SEARCH

SEARCH BY CITATION

Keywords:

  • tendon;
  • Scleraxis;
  • transgenic reporter;
  • connective tissue

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Defects in tendon patterning and differentiation are seldom assessed in mouse mutants due to the difficulty in visualizing connective tissue structures. To facilitate tendon analysis, we have generated mouse lines harboring two different transgene reporters, alkaline phosphatase (AP) and green fluorescent protein (GFP), each expressed using regulatory elements derived from the endogenous Scleraxis (Scx) locus. Scx encodes a transcription factor expressed in all developing tendons and ligaments as well as in their progenitors. Both the ScxGFP and ScxAP transgenes are expressed in patterns recapitulating almost entirely the endogenous developmental expression of Scx including very robust expression in the tendons and ligaments. These reporter lines will facilitate isolation of tendon cells and phenotypic analysis of these tissues in a variety of genetic backgrounds. Developmental Dynamics, 2007. © 2007 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

In the past 20 years, the ability to create gain- and loss-of-function alleles of genes in mice has led to a revolution in our understanding of embryonic development. The phenotypic analysis of transgenic mice has often focused on the skeleton because it is easy to visualize in cleared and stained preparations. Subsequently, as markers for various progenitor cell populations were identified (such as Sox9 for chondrocytes or the myogenic basic helix–loop–helix [bHLH] transcription factors for myoblasts), it also became routine to look for defects in various differentiation pathways when analyzing mutants. Until very recently, however, genetic effects on tendons and ligaments were largely ignored. These thin connective tissue elements are not easy to visualize by histological stains and require painstaking dissection to free them from surrounding soft tissues for direct examination. Moreover, no markers existed to illuminate the progenitor cell populations for these tissues. Hence, little was learned about their genesis and patterning. Nonetheless, the tendons are a key component of the musculoskeletal system, structurally linking the muscles to the skeleton, and transmitting the mechanical forces between them. Furthermore, because muscles atrophy when not activated, even during embryogenesis (Rong et al.,1992; Edom-Vovard et al.,2002), it is likely that tendon-specific malformations in mutants might be mischaracterized as conditions affecting muscle formation.

Several years ago, we identified the bHLH transcription factor Scleraxis (Scx) as a very specific marker for all the connective tissues attaching muscle and bone, throughout their development (Schweitzer et al.,2001). Early Scleraxis expression marks the progenitor cell populations for these tissues, and Scleraxis continues to be strongly expressed in the mature tendons and ligaments throughout embryogenesis. Examination of changes in the expression of Scleraxis, thus, offers an ideal tool for analyzing alterations in tendon differentiation or patterning. To facilitate such analysis, we have produced two transgenic reporters in which mice express the markers alkaline phosphatase (AP) and green fluorescent protein (GFP) driven by Scx regulatory sequences such that AP and GFP are produced at very high levels in a pattern that largely recapitulates endogenous Scx expression.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Generation of the Transgenic ScxGFP and ScxAP Mice

The Scx gene consists of two exons, with the majority of the coding sequence contained within exon 1 (Fig. 1). To build the transgenic constructs, we used an 11-kb genomic clone from the Scx locus that extends ∼4 kb upstream and ∼5 kb downstream of the Scx gene (Fig. 1 and Brown et al.,1999). The open reading frames of the GFP and AP genes were cloned into the first exon of Scleraxis within the genomic fragment, beginning with the initiator ATG of Scx and replacing most but not all of the first exon (Fig. 1, see the Experimental Procedures section). The constructs were designed so that the splice donor sequences of the first exon were retained and the transgene transcripts are spliced to include the second Scx exon.

thumbnail image

Figure 1. Schematic representation of the ScxGFP and ScxAP constructs. The transgenic constructs were based on Scxg12, an 11-kb genomic clone from the Scx locus (Brown et al.,1999). ScxGFP and ScxAP were subcloned such that the open reading frames (ORFs) for green fluorescent protein (GFP) and alkaline phosphatase (AP), respectively, were introduced at the endogenous ATG, and the ORF replaced most but not all of the first exon (see the Experimental Procedures section). Open boxes, untranslated regions (UTRs); stippled boxes, coding regions.

Download figure to PowerPoint

The transgene constructs were injected into fertilized eggs and transgenic lines of mice were derived (see the Experimental Procedures section for details). Two independent ScxGFP lines and one ScxAP line were obtained. All lines showed robust expression that clearly labeled the tendons and tendon progenitors. We initially performed a developmental comparison of the expression in the different transgenic lines. At every stage and in every tissue, the two ScxGFP lines and the ScxAP line gave comparable results, indicating that the expression observed reflects the activity of the transgenic promoter and not position effects relating to the site of transgene insertion in the genome. We further established that one of the ScxGFP transgenic lines was homozygous viable and showed no phenotypic consequences. We, therefore, focused on this line for further analysis.

ScxGFP and ScxAP Transgenic Reporters Recapitulate Scx Expression

To determine whether the transgenic reporters faithfully recapitulate the endogenous Scx expression domains, a detailed analysis was carried out of the ScxGFP and ScxAP lines at a variety of developmental stages (Figs. 2, 3, and not shown). Both lines show robust expression and both recapitulate Scx expression faithfully at most stages.

thumbnail image

Figure 2. ScxAP and ScxGFP recapitulate Scx expression in tendons and ligaments. Color reaction was performed on ScxAP embryos to detect the ScxAP signal. A:ScxAP at embryonic day (E) 10.5. B,C: Tendon progenitors in the branchial arches of embryos at E9.5 detected in an ScxAP embryo (B) and in a wild-type (WT) embryo using whole-mount in situ hybridization with a Scx probe (C). D,E:ScxGFP in a skinned head and calvarium at E18.5. F:ScxGFP in a rib section at E14.5. G:ScxGFP expression the tail of E14.5 embryo. H:ScxGFP in a skinned tail, 3 weeks after birth. Brightfield image of the tail highlights the junction between tail vertebrae (red arrows), in which the annulus fibrosis, the ligaments connecting tail vertebrae, is also expressing in ScxGFP mice. I: Cruciate ligaments of the knee in a skinned and dissected knee at E18.5.

Download figure to PowerPoint

thumbnail image

Figure 3. ScxGFP is useful for analyzing limb tendons in whole-mount and tissue sections. A,B:ScxGFP in a skinned forelimb at embryonic day (E) 18.5. C: Details of the skeletal junction of the patellar tendon. D: Longitudinal section of a forelimb of ScxGFP embryo at E15.5, stained with an anti myosin heavy chain mAb.

Download figure to PowerPoint

At embryonic day (E) 10.5 endogenous Scx expression is strongly detected in both the tendon progenitor domain of the developing limb buds (Schweitzer et al.,2001) and in a somitic compartment of tendon progenitors, called the syndetome, at the dorsolateral edge of the sclerotome (Brent et al.,2003). Much weaker expression can also be seen at this stage marking the cranial tendon progenitors in the branchial arches (Fig. 2C). The ScxGFP (not shown) and ScxAP (Fig. 2A) reporters robustly mark each of these domains, including quite noticeable expression in the branchial arches, which, in both lines, is clearly seen by E9.5 (Fig. 2B and not shown). Thus, the ScxAP and ScxGFP reporters closely approximate the normal expression of Scx in tendon progenitors, providing an exceptional tool for analyzing changes in tendon specification and development.

The ScxAP and ScxGFP transgenic animals continue to express the two markers at high levels at later stages of development, robustly highlighting the formation of all tendons, ligaments, and other connective tissue attachments between skeletal elements and between the skeletal elements and the muscles. For example, in the cranial region, ScxGFP clearly outlines the head tendons at E18.5 (Fig. 2D) as well as the connective tissue between the unfused embryonic cranial sutures (Fig. 2E). Similarly, the tendinous layer connecting the ribs to the intercostal muscles is easily visualized both in whole embryo specimens (not shown) and in sections at E18.5 (Fig. 2F). In the tail, the ScxGFP reporter displays the transformation of the syndetome as the progenitors start to form tendons (Fig. 2G). The transgene signal persists after birth and continues to highlight both the long tendons in the tail and the ligaments between the tail bones, the annulus fibrosis, in 3-week-old mice (Fig. 2H). All ligaments are easily visualized in ScxGFP embryos, as seen by examination of the anterior and medial cruciate ligaments of the knee shown at E18.5 (Fig. 2I).

The differences between the fluorescent signal of ScxGFP and the color reaction of ScxAP mice affect the choice of transgene to use in specific experiments. ScxAP provides better delineation and detail of expression in young embryos and soft tissues (Fig. 2A,B), while the fluorescence in ScxGFP is better suited for analysis of large tissues, precluding the need to perform any staining procedure, an advantage that is evident when evaluating the integrity of mature tendons. The robustness of the ScxGFP reporter is illustrated in dorsal and ventral views of the ScxGFP transgene in a skinned forelimb of an E18.5 embryo (Fig. 3A,B) where every tendon and ligament is strikingly displayed. The signal from the ScxGFP transgene is very strong, and analysis of these tissues can also be performed in thin sections. At high magnification of histological sections, these tools allow the tendon anatomy to be examined with cellular resolution, for example at the insertion of the patellar tendon (Fig. 3C). Moreover, combining ScxGFP or ScxAP staining with antibody detection allows the relationship between the tendons and neighboring tissues to be examined easily and in exquisite detail. For example, a histological section through the digit of an E15.5 ScxGFP embryo, counterstained with an antibody directed against myosin heavy chain, provides the possibility to follow entire tendons from their origin in the muscle and up to the skeletal insertion (Fig. 3D).

The ability to detect the GFP reporter in adult animals and thick tissues enables the detection of unexpected domains of expression through simple scanning of tissues. Indeed, in addition to the expression previously noted in tendons and ligaments, expression in a host of other tissues was evident in the ScxGFP mice. Each of these nontendon domains is also recapitulated by both the ScxGFP and ScxAP transgenes, and was also subsequently verified by in situ hybridization as an authentic site of endogenous Scx expression (Fig. 4C, and not shown). ScxGFP expression was detected in the epididimis and the Sertoli cells of the testes (Fig. 4A; Muir et al.,2005), the epithelium of the bronchi of the lungs (Fig. 4B,C), the kidneys, adrenal glands and ureter (Fig. 4D,E), and a distinct group of cell within the hair follicle starting 10 days after birth (Fig. 4F).

thumbnail image

Figure 4. Expression of ScxGFP in tissues outside the musculoskeletal system. A: Section through a testis and epididimis at embryonic day (E) 18.5. B,C: Section through a lung of an ScxGFP embryo at E18.5(B) and whole-mount in situ hybridization of lungs at E13.5 with a Scx probe (C). D,E: Kidneys of an ScxGFP embryo at E18.5 viewed in darkfield (D) and fluorescence (E). Red arrowheads, adrenal glands; purple arrowhead, ureter. F: Ventral view of tail skin from a 3-week-old ScxGFP mouse.

Download figure to PowerPoint

ScxGFP and ScxAP Reporter Overexpression in Early Limb Buds

While the transgenic tendon reporters recapitulate Scx expression very faithfully in most tissues, in early embryonic limb stages, expression patterns of the ScxAP and ScxGFP do not precisely coincide with the endogenous expression of Scx as assessed by in situ hybridization. The difference is most pronounced in E10.5–E12.5, before overt tendon formation (Schweitzer et al.,2001). The transgenic reporters show a broader, albeit overlapping, domain than that encompassed by the Scx-expressing cells identified by in situ hybridization. See, for example, the broader “progenitor” domain in the E12.5 limb as detected by in situ hybridization detecting the ScxGFP transcript, compared with the expression of the endogenous Scx gene (Fig. 5A,B). The same expanded domain of expression is seen in both transgenic reporters with their respective reporter signals. While the endogenous expression domains can still be clearly seen in the ScxGFP transgenic limb, these domains appear expanded compared with nontransgenic animals. This difference is specific to the limbs, and expanded reporter domains of Scleraxis expression are not as pronounced in axial tissues at the same time. One possibility, therefore, is that the expanded domain is simply a reflection of the ability to detect low level expression due to the high transcript level in the ScxGFP and ScxAP reporters.

thumbnail image

Figure 5. Overexpression of ScxGFP in early limbs and putative regulatory elements in the Scx locus. A,B: Whole-mount in situ hybridization with a full-length Scx probe (ScxFL) on forelimbs of wild-type (WT; A) and ScxGFP (B) embryos at embryonic day (E) 12.5. C: Comparison of the human and murine genomes in the Scleraxis locus reproduced from the VISTA Web site (Dubchak and Ryaboy,2006). The plot represents levels of similarity between the two genomes. Numbering of the conserved noncoding genomic sequences was arbitrarily assigned. Purple, coding sequences in the Bop1 and Scx genes; light blue, untranslated region sequences; pink, conserved noncoding genomic sequences. The ScxGFP construct was schematically drawn to scale, to represent the genomic sequences included in the transgenic construct.

Download figure to PowerPoint

Indeed, in situ hybridization on ScxGFP tissues with a probe that detects both the endogenous and the transgenic transcripts (see the Experimental Procedures section) results in much stronger signals in the transgenic embryo, that develop significantly faster than the endogenous transcripts in wild-type (WT) embryos. Although we did not directly quantitate transcript levels, we note that in section in situ hybridization on transgene and WT tissues of the same stage and with the same probe (see the Experimental Procedures section), we will need to stop the transgene reaction after 20–30 min, but carry the WT tissue reaction for 1–2 days. An alternative explanation for the broader Scx domains observed in the early transgenic limbs is that the Scx reporters are missing some regulatory elements required to faithfully represent Scx expression.

Structure of the Scleraxis Promoter

To further explore this latter possibility that the transgenic constructs are missing regulatory elements necessary for limiting the domain of Scx expression in early limbs, we examined the genomic locus of the Scleraxis gene more closely. Regulatory elements are frequently found in stretches of noncoding DNA that are highly conserved in evolution (Nobrega et al.,2003; Dermitzakis et al.,2005). To identify such domains in the Scx locus, we used the VISTA software that provides a graphic representation of comparisons of whole genomes (Frazer et al.,2004; Dubchak and Ryaboy,2006). We find nine blocks of sequence that are highly conserved in all mammals in stretches of between 100 bp and up to 400 bp (Fig. 5C). The conservation of these sequences in all mammals is highly suggestive that they may have indeed been maintained through positive selection due to a regulatory function. The Scleraxis gene is located within the fourth intron of a second gene called Bop1 which is transcribed in an inverse orientation (Fig. 5C). Interestingly, the conserved noncoding domains were grouped near the Scx gene, mostly within the Bop1 intron that includes the Scx gene. Additional conserved noncoding domains are found only outside the Bop1 gene, mostly again downstream to the gene (not shown). It is, therefore, likely that the nine conserved noncoding domains around the Scx gene represent Scx regulatory domains. Notably, two of these (blocks eight and nine) were not included in the transgenic constructs we designed (bottom, Fig. 5C), consistent with the possibility that they may represent repressor elements that are missing in the transgenic constructs. In the future, it will be interesting to add these domains to a transgenic construct and see if the expanded limb domain will indeed be reduced to resemble the in situ signal of Scx in early limb stages. A recent study attempting to use Scleraxis regulatory elements to drive reporter gene expression in tendons has shown that a construct encompassing sequences 5′ to the Scx gene and only 1.8 kb of genomic sequence 3′to the Scleraxis gene was sufficient to drive expression in tendons, but not all tendons were labeled in these mice and regions outside the endogenous domains of Scx expression were also labeled, further reinforcing the importance of additional 3′ regulatory elements included in our construct to faithfully recapitulate the regulation of the Scx gene (Perez et al.,2003).

We have generated two novel transgenic reporters, ScxAP and ScxGFP, that allow tendon and ligament progenitors to be followed with a high degree of fidelity, and which allow the mature connective tissue anatomy to be easily visualized. Both transgenic reporters are highly sensitive, giving extremely robust signals with virtually no background, from embryonic stages through the first 3 months after birth. As the tendons and ligaments become increasingly acellular postnatally, this expression fades and becomes undetectable by 8 months. Crossing various gain- and loss-of-function mutations into the ScxAP and ScxGFP lines provides a unique opportunity to assess the impact of genetic changes on the differentiation and patterning of tendons and ligaments. Furthermore, tendon and ligament cells can be readily isolated from these mice for a variety of experimental purposes.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Cloning of the ScxGFP and ScxAP Transgenic Constructs

The cloning strategy was based on a unique BglII fragment in the Scxg12, genomic clone encoding the Scx locus (Brown et al.,1999), that extends from 700 bp upstream of the ATG to 40 bp from the end of the first exon. The 700-bp fragment encoding the sequence immediately upstream of the Scx start ATG was subcloned, and an SmaI site was introduced at the 3′ end of the fragment to allow for blunt ligation into this site of any desired open reading frame. AP and GFP (gift of Connie Cepko) were cloned into the SmaI site, and the AP or GFP fused fragments were reintroduced into the Scxg12 genomic clone, with AP or GFP replacing most but not all of the first exon. NotI or SalI sites at the two ends of the original genomic clone can be used to isolate the transgenic construct. The ScxGFP transgene can be identified directly in all stages of development and, therefore, does not require polymerase chain reaction (PCR) genotyping. ScxAP was identified by PCR of genomic DNA using the primers 5′GGACATTGACGTGATCCTAG and 5′GCCGTCCTTGAGCAGATAG.

The transgenic construct does not disrupt the splicing donor at the end of the first exon of Scx, so that the second exon of Scx is spliced with the transgenic transcript. Consequently, the transgenic constructs are detected by the full length in situ hybridization probe used by most labs to detect Scx expression (ScxFL). To detect the endogenous expression of Scx in tissues from a ScxGFP mouse, we, therefore, generated an additional probe with sequences just from the first exon of Scx (Scxex1).

Histochemistry, Immunohistochemistry, and In Situ Hybridization

For whole-mount visualization of fluorescent signals, tissues were skinned and observed directly or after 2–6 hr fixation in 4% paraformaldehyde. Pictures were captured on an MZFLIII dissecting microscope with a DXM1200 camera (Nikon). For sections, the tissue was fixed as above and cryoembedded, and 12-μm sections were collected. Myosin heavy chain was detected with My32 (1:400, Sigma). Whole-mount in situ hybridizations was performed using the protocol found on the Tabin lab Web page: http://genepath.med.harvard.edu/∼cepko/protocol/ctlab/ish.ct.htm. The full-length Scx in situ hybridization probe was the Scx mRNA cloned in Brown et al. 1999.

For AP detection, embryos were fixed overnight in 4% paraformaldehyde and washed several times in phosphate buffered saline (PBS). To inactivate endogenous AP activity, embryos were heated for 1.5 hr at 70°C in PBS. Embryos were then cooled to room temperature, equilibrated in NTM staining buffer (100 mM NaCl, 100 mM Tris pH9.5, 50 mM MgCl2), and incubated from several hours to overnight in NTM with 1 mg/ml nitroblue tetrazolium and 0.1 mg/ml 5-bromo-4-chloro-3-indolyl phosphate at 4°C.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The authors thank Dr. Andy McMahon and Mike Rule for injection of the transgenic constructs and Tim Riordan and Spencer Watson for technical assistance. The research was supported in part by Research Grants from the Shriners hospitals for children and by the March of Dimes Birth Defects Foundation (Basil O'Connor Starter Scholar Research Grant, and NIH grant P01 DK56246 to C.J.T.).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES