Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Oct 1.
Published in final edited form as: Dev Biol. 2015 Jul 31;406(1):92–103. doi: 10.1016/j.ydbio.2015.07.022

Spatiotemporal regulation of GLI target genes in the mammalian limb bud

Jordan P Lewandowski 1, Fang Du 2, Shilu Zhang 2, Marian B Powell 1, Kristin N Falkenstein 1, Hongkai Ji 2, Steven A Vokes 1,*
PMCID: PMC4587286  NIHMSID: NIHMS714455  PMID: 26238476

Abstract

GLI proteins convert Sonic hedgehog (Shh) signaling into a transcriptional output in a tissue-specific fashion. The Shh pathway has been extensively studied in the limb bud, where it helps regulate growth through a SHH-FGF feedback loop. However, the transcriptional response is still poorly understood. We addressed this by determining the gene expression patterns of approximately 200 candidate GLI-target genes, and identified three discrete SHH-responsive expression domains. GLI-target genes expressed in the three domains are predominately regulated by derepression of GLI3 but have different temporal requirements for SHH. The GLI binding regions associated with these genes harbor both distinct and common DNA motifs. Given the potential for interaction between the SHH and FGF pathways, we also measured the response of GLI-target genes to inhibition of FGF signaling and found the majority were either unaffected or upregulated. These results provide the first characterization of the spatiotemporal response of a large group of GLI-target genes and lay the foundation for a systems-level understanding of the gene regulatory networks underlying SHH-mediated limb patterning.

Keywords: Sonic Hedgehog, GLI, Limb development, FGF, RNAseq, in situ hybridization, DNA motifs, cis-regulatory modules

INTRODUCTION

The Hedgehog (HH) signaling pathway regulates growth and patterning in multiple tissues in a variety of metazoan embryos (reviewed in Wilson and Chuang, 2010). Secreted HH ligands can spread over several cell diameters, eliciting both short and long-range effects (Chamberlain et al., 2008; Li et al., 2006; Nahmad and Stathopoulos, 2009; Sanders et al., 2013). HH-receiving cells respond by modulating the activity of the GLI transcription factors (GLI1–3, homologs of Ci in Drosophila). In the absence of HH ligand, GLIs are partially degraded by the proteasome, forming a truncated protein that functions as a transcriptional repressor (GLI-R). Conversely, in the presence of HH ligand, processing of GLI proteins is inhibited, permitting the formation of GLI activators (GLI-A) (Aza-blanc et al., 1997; Méthot and Basler, 1999; Pan et al., 2006; Wang et al., 2000).

GLI proteins activate or repress transcription of their target genes by binding to a similar sequence motif within a cis-regulatory module (CRM) (Hallikas et al., 2006; Peterson et al., 2012). Transcriptional responses to the HH pathway can be elicited either by de-repression of GLI-R or in other cases, by transcriptional activation through GLI-A (reviewed in Falkenstein and Vokes, 2014). Recent studies suggest that additional tissue-specific factors are necessary for activating appropriate GLI target genes (Biehs et al., 2010). In the neural tube, GLI-bound CRMs are enriched for Sox binding motifs, and SOX2 and SOXB1 proteins act as neural-specific GLI co-factors (Oosterveen et al., 2012; Oosterveen et al., 2013; Peterson et al., 2012). The mechanisms underlying transcriptional specificity in other HH-mediated developmental processes remain poorly understood. In several contexts, CRMs associated with GLI-target genes that are closest to the Hh signaling source have higher affinity Gli binding sites, while genes farther away are associated with CRMs that contain lower affinity Gli binding sites (Oosterveen et al., 2012; Parker et al., 2011; Peterson et al., 2012).

In the vertebrate limb bud, Sonic hedgehog (Shh) signaling regulates digit number and growth (Chiang et al., 1996; Towers et al., 2008; Zhu et al., 2008). The timing and duration of SHH is important for establishing polarity within the limb bud (Li et al., 2014; Zhulyn et al., 2014), and there is some evidence suggesting that cells retain a memory of their exposure to SHH (Harfe et al., 2004). In addition, studies have suggested that a relatively brief exposure to SHH specifies digit patterning, while longer exposures are needed for subsequent growth and expansion (Towers et al., 2008; Zhu et al., 2008). Shh expression in the limb bud is maintained by FGF proteins secreted from the apical ectodermal ridge (AER). Shh signaling regulates the transcription of the BMP inhibitor, Gremlin 1 (Grem1) (Zuniga et al., 1999, Panman et al., 2006, Zuniga et al., 2012; Li et al., 2014; Vokes et al., 2008). GREM1 inhibits localized BMP activity, thereby maintaining the apical ectodermal ridge (AER). Together, these interactions comprise a signaling loop between the mesoderm and the AER that regulates limb growth and digit number (Khokha et al., 2003; Laufer et al., 1994; Litingtung et al., 2002; Michos et al., 2004; Niswander et al., 1994; te Welscher et al., 2002; Verheyden and Sun, 2008; Zuniga, et al., 1999).

Here, we determine the expression patterns of a large set of predicted GLI-target genes in the mouse limb. Using this approach we find three distinct expression domains, which have different temporal SHH signaling requirements and are predominately regulated by derepression of GLI3-R. The GLI-bound CRMs associated with genes in each domain are enriched for both unique and common DNA motifs. Finally, we show that while some of these genes are downregulated when FGF signaling is inhibited, the majority of GLI-target genes are either unaffected or are upregulated. Collectively, these results provide the first characterization of the spatiotemporal response of a large candidate group of direct GLI-target genes that mediate Shh signaling in the limb bud.

MATERIAL AND METHODS

Mice and ethics statement

Experiments involving mice were approved by the Institutional Animal Care and Use Committee at the University of Texas at Austin (protocol AUP-2013–00168).

Whole-mount in situ hybridization

Antisense probes (supplementary material Table S2) were generated from plasmids using PCR templates as described previously (Yu et al., 2012). in situ hybridization was performed on a minimum of two Swiss-Webster embryos per stage at E10.5 and E11.5 using an Intavis system as described previously (Yu et al., 2012).

qRT-PCR

RNA was extracted using TRIzol (Invitrogen). 300ng of DNAse I treated RNA was used to synthesize cDNA using random hexamers and SuperScript II (Invitrogen). Primers used in qRT-PCR experiments are listed in supplementary material Table S7.

Culturing limb buds

Mouse forelimbs were cultured for 15 hours as previously described (Panman et al., 2006; Zuniga et al., 1999). To inhibit HH signaling, limb buds were cultured in 10 µM cyclopamine (Toronto Research) or in 0.125% ethanol for controls. To inhibit FGF signaling, contralateral forelimbs were cultured in 10 µM SU5402 (Tocris) or 0.125% DMSO for controls.

Shh−/− forelimb RNA-seq

Heterozygous Shhtm1amc mice (in previous generations mated to a Cre deleter strain to generate a null allele) (Dassule et al., 2000) were crossed, and E10.25 (33 to 35 somites) embryos were collected and genotyped for the wild-type and null allele. Forelimbs were collected and combined from three embryos of the same genotype, and RNA was isolated using TRIzol (Invitrogen) and treated with DNase I. Two biological replicates for each genotype were sequenced. The average Shh−/− somite number for replicate one was 34, and replicate two was 33.3. The average wild-type somite number for replicate one was 34, and replicate two was 33.7. Library construction was performed following Illumina manufacturer suggestions, and libraries were sequenced on the Illumina HiSeq platform using paired-end sequencing. Reads were aligned to the mouse reference genome mm10 using TopHat 2.0.9 (Trapnell et al., 2009) with default parameters and the option to incorporate genome annotation (parameter “-G”). Aligned reads were assigned to genes by HTSeq-count (Anders et al., 2010) using the default union-counting mode. Following HTSeq-count, edgeR 3.4.2 was used to conduct differential expression analysis (‘classic’ edgeR) (Robinson et al., 2010). Differentially expressed genes were identified based on an FDR of 0.05 and a mean fold change of 25% (supplementary material Table S3). The data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus and are accessible through GEO Series accession number GSE58222.

Cyclopamine treated wild-type forelimb RNAseq

Wild-type E10.25 (31–34 somites) forelimb pairs from seven wild-type embryos were cultured for 15 hours in 10 µM cyclopamine (Toronto Research) or in 0.125% ethanol for controls. Immediately after the incubation period, limb buds were separated from the adjacent tissue and RNA was isolated using TRIzol (Invitrogen) and treated with DNase I. Two biological replicates for each culture condition were sequenced. The average somite number for wild-type controls was 32 for replicate one, and 32.6 for replicate two. The average somite number for wild-type samples treated with cyclopamine was 32 for replicate one, and 32.8 for replicate two. Library preparations were generated following ABI manufacturer suggestions, and libraries were sequenced on an ABI SOLiD platform using paired-end sequencing. Reads were aligned to the mouse reference genome mm10 using TopHat 2.0.9 (Trapnell et al., 2009) with default parameters and the option to incorporate genome annotation (parameter “-G”). Aligned reads were assigned to genes by HTSeq-count (Anders et al., 2010) using the default union-counting mode. Following HTSeq-count, edgeR 3.4.2 was used to conduct differential expression analysis (‘classic’ edgeR) (Robinson et al., 2010). Differentially expressed genes were identified based on an FDR of 0.05 and a mean fold change of 25% (supplementary material Table S4). The data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus and are accessible through GEO series accession number GSE58222.

de novo motif discovery and Gli motif quality analysis

DNA motifs in GLI-bound CRMs were uncovered by a de novo motif discovery method. We mapped motif PWMs to GLI-bound CRMs in each category, background sequences were modeled as a third-order Markov chain (Ji et al. 2006). Then, we compared relative enrichment levels (r1) of the discovered motifs in high-quality binding regions versus matched control genomic regions. We chose a motif selection procedure to select enriched motifs by simultaneously requiring r1 ≥ 2, number of motif sites (n1B) ≥ max(1/5*(number of genes),5), motif score ≥ 1. We used TOMTOM motif comparison tool to visualize their sequence logos with their PWMs as input. The quality of Gli motifs was assessed by using a Gli motif with the highest score, and then mapped the PWM of the Gli motif to GLI-bound CRMs within each category. The Gli matrix was compared to a third-order background Markov model. A log-likelihood ratio for Gli motif quality was determined as described previously (Vokes et al., 2007). With the Gli consensus-binding pattern from each binding region, we calculated probability for each motif site in each category using a Welch’s t-test.

Sp1 motif meta analysis

Publically available ChIP binding regions were obtained from hmChIP and analyzed using CisGenome software. The binding regions from each dataset were compared to both matched genomic controls and independently to random controls. For each control we compared the relative enrichment (r1) of the Sp1 motif; r1>2 means Sp1 is enriched, while r1<1 means Sp1 is not enriched. Additional ChIP datasets were obtained from (Cotney et al., 2012; DeMare et al., 2013; Infante et al., 2013; Peterson et al., 2012; Visel et al., 2009). Each of these datasets were also intersected with previously generated GLI3 binding regions in the mouse limb (Vokes et al., 2008) and divided into two datasets: regions containing the Gli motif and regions without the Gli motif. The Sp1 enrichment analyses used the same binding regions from hmChIP.

RESULTS

A previous study assigned GLI-bound CRMs within the mouse limb to genes responsive to SHH signaling on the basis of their proximity to identify putative direct GLI target genes. This approach incorporated both distance to a SHH responsive gene as well as the gene density of the locus (Vokes et al., 2008). The expression patterns of 199 of the 205 previously predicted GLI target genes (Vokes et al., 2008) were determined at E10.5 and E11.5 (Fig. 1A). Of the 199 genes analyzed, 96% (n=191) were detected in embryos at E10.5, and of these, 90% (n=171) were detected in the limb (13 of 19 genes not detected in the limb at E10.5 were detected at E11.5) (supplementary material Table S1). We classified genes based on their expression pattern in the limb at E10.5 (supplementary material Fig. S1A,B and Table S1).

Figure 1.

Figure 1

Open in a new tab

GLI target genes cluster into three distinct domains within the SHH-responsive region. (A) in situ hybridization pipeline. (B) Schematic of E10.5 limb bud showing Shh expression (green) and the SHH-responsive region (dark gray). (C–E) Schematized expression patterns observed for GLI target genes in individual categories depicted in blue, red, and purple. Different intensity shading indicates expression variability within a domain. White arrows indicate expression in the posterior-proximal domain.

To identify genes that might primarily be regulated by SHH, we restricted our analysis to genes that were predominately expressed in the posterior limb bud, a region responsive to SHH signaling as defined by the expression of pathway target genes Gli1 and Ptch1 (Fig. 1B) (Litingtung et al., 2002; Marigo et al., 1996; te Welscher et al., 2002; Chiang et al., 2001; Ahn et al., 2004). Posteriorly expressed genes cluster into three broad domains (Fig. 1C–E). Category 1 contained 24 genes expressed in the posterior and posterior-distal limb bud (Fig. 1C). Genes in this category included the SHH pathway target genes Gli1, Ptch1, and Ptch2 (Chiang et al., 2001; Litingtung et al., 2002; te Welscher et al., 2002; Marigo et al., 1996; Motoyama et al., 1998). Category 2 comprised 12 genes expressed in the central limb (Fig. 1D), and category 3 contained 9 genes that were expressed in the posterior-proximal limb (Fig. 1E). Lastly, we identified 14 genes expressed in multiple domains, where at least one expression domain was spatially located within the SHH-responsive region. Because these may have more complex forms of regulation we excluded these genes from further analysis (supplementary material Fig. S1A and Table S1). Altogether, we identified 45 genes that were expressed in three domains in the limb bud. The genes in category 1 were associated with more GBRs (average of 2.9 GBRs per gene) compared to Categories 2 and 3 (average of 1.5 and 1.9 GBRs per gene respectively). There were no significant differences in the distribution of GBRs relative to the transcriptional start site for the genes in categories 1–3 (data not shown).

The identification of distinct expression domains in the limb suggested that these domains might have specific biological functions. We explored this possibility by determining enriched biological processes for each category using GO ontology term analysis (WebGestalt) (Zhang et al., 2005). Category 1 was notably enriched for transcriptional regulation and cell proliferation processes (P≤0.000001) (supplementary material Fig. S2). In contrast, category 2 was notably enriched for cell differentiation, cell adhesion, and ossification processes (P≤0.05), and category 3 was enriched for skeletal development processes (P≤0.05) (supplementary material Fig. S2). Differential enrichment within these categories suggests that the three distinct SHH-mediated domains in the limb may have unique functions.

Our categorization of expression patterns was based on E10.5 limb buds. Because SHH signaling within the limb bud is dynamic, it is possible that the expression of GLI target genes within these domains could also be dynamic. To assess this, we examined the temporal changes in the expression domains of GLI target genes by comparing their expression patterns between E10.5 and E11.5. Most genes expressed in a particular domain at E10.5 remain expressed in a broader version of the same domain at E11.5 (category 1, n=18/24; category 2, n=9/12; category 3, n=8/9) (Fig. 2A–C and supplementary material Fig. S1A). Taken together, these results indicate that the expression domains of GLI target genes are relatively stable despite the large changes in SHH signaling that occur during this period. In summary, we identified three stable domains of GLI target gene expression within the SHH-responsive region of the limb bud.

Figure 2.

Figure 2

Open in a new tab

GLI target genes maintain early expression domain. (A–C) Schematized expression domains in E10.5 and E11.5 limbs for GLI target genes. Expression domains at E11.5 are broader (indicated in gray). Number of genes in each category expressed in the same domain at E10.5 and E11.5: (A) category 1, n=18/24), (B) category 2 (n=9/12), and (C) category 3 (n=8/9).

GLI target genes are primarily regulated by GLI repression

GLI activators (GLI-A) and GLI repressors (GLI-R) regulate the expression of GLI target genes, and the transcriptional requirement for GLI-A and GLI-R has been evaluated for a limited number of GLI target genes (reviewed in Falkenstein & Vokes, 2014). Notably, Gli1 and Ptch1 require GLI-A while several additional target genes are activated in the absence of GLI-R (Litingtung et al., 2002; te Welscher et al., 2002). In order to clarify the role of GLI-A and GLI-R for genes in categories 1–3, we generated Shh−/−;Gli3−/− double mutant embryos and quantified gene expression in the forelimb buds using qRT-PCR. Consistent with previous reports (Litingtung et al., 2002; te Welscher et al., 2002), we find that Gli1 is greatly downregulated in the double mutants (Fig. 3), indicating a requirement for GLI-A. In agreement with previous studies, we find that Hand2, Hoxd13, and Grem1, which require only a loss of GLI-R (Litingtung et al., 2002; te Welscher et al., 2002), are expressed in the double mutant forelimbs at levels unchanged or higher than wild-type (Fig. 3). Among the 33 candidate GLI target genes assessed, we find that only three were significantly downregulated in the double mutant limb buds (indicating a role for GLI-A), while 30 genes were either unchanged or significantly upregulated (Fig. 3). We conclude that the loss of GLI-R is sufficient to confer robust expression to the majority of candidate GLI target genes analyzed.

Figure 3.

Figure 3

Open in a new tab

Most GLI target genes are activated by a loss of GLI derepression. Single forelimb pairs from E10.5 Gli3−/−;Shh−/− (red bars) and sibling wild-type (black bars) (34–36 somites) examined for expression of select candidate GLI target genes by qRT-PCR (n=4 embryos per genotype, both forelimbs). Boxes indicate the mean, dots indicate expression levels in individual forelimb pairs, error bars indicate the s.e.m, an asterisk indicates statistical significance determined using a two-tailed t-test (P<0.05), and ns indicates not statistically significant. Data normalized to a single pair of wild-type forelimbs using Gapdh expression across all samples.

Categories of GLI target genes have different temporal requirements for SHH

GLI target genes require SHH signaling for transcriptional activation or robust expression; however, subsequent requirements for SHH signaling have been determined for only a few GLI target genes. For example, Hoxd13, greatly reduced in Shh−/− fore- and hindlimbs (Chiang et al., 2001; Ros et al., 2003), requires SHH signaling for robust expression and continues to require SHH to maintain robust expression beyond E10.75 (Panman et al., 2006). In contrast, Jag1, requires exposure to SHH signaling through E10.25 before expression becomes independent of SHH signaling (Panman et al., 2006). We sought to determine the temporal requirement of SHH signaling for the group of 45 GLI target genes expressed in the posterior limb. To determine whether transient SHH activation was required for expression of GLI target genes, we cultured E10.25 wild-type limbs, which already express Shh, in the presence of the HH pathway inhibitor cyclopamine (Chen et al., 2002; Panman et al., 2006). If gene expression was maintained after transient SHH exposure (cyclopamine cultures), but downregulated in Shh−/−, then SHH is required only transiently for GLI target gene expression. In contrast, if genes required continued exposure to SHH, they would remain downregulated in wild-type limb buds that received only a brief exposure to SHH. Although previous studies have identified SHH-responsive genes at E10.5 using microarrays (Probst et al., 2011; Shah et al., 2009), we decided to perform a new analysis using RNA-seq on E10.25 limb buds in order to make direct quantitative comparisons (Fig. 4A).

Figure 4.

Figure 4

Open in a new tab

GLI target genes have distinct temporal requirements for SHH. (A) Schematic depicting RNA-seq pipeline. Forelimb samples with different SHH exposures indicated as WT, No SHH, and Brief SHH. (B) Heatmap showing expression for GLI target genes in categories 1–3 across eight RNA-seq samples: wild-type (WT), Shh−/−, control (Ctrl), and cyclopamine (Cyc). (C–E) Scatter plots showing expression (Log2FC) of candidate GLI target genes in Shh−/− (No SHH) and cyclopamine (Brief SHH) samples. Blue lines are at 0 and red lines are at −0.4105 (log2 expression). (C) ‘Sustained SHH’ indicates GLI target genes downregulated in both Shh−/− and cyclopamine samples (n=18) (Log2FC≤-0.4105 and FDR ≤-0.1). (D) ‘Transient SHH’ indicates GLI target genes (n=5) downregulated in the Shh−/− dataset (Log2FC≤-0.4105 and FDR ≤0.1) but unregulated in the cyclopamine dataset (Log2FC≥-0.1 and FDR ≤0.1). (E) ‘Other’ indicates GLI target genes (n=21) outside the parameters for ‘transient’ and ‘sustained SHH’ categories. Expression values for genes in B-E indicated in Table 1.

Using littermate E10.25 (32–35 somites) wild-type and Shh−/− forelimbs, we identified 297 SHH-responsive genes downregulated ≥25%, (Log2FC≤-0.4105) with a false discovery rate (FDR) ≤0.05 (supplementary material Table S3). Among these downregulated genes, were several known GLI target genes: Gli1, Ptch1, Ptch2, Hhip, HoxD13, Grem1, and Hand2 (Table 1 and supplementary material Table S3) (Bénazet et al., 2009; Chiang et al., 2001; Litingtung et al., 2002; Panman et al., 2006; Vokes et al., 2008; te Welscher et al., 2002). To determine which of these genes continue to require SHH, we cultured E10.25 (31–34 somites) wild-type forelimbs for 15 hours in either control media or media containing cyclopamine (Panman et al., 2006). Within this dataset, we identified 61 genes downregulated ≥25% (Log2FC≤-0.4105) with a FDR≤0.05. Consistent with previous reports, several genes that continue to require SHH at E10.25 were confirmed in our dataset: Gli1, Hoxd11, Hoxd12, and Hoxd13 (Table 1 and supplementary material Table S4) (Panman et al., 2006).

Table 1.

Classification and regulation of GLI target genes. (A)Schematized expression patterns of GLI target genes found in categories 1–3. (B) List of GLI target genes found in each category.

graphic file with name nihms714455t1.jpg graphic file with name nihms714455t2.jpg graphic file with name nihms714455t3.jpg
Gene FC Shh−
/− 		FC cyc Gene FC Shh−
/− 		FC cyc Gene FC Shh−
/− 		FC cyc
Calml4 −2.22 −1.27 Cldn11   0.55 −0.20 Adamts19 −0.15 −0.70
Cdk6* −1.61 −0.99 Cldn9   0.20   0.07 Cntfr‡ −0.80   0.10
Dock6* −0.49 −0.66 Fam181b −0.45   0.00 Col23a1 −0.37 −0.12
Gli1* −5.71 −3.14 Fam69c   0.17   0.92 Dlk1 −0.41   0.17
Grem1* −2.63 −0.89 Fbxo41 −0.69 −0.34 Ltbp1 −0.33 −0.17
Hand2* −2.16 −0.53 Fbxo8 −0.68   0.22 Lypd6* −1.27 −0.72
Hoxd11* −2.51 −0.91 Hhip* −3.00 −1.10 Osr1‡ −1.39 −0.10
Hoxd13* −6.01 −1.09 Ndnf −0.44   0.31 Prdm1 −1.15 −0.21
Ier2 −0.43 −0.15 Osr2‡ −4.01   0.02 Svep1‡ −0.63   0.11
Jag1 −0.86 −0.40 Rspo3‡ −1.29 −0.04
Klf9 −1.03 −0.43 Smoc1* −1.43 −1.06
Msi2* −0.95 −0.84 Whrn −0.16   0.10
Pam* −1.01 −0.61
Ptch1* −2.37 −1.42
Ptch2* −4.41 −3.52
Ralgps2 −0.32 −0.24
Rasgef1b* −1.44 −0.51
Sall1* −1.90 −1.26
Sall3* −1.79 −0.95
Sap30 −0.35 −0.05
Shroom3   0.05   0.01
Sox4   0.16 −0.01
Tpd52l1* −1.56 −0.82
Open in a new tab

Gene expression values in log2 fold change (FC) from RNA-seq experiments (shown in Fig. 4). Targets participating in the Hedgehog pathway response are underlined.

Genes downregulated ≥25% (Log2FC≤-0.4105 and FDR ≤0.1) in both Shh−/− limbs and cyclopamine treated limbs are indicated with an asterisk (*).

Genes downregulated ≥25% (Log2FC≤-0.4105 and FDR ≤0.1) in Shh−/− limbs but unregulated in the cyclopamine data set (Log2FC≥-0.1 and FDR ≤0.1) are indicated with a double dagger (‡).

Note that Calml4 and Klf9 were downregulated in both Shh−/− and cyclopamine datasets but they were excluded from the ‘sustained SHH’ category because their adjusted FDRs were >0.1.

Next, we asked if the 45 GLI target genes predominately expressed in the posterior limb have differential requirements for SHH signaling. To determine genes that require sustained SHH signaling, we identified genes that were downregulated ≥25% (Log2FC≤-0.4105) with a FDR≤0.1 in both datasets (Fig. 4B). One gene, Runx3, was excluded from the analysis because of a low read count in one dataset. Among the group, 18 out of 44 genes were downregulated in both Shh−/− limbs and limbs treated with cyclopamine (Fig. 4C and Table 1, indicated with an asterisk). Interestingly, nearly all of the genes found downregulated in both datasets (n=15) were expressed in category 1 (posterior and posterior-distal), making it significantly enriched for genes requiring sustained SHH (two-sided Fisher’s exact test, P=0.0018) (Table 1, indicated with an asterisk).

To determine the genes that require a transient exposure to SHH signaling, we identified genes downregulated ≥25% (Log2FC≤-0.4105) with a FDR≤0.1 in Shh−/− limbs that had unchanged expression relative to wild-type limbs cultured in cyclopamine. Among the group of 44 genes, 5 genes were downregulated in Shh−/− limbs but unchanged relative to controls in cyclopamine treated limb cultures (Fig. 4D and Table 1). All of these genes were expressed in categories 2 and 3, suggesting that in contrast to category 1, genes within categories 2 and 3 have more transient requirements for SHH. The expression values for 21 out of the 44 genes did not fall into the parameters of the two previous classes, and are therefore labeled as ‘other.’ (Fig. 4E and Table 1). We conclude that GLI target genes have differential requirements for SHH signaling, and that the transcriptional regulation of GLI target genes correlates with their expression domain.

Cis-regulatory modules associated with GLI target genes in different categories contain both unique and common DNA motifs

The 45 genes defined above are likely to be regulated by SHH signaling through their associated GLI-bound CRMs. To identify co-factors potentially involved in mediating the GLI response, we searched the GLI-bound CRMs associated with these genes for enriched DNA motifs using de novo motif analysis (Ji et al., 2008; Ji et al., 2006). We uncovered novel DNA motifs enriched in categories 1 and 2 (Fig. 5A,B). No enriched motifs were uncovered for CRMs associated with genes in category 3 due to the small sample size; this is also a likely explanation for why no Gli motif was uncovered for Category 2 genes. The most enriched motif in any category was Gli, which was substantially enriched in GLI-bound CRMs associated with category 1 genes (Fig. 5A). We also identified a GC rich motif which corresponded to the known binding motif for the SP1 transcription factor (Fig. 5A) (Gidoni et al., 1984; Ji et al., 2006). In addition to being enriched in category 1, the Sp1 motif was the most enriched motif in category 2 (Fig. 5B). We also identified several novel motifs that are highly sequence specific; however, do not correspond to any known transcription factor binding motifs (Fig. 5A,B). Thus, DNA motif analysis identified both unique and common motifs among two of the expression categories.

Figure 5.

Figure 5

Open in a new tab

DNA motifs enriched in GLI-bound CRMs. (A) Enriched DNA motifs in GLI-bound CRMs associated with category 1 and (B) category 2 genes. No statistically enriched DNA motifs were identified in category 3. (C) Assessment of the quality of Gli motifs in GLI-bound CRMs associated with genes found in categories 1–3 shown as log2 likelihood ratio. Reported p-values using a Welch’s t-test for comparing Gli motif quality between the categories: category 1 to 2 (P=0.083); category 1 to 3 (P=0.776); category 2 to 3 (P=0.144).

The binding affinity of Gli motifs have been shown to function in differential interpretation of HH signaling in the developing mouse neural tube and Drosophila imaginal wing disc (Oosterveen et al., 2012; Parker et al., 2011; Peterson et al., 2012). We asked if the different expression categories had differences in Gli consensus motifs. We chose the Gli motif with the highest motif score in our dataset and mapped the position weight matrix of the Gli motif to GLI-bound CRMs associated with genes expressed in categories 1, 2, and 3. Gli motifs associated with genes expressed in category 1 and 3 have higher means of log-likelihood, indicating higher quality Gli motifs (Fig. 5C). In contrast, Gli motifs associated with category 2 have a lower (although not statistically significant) mean of log-likelihood, indicating lower quality Gli motifs (Fig. 5C). Interestingly, genes in categories 1 and 3 have overlapping or adjacent expression to the Shh expression domain, while genes in category 2 are expressed further from Shh. This suggests that long-range SHH signaling could regulate category 2 genes. The same trends in Gli motif quality were also observed in the mouse neural tube and Drosophila imaginal wing disc for short and long range GLI target genes (Oosterveen et al., 2012; Parker et al., 2011).

FGF signaling co-regulates a subset of GLI target genes

Since the signaling activities of SHH and FGF intersect in the posterior mesoderm and co-regulate each other through a signaling loop, we sought to determine if FGF signaling influenced the expression of putative GLI target genes. We cultured contralateral E10.25-E10.5 limb buds in control media or media containing the FGF receptor antagonist SU5402 for 15 hours and assayed expression. The FGF target gene Spry4 (Minowada et al., 1999; Verheyden and Sun, 2008) was reduced by 88% (qRT-PCR) and nearly undetectable by in situ hybridization (Fig. 6B,C), indicating that FGF signaling was substantially inhibited. We then assayed a cohort of genes from the three expression categories by qRT-PCR. Surprisingly, the majority of putative GLI target genes assayed did not have reduced expression (Fig. 6A; n=18/34). However, as expected some putative GLI target genes were downregulated (n=6/34), including HoxD13, which has been shown to be dependent upon both SHH and FGF signaling (Fig 6A) (Laufer et al., 1994; Panman et al., 2006; Ros et al., 1996). Since the loss of FGF signaling leads to a gradual loss of Shh expression, these genes could be downregulated either due to a loss of FGF or SHH signaling (Laufer et al., 1994; Niswander et al., 1994; Zuniga et al., 1999) (Supplementary material Fig. 3A,B). Interestingly, a group of genes were significantly upregulated in SU5402 treated limb cultures (Fig. 6A; n=10/34). This result is consistent with previous reports demonstrating repressive roles for FGF in regulating Shh and Grem1 expression (Mao et al., 2009; Verheyden and Sun, 2008; Zhang et al., 2009). To determine if FGF inhibition caused changes in the expression domains of these genes, we compared the expression of a subset of them in contralateral cultured forelimbs treated with control media or media containing SU5402. Osr2, Smoc1, and Rspo3 had larger expression domains that expand into the posterior-distal forelimb while Osr1 has a broader domain in the proximal-posterior forelimb (Fig. 6C). In contrast, Cntfr, is not significantly expanded in SU5042 cultures (Fig. 6C). Taken together, we found that the majority of GLI target genes are not positively regulated by FGF signaling; however, a subset of GLI target genes are significantly upregulated upon loss of FGF activity.

Figure 6.

Figure 6

Open in a new tab

Most GLI-target genes are not downregulated by FGF inhibition. Assessment of the expression for GLI target genes in E10.25-E10.5 forelimbs cultured in the presence of an FGF inhibitor (SU5402). Wild-type forelimbs cultured in control media (black bars) and the contralateral forelimbs cultured in 10 µM SU5402 (red bars) for 15 hours. (A) qRT-PCR for 34 GLI-target genes (Gli1 shown in S3A). (B) qRT-PCR for Fgf8 and Spry4. Boxes indicate the mean value, error bars indicate the s.e.m, data points indicate expression levels from individual forelimbs (n=4–5 embryos), and gene expression is normalized to a single wild-type limb bud using Gapdh expression across all samples. (C) Contralateral forelimbs cultured in control or 10 µM SU5402 for 15 hours. in situ hybridization for Spry4 (n=5) and candidate GLI target genes, Osr2 (n=3), Osr1 (n=5), Smoc1 (n=4), Rspo3 (n=3), and Cntfr (n=10). Domain size was quantified as a ratio of expression area divided by limb area. In all panels an asterisk indicates statistical significance determined using a two-tailed t-test (P<0.05) and ‘ns’ indicates not statistically significant.

DISCUSSION

In this study, we have identified a set of genes that by multiple criteria are likely to be directly regulated by GLI proteins in the developing limb bud. These genes are expressed in three domains, which have distinct temporal requirements for SHH signaling (Fig. 7A). The GLI-bound CRMs associated with genes in two of the domains are enriched for additional DNA motifs that are unique and common, representing binding sites for potential transcriptional co-regulators (Fig. 7B). The identification of distinct expression domains has greatly improved the understanding of HH patterning in the Drosophila imaginal wing disc and vertebrate neural tube, and we hope that the identification of different domains in this study will facilitate the determination of the gene regulatory networks underlying SHH-mediated limb patterning.

Figure 7.

Figure 7

Open in a new tab

Model of GLI-mediated gene expression domains in the mouse limb. (A) SHH (green) establishes three domains (schematized in blue, red, and purple) in distinct regions in the limb. Two of the domains, categories 2 (red) and 3 (purple), require transient exposure to SHH (small, dashed green arrows) while many genes in category 1 (blue) require sustained SHH signaling (large, solid green arrow). (B) Schematic of a GLI-bound cis-regulatory module and the cognate GLI-target gene. The majority of GLI-target genes require depression of GLI-R (large black arrow and green circle) while a small subset of GLI-target genes require GLI-A (small black arrow and green circle). Category-specific DNA motifs uncovered in GLI-bound CRMs associated with each domain could potentially confer domain specific co-factors (indicated with a small dashed arrow and red, blue, and purple circles) while a common Sp1 DNA motif (dark yellow) represents a potential co-regulator for all domains (dark yellow circle).

Expression patterns of GLI target genes during limb development

Whole-genome approaches using DNA microarrays have been used to identify SHH-responsive genes during limb development (Bangs et al., 2010; Hu et al., 2012; McGlinn et al., 2005; Probst et al., 2011; Shah et al., 2009; Vokes et al., 2008). Combining this approach with three-dimensional spatial information from whole-mount embryo in situ hybridization has been successfully used to identify genes regulated in distinct domains (Bangs et al., 2010; Probst et al., 2011; Welten et al., 2011). Here, we focused on genes that were previously identified as likely GLI target genes and performed a comprehensive in situ hybridization screen to determine the limb expression patterns for nearly all of the 205 predicted GLI target genes at two developmental stages.

While our analysis identified a group of high confidence SHH regulated genes, many additional candidate GLI target genes did not show spatially restricted expression in the posterior limb. Many of these genes are likely have direct GLI regulatory inputs. For example, Rab34 has relatively uniform limb expression but is nonetheless SHH-responsive, containing a GLI-bound CRM with posterior-specific activity (Vokes et al., 2007). Similarly, we excluded many putative GLI target genes expressed in the distal limb from further analysis because they were not primarily expressed within the SHH-responsive region (supplementary material Table S1). A previous study classifying SHH-responsive genes suggested that Cyp26b1, a distally expressed genes, while reduced in Shh−/− limbs, is primarily regulated by FGFs and becomes reduced because of a breakdown in the SHH-FGF signaling loop (Probst et al., 2011). Since many of these distally expressed genes are in proximity to GLI-bound CRMs, it is possible that they are co-regulated by SHH and FGF. By restricting our analysis to spatially restricted genes in the posterior limb, we are therefore likely excluding many bona-fide GLI target genes that have more complex regulatory inputs.

We identified three broad expression domains of GLI target genes (Fig. 1C–E). The coordinated expression of developmentally expressed genes, termed syn-expression groups, has been proposed to be involved in regulating common biological pathways (Gawantka et al., 1998; Ramialison et al., 2012; Visel et al., 2007). Category 1 genes, expressed in the posterior and posterior-distal limb, are enriched for GO processes involved in transcriptional regulation and cell proliferation (supplementary material Fig. S2). Consistent with these processes, the posterior-distal domain undergoes significant proliferative expansion during limb bud development, ultimately giving rise to most of the developing digits (reviewed in Zeller et al., 2009). If SHH acts as a morphogen in the limb bud, it would likely be doing so through these category 1 genes. Interestingly, besides SHH pathway feedback components, we did not identify GLI target genes that had obvious graded expression in the limb. In contrast to category 1, GLI target genes expressed in category 2 (the central limb) are enriched for skeletal differentiation markers and BMP inhibitors (supplementary material Fig. S2) (Hsu et al., 1998; Khokha et al., 2003; Rainger et al., 2011). Similarly, genes expressed in category 3 are enriched for skeletal pathway genes (supplementary material Fig. S2). Lineage tracing experiments have established that cells within both of these domains primarily contributes to the presumptive forearm (zeugopod) (Vargesson et al., 1997).

GLI target genes have differential requirements for the duration of SHH signaling

Previous reports identified four GLI target genes within the limb bud that require sustained SHH signaling beyond E10.5 (Panman et al., 2006). Our analysis of the temporal requirement of SHH signaling for GLI target genes extends these results and provides quantitative measurements of the response, identifying a total of 18 GLI target genes that require sustained SHH signaling beyond E10.25 (Table 1). These genes are nearly all expressed in category 1, which might comprise a gene regulatory module important for mediating limb growth during the period when SHH signaling is required for expansion of the cartilage progenitors (Towers et al., 2008; Zhu et al., 2008). It is not clear why genes in categories 2 and 3 generally require a briefer exposure to SHH. One possibility is that since genes in categories 2 and 3 are expressed in more proximal regions, their transient need for SHH signaling reflects the earlier differentiation of more proximal limb elements compared to distal elements (Roselló-Díez et al., 2014).

Although the RNA-seq analysis was specifically applied to a group of previously defined GLI target genes, the analysis also identified a larger SHH-responsive group of genes that require transient or continued SHH signaling (supplementary material Table S5). In future studies it would be interesting to identify the expression pattern of this group of SHH-responsive genes and determine to what extent they conform to the expression categories as described in this work. Based on our finding that the regulation of GLI target genes correlates with where they are expressed, we predict that genes requiring sustained SHH signaling will be predominately expressed in the posterior and posterior-distal limb (category 1), while genes that require transient SHH signaling will be expressed in the central (category 2) or posterior-proximal limb (category 3).

Co-regulation of GLI target genes

Consistent with multiple previous studies (Bastida et al., 2009; Laufer et al., 1994; Niswander et al., 1994; Zuniga et al., 1999) we found that FGF is required to maintain normal levels of Shh expression (supplementary material Fig. 3A,B). The relatively few GLI-target genes downregulated upon FGF inhibition could either be responding to the reduced levels of Shh or alternatively could require direct FGF co-regulation. Interestingly, most GLI target genes are either unaffected or are repressed by FGF signaling notwithstanding the reduction in Shh expression (Fig. 6A). For some genes that were upregulated, FGF inhibition led to changes in spatial domain expression (Fig. 6C). Since FGF or FGF-responsive ETS/ETV transcription factors have previously been shown to repress expression of Shh and the GLI target Grem1 (Mao et al., 2009; Verheyden and Sun, 2008; Zhang et al., 2009) our results suggest a larger role for the FGF pathway in repressing GLI target genes. The molecular mechanisms of FGF repressive inputs at the cis-regulatory level are not known. One possibility is that the repressive inputs could potentially be integrated with GLI-bound CRM. However, since a GLI-bound CRM regulating Gremlin (GRE1) does not respond to FGF signaling inhibition (Li et al., 2014) and DNA motif analysis of GLI-bound regions (this study) does not uncover enriched ETS/ETV transcription factor binding sites (Fig. 5A,B), it is possible that FGF repressive inputs might coordinately affect additional GLI-independent CRMs or work by regulating an intermediate repressor.

Regulatory inputs in GLI CRMs

In current ChIP approaches, it is difficult to distinguish between transcriptionally relevant binding sites from the majority of inert sites in a particular tissue (Shlyueva et al., 2014). We previously estimated that only 15% of GLI-bound CRMs are likely be transcriptionally relevant (Vokes et al., 2008). The GLI-bound CRMs associated with genes in the domains identified in this study are a significant improvement on our previous approach, which only considered gene response irrespective of their spatial expression domains.

From this set of GLI-bound CRMs we identified both unique and common DNA motifs. The unique motifs do not correspond to known transcription factor binding sites and might represent binding sites for unidentified co-factors that confer specific expression. After the Gli motif, the Sp1 motif was the most common motif identified in categories 1 and 2 (Fig. 5A,B). While Sp1 is a ubiquitously expressed gene, its expression levels are dynamic in different tissues including the limb bud (supplementary material Fig. S3C) (Saffer et al., 1991). Despite their reputation as ubiquitous transcription factors, several individual Sp-family genes have specific loss-of-function phenotypes, suggesting that they have specific roles in development (reviewed in Zhao and Meng, 2005; Suske et al., 2005). The Sp1 motif is enriched in at least 17% of published mouse ChIP datasets (supplementary material Table S6), suggesting that any functional role might not be limited to GLI target genes. Interestingly, in Drosophila, Sp1 is broadly expressed in the imaginal disc and is required for leg development (Estella and Mann, 2010; McKay et al., 2009; Estella et al., 2003). In future studies it will be interesting to determine if Sp1 or other mesodermally-expressed Sp-family proteins might co-regulate GLI target genes during vertebrate limb development.

Supplementary Material

1
9
10
2
3
4
5
6
7
8
11

Highlights.

  • ▪

    GLI-target genes cluster into three domains in the limb bud

  • ▪

    GLI3-R predominately regulates expression

  • ▪

    GLI-bound cis-regulatory modules contain common and unique DNA motifs

  • ▪

    FGF signaling represses a subset of GLI-target genes

ACKNOWLEDGEMENTS

We thank Martha Vokes for identifying plasmids and designing primers for generating templates for antisense riboprobes. We thank Xia Li for performing the initial analysis of RNA-seq dataset on cyclopamine-treated embryos. We are grateful to Dr. Simone Probst and the other members of Dr. Rolf Zeller’s laboratory for teaching us the limb bud culture technique. We thank Dr. Jacqueline Tabler and Simone Giovanetti for providing critical comments on the manuscript. This work was supported by NIH R01HD073151 (to S.A.V and H.J.) and startup funds from the College of Natural Sciences and the Institute for Cellular and Molecular Biology at the University of Texas at Austin (to S.A.V).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

  1. Ahn S, Joyner AL. Dynamic changes in the response of cells to positive hedgehog signaling during mouse limb patterning. Cell. 2004;118:505–516. doi: 10.1016/j.cell.2004.07.023. [DOI] [PubMed] [Google Scholar]
  2. Aza-blanc P, Laget M, Schwartz C, Kornberg TB. Proteolysis That Is Inhibited by Hedgehog Targets Cubitus interruptus Protein to the Nucleus and Converts It to a Repressor. Cell. 1997;89:1043–1053. doi: 10.1016/s0092-8674(00)80292-5. [DOI] [PubMed] [Google Scholar]
  3. Bangs F, Welten M, Davey MG, Fisher M, Yin Y, Downie H, Paton B, Baldock R, Burt DW, Tickle C. Identification of genes downstream of the Shh signaling in the developing chick wing and syn-expressed with Hoxd13 using microarray and 3D computational analysis. Mech. Dev. 2010;127:428–441. doi: 10.1016/j.mod.2010.08.001. [DOI] [PubMed] [Google Scholar]
  4. Bastida MF, Sheth R, Ros MA. A BMP-Shh negative-feedback loop restricts Shh expression during limb development. Development. 2009;136:3779–3789. doi: 10.1242/dev.036418. [DOI] [PubMed] [Google Scholar]
  5. Bénazet J, Bischofberger M, Tiecke E, Goncalves A, Martin J, Zuniga A, Naef F, Zeller R. A Self-Regulatory System of Interlinked Signaling Feedback Loops Controls Mouse Limb Patterning. Science. 2009;1050:1050–1053. doi: 10.1126/science.1168755. [DOI] [PubMed] [Google Scholar]
  6. Biehs B, Kechris K, Liu S, Kornberg TB. Hedgehog targets in the Drosophila embryo and the mechanisms that generate tissue-specific outputs of Hedgehog signaling. Development. 2010;137:3887–3898. doi: 10.1242/dev.055871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chamberlain CE, Jeong J, Guo C, Allen BL, McMahon AP. Notochord-derived Shh concentrates in close association with the apically positioned basal body in neural target cells and forms a dynamic gradient during neural patterning. Development. 2008;135:1097–1106. doi: 10.1242/dev.013086. [DOI] [PubMed] [Google Scholar]
  8. Chen JK, Taipale J, Young KE, Maiti T, Beachy P. Small molecule modulation of Smoothened activity. Proc. Natl. Acad. Sci. U. S. A. 2002;99:14071–14076. doi: 10.1073/pnas.182542899. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chiang C, Litingtung Y, Lee E, Young K, Corden J, Westphal H, Beachy P. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. 1996:407–413. doi: 10.1038/383407a0. [DOI] [PubMed] [Google Scholar]
  10. Chiang C, Litingtung Y, Harris MP, Simandl BK, Li Y, Beachy Pa, Fallon JF. Manifestation of the limb prepattern: limb development in the absence of sonic hedgehog function. Dev. Biol. 2001;236:421–435. doi: 10.1006/dbio.2001.0346. [DOI] [PubMed] [Google Scholar]
  11. Cotney J, Leng J, Oh S, DeMare LE, Reilly SK, Gerstein MB, Noonan JP. Chromatin state signatures associated with tissue-specific gene expression and enhancer activity in the embryonic limb. Genome Res. 2012;22:1069–1080. doi: 10.1101/gr.129817.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dassule HR, Lewis P, Bei M, Maas R, McMahon aP. Sonic hedgehog regulates growth and morphogenesis of the tooth. Development. 2000;127:4775–4785. doi: 10.1242/dev.127.22.4775. [DOI] [PubMed] [Google Scholar]
  13. DeMare LE, Leng J, Cotney J, Reilly SK, Yin J, Sarro R, Noonan JP. The genomic landscape of cohesin-associated chromatin interactions. Genome Res. 2013;23:1224–1234. doi: 10.1101/gr.156570.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Estella C, Mann RS. Non-Redundant Selector and Growth-Promoting Functions of Two Sister Genes , buttonhead and Sp1 , in Drosophila Leg Development. PLoS Genet. 2010;6 doi: 10.1371/journal.pgen.1001001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Estella C, Rieckhof G, Calleja M, Morata G. The role of buttonhead and Sp1 in the development of the ventral imaginal discs of Drosophila. Development. 2003;130:5929–5941. doi: 10.1242/dev.00832. [DOI] [PubMed] [Google Scholar]
  16. Falkenstein KN, Vokes Sa. Transcriptional Regulation of Graded Hedgehog Signaling. Semin. Cell Dev. Biol. 2014:1–8. doi: 10.1016/j.semcdb.2014.05.010. In press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gawantka V, Pollet N, Delius H, Vingron M, Pfister R. Gene expression screening in Xenopus identifies molecular pathways , predicts gene function and provides a global view of embryonic patterning. Mech. Dev. 1998;77:95–141. doi: 10.1016/s0925-4773(98)00115-4. [DOI] [PubMed] [Google Scholar]
  18. Gidoni D, Dynan W, Tjian R. Multiple specific contacts between a mammalian transcription factor and its cognate promoters. Nature. 1984;312 doi: 10.1038/312409a0. [DOI] [PubMed] [Google Scholar]
  19. Hallikas O, Taipale J. High-throughput assay for determining specificity and affinity of protein-DNA binding interactions. Nat. Protoc. 2006;1:215–222. doi: 10.1038/nprot.2006.33. [DOI] [PubMed] [Google Scholar]
  20. Harfe BD, Nissim S, Tian H, McMahon AP, Tabin CJ. Evidence for an expansion-based temporal Shh gradient in specifying vertebrate digit identities. Cell. 2004;118:517–528. doi: 10.1016/j.cell.2004.07.024. [DOI] [PubMed] [Google Scholar]
  21. Hsu DR, Economides AN, Wang X, Eimon PM, Harland RM. Identifies a Novel Family of Secreted Proteins that Antagonize BMP Activities. Mol. Cell. 1998;1:673–683. doi: 10.1016/s1097-2765(00)80067-2. [DOI] [PubMed] [Google Scholar]
  22. Hu JK-H, McGlinn E, Harfe BD, Kardon G, Tabin CJ. Autonomous and nonautonomous roles of Hedgehog signaling in regulating limb muscle formation. Genes Dev. 2012;26:2088–2102. doi: 10.1101/gad.187385.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Infante CR, Park S, Mihala AG, Kingsley DM, Menke DB. Pitx1 broadly associates with limb enhancers and is enriched on hindlimb cis-regulatory elements. Dev. Biol. 2013;374:234–244. doi: 10.1016/j.ydbio.2012.11.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ji H, Vokes SA, Wong WH. A comparative analysis of genome-wide chromatin immunoprecipitation data for mammalian transcription factors. Nucleic Acids Res. 2006;34:146. doi: 10.1093/nar/gkl803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ji H, Jiang H, Ma W, Johnson DS, Myers RM, Wong WH. An integrated software system for analyzing ChIP-chip and ChIP-seq data. Nat. Biotechnol. 2008;26:1293–1300. doi: 10.1038/nbt.1505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Khokha MK, Hsu D, Brunet LJ, Dionne MS, Harland RM. Gremlin is the BMP antagonist required for maintenance of Shh and Fgf signals during limb patterning. Nat. Genet. 2003;34:303–307. doi: 10.1038/ng1178. [DOI] [PubMed] [Google Scholar]
  27. Laufer E, Nelson, Craig, Johnson R, Morgan B, Tabin C. Sonic hedgehog and Fgf-4 Act through a Signaling Cascade and Feedback Loop To Integrate Growth and Patterning of the Developing Limb Bud. Cell. 1994;79:993–1003. doi: 10.1016/0092-8674(94)90030-2. [DOI] [PubMed] [Google Scholar]
  28. Li Y, Zhang H, Litingtung Y, Chiang C. Cholesterol modification restricts the spread of Shh gradient in the limb bud. Proc. Natl. Acad. Sci. U. S. A. 2006;103:6548–6553. doi: 10.1073/pnas.0600124103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Li D, Sakuma R, Vakili Na, Mo R, Puviindran V, Deimling S, Zhang X, Hopyan S, Hui C-C. Formation of proximal and anterior limb skeleton requires early function of irx3 and irx5 and is negatively regulated by shh signaling. Dev. Cell. 2014a;29:233–240. doi: 10.1016/j.devcel.2014.03.001. [DOI] [PubMed] [Google Scholar]
  30. Li Q, Lewandowski JP, Powell MB, Norrie JL, Cho SH, Vokes SA. A Gli silencer is required for robust repression of gremlin in the vertebrate limb bud. Development. 2014b;141:1906–1914. doi: 10.1242/dev.104299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Litingtung Y, Dahn RD, Li Y, Fallon JF. Shh and Gli3 are dispensable for limb skeleton formation but regulate digit number and identity. Nature. 2002;418:979–983. doi: 10.1038/nature01033. [DOI] [PubMed] [Google Scholar]
  32. Mao J, McGlinn E, Huang P, Tabin CJ, McMahon AP. Fgf-dependent Etv4/5 activity is required for posterior restriction of Sonic Hedgehog and promoting outgrowth of the vertebrate limb. Dev. Cell. 2009;16:600–606. doi: 10.1016/j.devcel.2009.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Marigo V, Scott MP, Johnson RL, Goodrich LV, Tabin CJ. Conservation in hedgehog signaling: induction of a chicken patched homolog by Sonic hedgehog in the developing limb. Development. 1996;122:1225–1233. doi: 10.1242/dev.122.4.1225. [DOI] [PubMed] [Google Scholar]
  34. McGlinn E, van Bueren KL, Fiorenza S, Mo R, Poh AM, Forrest A, Soares MB, Bonaldo MDF, Grimmond S, Hui CC, et al. Pax9 and Jagged1 act downstream of Gli3 in vertebrate limb development. Mech. Dev. 2005;122:1218–1233. doi: 10.1016/j.mod.2005.06.012. [DOI] [PubMed] [Google Scholar]
  35. McKay DJ, Estella C, Mann RS. The origins of the Drosophila leg revealed by the cis-regulatory architecture of the Distalless gene. Development. 2009;136:61–71. doi: 10.1242/dev.029975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Méthot N, Basler K. Hedgehog controls limb development by regulating the activities of distinct transcriptional activator and repressor forms of Cubitus interruptus. Cell. 1999;96:819–831. doi: 10.1016/s0092-8674(00)80592-9. [DOI] [PubMed] [Google Scholar]
  37. Michos O, Panman L, Vintersten K, Beier K, Zeller R, Zuniga A. Gremlin-mediated BMP antagonism induces the epithelial-mesenchymal feedback signaling controlling metanephric kidney and limb organogenesis. Development. 2004;131:3401–3410. doi: 10.1242/dev.01251. [DOI] [PubMed] [Google Scholar]
  38. Minowada G, Jarvis LA, Chi CL, Neubüser A, Sun X, Hacohen N, Krasnow MA, Martin GR. Vertebrate Sprouty genes are induced by FGF signaling and can cause chondrodysplasia when overexpressed. Development. 1999;126:4465–4475. doi: 10.1242/dev.126.20.4465. [DOI] [PubMed] [Google Scholar]
  39. Motoyama J, Takabatake T, Takeshima K, Hui CC. Ptch2, a second mouse Patched gene is co-expressed with Sonic hedgehog. Nat. Genet. 1998;18:104–106. doi: 10.1038/ng0298-104. [DOI] [PubMed] [Google Scholar]
  40. Nahmad M, Stathopoulos A. Dynamic interpretation of hedgehog signaling in the Drosophila wing disc. PLoS Biol. 2009:7. doi: 10.1371/journal.pbio.1000202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Niswander L, Jeffrey S, Martin G, Tickle C. A positive feedback loop coordinates growth and patterning in the vertebrate limb. 1994:609–612. doi: 10.1038/371609a0. [DOI] [PubMed] [Google Scholar]
  42. Oosterveen T, Kurdija S, Alekseenko Z, Uhde CW, Bergsland M, Sandberg M, Andersson E, Dias JM, Muhr J, Ericson J. Mechanistic differences in the transcriptional interpretation of local and long-range Shh morphogen signaling. Dev. Cell. 2012;23:1006–1019. doi: 10.1016/j.devcel.2012.09.015. [DOI] [PubMed] [Google Scholar]
  43. Oosterveen T, Kurdija S, Ensterö M, Uhde CW, Bergsland M, Sandberg M, Sandberg R, Muhr J, Ericson J. SoxB1-driven transcriptional network underlies neural-specific interpretation of morphogen signals. Proc. Natl. Acad. Sci. U. S. A. 2013;110:7330–4335. doi: 10.1073/pnas.1220010110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pan Y, Bai CB, Joyner AL, Wang B. Sonic hedgehog Signaling Regulates Gli2 Transcriptional Activity by Suppressing Its Processing and Degradation. Mol. Cell. Biol. 2006;26:3365–3377. doi: 10.1128/MCB.26.9.3365-3377.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Panman L, Galli A, Lagarde N, Michos O, Soete G, Zuniga A, Zeller R. Differential regulation of gene expression in the digit forming area of the mouse limb bud by SHH and gremlin 1 / FGF-mediated epithelial-mesenchymal signalling. Development. 2006;3428:3419–3428. doi: 10.1242/dev.02529. [DOI] [PubMed] [Google Scholar]
  46. Parker DS, White Ma, Ramos AI, Cohen Ba, Barolo S. The cis-regulatory logic of Hedgehog gradient responses: key roles for gli binding affinity, competition, and cooperativity. Sci. Signal. 2011;4 doi: 10.1126/scisignal.2002077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Peterson KA, Nishi Y, Ma W, Vedenko A, Shokri L, Junker JP, Zhang X, Mcfarlane M, Oudenaarden A, Van Mikkelsen T, et al. Neural-specific Sox2 input and differential Gli-binding affinity provide context and positional information in Shh-directed neural patterning. Genes Dev. 2012;26:2802–2816. doi: 10.1101/gad.207142.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Probst S, Kraemer C, Demougin P, Sheth R, Martin GR, Shiratori H, Hamada H, Iber D, Zeller R, Zuniga A. SHH propagates distal limb bud development by enhancing CYP26B1-mediated retinoic acid clearance via AER-FGF signalling. Development. 2011;138:1913–1923. doi: 10.1242/dev.063966. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Rainger J, van Beusekom E, Ramsay JK, McKie L, Al-Gazali L, Pallotta R, Saponari A, Branney P, Fisher M, Morrison H, et al. Loss of the BMP antagonist, SMOC-1, causes Ophthalmo-acromelic (Waardenburg Anophthalmia) syndrome in humans and mice. PLoS Genet. 2011;7:1002114. doi: 10.1371/journal.pgen.1002114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Ramialison M, Reinhardt R, Henrich T, Wittbrodt B, Kellner T, Lowy CM, Wittbrodt J. Cis-regulatory properties of medaka synexpression groups. Development. 2012;139:917–9128. doi: 10.1242/dev.071803. [DOI] [PubMed] [Google Scholar]
  51. Ros MA, López-Martínez a, Simandl BK, Rodriguez C, Izpisú a, Belmonte JC, Dahn R, Fallon JF. The limb field mesoderm determines initial limb bud anteroposterior asymmetry and budding independent of sonic hedgehog or apical ectodermal gene expressions. Development. 1996;122:2319–2330. doi: 10.1242/dev.122.8.2319. [DOI] [PubMed] [Google Scholar]
  52. Ros MA, Dahn RD, Fernandez-Teran M, Rashka K, Caruccio NC, Hasso SM, Bitgood JJ, Lancman JJ, Fallon JF. The chick oligozeugodactyly (ozd) mutant lacks sonic hedgehog function in the limb. Development. 2003;130:527–537. doi: 10.1242/dev.00245. [DOI] [PubMed] [Google Scholar]
  53. Roselló-Díez A, Arques CG, Delgado I, Giovinazzo G, Torres M. Diffusible signals and epigenetic timing cooperate in late proximo-distal limb patterning. Development. 2014;141:1534–1543. doi: 10.1242/dev.106831. [DOI] [PubMed] [Google Scholar]
  54. Saffer JD, Jackson SP, Annarella MB. Developmental Expression of Spl in the Mouse. Mol. Cell. Biol. 1991;11 doi: 10.1128/mcb.11.4.2189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Sanders Ta, Llagostera E, Barna M. Specialized filopodia direct long-range transport of SHH during vertebrate tissue patterning. Nature. 2013;497:628–632. doi: 10.1038/nature12157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Shah A, Tenzen T, McMahon AP, Woolf PJ. Using mechanistic Bayesian networks to identify downstream targets of the sonic hedgehog pathway. BMC Bioinformatics. 2009;10:433. doi: 10.1186/1471-2105-10-433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Shlyueva D, Stampfel G, Stark A. Transcriptional enhancers: from properties to genome-wide predictions. Nat. Rev. Genet. 2014;15:272–286. doi: 10.1038/nrg3682. [DOI] [PubMed] [Google Scholar]
  58. Suske G, Bruford E, Philipsen S. Mammalian SP/KLF transcription factors: bring in the family. Genomics. 2005;85:551–556. doi: 10.1016/j.ygeno.2005.01.005. [DOI] [PubMed] [Google Scholar]
  59. Te Welscher P, Zuniga A, Kuijper S, Drenth T, Goedemans HJ, Meijlink F, Zeller R. Progression of vertebrate limb development through SHH-mediated counteraction of GLI3. Science. 2002;298:827–830. doi: 10.1126/science.1075620. [DOI] [PubMed] [Google Scholar]
  60. Towers M, Mahood R, Yin Y, Tickle C. Integration of growth and specification in chick wing digit-patterning. Nature. 2008;452:882–886. doi: 10.1038/nature06718. [DOI] [PubMed] [Google Scholar]
  61. Vargesson N, Clarke JD, Vincent K, Coles C, Wolpert L, Tickle C. Cell fate in the chick limb bud and relationship to gene expression. Development. 1997;124:1909–1918. doi: 10.1242/dev.124.10.1909. [DOI] [PubMed] [Google Scholar]
  62. Verheyden JM, Sun X. An Fgf / Gremlin inhibitory feedback loop triggers termination of limb bud outgrowth. Nature. 2008;454:638–642. doi: 10.1038/nature07085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Visel A, Carson J, Oldekamp J, Warnecke M, Jakubcakova V, Zhou X, Shaw Ca, Alvarez-Bolado G, Eichele G. Regulatory pathway analysis by high-throughput in situ hybridization. PLoS Genet. 2007;3:1867–1883. doi: 10.1371/journal.pgen.0030178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Visel A, Blow MJ, Li Z, Zhang T, Akiyama JA, Holt A, Plajzer-Frick I, Shoukry M, Wright C, Chen F, et al. ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature. 2009;457:854–858. doi: 10.1038/nature07730. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Vokes Sa, Ji H, McCuine S, Tenzen T, Giles S, Zhong S, Longabaugh WJR, Davidson EH, Wong WH, McMahon AP. Genomic characterization of Gli-activator targets in sonic hedgehog-mediated neural patterning. Development. 2007;134:1977–1989. doi: 10.1242/dev.001966. [DOI] [PubMed] [Google Scholar]
  66. Vokes Sa, Ji H, Wong WH, McMahon AP. A genome-scale analysis of the cis-regulatory circuitry underlying sonic hedgehog-mediated patterning of the mammalian limb. Genes Dev. 2008;22:2651–2663. doi: 10.1101/gad.1693008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Wang B, Fallon JF, Beachy PA, Hughes H. Hedgehog-Regulated Processing of Gli3 Produces an Anterior / Posterior Repressor Gradient in the Developing Vertebrate Limb. Cell. 2000;100:423–434. doi: 10.1016/s0092-8674(00)80678-9. [DOI] [PubMed] [Google Scholar]
  68. Welten M, Pavlovska G, Chen Y, Teruoka Y, Fisher M, Bangs F, Towers M, Tickle C. 3D expression patterns of cell cycle genes in the developing chick wing and comparison with expression patterns of genes implicated in digit specification. Dev. Dyn. 2011;240:1278–1288. doi: 10.1002/dvdy.22633. [DOI] [PubMed] [Google Scholar]
  69. Wilson CW, Chuang P-T. Mechanism and evolution of cytosolic Hedgehog signal transduction. Development. 2010;137:2079–2094. doi: 10.1242/dev.045021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Yu J, Valerius MT, Duah M, Staser K, Hansard JK, Guo J-J, McMahon J, Vaughan J, Faria D, Georgas K, et al. Identification of molecular compartments and genetic circuitry in the developing mammalian kidney. Development. 2012;139:1863–1873. doi: 10.1242/dev.074005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Zeller R, López-Ríos J, Zuniga A. Vertebrate limb bud development: moving towards integrative analysis of organogenesis. Nat. Rev. Genet. 2009;10:845–858. doi: 10.1038/nrg2681. [DOI] [PubMed] [Google Scholar]
  72. Zhang B, Kirov S, Snoddy J. WebGestalt: an integrated system for exploring gene sets in various biological contexts. Nucleic Acids Res. 2005;33:741–748. doi: 10.1093/nar/gki475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Zhang Z, Verheyden JM, Hassell JA, Sun X. FGF-Regulated Etv Genes Are Essential for Repressing Shh Expression in Mouse Limb Buds. Dev. Cell. 2009;16:607–613. doi: 10.1016/j.devcel.2009.02.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Zhao C, Meng A. Sp1-like transcription factors are regulators of embryonic development in vertebrates. Dev. Growth Differ. 2005;47:201–211. doi: 10.1111/j.1440-169X.2005.00797.x. [DOI] [PubMed] [Google Scholar]
  75. Zhu J, Nakamura E, Nguyen M-T, Bao X, Akiyama H, Mackem S. Uncoupling Sonic hedgehog control of pattern and expansion of the developing limb bud. Dev. Cell. 2008;14:624–632. doi: 10.1016/j.devcel.2008.01.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Zhulyn O, Li D, Deimling S, Vakili NA, Mo R, Puviindran V, Chen M-H, Chuang P-T, Hopyan S, Hui CC. A switch from low to high shh activity regulates establishment of limb progenitors and signaling centers. Dev. Cell. 2014;29:241–249. doi: 10.1016/j.devcel.2014.03.002. [DOI] [PubMed] [Google Scholar]
  77. Zuniga A, Haramis A, McMahon AP, Zeller R. Signal relay by BMP antagonism controls the SHH/FGF4 feedback loop in vertebrate limb buds. Nature. 1999;401:598–602. doi: 10.1038/44157. [DOI] [PubMed] [Google Scholar]
  78. Zuniga A, Laurent F, Lopez-Rios J, Klasen C, Matt N, Zeller R. Conserved cis-regulatory regions in a large genomic landscape control SHH and BMP-regulated Gremlin1 expression in mouse limb buds. BMC Dev. Biol. 2012;12:23. doi: 10.1186/1471-213X-12-23. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS714455-supplement-1.tif (24.1MB, tif)
9
NIHMS714455-supplement-9.xlsx (10.7KB, xlsx)
10
NIHMS714455-supplement-10.xlsx (1.2MB, xlsx)
2
NIHMS714455-supplement-2.jpg (730.7KB, jpg)
3
NIHMS714455-supplement-3.jpg (682.8KB, jpg)
4
NIHMS714455-supplement-4.xls (37KB, xls)
5
NIHMS714455-supplement-5.xlsx (20.6KB, xlsx)
6
NIHMS714455-supplement-6.xlsx (1MB, xlsx)
7
NIHMS714455-supplement-7.xlsx (24.1KB, xlsx)
8
NIHMS714455-supplement-8.xlsx (56.2KB, xlsx)
11
NIHMS714455-supplement-11.pdf (106.2KB, pdf)

RESOURCES