Skip Navigation


ToxSci Advance Access originally published online on August 14, 2008
Toxicological Sciences 2008 106(1):223-232; doi:10.1093/toxsci/kfn169
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
106/1/223    most recent
kfn169v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Galdones, E.
Right arrow Articles by Hales, B. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Galdones, E.
Right arrow Articles by Hales, B. F.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org

Retinoic Acid Receptor Gamma-Induced Misregulation of Chondrogenesis in the Murine Limb Bud In Vitro

Eugene Galdones and Barbara F. Hales1

Department of Pharmacology and Therapeutics, McGill University, Montréal, Québec, Canada H3G 1Y6

1 To whom correspondence should be addressed at Department of Pharmacology and Therapeutics, McGill University, 3655 Promenade Sir-William-Osler, Room 110, Montréal, Québec, Canada H3G 1Y6. Fax: (514) 398-7120. E-mail: barbara.hales{at}mcgill.ca.

Received June 18, 2008; accepted August 8, 2008


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
Vitamin A derivatives modulate gene expression through retinoic acid and rexinoid receptor (RAR/RXR) heterodimers and are indispensable for limb development. Of particular interest, RAR{gamma} is highly expressed in cartilage, a target affected following retinoid-induced limb insult. The goal of this study was to examine how selective activation of RAR{gamma} affects limb development. Forelimbs from E12.5 CD-1 mice were cultured for 6 days in the presence of all-trans RA (pan-RAR agonist; 0.1 or 1.0µM) or BMS-189961 (BMS961, RAR{gamma}-selective agonist; 0.01 or 0.1µM) and limb morphology assessed. Untreated limbs developed normal cartilage elements whereas pan-RAR or RAR{gamma} agonist-treated limbs exhibited reductive effects on chondrogenesis. Retinoid activity was assessed using RAREβ2 (retinoic acid response element β2)-lacZ reporter limbs; after 3 h of treatment, both drugs increased retinoid activity proximally. To elucidate the expression profiles of a subset of genes important for development, limbs were cultured for 3 h and cRNA hybridized to osteogenesis-focused microarrays. Two genes, matrix GLA protein (Mgp; chondrogenesis inhibitor) and growth differentiation factor-10 (Gdf10/Bmp3b) were induced by RA and BMS-189961. Real-time PCR was done to validate our results and whole mount in situ hybridizations against Mgp and Gdf10 localized their upregulation to areas of cartilage and programmed cell death, respectively. Thus, our results illustrate the importance of RAR{gamma} in mediating the retinoid-induced upregulation of Mgp and Gdf10; determining their roles in chondrogenesis and cell death will help further unravel mechanisms underlying retinoid teratogenicity.

Key Words: matrix GLA protein; growth differentiation factor 10; retinoic acid receptor gamma; retinoic acid; limb development; chondrogenesis.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
All-trans retinoic acid (RA), the most bioactive metabolite of vitamin A (retinol), is a potent morphogen that plays crucial roles during embryo development (Kochhar, 1967Go). Retinoid concentrations are tightly regulated during embryogenesis. Both deficiency (Hale, 1930Go) and excess can cause malformations of various embryonic structures in a time- and dose-dependent manner (Collins and Mao, 1999Go).

Limb development is an orchestrated process; a multitude of signals converge that specify and sculpt the heterogeneous (i.e., bone, muscle, skin) limb structure. Retinoids actively regulate various signaling pathways during normal limb morphogenesis, and it is no surprise that retinoid imbalances alter such developmentally important processes (Romand et al., 2004Go). Additionally, the molecular mechanisms responsible for patterning the limb bear striking similarities to various other organ systems, making the limb an excellent model to help decipher how teratogens illicit their detrimental effects on organogenesis (Yamada, 2005Go).

As the limb develops, mesenchymal cells are recruited, migrate, and condense, initiating chondrogenesis, the earliest phase in skeletal development. These mesenchymal condensations differentiate and give rise to cartilaginous anlagen that eventually yield the ossified elements that comprise the skeleton (Olsen et al., 2000Go). Previous studies have demonstrated that excess retinoid disrupts chondrogenesis (Kochhar, 1973Go; Kwasigroch and Kochhar, 1980Go). Although these effects on cartilage formation have been attributed to aberrant proliferation, differentiation and apoptosis (Ali-Khan and Hales, 2006Go; Zakeri and Ahuja, 1994Go), many of the genes and their underlying downstream mechanisms that regulate the retinoid-induced reduction in chondrogenesis are still unresolved.

Retinoids elicit their biological effects by regulating gene transcription. In the presence of ligand, retinoic acid:rexinoid nuclear receptor (RAR:RXR) heterodimers bind to retinoic acid response elements (RARE) and recruit and activate transcriptional machinery (Bastien and Rochette-Egly, 2004Go). Various RAR and RXR isotypes (and variant isoforms) have been identified ({alpha}, β, {gamma}) yet only pharmacological activation of RARs, and not RXRs, causes teratogenic effects, illustrating the importance of RARs in mediating the teratogenic signal in the developing limb (Arafa et al., 2000Go). Expression of RARs in the limb is temporally and spatially dynamic. During limb outgrowth, although RAR{alpha} is expressed ubiquitously and RARβ is restricted to the webbing, the cartilage tissue which is highly sensitive to retinoid insult, expresses RAR{gamma} (Dolle et al., 1989Go).

Single RAR isoform ablation studies have uncovered the importance of RAR{gamma} in controlling axial development and patterning, but receptor redundancy has been linked to the lack of effect seen in the appendicular skeletons of RAR{gamma}-deficient mice (Chambon, 1994Go; Lohnes et al., 1995Go). Notwithstanding, previous work in our lab has shown that limbs deficient in RAR{gamma} (on an {alpha}1-null background) were less susceptible to retinoid-induced limb dysmorphogenesis. Chondrogenesis and overall limb growth in these transgenic mice were vastly improved following retinoid insult, when compared with their retinoid-treated wild-type counterparts, indicating the importance of RAR{gamma} in mediating limb teratogenesis (Galdones et al., 2006Go).

Contrary to RAR{gamma} ablation, the objective of the present study was to determine the outcomes of aberrant RAR{gamma} activation on limb morphogenesis. To do so, an in vitro limb bud culture system was employed (Kwasigroch and Skalko, 1983Go) and aberrant RAR{gamma} activation was induced pharmacologically using BMS-189961 (BMS961), a highly RAR{gamma}-selective agonist (Klaholz et al., 2000Go). Given the predominantly high expression of RAR{gamma} in cartilage during limb development, we hypothesized that retinoid-induced limb dysmorphogenesis is mediated by a RAR{gamma}-dependent misregulation of genes important for proper chondrogenesis.

From a pharmacological standpoint, RAR{gamma} is an extensively pursued drug target for the treatment of several medical conditions including polycystic acne, psoriasis, diabetes, and various cancers (Johnson and Chandraratna, 1999Go). Unfortunately, such treatments are contraindicated for women of childbearing age. Understanding the modes of RAR{gamma}-mediated toxicity during development will effectively aid in the development of safer, less teratogenic retinoid analogs.

In this study, we demonstrate the teratogenic action of a RAR{gamma}-selective agonist on the developing limb. Retinoid-responsive reporter mice (RARβ2-lacZ) were employed to assess the location and extent of retinoid activity following activation of RAR{gamma}. Gene expression analysis using chondrogenesis-focused microarrays has identified two RAR{gamma}-responsive genes (Mgp [matrix GLA protein] and Gdf10 [growth differentiation factor-10]) that may have novel uncharacterized roles in transducing the teratogenic signal in the limb. Furthermore, in situ hybridization (ISH) has localized Mgp to areas of developing cartilage and Gdf10 to areas of programmed cell death in the limb, suggesting independent functions for each of these candidate genes.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
Limb bud culture and drug treatments.
Pregnant gestational day (GD) 12 CD-1 mice were euthanized, embryos were explanted and forelimbs were excised just lateral to the somites and cultured in vitro in a chemically defined culture medium, as previously described (Ali-Khan and Hales, 2003Go). All-trans RA (Sigma, St Louis, MI) and BMS-189961, a RAR{gamma}-selective agonist (BMS961; a gift from Bristol-Myers Squibb, Wallingford, CT), were dissolved in 100% EtOH. Limbs were exposed to low and high concentrations of RA (0.1 and 1.0µM) or BMS961 (0.01 and 0.1µM) at the onset of culture. All animal studies complied with the guidelines established by the Canadian Council on Animal Care.

Limb morphology.
Limbs were cultured for 6 days, with one change of the medium that was not supplemented with RA or BMS961, on day 3, as previously described (Galdones et al., 2006Go). Briefly, limbs were fixed and stained with 0.1% toluidine blue (Fisher Scientific, Montreal, Canada) and examined under a dissecting microscope (Wild Heerbrugg 99067; Wild Leitz, Ottawa, Canada) and photographs acquired with a JVC digital camera (JVC GCQX3HD, Tokyo, Japan). The extent of proper limb development was quantified using a limb morphogenetic scoring system (Neubert and Barrach, 1977Go). Five separate replicates were completed and examined for morphological changes or markers.

RAREβ2-hsp68lacZ reporter mice and staining.
To examine the extent and localization of retinoid activity, RAREβ2-hsp68lacZ double-transgenic males were mated with wild-type CD-1 females to produce single copy transgene offspring. These embryos were explanted at E12.5 and limbs cultured for 3 and 12 h as described above, in the absence or presence of RA or BMS961. Tissues were fixed immediately with 0.25% glutaraldehyde then stained with X-gal for 2–4 h, as previously described (Rossant et al., 1991Go). Limbs were rinsed in phosphate-buffered saline (PBS) then visualized under a dissecting microscope. Results were acquired from three independent limb culture experiments.

RNA extraction.
After 3 h in vitro, each group of six to eight forelimbs was removed from culture, rinsed with PBS and stored at –80°C in RNAlater RNA Stabilization Reagent (Qiagen, Mississauga, Canada) until further processing. Total RNA was extracted using the RNeasy Micro Kit (Qiagen) and quality and purity/quantity assessed by agarose gel electrophoresis and spectrophotometric ultraviolet analysis in 10mM Tris, pH 8.0 buffer, respectively. Five separate independent cultures of each treatment group were completed and used for microarray and quantitative RT-PCR analysis.

cRNA synthesis and hybridization.
Starting with 2 µg of total mRNA, the TrueLabeling-AMP 2.0 Kit (Superarray Bioscience, Frederick, MD) was used to produce, amplify and biotinylate antisense cRNA for hybridization following the manufacturer's guidelines. The cRNA product was purified with ArrayGrade cRNA Cleanup (SuperArray Bioscience) and then hybridized overnight to Mouse Oligo GEArray Osteogenesis Microarrays (OMM-026, SuperArray Bioscience), using 3 µg of cRNA per array and following the manufacturer's suggested HybTube Standard Protocol. The following day, the membranes were stained with an alkaline-phosphatase-conjugated streptavidin antibody (1:8000, SuperArray Bioscience) and the chemiluminescence signal was developed using the CDP-Star Kit (SuperArray Bioscience). Five arrays, each from independent limb cultures, were completed for each treatment group.

Data acquisition and gene expression analysis.
The array images were recorded using X-ray film, digitized using a desktop scanner, saved as 16-bit TIFF files and uploaded onto GEArray Expression Analysis Suite (GEASuite, SuperArray Bioscience). Array alignment and raw intensity quantification were done using GEASuite, then the data were imported into Agilent GeneSpring 7.3 GX (Agilent Technologies, Palo Alto, CA) where background correction and all further analysis was completed. For each individual array, the gene detection threshold was set at a raw signal intensity of 2x the background intensity; to minimize array-to-array variation, expression values were normalized to the median of all measurements on that individual array. Upon identification of the uniquely and commonly expressed genes in each group, statistical comparisons between control and RA or BMS961 treated groups were done using Student's t-tests (p ≤ 0.05). Genes were only included in the significantly regulated group of genes when significance was detected in at least three out of five replicates, and the changes were all in the same direction (i.e., either all up- or all downregulated).

Quantitative real-time PCR.
RNA was diluted to a final working solution of 10 ng/µl and Quantitect One-Step SYBR Green RT-PCR (Qiagen) was done using the Roche LightCycler (Roche Diagnostics, Laval, Canada). The quantitative real-time PCR (qRT-PCR) cycling parameters were as follows: 95°C for 15 min (one cycle) and then 94°C for 15 s, 55°C for 30 s, and 72°C for 20 s (for 50 cycles) and each sample was measured in duplicate. Embryonic head tissue was also collected, and RNA extracted to provide 1, 10, and 50 ng/µl RNA stocks used to generate standard curves for quantification. The primers were generated with Primer3 software (http://frodo.wi.mit.edu) and produced by alpha DNA (Montreal, Canada; see Table 1 for sequence information). To normalize the output, the expression of each gene of interest was divided by 18S gene expression, a commonly used housekeeping gene whose expression remains unchanged in murine limbs following retinoid exposure (Ali-Khan and Hales, 2006Go). Melting-curve analyses were done following each PCR to determine the output and detection quality (i.e., the formation of primer-dimers). Each treatment consisted of RNA from four separate culture experiments and each sample was measured in duplicate.


View this table:
[in this window]
[in a new window]

  		TABLE 1 Primer Sequences Used for Quantitative RT-PCR

 
Whole mount ISH.
Limbs cultured for 3, 6, or 12 h were rinsed with PBS containing 0.1% Tween 20 (PBT) and fixed overnight at 4°C in 4% paraformaldehyde. The following day, the limbs were dehydrated through a methanol/PBT gradient and stored at –20°C until further use. For whole mount ISH, a DIG RNA labeling kit (Roche Diagnostics) was used to synthesize Mgp and Gdf10 antisense probes from plasmids kindly provided by G. Karsenty (Columbia University, New York, NY) and S.-J. Lee (Johns Hopkins University, Baltimore, MD), respectively. Whole mount ISH was done as previously described (Decimo et al., 1995Go). After hybridization, limbs were rinsed in PBT and photographed under a dissecting stereomicroscope (Wild Leitz). Three separate whole mount ISH experiments were done for each gene of interest.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
Aberrant Activation of RAR{gamma} is Teratogenic to the Developing Limb
Because retinoid insult in the limb is attenuated in the absence of RAR{gamma} (Galdones et al., 2006Go), our first goal was to determine if and how aberrant RAR{gamma} activation affects limb development. In vitro cultured murine limbs were treated with either RA (pan-RAR agonist) or BMS961 (RAR{gamma}-selective agonist) at E12.5. The treatment of limbs with either drug resulted in dose-dependent detrimental effects on gross morphology after culture for 6 days, with marked effects on chondrogenesis, as evidenced by decreases in toluidine blue stained cartilage (Fig. 1A). After exposure to low concentrations of RA or BMS961 (RA: 0.1µM, BMS961: 0.01µM), the zeugopod (long bones) appeared growth retarded, whereas the autopod (paw) remained intact. However, at high exposures (RA: 1.0µM, BMS961: 0.1µM) both limb regions were equally undifferentiated.


Figure 1
View larger version (41K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. The effects of pan-RAR and RAR{gamma}-selective activation on limbs following 6 days in culture. Limbs were collected from CD-1 mice at embryonic day 12.5 and cultured in vitro for 6 days in the presence or absence of pan-RAR agonist, all-trans RA (0.1, 1.0µM) or RAR{gamma}-selective agonist, BMS961 (0.01, 0.1µM). (A) Limbs were stained with 0.1% toluidine blue (wt/vol 70% ethanol). Both RA and BMS961 treatment result in marked effects on chondrogenesis, namely long bone outgrowth and digit formation. (B) The limb morphogenetic scores measured the extent of differentiation and chondrogenesis following treatment.

 
A limb scoring system was used to determine the quality and extent of morphogenetic differentiation of the cartilage elements that comprise the forming limb after exposure to the pan- or RAR{gamma}-selective agonist (Neubert and Barrach, 1977Go). In accordance with the gross morphology, limb scores were compromised following treatment with either the pan- or RAR{gamma}-selective agonist (Fig. 1B).

RAR{gamma}-Selective Agonism Induces Retinoid Activity in RAREβ2-hsp68lacZ Reporter Limbs
To determine the amount and distribution of retinoid activity in limbs following RA or BMS961 exposure, we treated transgenic limbs from a retinoid reporter strain containing a transgene under the control of the retinoid sensitive RARβ2 response element (Rossant et al., 1991Go). In the absence of retinoid, the lacZ transgene was strongly expressed in the interdigital zones and observable along the axes of the forming long bones after 3 h in vitro (Fig. 2). With the addition of 0.1 or 1.0µM RA for 3 h, transgene expression was strongly upregulated in the proximal mesenchyme, necrotic zones, dorsal ectoderm, and interdigital webbing. At 3 h, BMS961 treatment (0.01 or 0.1µM) increased transgene expression in the proximal mesenchyme as well, but only in a modest fashion; transgene expression in the webbing was affected similarly in BMS961 and RA-treated limbs.


Figure 2
View larger version (55K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 2. Localization of RAR activity, as measured using the RAREβ2-hsp68lacZ retinoic-responsive reporter mice.

 
After 12 h in vitro, vehicle-treated limbs exhibited minimal transgene expression. In contrast, both pan-RAR and RAR{gamma}-selective activation induced restricted transgene expression in the zone of polarizing activity, the limb signaling center responsible for anterior-posterior patterning (Tickle, 2006Go). Additionally, although RA did not induce transgene expression in the interdigital zones at this later time point, BMS961 exposure resulted in a faint induction of staining in the webbing. Thus, because maximal effects on lacZ transgene expression were observed after 3 h in vitro, subsequent experiments were done at that early time point.

Changes in Gene Expression Induced by RA or BMS961 Treatment
It is clear that retinoids have detrimental effects on chondrogenesis (Jiang et al., 1995Go) yet the various mechanisms underlying this process are still unresolved. We adopted a gene array approach to assess how the regulation of chondrogenesis-related genes in the limb was affected following pan-RAR or RAR{gamma}-selective insult. Limbs were treated for 3 h with either RA or BMS961 and RNA was hybridized to Mouse Oligo Osteogenesis Microarrays from GEArray containing approximately 100 genes important in various aspects of cartilage and bone development.

In comparing the vehicle-treated limbs to those treated with 0.1 or 1.0µM RA, 55 genes were commonly expressed in all groups (Supplementary Table 1). Four genes were expressed uniquely under control conditions, whereas one gene was expressed only after exposure to 1.0µM RA (Fig. 3A). Only two genes were expressed in two out of three treatment groups (0 and 0.1µM; genes listed in Fig. 3A). Limbs treated with BMS961 exhibited a similar outcome, with the majority of genes commonly expressed between all groups (58; Supplementary Table 2) although one gene was expressed solely after exposure to low concentrations; four genes were expressed in both control and 0.01µM BMS961-treated limbs, but not in 0.1µM exposed limbs (genes listed in Fig. 3B). All uniquely expressed genes were only expressed at minimal baseline levels, therefore our analysis was limited to the groups of commonly expressed genes.


Figure 3
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 3. Venn diagrams of gene expression. (A) The numbers of genes that were uniquely and commonly expressed following exposure to 0, 0.1, or 1.0µM RA for 3 h. (B) The numbers of genes that were uniquely and commonly expressed following exposure to 0, 0.01, or 0.1µM BMS961 for 3 h. (C) Genes that were significantly up- or downregulated when comparing control-treated limbs to those treated with either low and high RA and/or BMS961 treatment.

 
Activation of RAR{gamma} is Implicated in the Misregulation of Mgp and Gdf10 Gene Expression
To identify genes important in mediating retinoid-induced limb teratogenesis, the lists of commonly expressed genes in all RA or BMS961 treatment groups were analyzed for statistical significance (p < 0.05) using Agilent GeneSpring 7.3 GX software. Comparison between vehicle and the two RA-treated groups identified nine genes that were significantly up- or downregulated by treatment (Table 2). Our work identified a subset of genes that were significantly changed (either up- or downregulated) following RA exposure (Table 2). Several of the identified genes are related to bone morphogenetic protein (BMP)/transforming growth factor beta (TGF-β) signaling (Smad2, Bmp4, and Gdf10) and components of cartilage and extracellular matrix (Col11a1, Spp1, and Dcn), whereas others are linked to matrix breakdown (Ctsk), oxidative stress (Nfkb1), and regulation of chondrogenesis (Mgp).


View this table:
[in this window]
[in a new window]

  		TABLE 2 List of Genes Significantly Up- or Downregulated following RA or BMS961 Treatment

 
When comparing the control versus the two BMS961-treated groups, only two genes were identified as being upregulated by RAR{gamma} activation (Table 2). The lists of significantly altered genes were compared; both of the genes altered by RAR{gamma}-selective activation were also misregulated by RA exposure (Fig. 3C). These genes were Mgp, a potent inhibitor of chondrogenesis and osteogenesis (Price and Williamson, 1985Go) and Gdf10, a bone morphogenetic protein of uncharacterized function in the developing limb (Cunningham et al., 1995Go).

Array Verification by qRT-PCR
We used qRT-PCR to assess the gene expression profiles of the two RAR{gamma}-responsive genes, Mgp and Gdf10. After culture for 3 h, the expression of both Mgp and Gdf10 was significantly upregulated by exposure to low or high concentrations of RA as well as the high concentration of BMS961; the expression of these genes was not responsive to the low concentration of BMS961 (Figs. 4A and 4B). Interestingly, after 12 h in culture, the expression of both genes did not differ from control in the RA or BMS961 exposed limbs, illustrating that the effect of the retinoids on the expression of these genes in the cultured limbs is transient; indeed, the expression of Gdf10 was significantly downregulated in 0.1µM RA-treated limbs at 12 h.


Figure 4
View larger version (26K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 4. qRT-PCR validation of the changes in expression of genes responsive to pan-RAR and RAR{gamma}-selective activation. Limbs were cultured in vitro for 3 (open bars) or 12 (closed bars) h in the absence or presence of RA (0.1, 1.0µM) or BMS961 (0.01, 0.1µM). (A) Mgp gene expression is upregulated by RA and BMS961 treatment after 3 h in culture but returns to baseline expression levels by 12 h. (B) Gdf10 expression is upregulated in response to RA and BMS961 treatment following 3 h, but not 12 h in vitro. (C) Dcn and (D) Ctsk are both downregulated by RA but not BMS961 after 3 h in culture. Each bar represents the mean gene expression of four separately conducted limb cultures normalized against 18S mRNA expression. Statistical analysis was done using two-way ANOVA and the Holm-Sidak post hoc multiple-comparison test. *Significant increase, p < 0.05; #significant decrease, p < 0.05.

 
In addition, to further validate the results from the array experiments, we selected two genes that were pan-RAR, but not RAR{gamma}-responsive. Dcn (decorin), a proteoglycan associaed with collagen (Scott and Orford, 1981Go) and ctsk (cathepsin K), a cysteine protease expressed in bone (Garnero et al., 1998Go) were both downregulated by exposure to 1.0µM RA for 3 h (Figs. 4C and 4D). As anticipated from the array data, exposure to either 0.1µM RA or BMS961 had no effect on the expression of these genes in limbs for 3 h in vitro.

Distribution of Mgp and Gdf-10 Gene Transcripts Following pan-RAR and RAR{gamma}-Selective Activation
Whole mount ISHs were done to determine the distribution of Mgp and Gdf10, our gene transcripts misregulated by RAR{gamma} activation. Vehicle-treated limbs exhibited very little Mgp staining at 3 h; Mgp expression was induced by RA in a concentration-dependent manner. The induction of Mgp expression was restricted to the forming long bones at low RA concentrations and was extended distally into the rays at high RA levels. Mgp expression was upregulated strongly by BMS961 in the forming long bones, as well as the proximal mesenchyme, but was not induced in the digits by exposure to either concentration of BMS961 (Fig. 5A). At 6 h, Mgp transcripts were detected in the long bones in vehicle exposed limbs; exposure to either RA concentration further upregulated Mgp expression at this time point, whereas BMS961 did not. Interestingly, by 12 h in vitro, little Mgp expression was found in any treatment group, echoing the results seen by qRT-PCR where a transient decrease in expression was observed (Fig. 4A).


Figure 5
View larger version (61K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 5. Localization of RAR{gamma}-responsive gene transcripts following pan-RAR or RAR{gamma}-selective activation. Whole mount ISH of (A) Mgp and (B) Gdf10 in E12.5 limbs following exposure to RA or BMS961 for 3, 6, or 12 h in vitro.

 
Localization of the expression of Gdf10 was distinct from that of Mgp. Gdf10 expression was faintly observable in the interdigital webbing under control conditions at 3 h (Fig. 5B). Ectopic expression was induced by exposure to either the low or high concentration of RA and shown to be highly localized to the interdigital webbing as well as the anterior and posterior necrotic zones, all areas that are predestined to apoptose and help sculpt the developing limb. In addition, high concentrations of RA induced Gdf10 transcripts in the proximal regions of the forming digits. BMS961 treatment resulted in similar expression patterns, albeit not as intense as their RA-treated counterparts. At 6 h, when control limbs expressed very little Gdf10, RA and BMS961 induced ectopic expression in the regions of programmed cell death and proximal rays. By 12 h in vitro, in RA-exposed limbs only faint staining was observable in the digits; however, BMS961-treated limbs still expressed low levels of Gdf10 in the webbing, necrotic zones and proximal rays (Fig. 5B).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
Although the effects of excess retinoids on development have been described, the mechanisms by which normal limb chondrogenesis and patterning are disrupted, and specifically the roles of distinct RARs, are not completely understood. Given that RAR{gamma} is highly expressed in the forming cartilage (Dolle et al., 1989Go), we hypothesized that RAR{gamma} mediates retinoid-induced limb dysmorphogenesis. Specifically, the goal of this study was to examine how pharmacological activation of RAR{gamma} affects limb chondrogenesis.

Our results show that treatment of limbs with either RA (pan-RAR agonist) or BMS961 (RAR{gamma}-selective agonist) in vitro is detrimental to limb chondrogenesis and growth. Microarray-based expression analysis of chondrogenesis-related genes identified a group of RA-responsive genes and two RAR{gamma}-responsive genes (Mgp and Gdf10). The localization of upregulated Mgp transcripts to chondrocytes, and those of Gdf10 to interdigital webbing/necrotic zones, suggests that they play novel roles during retinoid-induced limb dysmorphogenesis; namely the misregulation of chondrogenesis and programmed cell death, respectively.

RAR{gamma}-Induced Limb Defects
Both pan-RAR and RAR{gamma}-selective activation are detrimental to limb morphogenesis, significantly affecting the proper chondrogenesis of long bones, carpalia, and digits (Fig. 1). Specifically, although RA-treated limbs exhibited marked reductive effects on all cartilage elements at 0.1 and 1.0µM RA concentrations, BMS961 illicited phenotypically similar effects at concentrations that were 10-fold less (0.01 and 0.1µM). Although there is evidence that BMS961 can exhibit weak affinity for RARβ, the effective concentration (EC50) at which BMS961 binds RARβ is threefold higher than our highest treatment (Klaholz et al., 2000Go), suggesting that BMS961 is indeed acting as a RAR{gamma}-selective agonist in our limb bud cultures.

Retinoid Activity
The pattern of retinoid-induced lacZ upregulation in areas of endogenous RARβ expression has been observed previously (Rossant et al., 1991Go). Given that transgene expression is driven by three copies of the highly retinoid-responsive RARE of RARβ2 (RAREβ2), the specific sequence of RARβ2 may not have a strong binding affinity for RAR{gamma}. Indeed, RAR{gamma}-associated retinoid activity is most likely higher than observed, because RAR{gamma}-selective activation induced gene expression changes in areas (i.e., chondrocytes) where no lacZ transgene expression was detected (Fig. 5A).

Pan-RAR Responsive Genes
Several groups, including our own, have studied global gene expression changes in the limb following retinoid excess (Ali-Khan and Hales, 2006Go; Qin et al., 2002Go). To refine our search, and because chondrogenesis is severely affected by exogenous retinoid exposure, we assessed the expression changes of genes known to participate in aspects of cartilage and bone development.

Interestingly, the direction in which the genes were regulated was mixed; this is contrary to previous work illustrating the positive unidirectionally of the gene changes following exogenous retinol (vitamin A) exposure (Ali-Khan and Hales, 2006Go), a discrepancy that may be explained by the kinetics of the drug used. The immediate bioactivity of RA treatment leads to direct effects on their respective targets whereas retinol must be activated into RA, thus it would be of particular interest to assess retinol-induced gene changes following time points later than 3 h in vitro.

Classically, in the absence of ligand, RARs remain associated with RAREs and transcriptionally repress target genes; however, once a ligand is bound, the RARs change conformation, release their associated corepressors and in turn recruit coactivators and the transcriptional machinery that drive gene expression (Weston et al., 2003Go). In our study, although several genes were upregulated by RA treatment (Mgp, Gdf10, Spp1, and Bmp4), others (Ctsk, Col11a1, Nfkb1, Smad2, and Dcn) were significantly downregulated. How this downregulation occurs is not well understood, yet at least two explanations exist: (1) RA may upregulate an intermediate gene upstream, which in turn inhibits the transcription of the observed target gene, or (2) the liganded RAR/RXR heterodimers may bind negative response elements (NRE) that repress rather than activate transcription. These NREs have been identified upstream of several retinoid-responsive genes, including Mgp (Kirfel et al., 1997Go). However, because we have observed an upregulation of Mgp in our system, the activity of the NRE must be cell and tissue dependent.

We propose that these RA-responsive genes may play important novel roles in mediating retinoid-induced limb dysmorphogenesis. Because the scope of this study was focused on the RAR{gamma}-mediated regulation of gene expression, future studies will be required to determine how the misregulation of these newly identified RA-responsive genes affects limb morphogenesis.

RAR{gamma} Responsive Genes
In contrast to limbs treated with RA, BMS961-treated limbs exhibited a significant upregulation of only two genes: Mgp and Gdf10 (Fig. 3C). Whole mount ISHs were performed to determine the extent and localization of gene upregulation; Mgp was expressed and significantly upregulated in chondrocytes (Fig. 5A). Of interest, Gdf10 upregulation was limited to the interdigital webbing and the necrotic zones (Fig. 5B). These differential locations of expression suggest that each of these genes performs unique functions following retinoid-induced misregulation.

Inhibition of Chondrogenesis by Mgp
MGP, a {gamma}-carboxyglutamic acid-rich, vitamin K–dependent protein (Price and Williamson, 1985Go), has been characterized as a potent inhibitor of chondrocyte maturation and mineralization (Yagami et al., 1999Go). Mice lacking Mgp are viable, but exhibit spontaneous calcification of arteries and cartilage (Luo et al., 1997Go). Our array experiments and follow-up qRT-PCR analysis show that Mgp expression is significantly upregulated by RA and BMS961. These results are consistent with those of Cancela and Price (1992)Go who identified a putative RARE upstream of Mgp bearing a striking resemblance to the RAREβ2 (de The et al., 1990Go) and showed the RA-dependent upregulation of Mgp in various human tissues, including chondrocytes (Cancela and Price, 1992Go). However, although the RAR{gamma}-selective agonist did significantly upregulate Mgp, it did not do so to the same extent as RA (Fig. 3A), suggesting that other RARs ({alpha} and β) may participate in its complete transcriptional regulation.

Although Mgp transcripts were not detected by whole mount ISH in untreated limbs, RA, and BMS961-treated limbs aberrantly expressed Mgp in areas of chondrocyte development. Because chondrogenesis is a major target in retinoid-induced limb dysmorphogenesis, the distribution of Mgp transcripts may correlate with the reductive effects associated with retinoid insult. This is the first report implicating Mgp in a teratological context. MGP has been identified as playing a role as a functional inhibitor of BMP-2 and BMP-4 function (Yao et al., 2007Go). Given that regulated BMPs are required for proper limb development, further work is needed to determine whether Mgp is indispensable for mediating specific retinoid-induced aberrations during development.

Gdf10 and Programmed Cell Death
GDF10 is a member of the TGFβ superfamily and is highly related to BMP-3 (Cunningham et al., 1995Go), hence Gdf10 is also commonly referred to as BMP-3b. Expression of Gdf10 during organogenesis is highly localized in uterus, adipose and brain and less so in bone; ablation of Gdf10 resulted in no apparent effects on developmental outcome (Zhao et al., 1999Go). Many TGF-β superfamily members have been linked to various aspects of limb development: BMP-2, -4, and -7, the most highly characterized, play crucial roles in precartilaginous and chondrocyte differentiation, apical ectodermal ridge-dependent limb outgrowth, anterior-posterior patterning and interdigital cell death (Kawakami et al., 1996Go), whereas TGF-β and GDF5 have been implicated in joint development (Storm and Kingsley, 1999Go). Gdf10 function in the limb has not been previously characterized (Zhao et al., 1999Go). Experiments targeting osteoblast function in vitro have recently revealed that GDF10 can inhibit BMP-2–dependent alkaline phosphatase activity, hence illustrating the negative regulation of Gdf10 on osteogenesis (Hino et al., 1999Go). We are the first to show that Gdf10 can be transcriptionally regulated by retinoids and that its upregulation is localized specifically to domains of programmed cell death in the limb, namely the webbing and necrotic zones (Fig. 5B). We postulate that Gdf10 plays an integral role in regulating the apoptosis associated with retinoid teratogenicity (Ali-Khan and Hales, 2003Go).


   SUMMARY
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 SUMMARY
SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
These results show the importance of RAR{gamma} agonists in mediating retinoid-induced limb dysmorphogenesis. Our chondrogenesis-focused gene expression analysis has identified a handful of retinoid-regulated candidate genes that may play novel roles in limb dysmorphogenesis. Notably, our work is the first indication that Mgp and Gdf10 are significantly upregulated following pan-RAR or RAR{gamma}-selective activation. The identification of these candidate genes and their association with RAR{gamma} will help to unravel new mechanisms underlying retinoid-induced insult during development.


   SUPPLEMENTARY DATA
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
FUNDING
 REFERENCES
 
Supplementary data are available online at http://toxsci.oxfordjournals.org/.


   FUNDING
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
REFERENCES
 
Canadian Institutes for Health Research (FRN 7078).


   ACKNOWLEDGMENTS
 
We should like to thank Bristol-Myers Squibb for providing us with the RAR{gamma}-selective agonist, BMS-189961 and Dr Gerard Karsenty and Dr Se-Jin Lee for the Mgp and Gdf10 cDNA probes, respectively. Also, we are grateful to Dr Glenn Maclean for his technical advice.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 SUMMARY
 SUPPLEMENTARY DATA
 FUNDING
 REFERENCES
 
Ali-Khan SE, Hales BF. Caspase-3 mediates retinoid-induced apoptosis in the organogenesis-stage mouse limb. Birth Defects Res. A Clin. Mol. Teratol. (2003) 67:848–860.[CrossRef][Web of Science][Medline]

Ali-Khan SE, Hales BF. Novel retinoid targets in the mouse limb during organogenesis. Toxicol. Sci. (2006) 94:139–152.[Abstract/Free Full Text]

Arafa HM, Elmazar MM, Hamada FM, Reichert U, Shroot B, Nau H. Selective agonists of retinoic acid receptors: Comparative toxicokinetics and embryonic exposure. Arch. Toxicol. (2000) 73:547–556.[CrossRef][Web of Science][Medline]

Bastien J, Rochette-Egly C. Nuclear retinoid receptors and the transcription of retinoid-target genes. Gene (2004) 328:1–16.[CrossRef][Web of Science][Medline]

Cancela ML, Price PA. Retinoic acid induces matrix Gla protein gene expression in human cells. Endocrinology (1992) 130:102–108.[Abstract/Free Full Text]

Chambon P. The retinoid signaling pathway: Molecular and genetic analyses. Semin. Cell Biol. (1994) 5:115–125.[CrossRef][Medline]

Collins MD, Mao GE. Teratology of retinoids. Annu. Rev. Pharmacol. Toxicol. (1999) 39:399–430.[CrossRef][Web of Science][Medline]

Cunningham NS, Jenkins NA, Gilbert DJ, Copeland NG, Reddi AH, Lee SJ. Growth/differentiation factor-10: A new member of the transforming growth factor-beta superfamily related to bone morphogenetic protein-3. Growth Factors (1995) 12:99–109.[Web of Science][Medline]

de The H, Vivanco-Ruiz MM, Tiollais P, Stunnenberg H, Dejean A. Identification of a retinoic acid responsive element in the retinoic acid receptor beta gene. Nature (1990) 343:177–180.[CrossRef][Web of Science][Medline]

Decimo D, Georges-Labouesse E, Dolle P. In situ hybridization to cellular RNA. In: Gene Probes 2, a Practical Approach—Hames BD, Higgins SJ, eds. (1995) New York: Oxford University Press.

Dolle P, Ruberte E, Kastner P, Petkovich M, Stoner CM, Gudas LJ, Chambon P. Differential expression of genes encoding alpha, beta and gamma retinoic acid receptors and CRABP in the developing limbs of the mouse. Nature (1989) 342:702–705.[CrossRef][Web of Science][Medline]

Galdones E, Lohnes D, Hales BF. Role of retinoic acid receptors alpha1 and gamma in the response of murine limbs to retinol in vitro. Birth Defects Res. A Clin. Mol. Teratol. (2006) 76:39–45.[CrossRef][Web of Science][Medline]

Garnero P, Borel O, Byrjalsen I, Ferreras M, Drake FH, McQueney MS, Foged NT, Delmas PD, Delaisse JM. The collagenolytic activity of cathepsin K is unique among mammalian proteinases. J. Biol. Chem. (1998) 273:32347–32352.[Abstract/Free Full Text]

Hale K. Hereditary deforming chondrodysplasia or multiple exostoses: Report of a father and two daughters showing similar multiple symmetrical exostoses and fifty other cases collected from the English Literature since 1917. Ann. Surg. (1930) 92:92–114.[Web of Science][Medline]

Hino J, Matsuo H, Kangawa K. Bone morphogenetic protein-3b (BMP-3b) gene expression is correlated with differentiation in rat calvarial osteoblasts. Biochem. Biophys. Res. Commun. (1999) 256:419–424.[CrossRef][Web of Science][Medline]

Jiang H, Soprano DR, Li SW, Soprano KJ, Penner JD, Gyda M 3rd, Kochhar DM. Modulation of limb bud chondrogenesis by retinoic acid and retinoic acid receptors. Int. J. Dev. Biol. (1995) 39:617–627.[Web of Science][Medline]

Johnson A, Chandraratna RA. Novel retinoids with receptor selectivity and functional selectivity. Br. J. Dermatol. (1999) 140(Suppl. 54):12–17.[CrossRef][Web of Science][Medline]

Kawakami Y, Ishikawa T, Shimabara M, Tanda N, Enomoto-Iwamoto M, Iwamoto M, Kuwana T, Ueki A, Noji S, Nohno T. BMP signaling during bone pattern determination in the developing limb. Development (1996) 122:3557–3566.[Abstract]

Kirfel J, Kelter M, Cancela LM, Price PA, Schule R. Identification of a novel negative retinoic acid responsive element in the promoter of the human matrix Gla protein gene. Proc. Natl. Acad. Sci. U. S. A. (1997) 94:2227–2232.[Abstract/Free Full Text]

Klaholz BP, Mitschler A, Belema M, Zusi C, Moras D. Enantiomer discrimination illustrated by high-resolution crystal structures of the human nuclear receptor hRARgamma. Proc. Natl. Acad. Sci. U. S. A. (2000) 97:6322–6327.[Abstract/Free Full Text]

Kochhar DM. Teratogenic activity of retinoic acid. Acta Pathol. Microbiol. Scand. (1967) 70:398–404.[Web of Science][Medline]

Kochhar DM. Limb development in mouse embryos. I. Analysis of teratogenic effects of retinoic acid. Teratology (1973) 7:289–295.[CrossRef][Web of Science][Medline]

Kwasigroch TE, Kochhar DM. Production of congenital limb defects with retinoic acid: Phenomenological evidence of progressive differentiation during limb morphogenesis. Anat. Embryol. (Berl) (1980) 161:105–113.[CrossRef][Medline]

Kwasigroch TE, Skalko RG. Development of mouse limbs in organ culture: Dose dependent retinoic acid-induced defects evaluated using image analysis. Prog. Clin. Biol. Res. (1983) 110(Pt A):335–344.[Medline]

Lohnes D, Mark M, Mendelsohn C, Dolle P, Decimo D, LeMeur M, Dierich A, Gorry P, Chambon P. Developmental roles of the retinoic acid receptors. J. Steroid. Biochem. Mol. Biol. (1995) 53:475–486.[CrossRef][Web of Science][Medline]

Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, Karsenty G. Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature (1997) 386:78–81.[CrossRef][Web of Science][Medline]

Neubert D, Barrach H-J. Techniques applicable to study morphogenetic differentiation of limb buds in organ culture. In: Methods in Pre-Natal Toxicology—Neubert D, Merker H-J, Kwasigroch T, eds. (1977) Stuttgart: Georg Thieme Publishers. 241–251.

Olsen BR, Reginato AM, Wang W. Bone development. Annu. Rev. Cell Dev. Biol. (2000) 16:191–220.[CrossRef][Web of Science][Medline]

Price PA, Williamson MK. Primary structure of bovine matrix Gla protein, a new vitamin K–dependent bone protein. J. Biol. Chem. (1985) 260:14971–14975.[Abstract/Free Full Text]

Qin P, Cimildoro R, Kochhar DM, Soprano KJ, Soprano DR. PBX, MEIS, and IGF-I are potential mediators of retinoic acid-induced proximodistal limb reduction defects. Teratology (2002) 66:224–234.[CrossRef][Web of Science][Medline]

Romand R, Niederreither K, Abu-Abed S, Petkovich M, Fraulob V, Hashino E, Dolle P. Complementary expression patterns of retinoid acid-synthesizing and -metabolizing enzymes in pre-natal mouse inner ear structures. Gene Expr. Patterns. (2004) 4:123–133.[CrossRef][Medline]

Rossant J, Zirngibl R, Cado D, Shago M, Giguere V. Expression of a retinoic acid response element-hsplacZ transgene defines specific domains of transcriptional activity during mouse embryogenesis. Genes Dev. (1991) 5:1333–1344.[Abstract/Free Full Text]

Scott JE, Orford CR. Dermatan sulphate-rich proteoglycan associates with rat tail-tendon collagen at the d band in the gap region. Biochem. J. (1981) 197:213–216.[Web of Science][Medline]

Storm EE, Kingsley DM. GDF5 coordinates bone and joint formation during digit development. Dev. Biol. (1999) 209:11–27.[CrossRef][Web of Science][Medline]

Tickle C. Making digit patterns in the vertebrate limb. Nat. Rev. Mol. Cell Biol. (2006) 7:45–53.[CrossRef][Web of Science][Medline]

Weston AD, Blumberg B, Underhill TM. Active repression by unliganded retinoid receptors in development: Less is sometimes more. J. Cell. Biol. (2003) 161:223–228.[Abstract/Free Full Text]

Yagami K, Suh JY, Enomoto-Iwamoto M, Koyama E, Abrams WR, Shapiro IM, Pacifici M, Iwamoto M. Matrix GLA protein is a developmental regulator of chondrocyte mineralization and, when constitutively expressed, blocks endochondral and intramembranous ossification in the limb. J. Cell. Biol. (1999) 147:1097–1108.[Abstract/Free Full Text]

Yamada G. Reproductive/urogenital organ development and molecular genetic cascades: Glamorous developmental processes of bodies. J. Biochem. (2005) 137:665–669.[Abstract/Free Full Text]

Yao Y, Nowak S, Yochelis A, Garfinkel A, Bostrom KI. Matrix GLA protein, an inhibitory morphogen in pulmonary vascular development. J. Biol. Chem. (2007) 282:30131–30142.[Abstract/Free Full Text]

Zakeri ZF, Ahuja HS. Apoptotic cell death in the limb and its relationship to pattern formation. Biochem. Cell. Biol. (1994) 72:603–613.[Web of Science][Medline]

Zhao R, Lawler AM, Lee SJ. Characterization of GDF-10 expression patterns and null mice. Dev. Biol. (1999) 212:68–79.[CrossRef][Web of Science][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
106/1/223    most recent
kfn169v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Galdones, E.
Right arrow Articles by Hales, B. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Galdones, E.
Right arrow Articles by Hales, B. F.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?