination of different transcription factors has already been demonstrated for different genes, such as the albumin gene, the fibroblast growth factor 4 gene, and the atrial natriuretic peptide gene (22-24). For example, the albumin gene requires the recruitment by hepatocyte nuclear factor-3 of several cis-activators and their corresponding DNA-binding proteins to be expressed in the liver (22). Similarly, Sox2 and Oct3 bind to adjacent sites and synergistically drive fibroblast growth factor-4 gene expression (23). Finally, the transcription factor Nkx-2.5 recruits other DNA-binding proteins such as GATA-4 by direct protein-protein interactions to activate the atrial natriuretic peptide gene (24). Similarly, type I collagen expression in fibroblasts and osteoblasts may also be the result of finely tuned regulation under the control of unique combinations of cis-acting elements and complex interactions between trans-acting factors. We speculate that their identification might help to better define the molecular mechanisms responsible for the differentiation of mesenchymal cells into fibroblasts.

	ACKNOWLEDGEMENTS

We thank F. Ruggiero for the generous gift of FV6M cells and M. Delauche for expert technical assistance. We are also grateful to J. Chambaz and C. Lasne for welcoming and helping us in the IFR 58 transgenic facility.

	FOOTNOTES

^* This work was supported by grants from the Association pour la Recherche sur le Cancer (to J. R.) and from the Poix Legacy to the University of Paris (to J. R.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Recipient of a fellowship from the Ministère de l'Education, de la Recherche, et de la Technologie.

^§ To whom correspondence should be addressed: INSERM U489, Hopital Tenon, 4 rue de la Chine, 75020 Paris, France. Tel.: 33-1-56-01-69-93; Fax: 33-1-56-01-69-99; E-mail: jerome.rossert@tnn.ap-hop- paris.fr.

Published, JBC Papers in Press, March 6, 2002, DOI 10.1074/jbc.M200125200

	ABBREVIATIONS

The abbreviations used are: X-gal, 5-bromo-4-chloro-3-indolyl--D-galactopyranoside; TSE, tendon-specific element.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				V. Lejard, G. Brideau, F. Blais, R. Salingcarnboriboon, G. Wagner, M. H. A. Roehrl, M. Noda, D. Duprez, P. Houillier, and J. Rossert Scleraxis and NFATc Regulate the Expression of the Pro-{alpha}1(I) Collagen Gene in Tendon Fibroblasts J. Biol. Chem., June 15, 2007; 282(24): 17665 - 17675. [Abstract] [Full Text] [PDF]

				T. T. Antoniv, S. Tanaka, B. Sudan, S. De Val, K. Liu, L. Wang, D. J. Wells, G. Bou-Gharios, and F. Ramirez Identification of a Repressor in the First Intron of the Human {alpha}2(I) Collagen Gene (COL1A2) J. Biol. Chem., October 21, 2005; 280(42): 35417 - 35423. [Abstract] [Full Text] [PDF]

				H. Nagato, N. Matsuo, H. Sumiyoshi, K. Sakata-Takatani, M. Nasu, and H. Yoshioka The Transcription Factor CCAAT-binding Factor CBF/NF-Y and Two Repressors Regulate the Core Promoter of the Human Pro-{alpha}3(V) Collagen Gene (COL5A3) J. Biol. Chem., November 5, 2004; 279(45): 46373 - 46383. [Abstract] [Full Text] [PDF]

				M. Ponticos, T. Partridge, C. M. Black, D. J. Abraham, and G. Bou-Gharios Regulation of Collagen Type I in Vascular Smooth Muscle Cells by Competition between Nkx2.5 and {delta}EF1/ZEB1 Mol. Cell. Biol., July 15, 2004; 24(14): 6151 - 6161. [Abstract] [Full Text] [PDF]

				O. Rahkonen, M. Su, H. Hakovirta, I. Koskivirta, S. G. Hormuzdi, E. Vuorio, P. Bornstein, and R. Penttinen Mice With a Deletion in the First Intron of the Col1a1 Gene Develop Age-Dependent Aortic Dissection and Rupture Circ. Res., January 9, 2004; 94(1): 83 - 90. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/21/19019    most recent M200125200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Terraz, C. Articles by Rossert, J. Search for Related Content PubMed PubMed Citation Articles by Terraz, C. Articles by Rossert, J. Social Bookmarking